Technological and Economic Adaptations in Aquaculture

Development in Taiwan

Q4SZTY
a NO
I

Ot

Thesis submitted for the degree of

Doctor of Philosophy

By
Shu-Ching Jeng, M. Sc.

Institute of Aquaculture
University of Stirling
Stirling FK9 4LA
Scotland, UK

November 2002
 Acknowledgements

I would like to expressmy sincere respect and gratitude to my supervisor Professor

James F. Muir for his continuous support, encouragement,guidance, and kindness

throughout my study period. I am extremely grateful to him for tirelessly reading and

correcting this thesis, making constructive suggestionsand comments, which helped

me refine my ideas about the kind of thesis I wanted to write

My local supervisor Professor David S. Liao, Institute of Fisheries Economics,

National Taiwan Ocean University, Taiwan, was generousin offering me advice and

help especially during my field visits to Taiwan.

I am grateful to the Ministry of Education, ROC (Taiwan) for generouslyfunding my

study through a governmentscholarshipwhich enabledme to study in Scotland.

It was Dr. Y. S. Chen who suggestedthat I should study at Stirling, for which I am

very thankful. I would like to extend my gratefulnessto the staff at the Institute of

Aquaculture- Dr. Rodney Wooten, Ms. Julia Farrinton, Ms. SarahWatson, Ms. Hazel

Gentles, Ms. PenelopeBeaton and Dr. Polly Douglas for their help. I would also like

to thank all my friends at the Institute of Aquaculture especially Ms SusanKueh, Mr.

Tayamen,Dr. Nasar Ahmed, Mr. Erkan Gozgozoglu, Dr. Duk Hyun Yoon, Dr. Wara

TaparhudeeDr. Stuart Bunting, Dr. Stuart Bell, Dr. Tariq Ezaz, Mr. Noe Sanches-

Perezand Mr. Brian Ross for their friendship, help and encouragement.

I would like to mention those who contributed the essential information I neededfor

this thesis. Dr. Yew-Hu Chen, Mr. S. H. Tseng,Mr. C. P. Jian, Mr. C. Y. Chang, Mr. J.

I
C. Chen, Mr, I. M. Chen, Mr. G R. Chen, Mr. W. C. Young, Mr. C. F. Shu, Mr. H. Cx

Kaou, Mr. Harry Hsiao, Dr. C. F. Chang and Mr B. H. Lin for help contacting fish

farmers and for their valuable information, when I visited Taiwan for this research. I

also thank Mr. Nobukazu Azuma and Mr. Kiyoshi Katsuyama for the information

about the Japanesemarket and Ms. CatherineHowie for the help on statistics.

I would like to express my appreciation to my external and internal examiner- Dr.

David Whitmarsh and Dr. Lindsay Ross for patiently reading this thesis, and gaving

me precious suggestionsand comments.

On a more personal note, special thanks go to Mr. Keith Baker, Mrs. Josephine Baker,

Miss Catherine Kirk, Miss Patricia Bascom Mn Edward Lee, Mr. D. J. Hwang, Mr. C.

I. Huang, Mr. Young Lee, Mr. Vinyu Veerayangkur,Mr. Raymond Bond and Mr.

Roman Jahnel for their encouragement.

Finally, I would like to thank my parent, sister, brothers, and my girl friend Eiko

Nakano. Their support and encouragementhelped me finish this thesis.

II
 Declaration

I hereby declare that this thesis has been composedentirely by myself and had not
beenpreviously submitted for any other degreeor qualification.

The work of which it is a record has been performed by myself, and all sourcesof
information have been specifically acknowledge.

Shu-Ching Jeng

III
 Abstract

In Taiwan, the history of aquaculturespansover three hundred yearsand the

breakthroughsin the artificial propagationof finfish and shrimp effectively reduced

the industries' reliance on wild fry, thereby stabilizing commercial operationsand

overcoming the barriers for expansion.Taiwan is located very close to Japan,one of

the biggest seafoodimporters in the world, which has also benefited the development

of aquaculture.However, the growing problems of water pollution and the

increasingly high environmental costsgeneratedby aquacultureventureshave made

Taiwan experiencea declining trend in recent years.

To overcomethose constraints,three main areasare described,which then form the

basis of this study.

(1).Adjustment of existing production practices- Milkfish culture, one of the most

vulnerable sectorssuffering from price fluctuation is usedas an example to

understandboth the production cost, market attributes and the ways in which impacts

of variations betweenproduction and price can be reduced.

(2). Improving existing systems-One of the methodsto reducethe use of

undergroundwater is to use superintensive culture in which high densitiesare stable

and water use minimised, and hasbeen tried in Taiwan for eels. However, the cost

and benefits must be evaluatedand as most eel products are exported to the Japanese

market, it is very important to examine the comparativeadvantagesagainstother

countries.

IV
(3). Develop new systems-One of the solutions to the constraintsof land-based

aquaculturein Taiwan is to develop seawater-basedcageculture. This has been

developed in a limited degree in Ping-Tong and Pen-Hu counties but the feasibility

and profitability have not been investigated.

Basedon 274 milkfish farms, 63 traditional eel farms, 5 intensive eel farms, 22 cage

culture farms and 133 consumersfrom different zones,constituted the primary data,

which combined with other secondarydata constructedthis investigation.

The milkfish sector was not economically sound. Farm size in the categories of 4- <5

ha could appear to be more profitable. Cold weather and unstable in price made this

industry more risky. The price was very unstable and strongly correlatedto seasonal

variation of production.

The various forms of averagefinancial appraisalhave shown that intensive eel culture

has a slight advantageover traditional eel culture. However, traditional eel culture has

a higher distribution and the financial advantageof intensive culture is primarily due

to the cheapereel seed.The massproduction of eel from China has causedTaiwan to

lose the comparative advantagein roastedeel for the Japanesemarket.

Cageculture is a new aquacultureventure in Taiwan. The structureof cages,feed and

other facilities still needto be improved. Although Dumerils's Amberjack (Seriola

dumerili) and red porgy (Pagrus major) can make higher profits than other species,

fish farmers still have great expectationfor cobia (Rachycentroncanadus).As

Taiwan's market is not big enough,there is great hope that the Japanesemarket can be

V
developedand cobia can becomea candidatefor sashimi (raw fish).

For sustainable development, aquaculture must be economically viable, ecologically

sound and socially acceptable. To attain these goals, production and marketing

groups, and production area were suggested. Proper administration and management

could help the industry to be sustainable.

VI
 List of Contents

Page
Acknowledgements I
Declaration III
Abstract IV
List of Contents VII
List of Tables XI
List of Figures XV
List of Annexes XVI

Chapter 1 General introduction 1

1.1 The role of aquaculture 1
1.2 Aquaculture backgroundin Taiwan 8
1.3 The researchobjective and structure 14

Chapter 2 Methodology 18
2.1 Introduction 18
2.2 Defining and locating the problems 19
2.3 Hypothesis 21
2.4 Data collection 22
2.4.1 Data source 22
2.4.2 Sampledesign 23
2.4.3 Quantitative and qualitative data 24
2.5 Analyzing and interpreting researchfindings 26
2.6 Researchlocations 26
2.6.1 Milkfish farm 26
2.6.2 Eel farms 26
2.6.3 Cageculture 27
2.6.4 Consumers'opinion 27
2.7 Discussion 27

Chapter 3 Milkfish culture 31

3.1 Introduction 31
3.2 Method of milkfish culture 37
3.2.1 Fry collection and distribution 38
3.2.2 Hatchery rearedfry 39
3.2.3. Generallayout of shallow-water milkfish culture
farms in Taiwan 43
3.2.4 Harvest 47
3.2.5 Overwintering 47
3.2.6 Deep-waterculture 48
3.3 Characteristicsof milkfish producers 49
3.4 Characteristicsof milkfish farms 51
3.5 Economic analysis 55
3.5.1 Cost categories 55
3.5.2 Capital cost characteristics 56
3.5.3 Operating cost characteristics 59
3.5.4 The benefit analysis 63

VII
 3.5.5 Cash-flow and discountedfinancial indicators 68
3.5.6 Price sensitivity 74
3.6 Marketing channels 75
3.7 Supply and price relationships 77
3.7.1 Seasonalvariation of production and price 77
3.7.2 Long-run variation of price and production 81
3.7.3 Annual trend of yield and price of milkfish 81
3.7.4 Actual apparent consumption 82
3.7.5 Consumer perspective 83
3.8 Discussion 88

Chapter 4 Eel culture 93

4.1 Introduction 93
4.1.1 Background 93
4.1.2 Eel seed 97
4.2 Traditional eel culture 101
4.2.1 Introduction 101
4.2.2 Facilities 102
4.2.3 Manipulation of glasseel 103
4.2.4 Managementof on-growing eels 104
4.2.5 Feeding on-growing eels 105
4.2.6 Characteristicsof traditional eel farmers 105
4.3 Super intensive eel culture 107
4.3.1 Culture tanks 108
4.3.2 The processof water flow 109
4.3.3 Management of water quality 110
4.3.4 Husbandry 114
4.4 Financial analysis of eel culture 115
4.4.1 Introduction 115
4.4.2 Cost analysis 116
4.4.3 Benefit analysis 121
4.4.4 Pay back period 123
4.4.5 Cash-flow and discountedfinancial indicators 123
4.4.6 Sensitivity of the price of glasseel in traditional eel
culture 125
4.4.7 Social cost 126
4.5 Post harvestingprocessand marketing of eel 127
4.5.1 Introduction 127
4.5.2 Post-harvest processing 129
4.5.3 Marketing channel 132
4.5.4 Import quantity of the Japanesemarket 133
4.5.5 RCA index 134
4.5.6 Seasonalvariation 136
4.5.7 Analysis of export market 137
4.5.8 Market survey 139
4.6 Discussion 144

Chapter 5 cageculture 150

5.1 Introduction 150
5.2 Develoment conditions 154

VIII
 5.2.1 Culture area 154
5.2.2 Legal rights for cage culture 155
5.3 Systemfeatures 156
5.3.1 Cage structure 156
5.3.2 Anchoring 159
5.3.3 Net management 159
5.3.4 Feeding 160
5.4 Cultured species 161
5.5 Markets 164
5.5.1 Marketing channel 164
5.5.2 Market characteristics 165
5.6 Economic evaluation 166
5.6.1 Introduction 166
5.6.2 Cost analysis 167
5.6.3 Benefit analysis 173
5.6.4 Cash-flow and discountedfinancial indicators 175
5.6.5 Sensitivity 177
5.7 Constraints 180
5.7.1 Environment 180
5.7.2 Diseases 181
5.7.3 Management 182
5.8 Discussion 183

Chapter 6 Sustainability 190

6.1 Introduction 190
6.2 Social and economic issues 193
6.2.1 Introduction 193
6.2.2 Improving financial performance 194
6.3 Environmental factors 208
6.3.1 Introduction 208
6.3.2 Key issues 209
6.3.3 Key biological impacts 211
6.3.4 Physicochemicalimpacts 214
6.3.5 Managementapproaches 216
6.4 Other issuesof sustainability 220
6.4.1 Risk 220
6.4.2 Role of the governmentand other institutions 220
6.5 Discussion 223

Chapter 7 Conclusions 228

7.1 Introduction 228
7.2 Milkfish culture 228
7.3 Eel culture 229
7.4 Cageculture 231
7.5 Sustainability 232
7.6 Further research 233
7.6.1 Economies of scaleand industry aggregation 233
7.6.2 Market research 233
7.6.3 External costs of aquaculture 234

Ix
References 235

Annexes 249

X
 List of Tables

Table 1.1 World aquacultureproduction statisticsand forecast. 5

Table 1.2 The average prices (US$/kg) of aquaculture products of
different continents. 6
Table 1.3 The quantities and values of distant water fisheries, offshore
fisheries, coastal fisheries and aquaculture in Taiwan. 10
Table 1.4 The population, national and average income in Taiwan. 11
Table 2.1 The strength and weaknesses of quantitative and qualitative
research. 25
Table 3.1. The production quantity and value of milkfish in three major
milkfish farming countries (Taiwan, the Philippines and
Indonesia). 33
Table 3.2 The ratios of production value to production quantity of
milkfish in three major milkfish farming countries. 34
Table 3.3 The ratios of polyculture areato monoculture areaand fresh
water areato brackish water areaof milkfish culture in
Taiwan. 36
Table 3.4 The ratio of production value to production quantity of
milkfish in Taiwan. 37
Table 3.5. Socioeconomic characteristics of milkfish farmers. 49
Table 3.6 Averages of education attainment, household size,
percentagesof farmers with outside income and percentage
of monoculture in different agecategories. 51
Table 3.7. The number of farm sizesand statusof milkfish farms
in
surveyed Taiwan. 52
Table 3.8 The frequencies of species in poly-culture with milkfish
ponds. 52
Table 3.9 The frequenciesof water sourcesthat were usedby farmers
for cultivating milkfish. 53
Table 3.10 The use of water sourcesby different farm sizes. 53
Table 3.11 The frequenciesof speciesin poly-culture with milkfish in
different water sources. 54
Table 3.12 The averageyield levels in different farm sizes. 55
Table 3.13. The averagecapital costsof monoculture milkfish farms of
different sizes 57
Table 3.14. The average capital costs of polyculture milkfish farms of
different sizes. 57
Table 3.15 The averagecapital cost for eachtonne of annual output of
monoculture milkfish farm of different size. 59
Table 3.16. The average operating costs of monoculture milkfish farms
of different sizes 60
Table 3.17. The averageoperating costs of polyculture milkfish farms of
different sizes. 61
Table 3.18. The average operating costs of monoculture milkfish farms
of different sizesin the production of 1kg of milkfish. 63
Table 3.19 Returns and benefit ratios for monoculture milkfish, by farm
size. 64
Table 3.20. Returns and benefit ratio for polyculture milkfish, by farm
size. 65

XI
Table 3.21 Gross profits and benefit for monoculture milkfish, by farm
size. 65
Table 3.22 Gross profits and benefit for polyculture milkfish, by farm
size. 66
Table 3.23 The ranges of gross profit for monoculture milkfish, by farm
size. 66
Table 3.24 The ranges of gross profit for polyculture milkfish, by farm
size. 67
Table 3.25 Nominal cash-flow projection for monoculture milkfish farm. 67
Table 3.26 Discounted cash-flow projection for monoculture milkfish
farm. 68
Table 3.27 Nominal cash-flow projection for polyculture milkfish farm. 69
Table 3.28 Discounted cash-flow projection for polyculture milkfish
farm. 70
Table 3.29 Nominal grosscash-flow projection for monoculture milkfish
farm. 71
Table 3.30 Discounted gross cash-flow projection for monoculture
milkfish farm. 72
Table 3.31 Nominal gross cash-flow projection for polyculture milkfish
farm. 73
Table 3.32 Discounted gross cash-flow projection for polyculture
74
milkfish farm.
Table 3.33 The sensitivity of profitability to price. 75
Table 3.34 The seasonalvariation of production of milkfish in Taiwan. 78
Table 3.35 The seasonalvariation of price of milkfish in Taiwan. 79
Table 3.36 Regressionequationsfor price of milkfish in Taiwan. 80
Table 3.37 The actual apparentconsumption (AAC) of milkfish. 83
Table 3.38 Preferencesfor milkfish product forms. 84
Table 3.39 Preferencesfor fresh milkfish product. 84
Table 3.40 Seasonalpreferencefor purchasingmilkfish. 84
Table 3.41. Preferredpurchasingsize of milkfish. 85
Table 3.42 The frequenciesof buying milkfish 85
Table 3.43 The quantity of milkfish purchasedeachtime. 86
Table 3.44. The reasonsthat consumersdo not buy milkfish. 86
Table 3.45 Opinions of respondentsconcerningthe price of milkfish. 86
Table 3.46 Opinions of respondents concerning the quality of milkfish. 87
Table 3.47 Evaluations of respondentsconcerningthe price and quality
of milkfish. 87
Table 3.48 Situations in which consumerswould buy more milkfish. 88
Table 4.1 Output and value of eel culture in Taiwan. 95
Table 4.2 The areafor aquacultureof Anguilla sp. 96
Table 4.3 Catchesof glasseel of different countries. 98
Table 4.4 The quantity and value of caught glasseel in Taiwan. 99
Table 4.5 Socioeconomiccharacteristicsof eel farmers. 106
Table 4.6 Averagesof educationattainment,experience,household
size, percentagesof farmers with outside income and yield
levels in different age categories. 107
Table 4.7 The average annual cost of eel production of per traditional
eel farm. 117
Table 4.8 Average capital cost of traditional eel farm for producing per

XII
 t of eel. 118
Table 4.9 Annual average operating cost of traditional eel farm for
producing 1kg of eel. 119
Table 4.10 The average annual cost of eel production for a super-
intensive eel farm. 120
Table 4.11 Average capital cost of super-intensive eel farm for
producing per t of eel. 120
Table 4.12 Annual average operating cost for a super-intensive farm to
produce 1 kg of eels. 121
Table 4.13 The benefit analysis of traditional eel farm and super
intensive eel farm. 122
Table 4.14 The distribution of profitability for traditional and super-
intensive eel farms 123
Table 4.15 Cash-flow projection for a traditional eel farm and an
intensive eel farm. 124
Table 4.16 Sensitivity analysisof eel fry price. 125
Table 4.17 Social prices of underground water, and shadow cost and
profit of traditional eel culture and intensive eel culture. 127
Table 4.18 The production, consumption and imports of eel in Japan. 129
Table 4.19 Market shareof Japaneseimports of fresh eels, by countries. 133
Table 4.20 Market share of Japan imports of roasted eels, by countries. 134
Table 4.21 RCA indices of fresh eel imports in the Japanesemarket from
1989 to 1998. 136
Table 4.22 RCA indices of processedeel imports in the Japanesemarket
from 1989 to 1998. 136
Table 4.23 Preference for product forms. 140
Table 4.24 Preferredsize of eel for domestic consumers. 140
Table 4.25 Seasonalpreferencefor purchasing. 141
Table 4.26 Preferredways of cooking. 141
Table 4.27 Situations in which consumerswould buy more eel. 141
Table 4.28 The frequenciesof buying eel. 142
Table 4.29 Purchasingquantities of eel at eachpurchase. 142
Table 4.30 The reasonsthat consumersdo not buy eel. 143
Table 4.31 Opinions of respondentsconcerning the price of eel. 143
Table 4.32 Opinions of respondentsconcerningthe quality of eel. 143
Table 4.33 Evaluation of respondentsconcerning the price and quality of
eel. 144
Table 4.34 The percentageof cost structure for producing Japaneseeel
in Taiwan. 137
Table 5.1 Production, value and area of cage culture in Taiwan, 1989 -
1998. 153
Table 5.2 The averageannual growth rate of production, value and area
of seacageculture in Taiwan, 1989 - 1998. 154
Table 5.3 Featuresof Zoning Fishery Right (ZFR) and Specification
Fishery Right (SFR). 156
Table 5.4 The size of fingerling, domestic demand,estimated
production potential and averageprice of different speciesfor
cageculture in Taiwan. 162
Table 5.5 The propagation season of different species for cage culture
in Taiwan. 163

XIII
Table 5.6 Size at harvest,stocking density and culture period of main
fishes cultured in offshore cagesin Taiwan. 163
Table 5.7 The production amount and percentageof cage aquaculture
fish in Pen-Hu, 1997. 164
Table 5.8 Fixed and variable cost structure(exclude feed and fingering)
for cage aquaculture in Taiwan. 169
Table 5.9 The average survival rate and feed conversion ratio (FCR) of
different species. 169
Table 5.10 The averagecost of different speciesof fingerlings for cage
aquacultureper year. 170
Table 5.11 The average cost of feed for different species of cage
aquacultureper year. 170
Table 5.12 The cost structureof the averageoperating cost of production
of cageaquaculturein Taiwan. 171
Table 5.13 The cost structureof the averageoperating cost for producing
1 kg of cage aquacultureproducts in Taiwan. 173
Table 5.14 The average cost of production, amount of production, price,
revenueand benefit for different cage aquaculturespecies. 174
Table 5.15 The average profit, benefit cost ratio and income ratio of
different cage aquaculture species. 175
Table 5.16 The range of profit, benefit cost ratio and income ratio of
different cage aquaculture species. 175
Table 5.17 Nominal cash-flow projection for cage culture 176
Table 5.18 Discounted cash-flow projection for cageculture. 177
Table 5.19 Sensitivity of profitability in changing market price of
products. 178
Table 5.20 Sensitivity of profitability in changing feed prices. 179
Table 5.21 Sensitivity of profitability to changing price of fingerling. 179
Table 5.22 Sensitivity of profitability in different survival rates. 180
Table.6.1 The major difficulties of production and marketing groups
for aquaculturein Taiwan. 198
Table 6.2 Speciescultivated in aquacultureproduction areas. 218

XIV
 List of Figures

Figure 1.1 Market interaction between common property fishery and

aquaculture. 7
Figure 2.1 The definition of research. 18
Figure 2.2 The five stepsof the researchprocess. 19
Figure 2.3 The three stages to fulfill the test of the study hypothesis. 22
Figure 2.4 Researchlocations. 30
Figure 3.1 The production procedure for milkfish culture in Taiwan. 37
Figure 3.2 Organization and operational systems involved in milkfish
fry production in Taiwan. 42
Figure 3.3 Chart of marketing channel for milkfish in Taiwan. 77
Figure 3.4 The average production quantity of milkfish from 1994 to
1998. 78
Figure 3.5 The averageprice of milkfish from 1994to 1998. 79
Figure 4.1 The production amount of eels in China, Japanand Taiwan. 97
Figure 4.2 The processof eel culture in Taiwan. 102
Figure 4.3 The culturing tank of superintensive eel culture. 109
Figure 4.4 Flow chart of superintensive eel culture system. 110
Figure 4.5 The required managementof water quality control. 111
Figure 4.6 The procedure of grading in super intensive eel culture
system. 115
Figure 4.7 The break-evenprices in different prices of eel fry. 126
Figure 4.8 The flow chart of the process for the frozen roasted eel
processin Taiwan. 130
Figure 4.9 The marketing channels of eel. 132
Figure 5.1 The development process of cage culture in Taiwan. 152
Figure 5.2 The traditional structure of cage in Taiwan. 158
Figure 5.3 The marketing channelsof products from cageaquaculture. 165
Figure 6.1 Sustainableaquaculturesystem. 192
Figure 6.2 Economic factors of aquaculture. 194
Figure 6.3 A model of industrial organization analysis. 195
Figure 6.4 The organization of aquaculture production and marketing
groupsin Taiwan. 199
Figure 6.5 Featuresof a demandpull aquaculturemarket. 204
Figure 6.6 Featuresof a supply push aquaculturemarket. 205

xv
 List of Annexes

Annex A. 1 The questionnairefor milkfish farmers. 249

Annex A. 2 The questionnairefor traditional eel farmers. 251
Annex A. 3 The questionnaire for super-intensive eel farmers. 253
Annex A. 4 The questionnairefor cagefarmers. 255
Annex A. 5 The questionnairefor consumersof milkfish. 257
Annex A. 6 The questionnaire for consumers of eel. 259
Annex B1 The culture area for milkfish in Taiwan from 1987 to 1997. 261
Annex B2 The production quantity and value of milkfish in Taiwan
from 1987to 1997. 263
Annex C. 1 The seasonalvariation of fresh eels imported from Taiwan
for the Japanesemarket from 1994 to 1998. 264
Annex C.2 The seasonalvariation of fresh eels imported from China for
the Japanesemarket from 1994 to 1998. 264
Annex C.3 The seasonal variation of processed eels imported from
Taiwan for the Japanesemarket from 1994 to 1998. 265
Annex C.4 The seasonal variation of processed eels imported from
China for the Japanesemarket from 1994 to 1998. 265
Annex D. 1 Production and value of different speciesused in cageculture
in Taiwan. 266

xv'
 Chapter 1

Generalintroduction

1.1 The role of aquaculture

Aquaculture is the farming of aquatic organisms, including aquatic animals and plants,

for food or for commercial purpose. Farming implies some forms of intervention in the

to
rearing process enhance production, such as regular stocking, feeding, maintaining

water quality, protection from predators, etc (Pullin 1993), that increase the yield to a

level above that naturally found in the environment. Therefore, a natural food-producing

is to
system changed a more productive, artificial, or manipulated ecosystem. Part of the

natural functions usually performed by ecosystems such as decomposition of wastes,

exchanges of gases and production of oxygen (by photosynthesis) are accomplished or

supplemented artificially by mechanical aeration or flushing.

Aquaculture may be undertakenwith full or partial culture operation. In full-culture

of
operation, all phases a life cycle, which include breeding, seedor larval rearing, and

feeding the various stagesfrom fry to fingerlings to adults are under control. In partial

culture, there are two varieties; stocking seedwithout attending to subsequentgrowth at

various levels of intensity, and caring for the cultivars to various levels of their life cycles

(Bardach 1997). Suitable environmentsfor aquacultureinclude fresh, brackish and

marine water. The historical evidenceof aquacultureis reportedto go as far back as 2500
BC in Egypt and 500 BC in China (Pillay 1990), although it is over the last few decades

that aquaculture become a rapidly expanding, and globally recognized food production

sector.

When comparedto alternative sourcesof protein, such as terrestrial livestock, fish have

severaladvantages.Firstly, they have a better feed conversionratio. According to Lovell

(1989), the ratio of weight gain per gram of dry feed consumedaverages0.84 for channel

catfish, compared to 0.48 for broiler chickens and 0.13 for cattle. Efficiency is greater, as

fishes are poikilothermic and do not have to maintain a constant body temperature. They

need less energy to maintain position and to move in water than animals do on land. They

also excretemost nitrogenouswastesas ammoniarather than urea and uric acid,

therefore, energy loss in protein metabolism is much lower (Ackefors et al. 1994).

Secondly, fish protein is of better quality as it is associatedwith a low calorie content, a

low saturatedfatty acid content, and a high content of poly-unsaturatedfatty acid (Q3).

Thirdly, many fish have a higher percentageof consumablelean flesh. In dressedcatfish

for example, the consumablelean flesh is 81% comparedto 60,50 and 65% for beef,

pork, and chicken, respectively (Lovell 1989).

The production and consumptionof fish has madelarger gains over the last four decades

(Williams 1997). However, the contribution of fish to sustainablefood security is now

undergoing a transition to increasingscarcity, and capturefisheries have beenrelatively

stable over recent years.Natural aquaticecosystemsare photosynthetically driven

systemswith relatively long food chains and have high energy losseswhich ultimately

2
result in a low yield per unit area.For example,overall biomassyield for the earth's

oceanis less than 2 kg per hectareper year (kg/ha/year) (Ackefores et al. 1994).

About 70% of the world's marine stock is fully exploited, overexploited, depleted or in

the process of rebuilding as a result of depletion (Karnicki. 1995), while most of the other

30% of stocks consist of low market-valuespecies.For the two decadesfollowing 1950,

world capture fisheries production increasedon averageby 6 percent annually. However,

during the 1970sand 1980s,the averagerate of increasedeclined to 2 percentper year,

falling to almost zero in the 1990s(FAO 2001). Many managementregimeshave not

been adequateto sustainthe fishery resource,or to counteractthe negative impacts

created by heavy fishing, habitat degradation and habitat loss. Encouraging more people

to catch fish for food or profits is no longer an option in most parts of the world. Instead,

encouragingpeople to grow their own fish or participate in artificial stock enhancement

is an increasingoption (Williams 1997).By contrast with the lower yields from capture

fisheries, by adding and controlling inputs the range of production from aquacultureis on

the order of 1,000 to 1,000,000kg/ha/year (Ackefores et al. 1994).

Many factors have actedpositively on the demandfor fish, e.g. growing affluence and

growing population numbers.Economic and social factors such as greaterdisposable

income, the price of fish relative to other animal proteins, trade opportunities, and dietary

and health preferencesall contributed to changing fish consumption patterns.

Furthermore, an ever-growing human population exacerbatesthe situation. The world's

population has been growing at a rate of 1.8 % per annum (FAO 2001) and it is estimated
that the population will attain 7 billion by the year 2010. To satisfy the world annual

demand,at an averageof 13 kg per capita, it will be necessaryto reach a supply level of

91 million tonnes of fish. It is widely considered that deficiencies of supply can be

compensated through aquaculture, i. e., from hunting to farming fish.

The aquacultureindustry has undoubtedly seengreat successover the last few decades,

with world production rising from 2.6 million tonnesin 1970 to more than 45 million

tonnesin 2000. Production is dominatedby Asia, in 2000, accounting for more than 91%

of total production and 82.1% of total value. From 1980to 2000, the annual average

growth rate of quantity and value were 9.5 and 10.2%. The highest growth rate of

quantity was in South America, attaining 19.5%, followed by Africa, Asia, North

America and Europe, attaining 13.8,10.0,6.5 and 4.1%, respectively.The highest growth

rate of value was in Africa, attaining 24.5%, followed by, South America, Asia, North

America and Europe, attaining 14.5,10.6,7.9 and 6.4%, respectively (Table 1.1). The

averageprice was highest in South America, followed by Europe, North America and

Africa, and lowest in Asia (Table 1.2).

4
Table 1.1 World aquaculture production statistics and forecast.
Unit: Millions of tonnes
Millions of US$
Year Asia Europe America, America, Africa Total
North South
Amount Value Amount Value Amount Value Amount Value Amount Value Amount Value
1980a 6.2 0.9 0.2 0.02 0.03 7.4
84.3 (12.5) (2.5) (0.2) 0.4
1982a 6.7 1.1 0.3 0.04 0.03 8.2
(82.1) 13.4 (3.4) (0.5) (0.4)
1984a 8.5 9.3 1.2 1.7 0.4 0.5 0.06 0.3 0.04 0.03 10.2 12.0
(83.6) (78.0) (11.8) (14.5) (3.5) (4.4) (0.6) (2.6) (0.4) (0.2)
1986a 10.7 13.3 1.4 2.4 0.4 0.6 0.07 0.4 0.06 0.04 12.7 16.7
(84.7) (79.5) (10.9) (14.3) (3.2) (3.4) (0.6) (2.4) (0.4) (0.3)
1988a 13.4 19.2 1.4 3.2 0.4 0.7 0.1 0.8 0.07 0.1 15.5 24.0
(86.6) (79.8) (9.2) (13.3) (2.7) (2.9 (0.9) (3.1 (0.5) (0.4)
1990a 14.5 21.4 1.6 4.0 0.4 0.8 0.2 0.7 0.08 0.2 16.8 27.2
(86.2) (78.7) (9.6) (14.8) (2.4) (3.0 (1.1) 2.5 (0.5) (0.6)
1992a 18.8 26.4 1.4 3.6 0.5 1.0 0.3 1.2 0.1 0.2 21.2 32.5
(88.9) (81.1) (6.5) (11.1) (2.4) (3.1) (1.4) (3.7) (0.5) 0.6)
1994a 25.3 34.3 1.5 3.7 0.5 1.2 0.3 1.3 0.1 0.2 27.8 40.9
(90.9) (83.8) (5.3) (9.1) (1.9) 2.9 (1.2) 3.2) (0.3) (0.5
1996a 30.9 39.7 1.7 3.9 0.6 1.3 0.6 2.0 0.1 0.3 33.9 47.4
(91.1) (83.8) (4.9) (8.2) (1.7) (2.7) 1.6) 4.2 (0.4) 0.6)
1998a 35.5 41.6 1.9 4.3 0.7 1.5 0.7 2.4 0.2 0.4 39.1 50.4
(90.8) (82.5) (4.9) (8.5) (1.7) (2.9) (1.7) (4.8) (0.5) (0.9)
2000 a 41.7 46.3 2.0 4.6 0.7 1.7 0.7 1 2.6 0.4 1.0 45.7 56.5
(91.3) (82.1) (4.4) (8.2) (1.5) (3.0) 1.6) (4.5) (0.8 (1.7)
2010 51.8

2025 92.6
2035 123.9
Growth 10.0 10.6 4.1 6.4 6.5 7.9 19.5 14.5 13.8 24.5 9.5 10.2
rate
Data source:a: FAO (2002). www. fao.org/fi/figis/tseries/index.jsp
b: New (1997)
The figures in this table are undeflated
Figures in parenthesesrepresentthe percentagesof annual total.
The growth rate of quantity is from 1980 to 2000, and value is from 1984 to 2000.
Table 1.2 The averageprices (US$/kg) of aquacultureproducts of different continents.

Year Asia Europe America, America, Africa Total

North South
1984 1.09 1.42 1.25 5.00 0.75 1.18
1986 1.24 1.71 1.50 5.71 0.67 1.31
1988 1.43 2.29 1.75 8.00 1.43 1.55
1990 1.48 2.50 2.00 3.50 2.50 1.62
1992 1.40 2.57 2.00 4.00 2.00 1.53
1994 1.36 2.47 2.4 4.33 2.00 1.47
1996 1.28 2.29 2.17 3.33 3.00 1.40
1998 1.17 2.26 2.14 3.43 2.00 1.29
2000 1.11 2.30 2.43 3.71 2.50 1.24
Data source:calculated from Table 1.1.
The figures in this table are undeflated

It is also proposedthat aquaculturecan mitigate againstoverfishing of certain species,for

example, Atlantic salmon.To demonstratein theoretical terms, in fig1.1, curve C

representsthe price-supply characteristicsof a common-propertyfishery. The supply

quantity increaseswhen the price increasesbut after the catch effort exceedsthe

maximum sustainableyield, the supply quantity will decreaseand the price will still

increase.Curve A, representsthe aquaculturesupply; curve A+C, is then the horizontal

sum of the aquaculturesupply and common property supply and Curve D, the demand

quantity. Without aquaculture,the common-propertyequilibrium output will be Q1 and

price will be P1. With such a high price, biological overfishing will take place. However,

the introduction of an aquaculturesupply to the market will lower the equilibrium price to

P2, raise the equilibrium quantity to Q2 and lower the risk of overfishing. At price P2, the

harvesting level of open accessfishing on Q3 is lower than the maximum sustainable

yield on Q4.

6
 P1

Price

P2

01 03 04 Q2
Quantity
Fig. 1.1. Market interaction between common property fishery and aquaculture (Adapted

from Ridler, N. B. and M. Kabir 1988).

Apart from generatingemployment,income and foreign exchange,aquacultureprovides

markets and raw material input to food manufacturing (Lee 1997).The demandof input

materials (backward linkages) and the supply of output commodities (forward linkages)

by aquaculturedrive a chain of economic activities, which enhancefurther growth of

economy. The demandof input materials like seeds,feeds,nets, fertilisers, chemicals,

etc. increasesthe income of input producersand entails the ancillary activities, such as

hatcheries,fingerling grow-out, fry collection from nature, feed processing,

pharmaceuticals,specialist engineeringservicesand suppliers, harvesters,post-harvest

processing,marketing and storage.The supply of fish products can depressthe prices of

fish to benefit the consuming public, increaseforeign exchangereservesthrough export

of fish to other countries and saveforeign exchangeby increasingthe consumption of

locally producedrather than imported fish. Against this generally positive background,
the evaluation of the aquacultureindustry in Taiwan, the focus of this study can be set

out.

1.2 Aquaculture backgroundin Taiwan

With the Tropic of Cancerpassingthrough its southernpart, Taiwan is a subtropical area.

Monsoons prevail from October to March, and from April to September,periodic

typhoon and thunderstormsbring abundantprecipitation. Over the period from 1949to

1986, the average precipitation was 2504 mm, about 3.5 times the world average (Hsiao,

1994a). The average temperature is 23°C and is higher than 20 °C for 9 months after

April (Hsiao 1994b). In the southern part of Taiwan, water temperature rarely drops

below 10 °C (Liao 1995a).

Though the climate is very suitable for developing aquaculture,Taiwan still facesa

shortageof water resource.Its high population density meansthat each individual can

only shareabout 4,500 m3 of water, about one-sixth of the world average,from

precipitation sources,per year (Hsiao, 1994a).The unevendistribution of precipitation

over different seasonsmakesthe problem of water shortageeven worse. Most of the

precipitation, about 78%, is from May to October. In the southernpart of Taiwan, the

shareof precipitation during this period can be as high as 90% (Hsiao 1994a).

Becauseof the deteriorating environment (Huang 1997), increasingresourceand energy

cost of exploitation, and, becauseof increasing application of control over Exclusive

Economic Zones (EEZ) in the world, there are not expectedto be any significant wild
catch increases for most commercially important species. In Taiwan also, current records

suggest that the wild fish catch has reached a plateau (Taiwan Fisheries Bureau 1987-

1998) (Table 1.3). Among the sources of fishery products, distant water fisheries* have

the highest production quantity, accounting for more than 50%, followed by aquaculture,

offshore fisheries and coastal fisheries, accounting for about 20%, 20% and 3%

respectively. Table 1.3 shows that except for slightly increased distant water catches,

outputs from other fisheries (offshore and coastal fisheries) have declined. The growth

1.2, in distant, offshore, coastal fisheries, and aquaculture,

rates were -3.8, -2.1 and -2.9

respectively. Thus, as in the global situation, aquaculture may also be promoted to

compensate and supplement fisheries supply.

Production from aquaculture is similar to other fisheries, and has been stable at about 250

to 340 thousand tonnes from 1990 to 1999. As a consequence of its development to date,

is
aquaculture a valued industry in Taiwan, is
and second only to distant water fisheries in

terms of its total value produced when compared with other fisheries. In 1999, it

accountedfor 19% of total quantity of fishery production, but 26% of its total value

(Table 1.3). The averageprices of products from aquacultureare between 90-130 NT$,

far higher than those from distant water fisheries and offshore fisheries, but similar to

those from coastalfisheries.

*Distant water fisheries, offshore fisheries and coastal fisheries refer to the fisheries which beyond 200
miles, between200 and 12 miles and within 12 miles respectively.

9
Table 1.3 The quantities and values of distant water fisheries, offshore fisheries, coastal
fisheries and aquaculturein Taiwan. Unit: 103tI 106NT$
1US$=32NT$
Year Distant Water Offshore Fisheries CoastalFisheries Aquaculture
Fisheries
V V/0 0 V V/ V V/ V V/
1990 767.0 35249 46.0 292.4 18235 62.4 48.4 3960 81.9 344.3 31531 91.6
(52.8) (39.6) (20.1) (20.5) (3.3) (4.5) (23.7) (35.4)
1991 714.3 32204 45.1 266.9 17457 65.4 41.2 3517 85.3 291.9 30256 103.7
(54.3) (38.6) (20.3) (20.9 (3.1) (4.2) (22.2) (36.3)
1992 737.6 34622 46.9 280.5 16394 58.4 45.4 3327 73.3 261.6 29292 112.0
(55.7) (41.4) (21.2) (19.6) (3.4) (4.0) (19.7) (35.0)
1993 835.0 42701 51.1 258.6 17286 66.8 43.4 3271 75.3 285.3 29816 104.5
(58.7) (45.9) (18.2) (18.6) (3.1) (3.5) (20.1 (32.0)
1994 683.8 36047 52.7 242.3 16084 66.4 39.8 3430 86.2 288.0 33566 116.6
(54.5) (40.4 19.3) 18.0 (3.2) (3.8 (23.0 37.7
1995 709.5 43084 60.7 256.0 16931 66.1 43.5 3976 91.4 286.6 36514 127.4
(54.8) (42.9) (19.8) 16.8 (3.4) 4.0 (22.1 (36.3)
1996 669.0 43828 65.5 256.7 16586 64.6 41.0 4256 103.7 272.5 32727 120.1
(54.0) (45.0 (20.7) 17.0 (3.3) (4.4) (22.0) (33.6)
1997 748.3 49041 65.5 247.6 16673 67.3 40.6 4524 111.5 270.2 26944 99.7
(57.3) (50.5 (18.9 17.2 (3.1) (4.7) (20.7 (27.8)
1998 839.2 49205 58.6 209.7 14504 69.2 43.6 4382 100.4 255.2 27382 107.3
(62.3) (51.5) (15.6) (15.2) (3.2) (4.6) (18.9) (28.7)
1999 856.7 48914 57.1 205.6 13182 64.1 39.9 4181 104.8 263.1 23508 89.4
(62.7) (54.5) (15.1) (14.7) (2.9) (4.7) (19.3) (26.2)
Gro- 1.2% 3.7% -3.8% -3.5% -2.1% 0.6 -2.9% -3.2%
wth
rate
The figures in parenthesesare percentageof total quantities/values.
Data source:FisheriesAdministration (2000).
The figures in this table are undeflated

Although production from aquacultureand other fisheries has broadly stabilisied, the

population, national and averageincome is still increasing.From 1987 to 2000, the

population increasedby more than 2 millions (0.9% per year), and the national (5.96%

per year) and averageper capita income (4.96% per year) increasedby 112.2 and 87.6%,

respectively (Table 1.4). The potential demandfor fishery products may be increasedby

10
thesefactors. However, the total output (-0.7% per year) and value (0.1% per year) of

fisheries did not change to a similar degree.

Table 1.4 The population, national and average income in Taiwan.

Year Population National income Average income

(Thousandpeople) (Million NT$) (NT$)
1987 19,725 4,116,645 210,420
1988 19,954 4,445,103 224,636
1989 20,157 4,757,383 237,798
1990 20,401 4,999,236 247,120
1991 20,606 5,407,640 264,368
1992 20,803 5,724,912 277,173
1993 20,995 6,128,190 293,943
1994 21,178 6,448,149 306,546
1995 21,357 6,734,717 317,451
1996 21,525 7,142,414 333,948
1997 21,743 7,690,149 356,398
1998 21,929 8,159,200 374,669
1999 22,092 8,475,919 386,103
2000 22,277 8,736,165 394,853
Growth rate 0.9% 5.96% 4.96%
Data source: Directorate General of Budget Accounting and Statistics Executive Yuan,
R. O. C. (www. dgbas. ogv. tw)
The figures are inflation-adjusted and the baseyear is 1996.

In Taiwan, the history of aquaculturespansover three hundred years.In its development,

three major breakthroughshave been commonly recognized.First, the techniquefor

propagation of freshwater finfish was set up after the successof induced spawning of

grasscarp (Ceteuopharyngodonidellus) and silver carp (Hypopthalmichthysmolitrix) in

1963. Second,in 1968, the developmentof the artificial propagation of tiger prawn

(Penaeusmonodon) provided a good foundation for the culture of other prawn species.

Finally, the successfulinducing of spawning and larvae rearing of grey mullet (Mugil

cephalus) in 1969 establisheda solid foundation for artificial propagation techniquesfor

11
other marine finfish species (Liao et al., 1995a). These breakthroughs in the artificial

propagation of finfish and shrimp effectively reduced the industries' reliance on wild fry,

thereby stabilizing commercial operations and overcoming the barriers for expansion.

Advanced techniquesallowed the industry to diversify culture species.In 1970,the

industry was already successfullyculturing 44 species,including freshwater and marine

finfish, crustaceans and shellfish. This reached 105 species in 1991 (Liao 1991). This

diversification of aquaculture also helped to satisfy the growing needs of the domestic

seafoodmarket.

One of the reasonssuggestedfor the successof aquaculturein Taiwan over the past few

decadeshas been the diligence of the producers.According to a survey on fisheries and

aquaculture households in 1992, around 59.1% of aquaculture operators held only an

elementary school diploma, while 36.7% graduatedfrom junior and senior high school

(Taiwan Fisheries Bureau 1994). Despite this, they were quick to learn and adapt the

techniques transferred through training programs and technical support provided by the

extension services. These are currently available from a range of sourcesincluding the

Taiwan Fisheries Research Institute (TFRI), Academica Sinica, some colleges and

universities, the national Fisheries Administration and fishermen's associations.Apart

from this external technical delivery, fish farmers also developed innovations of their

own. For example, the first natural spawning of milkfish was performed by a fish farmer

(Lin 1984). The application and innovation of farmers was thus an important factor in

the rapid growth of the industry (Liao et al. 1995a).

12
Taiwan is also located very close to Japan, one of the biggest seafood importers in the

world. In 1999, Taiwan exported 418,755 t of fishery products to Japan. It accounted for

32,610 million NT$ (-. 1,000 US$) (32NT$=1US$) (Fisheries Administration 2000).

Therefore, the developmentof aquaculturein Taiwan is very dependenton the changesof

Japaneseseafoodmarket.

However, in spite of thesepositive factors, aquaculturehas experienceda declining trend

in recent years. This could be part attributed to increasing problems of water pollution

and the increasingly high environmental costs generatedby aquacultureventures(Huang,

1990).

Becausethe water quality in major rivers and damshas significantly deteriorated,the

main fresh-water resourcefor aquacultureis now ground water. For example,the length

of severelypolluted rivers increasedfrom 5.69% of the total river length in 1983to

10.4% in 1992 (Huang 1997). Aquaculture has becomea significant fresh water

consuming industry and accountedfor 25% of the total ground water consumption in

Taiwan, while the other agriculture ventures accounted for 48% (Huang 1997). Due to

too much use of undergroundwater, aquaculturegenerateda huge environmental cost,

with seriousland subsidenceresulting in the late 1980sand early 1990s(Lee 1997).

Basedon field investigations in 1992,in the southernpart of Taiwan, the areaof land

subsidencewas estimatedto extend to1010 km2. The most seriouscaseswere located in

13
the Ping-Tong area with a local subsidence of 2.5 m vertically (Lee 1997). This very high

environmental cost has required limitations to be imposed.

Apart from the environmentalproblems,marketing constrainsand production instability

have hamperedthe developmentof aquaculturein Taiwan. For example,the Japanese

market for eels has stimulated competition from other Asian countries, especially from

China, while erratic production of milkfish has induced significant price fluctuation. Such

features are considered a hindrance to the orderly growth of the industry. Thus as the

domestic market is the major destinationfor milkfish, if the quantity supplied is too large,

the prices may be too low to offset the cost of production. It is clear thereforethat if the

aquacultureindustry in Taiwan is to be sustained,and to grow in the directions which

might be suggested,a range of potentially seriousimpedimentsmay have to understood

and overcome.

1.3 The researchobjective and structure

A number of approacheshave beenconsideredfor overcoming constraintsin the

aquaculturesector.Three main areascan be described,which form the basis of this study.

(1).Adjustment of existing production practices- e.g. diversifying productive and

marketing strategiesin coastalponds. All the farmers and processorsin the

aquaculturesector want to predict supply, forecastprice and understandwhat are

the most important forces that determineboth the quantities supplied and prices.

Milkfish culture, one of the most vulnerable sectorssuffering from price

14
 fluctuation is used as an example to understand both the production cost, market

attributes and the ways in which impacts of variations between production and

price can be reduced.

(2). Improving existing systems-e.g. intensification of eel culture. One of the

methodsis to use superintensive culture, which can support high densitiesand

minimise water exchange,reducing the use of ground water. This has been tried in

Taiwan for eels,but the cost and benefits of intensifying from traditional methods

be As
must evaluated. most eel products are exported to the Japanese
market, it is

very important to examine the comparative advantages against other countries.

(3). Develop new systems-e.g. cage culture. One of the solutions to addressing

the constraintsof land-basedaquaculturein Taiwan is to develop seawater-based

cageculture. This hasbeencarried out to a limited degreein Ping-Tong and Pen-

Hu counties but the feasibility and profitability have not been investigated.

There are therefore three casesassessingaquaculturedevelopmentin Taiwan, i. e.

milkfish, eel and cage culture. These three casesare describedin three chapters.In each

chapter,there are three major sections- outline of situation, financial and economic

and
perspectives, marketing issues.The thesis is divided into sevenchaptersand. the

layout of the study is as follows:

15
Chapter 1: generalisethe situation of world aquacultureand aquaculturebackgroundin

Taiwan, including its environment, situation, process of development and constraints

confronted. Finally, the research objective and structure are described.

Chapter2: presentsthe problems and researchhypothesis,describesthe research

methods,including data collection, sample design, data analysis and interpretation, and

the selection of researcharea.

Chapter3: presentsthe outline of milkfish culture, including the situation of milkfish

culture, characteristic of milkfish producers and farms, financial analysis, marketing

channels,relationship between supply and price, consumption level and consumer

perspective.

Chapter4: presentsthe outline of eel culture, including the situation of eel culture,

managementof traditional and superintensive eel culture, characteristicof farmers,

financial analysis,post-harvestingprocess,marketing channel, analysis of Japanese

market, and Taiwanesemarket survey.

Chapter5: presentsthe outline of cageculture, including the situation of cageculture,

legal rights, culture system,financial analysis,marketing channelsand constraints.

Chapter6: brings the findings from eachof thesecasestogether to make an overall

of
assessment sustainability. In this chapter, the sustainability of aquaculturein Taiwan is

16
examined and the ways and methods that might be suitable for the development of

aquaculture in Taiwan are suggested.

Chapter7: provides conclusionsand recommendationsfor further researchbasedon the

results from the previous chapters.

17
 Chapter 2

Methodology

2.1 Introduction

The purpose of this research is concerned with diagnosing the current situation,

setting objectives, and generatingpotential alternative strategiesfor the development

of aquaculturein Taiwan. To achievethis, it is necessaryto understandthe situation,

strength,and weaknessof aquaculturein Taiwan, from which the future trends and

issuescan be proposedand proper objectives and potential strategiescan be

developed.To attain theseaims, the collection, analysis and interpretation of data

were carried out, from which information could be developedto assistin diagnosing

and deciding objectives and potential strategies.The definition of this researchis

illustrated in Fig. 2.1.

Collect
Diagnose
Summar
i Information

Planning
Interpret
Fig 2.1 The definition of research(modified from Kent. 1999).

This chapterdescribesthe researchstrategyand methodology followed to achievethe

objective of the study.There were five basic stepsin the researchprocess(Fig.2.2).

They are

(1) defining and locating problems,

18
 (2) developing hypotheses,

(3) collecting data with which to test and modify the hypotheses,

(4) analyzing and interpreting research findings, and

(5) reporting researchfindings.

Thesefive stepswere viewed as an overall approachto conducting the researchrather

than a rigid set of rules. In conducting this research,each of the stepsand how they

can best be adjustedwere considered.It also describesthe selection of the research

sites

I Defining and locating problems

Developing hypotheses

--- Collecting data

Reporting researchfinding

Analyzing and interpreting researchfinding

Fig 2.2 The five stepsof the researchprocess.

2.2. Defining and locating the problems

Taiwan has had a history of Chinesetraditional fish farming for at least three

centuries.The developmentcan be divided into three distinct stages.In the traditional

stage 1661-1962(Liao, 1992), the first recordedcultured fish was milkfish (Chanos

chanos).In this period, production from aquaculturewas low with total output of no

19
more than 20,000 t per annumin the 1930sand 1940s(Sheeks,1989). Production

gradually increased,peaking at 53,453 tin 1961.A more prosperousstagedeveloped

from 1963to 1991.With systematicpromotion, diversification (Liao, 1992) and

modernization in the 1960s and 1970s, output increased more than ten-fold, to reach

201,925 tons in 1981, accounting for 22 percent of the total weight and 33 percent of

the total value of all fisheries production. In 1990,the peak year to date,total

production reached344,263 t, valued at 31,530,575,000NT$, with an averagesale

value per t of about 92,000NT$ (2875 US$).

The rise in output was largely due to severalrevolutionary breakthroughsin research

during the 1960sand in subsequentdecades,as earlier described.However, the

collapse of tiger prawn rearing, the competition from other countries (such as China

and Malaysia) in exporting eel to Japanesemarket and the shortageof eel fingerlings

(elver) made theseindustries decline. Limited land and water resourcesalso startedto

restrict developmentof aquaculturein Taiwan. For example, the excessivedraw down

of undergroundwater had causeda seriousproblem with the water table, resulting in a

partial settling of land and salinisation of undergroundwater in the vicinity of

aquacultureareas(Liao 1992).

In 1990, Taiwan commenced a process of transformation to address these constraints,

and to attempt to sustainthe industry. There have been various directions for the

aquaculturein Taiwan to change.The first direction has been to go back to traditional

fish culture, such as milkfish and Chinesecarps.The seconddirection has beento

apply new techniquesfrom other countries, such as using water recirculating systems

for eel culture to reducethe usageof fresh water.The third direction has been to shift

20
the culture areatoward the ocean,using systemssuch as seacages.The aim of this

researchis to examine eachof thesestrategies,to assesstheir relative potential. As

potential directions, 3 cases were identified, i. e. milkfish culture in traditional style,

eel culture in recirculating style and marine fish culture in sea cages.

2.3 Hypothesis

To develop and structurethe researchwork, an objective statementwas set out

including hypothesesdrawn from both previous researchand expectedresearch

findings. In this sense,the hypothesiscan be consideredas an informed proposal or

assumptionabout a certain problem or set of circumstanceswhich can then be subject

to test and evaluation, following which a clearer range of actions or potentials could

be recognised(Dibb et al., 1997).The basic hypothesisof this researchconcerns

whether aquacultureis capableof continuing and developing, and whether the change

of structure and operation in aquacultureis feasible.The fundamental hypothesisis

that it is feasible to go back to traditional fish culture, to apply new technology from

other countries and to shift the culture areatoward the ocean.Three casesare used to

test this linked hypothesis.To do this, somefactors must be understood.They are the

biotechnical, social economic background,and from these,the definition of

opportunities, goals, strengths,weakness,threatsand strategies.The three stagesfor

attaining this are illustrated in Fig. 2.3. In the first stage,must be understoodthe basic

situations, which include the input, market, environment, infrastructure and structure

of aquaculturein Taiwan. Theseinclude what, where, how much (capacity), who and

how to operate.The secondstageis to understandhow are they changing, why are

they changing and when are they changing?The third stageis to predict how will it go

in the future and how to adjust?

21
 Hypothesis

First stage
To understand input, market, environment,
infrastructure (support framework), and structure.

Second stage
To understand how, why and when these factors are
changing.

Third stage
To predict how thesesfactors will go in the future
and how to adjust.

Fig 2.3 The three stagesto fulfill the test of the study hypothesis.

2.4 Data collection

2.4.1 Data source

Both primary and secondarydata were collected. Primary data were observedand

recordedor collected directly from respondents.This included interview by

questionnaire,discussingwith key informants (fish farmers, marketing traders and

school teachersetc.) and personalcontacts.There are three main categoriesof

questionnaire-interview, telephoneand postal surveys.To avoid misunderstandingof

questionnaires,interview surveyswere used. However, some farmers only accepted

telephonesurveys.Four versions of questionnaireswere usedfor each producer

group: milkfish farmers, traditional eel farmers, superintensive eel farmers and cage

farmers.

22
Secondarydata refers to information and statisticsthat are already collected, and

processedto varying degree.Theseinclude governmentreports, such as the Year

Book of the Taiwan Fishery Bureau; unofficial published statistics, the result of

previous in-depth research, and the statistics from internet (such as global production

of milkfish; population; national and averageincome in Taiwan) etc. Secondarydata

can be unevenin coverage,availability and accessibility.It may focus only on specific

topic and sometime,it can also be unreliable. However, secondarydata still have

advantages,often providing information that is not available elsewhereand that

cannot be collected in a project context.

2.4.2 Sampledesign

Two kinds of sample design can be identified- purposive and representative samples.

Purposive samplesare generatedwhen the caseis chosenby the researchers'own

judgement. The selection may be deemedto be the most important, reflect a variety of

extremes,or those they are typical. Representativesamplesare chosenin a way that

reproducesthe structure and featuresof the population through casesfrom which the

sampleswere drawn. They are used to make estimatesof a population's

characteristics (Kent 1999). In this research, representative samples were chosen.

Lists of fish farmers were obtained from fishery agenciesof local governmentsand

local fishermen associations,and respondentswere selectedby simple random

sampling.The sampling points are carefully chosenin such a way to obtain a

representativecross-sectionof types of areas.Before interviews, they were contacted

by telephonein advance.Face-to-faceand telephoneinterviewing methodswere used.

23
In the areaof marketing research,three approachescan be defined- exploratory,

descriptive and causalmarketing research.Exploratory researchis aimed at

generating insight, idea and hypotheses rather than measuring or testing them.

However, descriptive research is characterized as being concerned with measuring or

estimating the size, quantities or frequencies.Causalresearchis typically seenas

concernedwith establishingcause-and-effectrelationshipsin an attempt to explain

why things happen(Kent, 1999).In this study,consumersurveyswere carried out and

were descriptive in nature. 2 kinds of questionnaireswere used for consumers'

survey- for milkfish and for eel The

consumers. reasonthat the consumers'survey for

cage culture was not performed is that some species from cage culture (such as

grouper) overlappedwith those from land basedaquaculture.

Before the survey,pre-testswere implementedon pilot groups of respondentsto see

how they worked. As a consequenceof the pre-test, someredesignsof questionnaires

were conducted.The questionnairesare presentedin Annex Al- Annex A6.

2.4.3 Quantitative and qualitative data

Both qualitative and quantitative information was sought. Qualitative researchdoes

not seekto establish absolutevalues but rather to build up an accurateinterpretation

of what is being researched,through many different descriptive sources.They are

therefore,non-numerical records and arise as words, phrases,statements,narrative,

text or pictures (Kent. 1999). Qualitative data can be obtained from interviews,

discussionand observation.Qualitative researchis usually used for exploratory

purposes,in areaswhere little researchhas been done, and may be usedto diagnose

problems in detail.

24
Quantitative researchseeksas far as possible to place firm, absolutelevels or values

on the investigation. Quantitative data are numerical records arising from a processof

measurement.This may be done by using simple count (such as membersin family,

numbersof farms), economic calculation (such as cost and benefit analysis,net

income) and statistical inference techniques(such as inference of the trend of

price/production). Table 2.1 showsthe comparative strengthsand weaknessesof

quantitative and qualitative research.

Table 2.1 The strength and weaknesses of quantitative and qualitative research.
Type of research Strengths Weaknesses
Qualitative Provides the initial basis for More prone to bias because
further quantitative work (may of reliance on interpretation
be sufficient on its own) Difficult to infer population
More participatory characteristicsfrom a small
Can be quick and low cost sample
Good for social processesand Can be very time consuming
context
Can explain causesof
quantitative finding
Quantitative Can be more concrete, Concretenesscan be misled
systematic Can be very extractive
Can infer population Tendencyto collect too much
characteristicsfrom a small data and to produce over-
sample complex analysis
Can test the significance of
quantitative findings
Adapted from DFID (Department for International Development). In the website:
www. livelihoods. org

Effective researchin this context needsa combination of qualitative and quantitative

researchmethods,the combination of which will vary according to the tasks. Here,

eachof the three casescontains qualitative and quantitative data, with qualitative data

being used for the basis of eachcaseand quantitative data for economic and statistic

inference.

25
2.5 Analyzing and interpreting researchfindings

Three activities can be defined in data analysis - display, reduction and statistical

inference (Kent, 1999). For qualitative data display, quoted text extractschecklists or

tables were used, while for quantitative data, table or chart formats were used.

Qualitative data reduction involved paraphrasingand summarizing statement,

classifying responsesinto categories,using quasi-statistics(e.g. the data from

questionnairesof consumers'opinion transferredto frequencies).Quantitative data

reduction used statistical methods,such as calculating averages,to reducethe data to

a few key summary measures,

after they were analyzed.Statistical inference focused

on what was typical or what deviatedfrom the average.Variablesincluded

descriptors,independentvariables and dependentvariables, indicating how widely

responsesvaried (such as Chi-squaretest) and how they were distributed in relation to

the variables (such as regression)being measured.In this study Excel, Minitab and

SPSSwere usedto perform the statistical analysis.

2.6. Researchlocations

2.6.1 Milkfish farms

Becausemilkfish are better acclimatedto warmer water, most milkfish farms are

located on southwestpart of Taiwan, which has therefore beenthe major research

area.This includes Cha-I and Tainan county, Tainan City, and Kaoshung and Ping-

Tong county (Fig 2.4). A total of 286 milkfish farms were surveyed.

2.6.2 Eel farms

Similar to milkfish farms, most traditional eel farms are located on southwestpart of

26
Taiwan. The researchareasfor eel farms were in Jang-Hwa,Yun-Lin, Cha-I, Tainan,

Kaoshungand Ping-Tong county (Fig 2.4). In total, 63 traditional eel farms were

surveyed.The indoor intensive eel culture is an infant industry in Taiwan, and

therefore, only 5 farms were surveyed, located in Taipei, Tao-Yen and Tainan county.

2.6.3 Cageculture

Most cagefarms are located in Pen-Hu Island and Ping-Tong county (Fig 2.4), and

therefore, only these2 areaswere surveyed.A total of 17 traditional fish farms and 5

submersiblefish farms were surveyed.

2.6.4. Consumers' opinion

In 1999,132 consumerswere surveyedto understandtheir opinion on milkfish and

by The
eel questionnaires. consumers'opinions from 3 areasare surveyed.The three

areaswere Taipei City (52 consumers),Taichung City (45 consumers)and Tainan City

(35 consumers).Theserepresentedthe northern, central and southernparts of Taiwan

respectively (Fig.2.4).

2.7 Discussion

In Taiwan, people dislike to be interviewed and are becoming increasingly awareof

their rights and are sensitive about the invasion of their privacy. Some interviewees

only acceptedtelephoneinterviewing and somejust rejected the interviews. The most

seriousdrawback of random samplesperhapsis there is a degreeof non-response.The

disadvantageof postal questionnairesis the low responserate. Some people are

reluctant to answersome sensitive questions.In thesecircumstances,broad categories

may be more suitable for thesequestions,and more personaldemographicquestions

27
may be better asked at the end of questionnaire. When drafting questionnaires there is

a tendency to put in all questions that might seem relevant. This can result in a very

long questionnaire that may have an impact on the respondent's willingness to finish.

Therefore, a pre-test was carried out to ensure that questionnaires were appropriate.

These were also helpful in understanding the real situation and in wording the

questionnaire properly. Questionnaire wordings need to be clear, to help respondents

understand the questions and be willing to provide answers. In spite of these

drawbacks there are some advantages in using questionnaires.

- Detailed information can be collected from small samplesabout a population, so

minimizing costs and sourcerequirements.

Standardization of questions and answers allows for comparisons to be made.

-

- They help reveal whether sample populations are relatively uniform or highly

heterogeneous, thereby improving the design of programmes.

However, there are also other disadvantages.

- Researchersmay underestimatethe time taken to processand derive results from

that data.This can result in much unutilized data and great delays between

collecting the information and being able to act on the data.

- Researchersmay feel impelled to skip simple data analysis in favor of

sophisticatedstatistical routines. In the process,more obvious insights can be

overlooked and valuable interpretation may be neglected.

- Asking for information about incomes, some assetsand intra-householdissues

can be very sensitive (and sometimescannot be done at all).

Where researchis basedon asking people questions,responseerrors may arise.These

may be becauseof dishonesty,forgetfulness, faulty memories,unwillingness or

28
misunderstanding the questions being asked. In addition to response errors, sampling

errors might be made (Kent 1999). These errors might be reduced by increasing the

sample size. However, larger samples will cost more and take more time to complete

in a very practical issue.

29
 fTaipei City

ATaipei county

OTao-Yencounty
4

fTaichung City
9

*Jang-Hwa

*Yun-Lin County
Pen-Hu Island

"*Cha-I
go County
'Pr

"*OTainan County

"1 Tainan City

"*Kaosung County

"* Ping-Tong

. Locations for milkfish farms

*Locations for traditional eel farms
A Locations for superintensive eel farms
Locations for cageculture farms
f Locations for consumers' opinion
Fig 2.4 Research locations.

30
 Chapter 3

Milkfish culture

3.1 Introduction

The Milkfish (Chanoschanos)is one of the most extensively farmed marine bony

fishes in the world and is an important culture speciesin SoutheastAsia (Liao, 1993).

It has a broad geographicdistribution, existing virtually throughout the entire tropical

Indo-Pacific Oceanand is found as far eastas the Pacific waters off Central America

(Lee, 1995).The large-scaleculture of milkfish is carried out in the Philippines,

Indonesiaand Taiwan (Boonyaratpalin 1997), where it has had a long history. It is

generally statedthat milkfish farming startedin Indonesia at least 500 yearsago

(Schuster,1952), though Ronquillo (1975) considersthat the culture can be traced

back 700 years.In the 16th century, milkfish culture was introduced to the Philippines

and Taiwan, and in the 17th century in Hawaii (Ling, 1977).

Milkfish have a number of advantagesin coastal aquaculture.They can tolerate

salinity changesfrom 0 to 150 ppt (Crean,1980), and are resistantto diseaseand

handling. According to FAO (2001), there are only 6 countries (Guam, Indonesia,

Kiribati, the Philippines, Singapore and Taiwan) recording the aquaculture production

of milkfish. Indonesia,the Philippines and Taiwan are the major producing countries,

accountingfor more than 95% of world production. Total annual production from

milkfish culture has exceeded300,000 tonnes since 1981 (Lee, et al 1997) and in

2000, culture in the three major milkfish farming countries yielded 461,857 tonnes

(Table 3.1). The annual production fluctuates, and was more than 400,000mt in 1990,

1991,1999 and 2000.

31
The production quantity and value from Taiwan accounted for - 6.3-20.9% and 7.5-

17.4% of global production quantity and value, respectively. The highest production

from Taiwan was in 1990 and 1994, reaching 90,000 and 66,000 t, accounting for 20.9

and 17.5% of world production, respectively. However, the production has gradually

declined since 1994, and in 2000, production from Taiwan reduced to 39,700 t,

accounting for 8.6% of world production. The Philippines and Indonesia produced

more than 80% of global production quantity and value (Table 3.1), while that from

the Philippines was unstable, the highest quantity reaching 23,400 t, in 1991

accounting for 52.2% of world production. Production from Indonesia showed an

increasing trend and reached 21.7 thousand tin 2000, accounting for 47.0% of world

production.

The average prices of milkfish fluctuated and were 1.15-2.32 US$/kg in Taiwan,

and1.23-2.22 US$/kg in the Philippines from 1988 to 2000. The ratios of highest to

lowest price were 0.50 and 0.55 in Taiwan and the Philippines, respectively. However,

the average price was stable in Indonesia and ranged froml. 50-1.90 US$ (Table 3.2).

The ratio of highest to lowest price was 0.79. Compared to Taiwan and the

Philippines, year to year price fluctuation in Indonesia were insignificant. (Table 3.1

and Table 3.2).

32
Table 3.1 The production quantity and value of milkfish in three major milkfish
farming countries (Taiwan, the Philippines and Indonesia). Unit: mt/ US$
Year Taiwan Phili pines Indonesia Global Total
V V V V
1988 39853 68428 187877 237818 118001 177001 345823 483460.8
(11.5) 14.2 (54.3) (49.2) (32.1 (36.6)
1989 21157 38047 192896 252693 119339 214810 333495 505789.3
(6.3) (7.5) (57.8) (50.0) (35.8) (42.5)
1990 90716 104560 210882 294628 132432 238377 434123 637808.3
(20.9) (16.4) (48.6) (46.2) (30.5) (37.4)
1991 41298 52535 234123 286969 141024 267945 416520 607651.5
(9.9) (8.6) (52.2) (47.2) (33.9) (44.1)
1992 25146 58263 171116 266607 147032 279360 343359 604410
(7.3) (9.6) (49.8) (44.1) (42.8) (46.2)
1993 45524 72819 148965 239373 164448 296006 359012 608398
(12.7) (12.0) (41.5) (39.3) (45.8) 48.7
1994 66784 110616 161006 312469 153093 275567 380938 698812
(17.5) (15.8) 42.3 (44.7) (40.2) (39.4)
1995 63254 106819 150858 315004 151256 272260 365408 694211.5
(17.3) (15.4) (41.3) (45.4) (41.4) (39.2)
1996 58453 129998 150151 333684 162127 283722 370765 747522
(15.8) (17.4) (40.5) (44.6) (43.7) (38.0)
1997 62749 93287 161426 309437 142709 256876 367286 660463.3
(15.4) (14.1) (42.8) (46.9) (38.9) (38.9)
1998 58349 74807 162401 209766 158666 285598 379593 570473.9
(15.4) (13.1) (42.8) (36.8) (41.8) (50.1)
1999 50824 72427 170677 238949 209758 377564 431678 689501.8
(11.8 (10.5 (39.5) (34.7) 48.6 (54.8)
2000 39730 64585 204204 258644 217208 390974 461857 715090.1
(8.6) (9.0) (44.2) (36.2) (47.0) 54.7
Data source: FAO, 2001 (www. fao. org/fi/figis/tseries/index. jsp)
The figures are undeflated.
Figures in the parenthesis are the ratios to total global production

33
Table 3.2 The ratios of production value to production quantity of milkfish in three
major milkfish farming countries. Unit: US$/kg
Year Taiwan Philippines Indonesia Global Total
1988 1.72 1.27 1.50 1.40
1989 1.80 1.31 1.80 1.52
1990 1.15 1.40 1.80 1.47
1991 1.27 1.23 1.90 1.46
1992 2.32 1.56 1.90 1.76
1993 1.60 1.61 1.80 1.69
1994 1.66 1.94 1.80 1.83
1995 1.69 2.09 1.80 1.90
1996 2.22 2.22 1.75 2.02
1997 1.49 1.92 1.80 1.80
1998 1.28 1.29 1.80 1.50
1999 1.43 1.40 1.80 1.60
2000 1.63 1.27 1.80 1.55
Data source:Calculated from Table 3.1.

In Taiwan, although a range of other specieshas now been developed,milkfish

remains one of the most important species,with farms distributed along the SW coast

of the island. It is generally believed that the culture was already practisedin Taiwan

during the reign of Cheng, Cheng-Kung (known in the west as Koxinga) in the 1640s

who set up his court and governmentnear Tainan , in SW Taiwan, where a milkfish

farm was built and named in his honour (Liao 1992).The areaof milkfish farming

had increasedto 12,545ha by 1990 (Annex B. 1), though this was due to the collapse

of tiger shrimp culture, after which some shrimp farmers adaptedtheir ponds to

cultivate milkfish. In 1990,output reached90,673 mt i. e. about 26.3% of total

aquacultureproduction (Annex B.2).

34
However, overproduction had createdsomeeconomic distressfor milkfish farmers

with market price dropping from 60-100NT$/kg to 40-45NT$/kg (32NT$=1US$) i. e.

1.9-3.1, to 1.3-1.4 US$/kg (Liao. 1993), as a results of which, some producers started

to de-commission their farms. In 2000, the remaining actual culture area was about

13,000 ha and production was about 40,000mt (Annex B1 and Annex B2). Compared

with 1990, the highest year of production, the real culture areaincreased4.0% by

2000, and this was an increaseof 60.7% over the areain 1987,before the collapseof

tiger shrimp. Regarding output, comparedwith 1990,the highest year of production,

this decreased56.16% by 2000. Though low, this was an increaseof 37.8% over the

production level in 1987.The reasonwhy the production reducedwhile culture area

increased (i. e. yields dropped) was due to cold weather in 2000.

Table 3.3 shows that there is an increasingtrend in the ratio of polyculture areato

monöculture area.In 1987, this was only 20.79%, increasing to 68.72% in 1996 and

105.27%in 1997,but returning to 70.65% in 2000. The ratio of fresh water to

brackish water areahad also beenincreasing.In 1987,the ratio was 19.42%, attaining

56.17% in 1997,but returning to 36.3% in 2000 (Table 3.3). It appearsthat the ratio

of polyculture areato monoculture area, and the ratio of fresh water areato brackish

water areahave had a positive relationship. The changeof the ratio of production

value to quantity describesthe changein averageprice. Table 3.4 demonstratesthat

this ratio is very unstable.Comparing averagevaluesper t betweenbrackish and fresh

water systemsthat from brackish water is not always higher. Although people

consideredthere to be risks of off-flavor from fresh water ponds, milkfish cultivated

in fresh water using feeds can be fatter and look better (Ding 1994), and can therefore

fetch higher price.

35
Since 1993,the ratio of milkfish culture areato total national aquacultureareahas

ranged from 11.2-21.3% and that of milkfish to total aquaculture output has ranged

from 15.5-23.2% (Annex B1 and Annex B2), indicating its comparatively high

productivity. This is possibly due to the introduction of the deepwater system(see

later) in the 1980s,in which farmers intensified production by using feed insteadof

just using fertilizers. In 1987, averageproductivity was about 3.47 t ha lyr 1,rising to

6.73 t ha lyr 1,in 1997.By contrast,the ratio of milkfish to total aquacultureoutput

value rangedfrom 6.4 -10.9% since 1993 (TableAnnex B2), and the ratios of

production value to quantity were far lower than that of total aquacultureproducts

(Table 3.4), confirming that, milkfish is not a high value species in Taiwan.

Table 3.3 The ratios of polyculture areato monoculture area and fresh water areato
brackish water areaof milkfish culture in Taiwan.
Year Ratios of polyculture area to Ratios of fresh water area
monoculture area to brackish water area
1987 20.79% 19.42%
1988 31.10% 31.65%
1989 51.95% 41.45%
1990 38.99% 42.89%
1991 50.54% 43.28%
1992 52.59% 33.86%
1993 45.21% 41.64%
1994 44.96% 39.14%
1995 49.45% 41.09%
1996 68.72% 49.40%
1997 105.27% 56.17%
1998 60.28% 53.10%
1999 55.30% 43.52%
2000 70.65% 36.29%
Data source:Calculated from Annex B1 and are real culture area.

36
Table 3.4 The ratio of production value to production quantity of milkfish in Taiwan.
Year Brackish water and Fresh water and Total milkfish Total A uaculture
1987 54.34 54.53 54.40 115.35
1988 50.44 47.05 49.03 114.56
1989 45.80 49.24 47.19 106.20
1990 30.99 31.22 31.03 91.59
1991 36.16 30.29 34.15 103.66
1992 60.74 54.32 58.30 111.95
1993 45.14 40.39 42.15 104.52
1994 43.86 43.82 43.83 116.56
1995 51.87 54.00 53.05 127.39
1996 51.68 69.60 61.08 120.09
1997 42.26 42.42 42.34 100.32
1998 45.40 40.50 42.88 106.75
1999 48.76 46.88 47.72 90.40
2000 54.13 48.53 50.82 101.06
Data source:Calculated from Annex B 2.

3.2 Methods of milkfish culture

The procedurefor milkfish culture includes fry acquisition, nursery stageproduction,

grow-out, overwintering and harvest.The procedureis shown in Fig 3.1. and will be

discussedin the following paragraphs.

Brood stock farms

Fry from import Fry from collection Fry from hatchery

Nursery ponds
I Overwintering ponds (fingerlings)

Production ponds

Harvest and marketing

Fig 3.1 The production procedurefor milkfish culture in Taiwan.

37
3.2.1 Fry collection and distribution

Traditionally, milkfish fry were caught from the wild, annually during natural

spawning.This was the sole sourcefor culture until natural spawning in captivity and

mass larval rearing in hatcheries were achieved in the late 1980s. Milkfish larvae are

pelagic. Younger larvae (less than 10mm TL) occur mostly near the surface, but also

down to 20-30 m, while older larvae (? 10mm TL) occur only near the surface.

Younger larvae are found both far from and near the shore,but older larvae occur only

inshore (Bagarinao, 1994). In the past, milkfish fry collection was an important source

of extra income for many fishermen. Usually, wild fry are consideredstrongerthan

cultured fry and command a higher price (Chang et al., 1993). However before the

successof natural spawning in captivity, fry supply from wild sourcewas

unpredictableand fluctuated significantly, with fry shortagesoccurring in some years,

in which case,Taiwanesefarmers imported fry from the Philippines, Indonesiaand

Malaysia.

Milkfish fry can be collected from March to October,but the peak seasonis between

April and July. Newly caught milkfish fry of about three weeks ageare 12-16 mm in

length, 0.03-0.11g in weight and are nearly transparent.The traditional gear,still

commonly used,is a triangular scoopnet, pushedforward in chest-high water (Chen

1990).A more efficient method usesa modified seinewith an open-endnet bag,

pulled at the samedepth of water by two fishers, or by one in the water and another

on a bamboo raft. Fishermenstop their operationsperiodically to dip the fry out

through the opening at the end of the net bag, which is suspendedon the water surface

by a float. Since the late 1960s,motorised boats have also beenused to tow the scoop

38
net or seine net, extending the fishing areafurther off shore and increasingcapacity.

Collectors sell fry to buyers who accumulatethem in shadedtanks, with 10cm water

depth, sited near the collecting area. Fry are then sealed in oxygen-filled plastic bags,

one-third filled with 10-15 ppt salinity water, and delivered to fry dealers. The dealers

then hold the fry in concretetanks, and feed twice daily with wheat flour and

occasionally egg yolk. The water in the tanks has to be changedat least daily (Chen

1990).However, after the successof natural spawning, the availability of cheaper

hatchery-producedfry has made somefishermen lose interest in collecting wild fry. In

turn fish farmers have fewer options but to acceptcultivated fry.

3.2.2 Hatchery rearedfry

Boodstock

The prime market size for milkfish throughout most countries in Asia is about 300 to

400 g, usually less than one year old. Liao and Chen (1984) reported that milkfish in

captivity showed gonadalmaturation in 5-year-old males and 6-year-old femalesbut

satisfactory spawning
spontaneous only occurred from the age of 10 years (Lin,

1984).Thus, traditional milkfish farming has had little emphasison producing

sexually mature and reproductively active fish in captivity. In Taiwan, the first

successful induced spawning was in 1979 (Tseng and Hsiao 1979; Hsiao and Tseng

1979, and the first successof spontaneousspawning was achievedin 1983 (Lin 1984).

Although spawning can be induced by hypophysation or hormone implantation (Lee

et al. 1986,Liao and Chen 1984,Lin 1982, Marte et al. 1987), there are drawbacks,

with inconsistent successin fertilization. Frequently, males do not releasesperm at the

sametime that females spawn, and stripping the ovary is not very effective, as it can

39
injure or even kill the fish and, the resultant fertilization rates are relatively low.

Spontaneousspawningis therefore preferred.

To reducecost, brood stock milkfish are commonly cultured in ponds in lower

densities with other main culture species in ponds, such as mullet, grouper or tilapia,

before they reach maturity. Becausethe spawnerseasily get injured and may even die

after poor handling, the harvestof other speciesmust be done with extreme caution.

Spawnerswith well-developed gonadsare very sensitive to low dissolved oxygen and

bigger fish usually die faster than others when oxygen is sharply depleted(Chang et

al. 1993). Spawnersare usually transferredinto spawning ponds two months before

the spawning season,with a recommendedstocking density of 0.5 fish/m2 surface

areaand a sex ratio of 3 female to 1 male (Lee 1995). Spontaneousspawningusually

occurs after midnight and seemsto be a significantly related to the intensity of

moonlight. Mating activity such as chasingbefore sunsetis a strong predictor of

spawningafter midnight. During strong chasing behavior, the dorsal and caudal fins

of spawnerscome out of the water (Chang et al 1993).The first spawning usually

startsin early April and the peak of spawning frequency and egg production occurs

from May to July, declining betweenAugust and October.

Generally,milkfish can spawn two to four times in one season(Kelly and Lee 1991)

and have asynchronousspawningbehavior in captivity (Lin 1987).This is different

from the observationof wild milkfish in Hawaii by Kuo and Nash (1979). Water

temperatureand salinity are important environmental factors affecting milkfish

maturation and egg quality, spawning occurring within 26 to 34.5°C. At the peak

season,water temperaturesare usually within the range of 29 to 33°C. Egg quality

40
was found to be better at salinity higher than 30 ppt. Although artificial fertilization

can be achievedat 5ppt (Lee 1985),effective spawningwas difficult to observeat

salinities lower than 26 ppt as eggs sink gradually to the bottom of the pond a few

hours after spawning (Chang et al 1993).

Egg collection and incubation

After spawning, a 0.8 mm meshplankton-net is stretchedacrossa corner of the pond.

Paddlewheelsare then usedto generatecurrents that passthrough the plankton nets,

trapping the eggsdrifting in the current inside the plankton nets. If the salinity is

lower or the price of fry is higher, guarding is done rotationally and the eggs are

collected very soon after spawning.Most of the fertilized eggsfloat at salinities of

over 30 ppt. Live fertilized eggs are about 1.2 mm in diameter, translucent with some

yellow tinges and are suspendedin the water column, while deadeggsnormally sink

to the bottom and are opaque.The fertilized eggs are usually collected before dawn to

prevent them being eatenby other small fish or by the spawnersthemselves.

After collecting the fertilized eggsand separatingand removing detritus, the eggs are

transferredto well-aeratedtanks. They are then stockedinto plastic bags and

distributed to larval rearing farmers for incubation. Becausebroodstock are rarely

caught in the wild and are not available from commercial grow-out ponds, fry

producersmust raise their own. Becausebroodstock farmers do not have enough

facilities and labour to hatch all the spawnedeggs, and want to make efficient use of

eggs,one central broodstock farmer usually keepsthe broodstock and provides

fertilized eggs to severalsatellite hatcheries.The organization and operating systems

are shown in Fig. 3.2.

41
Broodstock farmers offer fertilized eggs,technology, financial support and marketing

serviceto the satellite hatchery farmers and the hatchery farmers provide facilities and

labour for hatching. Total earnings from fry sales are typically shared by broodstock

farmers and satellite hatchery farmers at a ratio of 4: 6 (Lee et al 1997 and Chang et al

1993). The stocking density of fertilized eggs for incubation is about 2 kg eggs m 3, or

1.6 million eggsni-3(1600 eggs/cubic)(Chang 1993). Key managementconsistsof

the removal of various suspendedsubstancesand deadeggs from the incubation tank.

The fertilized eggswill hatch in less than 24 hours at 30°C. Below 30 ppt, increased

aerationis required to keep eggs suspendedin the water column, and so, the preferred

salinity range is about 30-40 ppt.

Broodstock farms

Supplying eggs,
technology,money, and
marketing service

Hatchery Hatchery I I Hatchery I Hatchery

____ý_

Supply of fry from import

Nurseries

Fig 3.2 Organization and operational systemsinvolved in milkfish fry production in

Taiwan (Modified from Chang et al 1993 and Lee et al 1997).

42
Larval rearing

Outdoor (semi-intensive)ponds coveredwith black plastic shading are normally used

for larvae rearing. The larval density is about 2-3 larvae 1-1and aeration is provided at

every 2.5 m interval. After hatching, the fry start feeding in about two days and to

ensure survival, exogenous feed must be provided before yolk absorption. Green

water is introduced into the larvae rearing ponds on the secondday after hatching to

reducetransparency,since larvae have a phototactic behavior, and it is harmful for

them to aggregatein schoolsand swim up to high light intensity (Chang et al 1993).

Larvae can be fed on oyster eggs and smaller rotifers after first feeding, but survival

rate can be increasedby supplying oyster eggs directly at first feeding. Oyster eggs

are supplied for seven days after first feeding and rotifers are supplied on the fifth day

or later after hatching. Rotifers can be cultured by using minced trash fish, yeast

powder and chicken dung. On the fifteen day after hatching, a start can be made in

using fishmeal, eel feed in powder form, or micropellet to feed larvae, and the rotifers

gradually replaced.

3.2.3 Generallayout of shallow-water milkfish culture farms in Taiwan

Traditionally, milkfish farmers usedfertilizer (such as chicken manure) to produce

benthic algae as food for milkfish. For producing benthic algae,the water depth

cannot be too deep (about 30 cm), to prevent the growth of phytoplankton. This

method of milkfish culture referred to as the shallow water milkfish culture system,

startedto be amendedin the 1980s,as some milkfish farmers applied the deepwater

system,in which, the water depth is usually more than 1m and feed is used (Ding

1994).

43
Production ponds

Production ponds are usually 3 to 8 hectaresin size, rectangular,with the long axis

running from east to west to reduce wind-driven waves. Ponds are usually built in the

coastal area, and water is supplied by tide or pumped, and salinity changed with the

rainfall and evaporation.Water depth is 30 to 40 cm and even in the deepestarea;only

45 cm deep.To prevent water being spilled out by waves or heavy rain, the height of

the dyke is about 80cm. The bottom of the pond is flat and has a slope of about 3cm

per 100m (0.03%) from inlet to outlet.

Overwintering canals

Except in Pingtung, the southernmost county in Taiwan, overwintering canals are

necessary components of milkfish farms. These are 100-300m long ditches, 5-8m

wide on the surface and 1.2-1.5 on the bottom, with water depth of 1.5-2m, also

aligned acrossthe direction of the prevailing wind. Traditionally, windbreaks made of

straw, canvasor polyethylene plastic, held togetherby bamboo sticks, are constructed

on the north side of the canals during the winter period. The windbreak has an angle

of 30-60 degreesfrom the horizontal to deflect the winter wind. There are generally

1mxO.5m windows in the windbreak at 20-25m intervals, kept open in wann days to

facilitate air circulation and enhanceinput of air into the water.

Nursery ponds

One nursery pond is connectedto each overwintering canal with a gate allowing fish

to swim betweenthem. The nursery pond is about 18-25m wide and 20 cm deep,and

can serve as a grazing and swimming ground for overwintering fingerlings on a warm

day, and can be used as a rearing pond for new fry. When the cold weatheris coming,

44
farmers drive the fish into overwintering canalsand close the gatesbetweenthe canals

and nursery ponds.

Water supply canals

Every production pond has sub-canal,connectedto the main canalsfor filling and

draining water.These serveas acclimatisatonareasfor fingerlings and temporary

emergencysheltersfor fish when the production ponds are being dried in the sun.The

water canalsare 30-40 cm deeperthan the production ponds and about 6-8m wide.

Pond preparation

After harvestin late November,the fish of under market size are driven into

overwintering canalsand the main pond is preparedfor the next growing season.The

organic debris in the ponds is raked and spreadevenly over the bottom and the ponds

left to be sun-dried till the mud on the bottom startscracking. After sun-drying,

fertilizers are applied to develop the benthic algae.These include chicken or pig

manureand rice bran. Next, the canalsare also drained, cleanedand dried. At the end

of February,tea seedmeal is usedas a pesticide to kill the pestsand predators,and

finally the ponds are left to let the benthic algae grow.

Nursery management

Before milkfish fry are stockedin the grow-out ponds, they are usually kept in

nursery ponds.To recover from transport stressor to be acclimatized to the new

culture environment, fry are first put into an acclimatization pool with a gate

connectedto the nursery pond. The gatesare openeda few hours after stocking and

the fry gradually swim into the nursery pond after they have becomeusedto its

45
salinity. Nursery ponds vary in size from 100 to 5000 m2 or 1-3% of the total farm

water surface.The stocking density is about 40-50 fry/ m2. Usually, nursery ponds

have easy accessto water and aerationin caseof emergency.Eel feed and wheat flour

are supplied to the fry. After 4-6 weeks,the fry can grow to 5-8cm fingerlings, the

ideal size for releasing to grow-out ponds.

Pond management

At the end of March or the start of April, fish in the overwintering canals are driven in

to the production ponds.Thesefingerlings are of different sizes and the density is

about 3500-4500 ha-1.If the fingerlings available are not sufficient, the shortageis

purchasedfrom an outside source.Then, wild caught or hatchery rearedfry are

stockedin the nursery ponds and releasedto the production ponds after they attain

fingerling size. After that, the nursery pond is restockedimmediately, and this cycle

repeatedmonthly 3 or 4 times until July. The different sizes of fingerling are

cultivated together to avoid all the fish attaining market size at the sametime, in

which case,the amount of fish could exceedthe carrying capacity of the on-growing

pond. The key point in shallow-water milkfish culture is to maintain the benthic algae.

If this is overgrown, the algaein the bottom layer will die and the algal mat may

detachfrom bottom of the pond, after which, the phytoplankton may dominate,

followed by the production of zooplankton, making the water in the pond yellowish-

green.This will inhibit the growth of benthic algae and could lead to anoxic

conditions and massfish kills. To avoid overgrowth of the algal mat, ponds must be

stockedwith enough fish to grazeon it. However, overstocking can depletethe algae

mat, and to avoid overgrazing, supplementalfeeding (rice bran) is used,typically at

30-50 kg every two days per hectare(Ding 1994). If the phytoplankton take over the

46
benthic algae,fish are driven from the production pond into the canals.The

production pond is then drained, manureapplied and sun-dried to grow anothercrop

of benthic algae before the fish are reintroduced into the pond.

3.2.4 Harvest

The market size of milkfish in Taiwan is about 200-500g. When this size is attained,

partial harvesting can be done. This is usually done once a month, usually 4 times,

from late May or early Juneto October.By late May or early June,the biomasswill

reach700-800Kg ha-1which is about the carrying capacity in the pond. After partial

harvest,the biomasscan be reducedto about 250Kg/ha, after which, supplemental

restocking and partial harvest can be repeated.

To prevent the algal mat from being damaged,a gill net is usedfor partial harvesting.

A gill net with a very large mesh red twine is used first to threatenthe milkfish and to

make them empty the contentsof their digestive tract. About 2-3 hours later, a smaller

mesh gill net is used for harvesting.At the end of November, a complete harvest is

made and the undersizedfish driven into the overwintering canals.

3.2.5 Overwintering

Prior to the year-endharvest at the end of October,fish farmers start preparationof

the overwintering canals.This is similar to that for the production ponds, and includes

draining, cleaning, sun drying, applying manure,killing pestsand then growing algae.

Then, the windbreak is repaired and installed. After driving the fish into the

overwintering canals,the fish can be fed with rice bran or fish meal/ eel feed meal on

the warmer days. Stocking density is usually less than 1.3 Kg m3 and the temperature

47
in the overwintering canalscan be 3-5°C higher than the water in the production

ponds during a cold spell.

3.2.6. Deep-waterculture

In responseto declining profit, and the limited and increasing value of land and

manpowerresources,deep-waterculture of milkfish was developed.Most deep-water

milkfish ponds are createdby converting the larger shallow-water ponds by

subdividing them into smaller and deeperponds or converting them into small fresh

water ponds.The managementof deep-waterponds is similar to that of traditional

pond procedures,and includes pond preparation,fry stocking, harvesting and

overwintering. After harvest,ponds are sun-dried and limed. Deep-waterponds are

typically about 2m in depth. The averagestocking density is about 22,000 fish ha 1

Every hectareis equippedwith 2-4 paddlewheelsto increasethe dissolved oxygen.

Benthic algae are presentduring the initial culture period, but feed pellets are the

main sourceof nutrition during the main growing phase.One to two blower-type

stationary automatic feedersare usedto distribute pelletized feed onto the surfaceof

eachpond from 0800 to 1800.

Daily feeding rate is about 3-5% of fish biomass,the protein content is about 23-27%

and food conversion ratio about 1.3-1.6.If fresh water is used,fry must be

acclimatized or else packs of coarsesalt put into the ponds near the bank before

releasingfry into the ponds.As with the shallow-water system,milkfish are

selectively harvestedwith a gill net. Ideally, to fetch higher prices, the harvesting

schedulewould differ from the peak production seasonfrom August to November.

Milkfish usually can command a higher price during the Decemberto May period,

48
especially in April. Using feed, the milkfish can continue to be grown during the

winter months, and can be harvestedlater. If the survival rate is higher, the annual

productivity the deep-waterpond can be 12 ton/ha.

Table 3.5 Socioeconomic characteristics of milkfish farmers.

Socioeconomic characteristics Number Percentage
Age (years)
21-30 9 3.14%
31-40 71 24.83%
41-50 99 34.62%
51-60 85 29.72%
61 and above 22 7.69%
Average 46.54 years

Education attainment
None 14 4.91%
Elementary 131 45.96%
Junior high school 55 19.30%
Senior high school 73 25.61%
College 12 4.21%
Averageyearsof schooling 8.24 years

Experiencein milkfish culture

1-10 years 48 16.84%
11-20 years 110 38.60%
21-30 years 64 22.46%
31-40 years 47 16.49%
41 yearsand above 16 5.61%
Average years of experience 22.70 years

Sourceof family income

Milkfish production only 69 24.13%
Milkfish production and other source 217 75.87%

Household size
1-5 180 63.16%
6-10 100 35.09%
11 and above 5 1.75%
Averagehouseholdsize 4.95 peoples

3.3. Characteristicsof milkfish producers

In 1999,286 fish farmers who also managedtheir farms were surveyedby

questionnaire.Most farms (> 97.6% ) were owned by single proprietors. On average,

49
milkfish producerswere 46.5 yearsold. The majority of the respondentsachieved

elementaryeducation(about 46%) and the averageyearsof schooling was about 8.2

years (Table 3.5). Most had many yearsof experiencein milkfish culture, averaging

22.7 years.Less than one-fourth (24.1%) statedthat milkfish farming was their only

sourceof family income. More than half of respondents(63.2%) had a householdsize

of <_5 people, the averagehouseholdsize being 4.95.

The education attainment is correlated to the age. Younger group had higher education

attainment.The averageeducationattainmentsin different agecategorieswere 11.3,

10.2,8.1,7.0 and 5.7 years in the categoriesof 20-<30,30-<40,40-<50,50-<60 and

60->60 yearsold, respectively.The averagehouseholdsizeswere similar in different

age groups,being 4.8,4.6,5.0,5.0 and 5.2 in the categoriesof 20-<30,30-<40,40-

<50,50-<60 and 60->60 years old, respectively.The percentageof eachgroup with

outside income was correlated to the age.The younger group except 20-<30 age

group, had a higher percentagewith an outside income, accounting for 55.0%, 92.0%,

79.6%, 72.5% and 68.1% in the categoriesof 20-<30,30-<40,40-<50,50-<60 and

60->60 yearsold, respectively.The high percentageof farmers with outside income

suggeststhat milkfish farming might be generally a part-time occupation.The

percentageof each group using monoculture for milkfish production was correlatedto

the age.The younger group, with the exception of the 20-<30 age group, had a higher

percentageof monoculture use for milkfish production, accounting for 37.5%, 70.4%,

56.7%, 46.9% and 46.2% in the categoriesof 20-<30,30-<40,40-<50,50-<60 and

60->60 yearsold, respectively (Table 3.6)

50
Table 3.6 Averagesof education attainment,householdsize, percentagesof farmers
with outside income and percentages of monoculture in different age categories.
Age category 20-<30 30-<40 40-<50 50-<60 60->60
Education attainment 11.3 10.2 8.1 7.0 5.7
Household size 4.8 4.6 5.0 5.0 5.2
Outside income 55.0% 92.0% 79.6% 72.5% 68.1%
% of monoculture 37.5% 70.4% 56.7% 46.9% 46.2%

3.4. Characteristicsof milkfish farms

According to the survey,most of the traditional shallow milkfish farms have been

transferredto deep-watersystemsand somefish farmers also use polyculture, i. e. with

other speciesin the ponds. Those specieswere fed by additional feed or the detritus of

milkfish feed.

Three types of milkfish culture could be defined, i. e. monoculture, polyculture and a

mix of monoculture and polyculture. In the first two casesthe fish farmers used all

their ponds either for monoculture or for polyculture, respectively.In the mixed

approach,farmers used someponds for monoculture and others for polyculture. The

survey showedthat 58.0 % of farms usedmonoculture, 33.6% polyculture and 8.4%

usedmixed culture (Table 3.7). Most (69.93%) of farm sizes are <3 hectare(Table

3.7). Speciesfor polyculture with milkfish include tiger prawn (Penaeusmonodon),

sandprawn (Metapenaeusensis), fresh water prawn (Macrobrachium rosenbergii),

tilapia (Oreochromisspp), mullet (Mugil cephalus) and other fishes.Among these,

tiger prawn and sandprawn are the most popular (Table 3.8).

51
Table 3.7 The number of farm sizesand statusof milkfish farms surveyedin Taiwan.
Farm size Mono-culture Poly-culture Mixed mono- Subtotal
and poly-culture
<1 ha 48 13 3 64 (22.38
1-< 2 ha 44 31 1 76 (26.57
2-< 3 ha 38 17 5 60 20.98
3 -< 4 ha 13 14 6 33 11.54
4 -< 5 ha 9 5 3 17 (5.94
5->5ha 14 16 6 36(12.59
Subtotal 166 58.0 96 33.6 24(8.4) 286(100
Figures in parenthesesare the percentagesof ratio to total surveyednumber.

Table 3.8 The frequencies of species in poly-culture with milkfish ponds.

Number Percentage
Tiger prawn 54 43.9%
Sandprawn 45 36.6%
Fresh water prawn 7 5.7%
Tilapia 5 4.1%
Mullet 5 4.1%
Other fishes 7 5.69%

Three kinds of water sourcewere observed,seawater,ground water and river or

reservoir water. For faster growing, fish farmers usually used fresh water to mix with

or
seawater, even fresh water alone (Table 3.9) as the fish do not needto expend so

much energyin osmotic regulation. When farm size is <1 ha, most farms tried to use

more fresh water (ground, and river or reservoir water), and accounting for more than

39% of farms. However, when the farm size was >1 ha, the percentageof farms using

fully fresh water reduced,accounting for less than 25% of farms (Table 3.10).

52
Table 3.9 The frequenciesof water sourcesthat were usedby farmers for cultivating
milkfish.
Water source Number Percentage
Seawater 76 26.6%
Ground water 47 16.4%
River or reservoir water 31 10.8%
Mixed seawater and ground water 40 14.3
Mixed seawaterand river or reservoir water 59 20.6%
Mixed ground water and river or reservoir water 8 2.8%
Mixed all 3 kinds of water 25 8.7%

Table 3.10 The use of water sources by different farm sizes.

Farm Seawater Ground River or Mixed Mixed Mixed Mixed all
size water reservoir seawaterseawaterand ground 3 kinds of
water and river or water and water
ground reservoir river or
water water reservoir
water
<1 ha 11 (17.2) 22(34.4) 3 (4.7) 2 (3.1) 17 (26.6) 1 (1.6) 8(12.5 )
1-< 2 ha 24 31.6) 10(13.2) 8(10.5) 11 14.5 14 18.4 2(2.6) 7(9.2)
2-< 3 ha 15 25.0) 4 (6.7) 10 16.7) 12 20.0 10 16.7 4(6.7) 5(8.3)
3 -< 4 ha 11 (33.3) 5(15.2) 3(9.1) 7(21.1) 6(18.2) 1(3.0) 0(0.0)
4 -< 5 ha 3(17. ) 2(11.8) 2(11.8) 1(5.9) 7(41.2) 0(0.0) 2(11.8)
5 -> 5 ha 12 (33.3) 4(11.1) 5(13.9) 7(19.4) 5(13.9) 0(0.0) 3(8.3)
Subtotal 76 47 31 40 59 8 25
The figures in parenthesesare the percentagesof water sourcesin different categories
of farm sizes.

Usually, tiger prawn, mullet and sandprawn were rearedin the water sourceswith

higher salinity. In the 54 farms using tiger prawn for polyculture with milkfish, only

two farms did not use seawateror mix seawater.Of 45 farms using sandshrimp for

polyculture 14 did not use seawateror mix seawater.All the farms using mullet for

polyculture used seawateror mixed seawater.However, except for one farm mixing

seawater for fresh water prawn, all the farm using fresh water shrimp, tilapia and

other fish for polyculture used fresh water as water sources(Table3.11).

53
Table 3.11 The frequenciesof speciesin poly-culture with milkfish in different water
sources.
Water source Tiger Sand Fresh water Tilapia Mullet Other
prawn prawn Prawn fish
Seawater 38 14 0 0 2 0
Ground water 2 9 6 3 1 2
River or reservoir water 2 5 0 2 0 2
Mixed seawaterand 7 11 0 0 2 3
ground water
Mixed seawater and river 2 3 0 0 0 0
or reservoir water
Mixed ground water and 2 2 0 0 0 0
river or reservoir water
Mixed all 3 kinds of 1 1 1 0 0 0
water
Subtotal 54 45 7 5 5 7

When monoculture farms were used to comparethe averageyield level in different

farm sizes,it shows that smaller farms had higher productivity, with higher yield level

at farm sizesof less than 3 ha. The highest averageyield levels were found in farm

sizesof < 1ha, attaining 18.6 t/ha, followed by 1-< 2 ha and 2-< 3 ha, attaining 11.7

and 11.3t/ha, respectively.However, when farm sizes were bigger than 3 ha, the

relationship betweensize and yield was not so clear, with insignificant difference of

averageyield levels. The averageyield levels of farm sizes in the categoriesof 3 -< 4

ha, 4 -< 5 ha and 5 -> 5 ha were 8.2,7.3 and 8.8 t/ha (Table 3.12). Although farm

sizesof <1 ha, 1-< 2 ha and 2-< 3 ha had higher yield level, the variation was

relatively larger, the highest and lowest yield level ranged from 4.6-33.3,3.4-14.9 and

4.7-18.0 t/ha respectively.

The variation of yield levels of farm sizes of 3 -< 4 ha, 4 -< 5 ha and 5 -> 5 ha were

relatively smaller, rangedfrom 5.7-12.9,4.8-13.5 and 4.9-15.0 t/ha, respectively.

These suggestedthat when farms were <3 ha, more intensive managementmight be

used,however the differencesbetweenbetter and poorer performance were greater.

54
On the contrary, when farms were >3 ha, more extensivemanagementwere used,and

the variation of performancewere smaller.

Table 3.12 The averageyield levels in different farm sizes.

Farm size Productivity (t/ha
<1 ha 18.6 4.6-33.3
1-< 2 ha 11.7(3.4-19.6)
2-< 3 ha 11.3(4.7-18.0)
3 -< 4 ha 8.2 (5.7-12.9)
4 -< 5 ha 7.3 4.8-13.5)
5->5ha 8.8 4.9-15.0
The figures in the parentheses are the range of highest and lowest yield levels.

3.5 Economic analysis

To assessthe profitability of milkfish farming in Taiwan, 262 milkfish farmers were

surveyedin 1999, excluding those which mixed mono- and poly-culture. In this

section,the cost and benefit analysis of different sizesof farms and different styles of

culture (monoculture and polyculture) are compared.

3.5.1.Cost catagories

Two cost componentscan be identified, capital and operating cost. Capital costs

comprisecosts of buildings, pond construction, power generator,feeder,paddlewheel

and pump (Table 3.13 and Table 3.15). Operating costsconsist of costs of fingerling,

feed, electricity, chemicals, wages,miscellaneous,land rent, depreciation and interest

(Table 3.16). As most fish farmers have their own land, land rent is estimatedat

100,000NT$ (3,100US$) hä 1 1.
yr Usually, fish farmers do not hire labor for routine

work, wagesonly being paid for harvesting.

Three important assumptionsare:

(1) investment costs are covered by a loan at an annual interest rate of 8%

55
 (2) facilities and equipment are taken as having a straight line depreciation

during their useful lifetime;

(3) the useful lifetime of the building and pond construction is 20 years,the

power generator and feeder is 10 years, and the paddle wheel and pump

is 5 years.

Becausespendingon operating costs is spreadthroughout the culture period, the

interest is chargedon only 50% of the total amount.Total expenditureon variable

costsis not paid out at the beginning, so will not incur full interest costs for the entire

period. However, the interest is chargedon 100% of the capital cost.

3.5.2 Capital cost characteristics

The averagecapital cost of monoculture systemwas shown in Table 3.13. Comparing

capital cost of different farm sizes,averagecapital cost per hectarereducedwith

increasedfarm size up to <4 hectare,suggestingthat the 3 -< 4 ha size categoryis the

most capital efficient for monoculture. In contrast,capital cost per hectareof

polyculture is highest at >1 ha and is very similar at sizes from 2 to 5 hectare,only

reducing at >5 ha, suggestingpossible economiesof scalebeyond that level (Table

3.14).

Comparing monoculture with polyculture, capital cost for the latter is higher by

average58.8% The capital cost of polyculture were especially higher in the items of

pond construction and paddlewheelthan that of monoculture. It implies that more

capital is neededto modify the ponds and paddle wheel to accommodateother

species.

56
Table 3.13 The averagecapital costs of monoculture milkfish farms of different sizes.
Unit: NT$, 32 NT$ =1 US$
Cost

<1ha 1-<2 2-<3 3-<4 4-<5 5->5ha

Item ha ha ha ha
No. sampled 48 44 38 13 9 14
Building 125350 157280 199740 133500 217140 370420
Pond construction 176670 225770 261110 283000 378400 900000
Power generator 100910 111770 121150 122730 135200 240420
Paddlewheel 48540 139630 142960 188540 211110 499190
Feeder 15940 20940 42860 45080 101670 119580
Pump 34580 34980 47635 53540 116890 159310
Total 501980 690360 815440 826380 1160410 2288910
Cost per hectare 910390 538680 355960 269520 278720 346090

Table 3.14 The average capital costs of polyculture milkfish farms of different sizes.
Unit: NT$, 32NT$ =1 US$
Cost

<1ha 1-<2 2-<3 3-<4 4-<5 5->5ha

Item ha ha ha ha
No. sampled 13 31 17 14 5 16
Building 155560 137830 179333 245390 260000 326430
Pond construction 387500 334000 775000 1175000 1250000 1263330
Power generator 120000 121920 126000 156090 194000 195000
Paddlewheel 99580 114130 151530 213360 510800 343380
Feeder 15460 26570 35530 41570 116400 88930
Pump 34540 33690 58270 113180 56000 78130
Total 812640 768140 1325660 19445802387200 2295200
Cost per hectare 1345100 584210 588410 607680 568380 352090

The averagecapital cost for producing per tonne of milkfish of monoculture is shown

in Table 3.15. The averagecapital costs to produce per tonne of milfish reducedwith

increasedfarm size up to <4 hectare,after that the costs increasedwith increased

farms size. The averagecost for producing one tonne of milkfish were 85.3,64.9,

57
41.3,32.7,33.4 and 50.3 thousandNT$/t in the size categoriesof < 1,1 -<2,2 -<3,

3-<4,4 -<5 and 5->5 ha, respectively. Pond construction was the highest cost

componentin production, per tonne of milkfish, the averagecost in producing one

tonne of milkfish reducedwith increasedfarm size up to > 5ha. The averagecostsof

pond construction in producing one tonne of milkfish were 31.1,25.7,14.1,10.3,6.7

and 21.8 thousand NT$/t in the size categories of < 1,1 -<2,2 -<3,3 -<4,4 -<5

and 5->5 ha, respectively.

The rangesof the highest and lowest cost of pond construction were 15.9-66.7,6.7-

54.5,3.8-20.0,8.6-13.9,3.5-12.5 and 13.3-30.0thousandNT$/t in the size categories

of<1,1 -<2,2-<3,3 -<4,4-<5 and 5->5 ha, respectively.Next to pond

construction was building, the averagecost in producing one tonne of milkfish

reducedwith increasedfarm size up to > 4ha, after that the cost increasedwith

increasedfarm size. The averagecosts of building in producing one tonne of milkfish

were 24.7,14.5,8.5,5.5,5.8 and 7.5 thousandNT$/t in the size categoriesof < 1,1 -

< 2,2 -<3,3 -<4,4 -<5 and 5->5 ha, respectively.The rangesof the highest and

lowest cost of building were 5.2-33.3,3.1-31.3,2.8-17.8,1.7-10.0,3.5-9.6 and 2.7-

16.7 thousandNT$/t in the size categoriesof < 1,1 -<2,2 -<3,3 -<4,4 -<5 and 5

->5 ha, respectively. The variation of the cost of pond construction in producing one

tonne of milkfish may be due to the site of farms and performanceof yield level

(Table 3.12). The variation of the cost of building in producing one tonne of milkfish

may be due to the sourcethe farmers bought them from, their size, the year they have

beenused and performanceof yield level.

58
Table 3.15 The averagecapital cost for eachtonne of annual output of monoculture
milkfish farm of different size. Unit: ThousandNT$
Cost

Item <1ha 1-<2ha 2-<3ha 3-<4ha 4-<5ha 5->5ha

No. sampled 48 44 38 13 9 14
Building 24.7 14.5 8.5 5.5 5.8 7.5
(5.2-33.3) (3.1-31.3) (2.8-17.8) (1.7-10.0) (3.5-9.6) (2.7-16.7)
Pond 31.1 25.7 14.1 10.3 6.7 21.8
construction (15.9-66.7) (6.7-54.5) (3.8-20.0) (8.6-13.9) (3.5-12.5) (13.3-30.0)
Power 10.6 10.1 5.3 4.9 5.4 5.2
generator (6.4-14.6) (6.5-15.6) (2.2-11.1) (2.5-6.3) (2.9-7.5) (2.1-10.0)
Paddlewheel 8.7 8.8 8.3 8.1 6.9 10.1
(2.4-15.4) (3.3-13.3) (3.1-12.5) (4.3-10.1) (2.7-10.9) (5.2-14.0)
Feeder 3.3 2.1 2.6 1.9 3.7 2.4
(1.1-7.4) (0.9-4.7) (0.9-4.3) (1.1-2.8) (1.4-6.0) (1.0-4.8)
Pump 6.9 3.7 2.4 2.1 4.8 3.4
(1.0-15.6) (0.8-9.1) (0.6-6.3) (0.8-3.8) (1.1-6.9) (0.8-9.1)
Total 85.3 64.9 41.3 32.7 33.4 50.3
48.7-132.0 (32.6-102.9) (28.9-64.5 (28.2-42.0) (22.3-48.0) (35.2-73.0)
The figures in the parenthesesare the range of highest and lowest cost.

3.5.3 Operating cost characteristics

The averageoperating cost of monoculture is shown in Table 3.16. Within this, the

cost of feed is highest, accountingfor 30.3% to 35.4%, after which is land rent,

accountingfor12.4% to 20.3%. The third highest cost is fingerling, accountingfor

10.31%to 14.83%.As with monoculture, the highest operating cost of polyculture is

feed, accounting for 31.6% to 42.5% (Table 3.17), after which is fingerling, at 12.2%

to 23.6% and land at 9.1% to 16.4%.The operating costs of polyculture included the

costs of the other speciesin the polyculture system.

59
Table 3.16 The averageoperatingcosts of monoculture milkfish farms of different
sizes. Unit: NT$
Cost
<lha 1-<2 2-<3 3-<4 4-<5 5->5ha
Item ha ha ha ha
No. sampled 48 44 38 13 9 14
Fingerling 45920 86490 135660 190120 304280 476450
(10.31) (11.62) (11.45) (12.04) (14.83) (11.94)
Feed 141170 255070 419290 551620 622940 1316220
(31.70) (34.26) (35.38) (34.94) (30.36) (32.98)
Electricity 57460 65560 100860 128860 205780 474180
(12.90) (8.80) (8.51) (8.16) (10.03) (11.88)
Chemicals 5190 8620 9280 17100 10080 23180
(1.16) (1.16) (0.78) (1.08) (0.49) (0.58)
Wage 21200 19090 33870 77690 65830 272000
(4.76) (2.56) (2.86) (4.92) (3.21) (6.82)
Land rent 55140 128160 229080 306620 416330 661360
(12.38) (17.21) (19.33) (19.42) (20.29) (16.57)
Miscellaneous 21760 35130 74090 99640 143720 215850
(4.89) (4.72) (6.25) (6.31) (7.01) (5.41)
Depreciation 43410 67350 77560 86020 119060 231220
(9.75) (9.04) (6.55) (5.45) (5.80) (5.79)
Interest 54070 79150 105320 120980 163590 320680
(12.14) (10.63) (8.89) (7.66) (7.97) (8.03)
Total 445310 744630 118500015786402051610 3991140
The figures in parenthesesare the percentagesof total operation cost.

60
Table 3.17 The averageoperating costs of polyculture milkfish farms of different
sizes. Unit: NT$, 32 NT$ =1 US$
Cost
<1ha 1-<2 2-<3 3-<4 4-<5 5->5ha
Item ha ha ha ha
No. sampled 13 31 17 14 5 16
Fingerling 99720 118500 171140 533630 383500 732410
(15.00) (13.54) (12.47) (23.64) (12.18) (16.15)
Feed 224000 293870 503180 713690 1226800 1926590
(33.69) (33.58) (36.67) (31.61) (38.98) (42.48)
Electricity 50790 62010 86440 112660 278000 240040
(7.64) (7.09) (6.30) (4.99) (8.83) (5.29)
Chemicals 10550 13030 12450 37590 32000 44070
(1.59) (1.49) (0.91) (1.66) (1.02) (0.97)
Wage 31540 43510 45410 44640 142000 195060
(4.74) (4.97) (3.31) (1.98) (4.51) (4.30)
Land rent 60420 131480 225290 320000 420000 651880
(9.09) (15.03) (16.42) (14.17) (13.34) (14.37)
Miscellaneous 34950 54510 71740 108950 149000 209720
(5.26) (6.23) (5.23) (4.83) (4.73) (4.62)
Depreciation 67520 68010 105830 156090 219900 192180
(10.15) (7.77) (7.71) (6.91) (6.99) (4.24)
Interest 85490 90130 150680 230410 296230 343610
(12.86) (10.30) (10.98) (10.21) (9.41) (7.58)
Total 664960 875050 11372170122576701314743014535550
The costs of fingerling and feed include the species for polyculture.
The figures in parentheses are the percentages of total operation cost.

The averageoperating cost of monoculture milkfish farms of different sizesin

producing 1 kg of milkfish is detailed in Table 3.18. This shows the most effective

size of farm is 4-<5 ha, the operating cost for producing 1 kg of milkfish was only

27.2 NT$. Followed by <1 ha, 2-<3 ha and 1-<2 ha, the averageoperating costs

are 44.1,45.2 and 48.2 NT$, respectively.The most inefficient farm sizes in operating

cost are 3-<4 ha and 5->5 ha, the averageoperating costs are 63.1 and 66.7 NT$,

respectively.The rangesof the highest and lowest operating cost for producing 1 kg of

milkfish were 29.6-58.8,32.6-62.2,30.1-58.0,50.1-79.4,20.4-39.9 and 50.7-81.1

NT$/kg in the size categoriesof <1 ha, 1- <2 ha, 2-<3 ha, 3-<4 ha, 4-<5 ha and

5->5 ha, respectively.

61
The key factors contributing to the operating cost for producing 1kg of milkfish were

feed, land rent and fingerling. The costsof feed for producing 1kg of milkfish were

14.0,16.5,16.0,22.1,8.3 and 22.0 NT$/kg, the highest and lowest cost of feed ranged

from 10.0-18.8,11.5-23.7,10.0-23.1,18.2-29.2,5.6-12.5 and 19.1-28.3in size

categoriesof< 1 ha, 1-<2ha, 2-<3ha, 3-<4ha, 4-<5haand5->5ha,

respectively.In the farm size of 4-< 5 ha had the most efficient FCR, it might be

becauseof the economiesof scaleand the supply of natural food. However, when a

farm size exceeded5ha it becomemore difficult for a family to manage.

The costs of land rent for producing 1 kg of milkfish were 5.5,8.3,8.7,12.3,5.5 and

11.1NT$/kg, the highest and lowest cost of land rent ranged from 1.4-11.0,2.6-13.9,

3.6-13.9,8.6-19.3,3.6-10.2 and 6.2-18.9 NT$/kg in size categoriesof <1 ha, 1- <2

ha, 2-<3 ha, 3-<4 ha, 4-<5 ha and 5->5 ha, respectively.As most farmers have

their own land, the land is

rent estimatedat 100,000NT$ ha-1 1,
yr therefore the costs

of land rent for producing 1 kg of milkfish were related to the yield levels (Table

3.12).

The costs of fingerling for producing 1 kg of milkfish were 4.5,5.6,5.2,7.6,4.0 and

8.0 NT$/kg, the highest and lowest cost of fingerling rangedfrom 1.9-8.9,2.5-10.1,

2.5-9.2,3.7-11.8,2.5-9.6 and 4.3-11.7 NT$/kg in size categoriesof <1 ha, 1- <2 ha,

2-<3 ha, 3-<4 ha, 4-<5 ha and 5->5 ha, respectively.The costs of fingerling

were related to the survival rate, therefore better managementmay lower costs of

fingerling for producing 1 kg of milkfish.

62
Table 3.18 The average operating costs of monoculture milkfish farms of different
sizes in the production of 1kg of milkfish. Unit: NT$
Cost

<1ha 1-<2ha 2-<3ha 3-<4ha 4-<5ha 5->5ha

Item
No. sampled 48 44 38 13 9 14
Fingerling 4.5 5.6 5.2 7.6 4.0 8.0
(1.9-8.9) (2.5-10.1) (2.5-9.2) (3.7-11.8) (2.5-9.6) (4.3-11.7)
Feed 14.0 16.5 16.0 22.1 8.3 22.0
(10.0-18.8) (11.5-23.7) (10.0-23.1) (18.2-29.2) (5.6-12.5) (19.1-28.3)
Electricity 5.7 4.2 3.8 5.2 2.7 7.9
(3.3-9.1) (3.2-6.4) (2.3-5.9) (3.9-7.7)) (1.8-5.9) (4.6-12.5)
Chemicals 0.5 0.6 0.4 0.7 0.1 0.4
(0.2-0.7) (0.2-0.8) (0.2-0.6) (0.3-1.0) (0.1-0.2) (0.2-0.6)
Wage 2.1 1.2 1.3 3.1 0.9 4.5
(0.9-3.6) (0.7-1.8) (0.7-1.8) (2.3-3.9) (0.6-1.2) (2.9-5.6)
Land rent 5.5 8.3 8.7 12.3 5.5 11.1
(1.4-11.0) (2.6-13.9) (3.6-13.9) (8.6-19.3) (3.6-10.2) (6.2-18.9)
Miscellaneous 2.2 2.3 2.8 4.0 1.9 3.6
(1.0-3.4) (1.1-3.4) (1.3-4.1) (2.7-5.1) (0.9-2.8) (2.5-4.8)
Depreciation 4.3 4.4 3.0 3.4 1.6 3.9
(2.1-6.0) (2.4-5.9) (1.8-4.9) (1.9-5.2) (0.9-2.4) (2.0-5.3)
Interest 5.3 5.1 4.0 4.8 2.2 5.4
(3.8-7.3) (3.8-7.2) (3.2-5.6) (3.6-6.2) (1.3-3.1) (3.9-7.3)
Total 44.1 48.2 45.2 63.1 27.2 66.7
(29.6-58.8) (32.6-62.2) (30.1-58.0) (50.1-79.4) (20.4-39.9) (50.7-81.1)
The figures in parenthesesare the rangesof highest and lowest cost.

3.5.4 Benefit analysis

When opportunity cost is considered,the profit (P) is equal to the net revenue(MI)

(excluding depreciation and interest) minus operating cost (C). Profitability can be

estimatedby the benefit-cost ratio (BCR) and the income ratio (IR), where formulas

are as follows:

BCR=P/C

IR=P/MI

Where P= Profit

63
 C= Production cost

MI = Revenue

The higher are thesevaluesthe more financially soundis the operation.This also

indicatesthat the operation is economically soundand further developmentmay be

considered.

Table 3.19 and Table 3.20 reveal that in monoculture, the IR values are similar in all

size categories,ranging from -35.2% (3- 4ha) to -43.1% (1- 2ha). The IR values of

polyculture were even poorer than monoculture, and at pond sizes 3- 4ha were even

below -100% (Table 3.19). However, the IR value in size categoriesof 4 ha and

abovewere higher in polyculture than in monoculture.The costs and revenuesin

polyculture include the costs and revenuesof the other speciesin the polyculture

system.Basedon the survey,milkfish culture was not financially sound,with costs

exceedingrevenuesin all size categoriesand both types of culture. However, most

facilities of farms have existed, farmers have their own farm and farmer usually can

get the loan with lower interest rate (about 6.5%), therefore, the costs of capital

interest, depreciation and land costs can be neglected.

Table 3.19 Returns and benefit ratios for monoculture milkfish, by farm size.
<I ha 1-<2ha 2-<3ha 3-<4ha 4-<5ha 5->5ha
Cost (NT$) 445307 744629 1185004 1578640 2051610 3991142
Revenue NT$ 322649 520443 858774 1167923 1501778 2804286
Profit (NT$) -122658 -224186 -326230 -410717 -549832 -1186856
BCR -27.54% -30.10% -27.53% -26.02% -26.80% -29.74%
IR -38.02% -43.08% -37.99% -35.17% -36.61% -42.32%
BCR is the ratio of profit to cost and IR is the ratio of profit to revenue.

64
Table 3.20 Returns and benefit ratio for polyculture milkfish, by farm size.
<lha 1-<2ha 2-<3ha 3-<4ha 4-<5ha 5->5ha
Cost (NT$) 664959 875050 1372171 2257665 3147428 4535545
Revenue(NT$) 432154 516113 911276 1000871 2631200 3627438
Profit (NT$) -232805 -358937 -460895 -1256794 -516228 -908107
BCR -35.01% -41.02% -33.59% -55.67% -16.40% -20.02%
IR -53.87% -69.55% -50.58% -125.57% -19.62% -25.03%
BCR is the ratio of profit to cost and IR is the ratio of profit to revenue.

If the grossprofit was considered,i. e., the opportunity costs (interest, depreciation

and land cost) are excluded and only expenditurecosts were considered,all the

monoculture size categoriescould make profit and the highest levels are in the 4-<

5ha size at 149160NT$ (Table 3.21). However, in polyculture, only the 2-<3 ha, 4-

<5 ha and >5 ha size categoriescould make profit at 20906,419900 and 279558 NT$

respectively,though, the size had higher profits than any monoculture size (Table

3.22).

The lowest cost in the farm size of 4-< 5 ha might permit this category to have the

highest profit. Table 3.18 shows that farm size of 4-< 5 ha had lower costs to produce

1 kg of milkfish. Feed is an important factor to reducecost becauseit accountsmore

than 30% of production, thereforebetter managementand lower FCR might be the

factors of reducing cost and causehigher profit.

Table 3.21 Gross profits and benefit for monoculture milkfish, by farm size.
<1ha 1-<2ha 2-<3ha 3-<4ha 4-<5ha 5->5ha
Cost (NT$) 292685 469971 773045 1065028 1352622 2777882
Revenue(NT$) 322649 520443 858774 1167923 1501778 2804286
Gross profit 29964 50472 85729 102895 149156 26404
(NT$)
BCR 10.24% 10.74% 11.09% 9.66% 11.03% 0.95%
IR 9.29% 9.70% 9.98% 8.81% 9.93% 0.94%
BCR is the ratio of grossprofit to cost and IR is the ratio of gross profit to revenue.

65
Table 3.22 Grossprofits and benefit for polyculture milkfish, by farm size.
<1ha 1-<2ha 2-<3ha 3-<4ha 4-<5ha 5->5ha
Cost (NT$) 451532 585433 890370 1551159 2211300 3347880
Revenue(NT$ 432154 516113 911276 1000871 2631200 3627438
Grossprofit -19378 -69320 20906 -550288 419900 279558
(NT$)
BCR 2.35% 1 18.99% 8.35%
-4.29% -11.84% -35.48%-
IR -4.48% -13.43% 2.29% -54.98% 15.96% 7.71%
BCR is the ratio of gross profit to cost and IR is the ratio of gross profit to revenue.

When the rangesof grossprofit were considered,the highest profit of monoculture

was in the size category of 4-<5 ha at 180,300NT$. However the size category of 3

ha and 2-<3 ha had higher lowest gross profit than that of the size category of
-<4

4-<5 ha (Table 3.23).

Table 3.23 The rangesof gross profit for monoculture milkfish, by farm size.
G ross prof it BCR IR
Hi hest Average Lowest Highest Average Lowest Highest Average Lowest
<1 ha 52436 29964 9117 19.4% 10.24% 2.9% 16.3% 9.29% 2.8%
1-<2 ha 79227 50472 31027 18.0% 10.74% 6.3% 15.2% 9.70% 6.0%
2- <3 ha 126012 85729 65975 17.2% 11.09% 8.3% 14.7% 9.98% 7.7%
3-<4 ha 177208 102895 71381 17.9% 9.66% 6.5% 15.2% 8.81% 6.1%
4-<5 ha 180300 149156 41514 13.6% 11.03% 2.8% 12.0% 9.93% 2.8%
5->5 ha 32626 26404 -16364 1.2% 0.95% -0.6% 1.2% 0.94% -0.6%

The rangesof grossprofit for polyculture were wider than monoculture.Although the

size categoriesof 4-<5 ha had highest averagegross profit, the highest profit was in

the size category of 5->5 ha at 1,283,900NT$ and only in the size category of 3-

4 ha could no profit be made. However, all the size categoriesof polyculture could not

make profit in the lowest range (Table 3.24).

66
Table 3.24 The rangesof grossprofit for polvculture milkfish, by farm size.
G ross profit BCR IR
Hi hest Average Lowest Hi hest Average Lowest Highest Average Lowest
<1 ha 116082 -19378 -177414 36.7% -4.29% -29.1% 26.9% -4.48% -41.1%
1 -<2 ha 106310 -69320 -244950 25.9% -11.84% -32.2% 20.6% -13.43% -47.5%
2- <3 ha 288017 20906 -335242 46.2% 2.35% -26.9% 31.6% 2.29% -36.8%
3-<4 ha -84940 -550288 -860520 -7.8% -35.48% -46.2% -8.5% -54.98% -85.9%
4-<5 ha 1083290 419900 -22360 70.0% 18.99% -0.8% 41.2% 15.96% -0.9%
5 -> 5 ha 112839221279558 1-557412 54.8% 8.35% -13.3% 35.4% 7.71% -15.4%

Table 3.25 Nominal cash-flow projection for monoculture milkfish farm.

Unit: Thousand NT$
Year0 Yearl Year2 Year3 Year4 Year5
Cashflow
Capital cost
<1 ha 502.0 0 0 0 0 0
1- < 2ha 690.4 0 0 0 0 0
2- < 3ha 815.4 0 0 0 0 0
3- < 4ha 826.4 0 0 0 0 0
4- < 5ha 1160.4 0 0 0 0 0
5->5ha 2288.9 0 0 0 0 0
Operatingcost
<1 ha 0 347.8 347.8 347.8 347.8 347.8
1- < 2ha 0 597.9 597.9 597.9 597.9 597.9
2- < 3ha 0 1002.1 1002.1 1002.1 1002.1 1002.1
3- < 4ha 0 1371.6 1371.6 1371.6 1371.6 1371.6
4- < 5ha 0 1769.0 1769.0 1769.0 1769.0 1769.0
5->5ha 0 3439.2 3439.2 3439.2 3439.2 3439.2
Revenue
<1 ha 322.6 322.6 322.6 322.6 322.6
1- < 2ha 520.4 520.4 520.4 520.4 520.4
2- < 3ha 858.8 858.8 858.8 858.8 858.8
3- < 4ha 1167.9 1167.9 1167.9 1167.9 1167.9
4- < 5ha 1501.8 1501.8 1501.8 1501.8 1501.8
5->5ha 2804.3 2804.3 2804.3 2804.3 2804.3
Net cash flow
<1 ha -502.0 -25.2 -25.2 -25.2 -25.2 -25.2
1- < 2ha -690.4 -77.5 -77.5 -77.5 -77.5 -77.5
2- < 3ha -815.4 -143.3 -143.3 -143.3 -143.3 -143.3
3- < 4ha -826.4 -203.7 -203.7 -203.7 -203.7 -203.7
4- < 5ha -1160.4 -267.2 -267.2 -267.2 -267.2 -267.2
5->5ha -2288.9 -635.0 -635.0 -635.0 -635.0 -635.0

67
3.5.5 Cash-flow and discountedfinancial indicators

The pattern of cashflow includes capital cost, operatingcost (excluding interest and

depreciation)and revenue.A 5-year nominal and a discountedcash flow analysis(at

10% discount rate) of monoculture milkfish farm reveal that investment in milkfish is

not financially viable (Table 3.25 andTable 3.26). Both nominal and discountedcash

flow shows monoculture milksish farms have negative cash flow.

Table 3.26 Discounted cash-flow projection for monoculture milkfish farm.

The discount rate for NPV is 10%. Unit: ThousandNT$
Year0 Yearl Year2 Year3 Year4 Year5
Cash outflow
<1 ha 502.0 313.0 281.7 253.6 228.2 205.4
1- < 2ha 690.4 538.1 484.3 435.9 392.3 353.1
2- < 3ha 815.4 901.9 811.7 730.5 657.5 591.7
3- < 4ha 826.4 1234.5 1111.0 999.9 899.9 809.9
4- < 5ha 1160.4 1592.1 1432.9 1289.6 1160.6 1044.6
5->5ha 2288.9 3095.3 2785.8 2507.2 2256.5 2030.8
Revenue
<1 ha 0 290.4 261.3 235.2 211.7 190.5
1- < 2ha 0 468.4 421.6 379.4 341.5 307.3
2- < 3ha 0 772.9 695.6 626.0 563.4 507.1
3- < 4ha 0 1051.1 946.0 851.4 766.3 689.6
4- < 5ha 0 1351.6 1216.4 1094.8 985.3 886.8
5->5ha 0 2523.9 2271.5 2044.3 1839.9 1655.9
Net cash flow
<1 ha -502.0 -25.2 -22.7 -20.4 -18.4 -16.5
1- < 2ha -690.4 -77.5 -69.7 -62.8 -56.5 -50.8
2- < 3ha -815.4 -143.3 -129.0 -116.1 -104.5 -94.0
3- < 4ha -826.4 -203.7 -183.3 -165.0 -148.5 -133.7
4- < 5ha -1160.4 -267.2 -240.5 -216.4 -194.8 -175.3
5->5ha -2288.9 -635.0 -571.5 -514.3 -462.9 -416.6
NPV
<1 ha -605.1
1- < 2ha -1007.6
2- < 3ha -1402.5
3- < 4ha -1660.6
4- < 5ha -2254.5
5->5ha -4889.1

68
As for monoculture milkfish farms, a 5-year nominal and a discounted cash flow

analysis (at 10% discount rate) of polyculture milkfish farm revealed that investment

in milkfish is not financially viable (Table 3.27 and Table 3.28). Both nominal and

discountedcashflow showspolyculture milkfish farms to have negative cashflow.

However, comparedwith monoculture, polyculture milkfish farm had better financial

performance.

Table 3.27 Nominal cash-flow projection for polyculture milkfish farm.

Unit: Thousand NT$
YearO Yearl Year2 Year3 Year4 Years
Cashflow
Capital cost
<1 ha 812.6 0 0 0 0 0
1- < 2ha 768.1 0 0 0 0 0
2- < 3ha 1325.7 0 0 0 0 0
3- < 4ha 1944.6 0 0 0 0 0
4- < 5ha 2387.2 0 0 0 0 0
5->5ha 2295.2 0 0 0 0 0
Operatingcost
<1 ha 0 512.0 512.0 512.0 512.0 512.0
1- < 2ha 0 716.9 716.9 716.9 716.9 716.9
2-<3ha 0 1115.7 1115.7 1115.7 1115.7 1115.7
3- < 4ha 0 1871.2 1871.2 1871.2 1871.2 1871.2
4- < 5ha 0 2631.3 2631.3 2631.3 2631.3 2631.3
5->5ha 0 3999.8 3999.8 3999.8 3999.8 3999.8
Revenue
<1 ha 432.2 432.2 432.2 432.2 432.2
1- < 2ha 516.1 516.1 516.1 516.1 516.1
2- < 3ha 911.3 911.3 911.3 911.3 911.3
3- < 4ha 1000.9 1000.9 1000.9 1000.9 1000.9
4- < 5ha 2631.2 2631.2 2631.2 2631.2 2631.2
5->5ha 3627.4 3627.4 3627.4 3627.4 3627.4
Net cash flow
<1 ha -812.6 -79.8 -79.8 -79.8 -79.8 -79.8
1- < 2ha -768.1 -200.8 -200.8 -200.8 -200.8 -200.8
2- < 3ha -1325.7 -204.4 -204.4 -204.4 -204.4 -204.4
3- < 4ha -1944.6 -870.3 -870.3 -870.3 -870.3 -870.3
4- < 5ha -2387.2 -0.1 -0.1 -0.1 -0.1 -0.1
5->5ha -2295.2 -382.3 -382.3 -382.3 -382.3 -382.3

69
Table 3.28 Discounted cash-flow projection for polyculture milkfish farm.
The discount rate for NPV is 10%. Unit: Thousand NT$
YearO Yearl Year2 Year3 Year4 Year5

Cashoutflow
<1 ha 812.6 460.8 414.7 373.2 335.9 302.3
1- < 2ha 768.1 645.2 580.7 522.6 470.4 423.3
2- < 3ha 1325.7 1004.1 903.7 813.3 732.0 658.8
3- < 4ha 1944.6 1684.1 1515.6 1364.1 1227.7 1104.9
4- < 5ha 2387.2 2368.2 2131.4 1918.2 1726.4 1553.8
5->5ha 2295.2 3599.8 3239.8 2915.8 2624.2 2361.8
Revenue
<1 ha 0 388.9 350.0 315.0 283.5 255.2
1- < 2ha 0 464.5 418.1 376.2 338.6 304.8
2- < 3ha 0 820.1 738.1 664.3 597.9 538.1
3- < 4ha 0 900.8 810.7 729.6 656.7 591.0
4-< 5ha 0 2368.1 2131.3 1918.1 1726.3 1553.7
5->5ha 0 3255.7 2930.1 2637.1 2373.4 2136.1
Net cash flow
<1 ha -812.6 -71.8 -64.6 -58.2 -52.4 -47.1
1- < 2ha -768.1 -180.7 -162.6 -146.4 -131.7 -118.6
2- < 3ha -1325.7 -183.9 -165.6 -149.0 -134.1 -120.7
3- < 4ha -1944.6 -783.3 -704.9 -634.4 -571.0 -513.9
4- < 5ha -2387.2 -0.1 -0.1 -0.1 -0.1 -0.1
5->5ha -2295.2 -344.1 -309.7 -278.7 -250.8 -225.8
NPV
<1 ha -294.9
1- < 2ha -740.8
2- < 3ha -754.6
3- < 4ha -3209.6
4- < 5ha -2.7
5->5ha -1411.4

If the grosscash flow was considered,i. e., the opportunity cost (interest, depreciation

and land cost) are excluded and only expenditure costs were considered,a 5-year

nominal and a discountedcash flow analysis (at 10% discount rate) of monoculture

milkfish farm reveal that investment in milkfish is not financially viable (Table 3.29

70
and Table 3.30). Both nominal and discounted gross cash flow shows monoculture

milkfish farms have negative cashflow.

Table 3.29 Nominal gross cash-flow projection for monoculture milkfish farm.
Unit: Thousand NT$
YearO Yearl Year2 Year3 Year4 Year5

Cashflow
Capital cost
<1 ha 502.0 0 0 0 0 0
1- < 2ha 690.4 0 0 0 0 0
2- < 3ha 815.4 0 0 0 0 0
3- < 4ha 826.4 0 0 0 0 0
4- < 5ha 1160.4 0 0 0 0 0
5->5ha 2288.9 0 0 0 0 0
Operatingcost
<1 ha 0 292.7 292.7 292.7 292.7 292.7
1- < 2ha 0 469.8 469.8 469.8 469.8 469.8
2- < 3ha 0 773.0 773.0 773.0 773.0 773.0
3- < 4ha 0 1065.0 1065.0 1065.0 1065.0 1065.0
4- < 5ha 0 1352.6 1352.6 1352.6 1352.6 1352.6
5->5ha 0 2777.9 2777.9 2777.9 2777.9 2777.9
Revenue
<1 ha 322.6 322.6 322.6 322.6 322.6
1- < 2ha 520.4 520.4 520.4 520.4 520.4
2- < 3ha 858.8 858.8 858.8 858.8 858.8
3- < 4ha 1167.9 1167.9 1167.9 1167.9 1167.9
4-< 5ha 1501.8 1501.8 1501.8 1501.8 1501.8
5->5ha 2804.3 2804.3 2804.3 2804.3 2804.3
Net cash flow
<1 ha -502.0 30.0 30.0 30.0 30.0 30.0
1- < 2ha -690.4 50.7 50.7 50.7 50.7 50.7
2- < 3ha -815.4 85.7 85.7 85.7 85.7 85.7
3- < 4ha -826.4 102.9 102.9 102.9 102.9 102.9
4- < 5ha -1160.4 149.1 149.1 149.1 149.1 149.1
5->5ha -2288.9 26.4 26.4 26.4 26.4 26.4

71
Table 3.30 Discounted gross cash-flow projection for monoculture milkfish farm.
The discount rate for NPV is 10%. Unit: Thousand NT$
Year0 Yearl Year2 Year3 Year4 Years
Cashoutflow
<1 ha 502.0 263.4 237.1 213.4 192.0 172.8
1- < 2ha 690.4 422.8 380.5 342.5 308.2 277.4
2- < 3ha 815.4 695.7 626.2 563.5 507.2 456.5
3- < 4ha 826.4 958.5 862.7 776.4 698.8 628.9
4- < 5ha 1160.4 1217.4 1095.6 986.1 887.4 798.7
5->5ha 2288.9 2500.1 2250.1 2025.1 1822.6 1640.3
Revenue
<1 ha 0 290.4 261.3 235.2 211.7 190.5
1- < 2ha 0 468.4 421.6 379.4 341.5 307.3
2- < 3ha 0 772.9 695.6 626.0 563.4 507.1
3- < 4ha 0 1051.1 946.0 851.4 766.3 689.6
4- < 5ha 0 1351.6 1216.4 1094.8 985.3 886.8
5->5ha 0 2523.9 2271.5 2044.3 1839.9 1655.9
Net cashflow
<1 ha -502.0 27.0 24.3 21.8 19.7 17.7
1- < 2ha -690.4 45.6 41.0 36.9 33.2 29.9
2- < 3ha -815.4 77.2 69.4 62.5 56.3 50.6
3- < 4ha -826.4 92.6 83.4 75.0 67.5 60.8
4- < 5ha -1160.4 134.2 120.8 108.7 97.9 88.1
5->5ha -2288.9 23.8 21.4 19.2 17.3 15.6
NPV
<1 ha -391.5
1- < 2ha -503.6
2- < 3ha -499.5
3- < 4ha -447.1
4- < 5ha -610.8
5->5ha -2191.6

As with monoculture milkfish farm, a 5-yaer nominal and a discountedgross cash

flow analysis (at 10% discount rate) of polyculture milkfish farm revealedthat

investment in milkfish is not financially viable (Table 3.31 and Table 3.32). Both

nominal and discountedcashflow shows polyculture milkfish farms have negative

cashflow. Comparedwith monoculture, polyculture milkfish farm did not have better

financial performance.

72
Table 3.31 Nominal grosscash-flow projection for polyculture milkfish farm.
Unit: ThousandNT$
Year0 Yearl Year2 Year3 Year4 Year5
Cash flow
Capital cost
<1 ha 812.6 0 0 0 0 0
1- < 2ha 768.1 0 0 0 0 0
2- < 3ha 1325.7 0 0 0 0 0
3- < 4ha 1944.6 0 0 0 0 0
4- < 5ha 2387.2 0 0 0 0 0
5->5ha 2295.2 0 0 0 0 0
Operatingcost
<1 ha 0 451.5 451.5 451.5 451.5 451.5
1- < 2ha 0 585.4 585.4 585.4 585.4 585.4
2- < 3ha 0 890.4 890.4 890.4 890.4 890.4
3- < 4ha 0 1551.2 1551.2 1551.2 1551.2 1551.2
4- < 5ha 0 2211.3 2211.3 2211.3 2211.3 2211.3
5->5ha 0 3347.9 3347.9 3347.9 3347.9 3347.9
Revenue
<1 ha 432.2 432.2 432.2 432.2 432.2
1- < 2ha 516.1 516.1 516.1 516.1 516.1
2-<3ha 911.3 911.3 911.3 911.3 911.3
3- < 4ha 1000.9 1000.9 1000.9 1000.9 1000.9
4-< 5ha 2631.2 2631.2 2631.2 2631.2 2631.2
5->5ha 3627.4 3627.4 3627.4 3627.4 3627.4
Net cash flow
<1 ha -812.6 -19.4 -19.4 -19.4 -19.4 -19.4
1- < 2ha -768.1 -69.3 -69.3 -69.3 -69.3 -69.3
2- < 3ha 20.9 20.9 20.9 20.9 20.9
-1325.7
3- < 4ha -1944.6 -550.3 -550.3 -550.3 -550.3 -550.3
4- < 5ha 419.9 419.9 419.9 419.9 419.9
-2387.2
5->5ha 279.6 279.6 279.6 279.6 279.6
-2295.2

73
Table 3.32 Discounted gross cash-flow projection for polyculture milkfish farm.
The discount rate for NPV is 10%. Unit: Thousand NT$
Yeah Yearl Year2 Year3 Year4 Year5

Cashoutflow
<1 ha 812.6 406.4 365.7 329.2 296.2 266.6
1- < 2ha 768.1 526.9 474.2 426.8 384.1 345.7
2- < 3ha 1325.7 801.3 721.2 649.1 584.2 525.8
3- < 4ha 1944.6 1396.1 1256.4 1130.8 1017.7 916.0
4- < 5ha 2387.2 1990.2 1791.2 1612.0 1450.8 1305.8
5->5ha 2295.2 3013.1 2711.8 2440.6 2196.5 1976.9
Revenue
<1 ha 0 388.9 350.0 315.0 283.5 255.2
1- < 2ha 0 464.5 418.1 376.2 338.6 304.8
2- < 3ha 0 820.1 738.1 664.3 597.9 538.1
3- < 4ha 0 900.8 810.7 729.6 656.7 591.0
4- < 5ha 0 2368.1 2131.3 1918.1 1726.3 1553.7
5->5ha 0 3264.7 2938.2 2644.4 2380.0 2142.0
Net cashflow
<1 ha -812.6 -17.4 -15.7 -14.1 -12.7 -11.4
1- < 2ha -768.1 -62.4 -56.1 -50.5 -45.5 -40.9
2- < 3ha -1325.7 18.8 16.9 15.2 13.7 12.3
3- < 4ha -1944.6 -495.3 -445.7 -401.2 -361.1 -324.9
4- < 5ha -2387.2 377.9 340.1 306.1 275.5 247.9
5->5ha -2295.2 251.6 226.4 203.8 183.4 165.1
NPV
<1 ha -884.1
1- < 2ha -1023.6
2- < 3ha -1254.9
3- < 4ha -3972.8
4- < 5ha -839.7
5->5ha -1264.9

3.5.6 Price sensitivity

The profitability is sensitiveto the changesin the selling price. Here, the relationship

betweengrossprofit of monoculture to different selling price is compared.The

74
averagefarm gate price was 33.8 NT$/kg. The averagebreak even prices were 29.0,

30.4,29.5,42.6,17.9 and 46.4 NT$/kg in the size categoriesof <1 ha, 1- <2 ha, 2-

3 ha, 3-<4 ha, 4-<5 ha and 5->5 ha, respectively.Prices of 25,30,35,40 and 45

NT$ were used to compare the sensitivity of profitability to price. When price

dropped down to 25 NT$/kg, only in the farm size of 4-<5 ha could make profit.

With the exception of farm size of 4-<5 ha, all the farms could make profit when the

price rose to 45 NT$/kg, the profit of farm size of 4-<5 ha could reach 2,041,600

NT$ (Table 3.33)

Table 3.33 The sensitivity of profitability to price. Unit: NT$

Price 25 30 35. 40 45
< 1ha -40243 10245 60733 111222 161710
1- <2 ha -83753 -6509 70735 147978 225222
2-<3 ha -117622 13462 144547 275631 406716
3-<4 ha -64305 60785
-439576 -314486 -189396
4-<5 ha 533049 910183 1287317 1664452 2041586
5 -> 5 ha -1281952 -982766 -683580 -384394 -85208

3.6 Marketing channels

The marketing channel for milkfish is very complex (Fig. 3.3). Usually, there are five

ways for fish farmers to sell their product and the money was paid by either cashor

short term cheque.

f Sell product directly to the retail market or restaurantsby themselves.Profits can be

higher but this is more complicated and time-consuming.Usually, farmers may only

sell part of their products directly to retail markets.

f Sell product directly to the processingplant. Usually, the processingplant will sign

a contract with the fish farmer and this is often preferred by producers.However,

75
becauseof factors such as the capacity of the processingplant, the required market

size of the processedproduct or the distancebetweenthe plant and the production

site, not all the fish farmers can use this method regularly.

f Sell product directly to an auction market. In this situation the fish farmer must take

the risk of price fluctuation directly, as there is no contract for price. In the auction

market, the prices are decided by bidding and sales are organised through agents.

f Sell to a middleman, who collect products from farmers or auction markets and re-

sell the products to retail markets or consumers(including Middlemen

restaurants).

may pay a higher price to the producers.However, becausethe middlemen usually

deal not only with milkfish and cannot buy large quantities, except for smaller farms,

most producersdo not sell their product this way. However somewill sell to a

middleman becauseof the pressureof friendship.

f Sell to a collector. This is a traditional method and is the most common because

although price may not be so high, it is more convenient and farmers can obtain their

money quickly. There were anothertwo reasonsfor farmers usually selling their

product to collectors. First, they do not know the marketing channel and are only used

to the traditional method of selling their fish. Second,they do not have vehicles to

transport their product to the auction market and therefore, they hire the vehicles

belonging to the collector for this purpose.However, the fish farmers must take the

risk in the sale to the collector.

76
 <1%

-30% Middleman
-1

T-, 30%
Fish farmers Consumers
Collector

Auction market Retail market

-30% II
-10% Processplant

Fig 3.3 Chart of marketing channel for milkfish in Taiwan.

Data source:Tzeng (personalcommunication 1999).

3.7 Supply and price relationships

For understandingthe factors which influence price, data were collected from Chia-I

fisheries market (one of the biggest fisheries auction markets in Taiwan). The data is

provided in the form of average by

prices month. Production data was obtained from

the Year Book of the Taiwan FisheriesBureau (1994-1998) to assessthe relation

betweenproduction quantity and price.

3.7.1 Seasonalvariation of production and price

Table 3.26 shows the outlines of production of milkfish in Taiwan over the years

1994-1998and presentsan index of seasonalvariation, basedon averagetrends.This

shows that the seasonof highest production centerson June to Decemberwhen the

indices are over 100%,the averagepeak being in October, with an index of 143%

(Table 3.34 and Fig. 3.4.). However, in different years,the peaksof production were

77
in different months, suggesting that farmers have changed the practice of harvesting

in different months. In general, the higher production season is from summer to early

winter. The index of seasonal price (Table 3.35 and Fig. 3.5) shows that the season of

average higher prices centers on January to June, with indices at over 100%, the

highest being in April, where the average index is 120.9%. The seasonal indices of

prices have a negative relationship with indices of production. The indices of price

became higher when the indices of production were lower.

Table 3.34 The seasonal variation of production of milkfish in Taiwan. Unit: ton

Year Jan. Feb. Mar. Apr. May Jun. Jul. Aug Sep. Oct. Nov. Dec. Ave
1994 3780 3673 3419 3800 4098 5883 8516 7137 7903 10499 4581 2515 5484
(69) (67) (62) (69) (75) (107) (155) (130) (144) (191) (84) (46) (100)
1995 3848 3983 3609 4613 4024 4786 4297 6819 6369 7524 9228 4424 5294
(73) (75) (68) (87) (76) (90) (81) (129) (120) (142) (174) (84) (100)
1996 3974 3459 4037 2752 3838 6078 3933 4279 6541 7153 7235 5269 4879
(81) (71) (83) (56) (79) (125) (81) (88) (134) (147) (148) (108) (100)
1997 3876 4061 2142 4607 4292 7870 6966 4589 5448 6613 3656 8625 5229
(74) (78) (41) (88) (82) (151) (133) (88) (104) (126) (70) (165) (100)
1998 3644 3663 3642 4265 4528 7391 7065 5468 5696 5141 2845 5002 4863
(75) (75) (75) (88) (93) (152) (145) (112) (117) (106) (59) (103) (100)
Ave. 3824 3768 3370 4007 4156 6402 6155 5658 6391 7386 5509 5167 5150
(74) (73) (65) (78) (81) (123) (120) (110) (124) (143) (107) (100) (100)
Data source: Year Book of Taiwan Fisheries Bureau (1995-1999).
The figures in the parentheses are the indices of seasonal variation.

8000

6000
Z
4000
U

C, ? 000

Ali. May Jun Jul Aug Sc 1). Oct. Nov. I)cc.

Jim.
. hab. Mai
Month

Fig. 3.4 The average production quantity of milkfish from 1994 to 1998.

78
Table 3.35 The seasonal variation of price of milkfish in Taiwan. Unit: NT$

Year Jan. Feb. Mar. Apr. May Jun. Jul. Au Sep. Oct. Nov. Dec. Ave
1994 56.9 57.3 55.8 59.2 55.9 46 43.3 47.7 43.2 45 48.9 53 51.0
(112) (112) (109) (116) (110) (90) (85) (93) (85) (88) (96) (104) (100)
1995 53.7 63.5 71.3 68.8 71.5 66.1 62.8 66.3 59.1 54.4 52.7 61.4 62.6
(86) (101) (114) (110) (114) (106) (100) (106) (94) (87) (84) (98) (100)
1996 71 77 91.6 101.1 104 90.2 66.1 66.4 54.4 55.5 55 61.6 74.5
(95) (103) (123) (136) (140) (121) (89) (89) (73) (75) (74) (83) (100)
1997 69.9 75.1 72 71.3 65.1 60.1 48.1 48.4 47.2 44.8 47.7 51.7 58.5
(120) (128) (123) (122) (111) (103) (83) (83) (81) (77) (82) (88) (100)
1998 52.7 53.3 51.3 52.3 47.4 42.3 39.9 45.7 40 40.8 39.5 37.9 45.3
(116) (118) (113) (116) (105) (93) (88) (101) (88) (90) (87) (84) (100)
Ave. 60.8 65.2 68.4 70.5 68.8 60.9 52.0 54.9 48.8 48.1 48.8 53.1 58.4
(104) (112) (117) (121) (118) (104) (89) (94) (84) (82) (84) (91) (100)
Data source: Chai-I Fisheries Market (1999).
The figures in the parentheses are the indices of seasonal variation.

80

60
-
Z 40
U

20

Jan. Feb. Mar. Apr. May Jun. Jul. Aug Sep. Oct. Nov. Dec.

Month

Fig. 3.5 The average price of milkfish from 1994 to 1998.

As indicated, sales levels are higher in summer and autumn and lower in spring, while

price is highest in spring and lowest in the summer and autumn, suggesting a negative

correlation between the price and output. To test the relationship between price and

79
production of milkfish, including lagged quantity and seasonalvariation, regressions

were determinedbasedon the following function form.

Pc =f (Qt, Qt-1, Qt-2, Qt-3, D1, D2, D3)

In this function Pt is the undeflatedprice of milkfish in month t, in NT$ per kilogram;

Qt-19Qt-29Qt_3,are production quantity in months t, t-1, t-2 and t-3, in respectively, in

thousandkilogram. D1 to D3 are dummy variables which representspring, summer

and autumn.The results of regressionwere shown in Table 3.36, from which it can be

shown that the price of milkfish is strongly affectedby seasonalvariation and by

production quantity in the samemonth (Qt). The coefficients of Qt in different

regressionsare from -0.0011 to -0.0024, that meanswhen the national production

increase 1 mt, the price of milkfish might decrease 0.0011 to 0.0024 NT$.

Table 3.36 Regressionequationsfor price of milkfish in Taiwan.

Equation Constant t t-1 Qt-2 Qt-3 D1 D2 D3 R F
A 72.643 -0.00244 0.179 33.47
(31.72) (-5.79)
B 75.372 Tr-
0.240 24.03
-0.001211 -0.0019
(31.91) (-2.27) (-3.57)
C 77.738 -0.001341 -0.000830 -0.0015 0.278 19.24
(31.35) (-2.56) (-1.30) (-2.90)
D 79.884 -0.00154 -0.000789 -0.000594 -0.00131 0.302 16.02
30.21 (-2.94) -1.26 (-0.95) (-2.51)
E 61.9 -0.00182 17.9 137 1.07 0.317 17.55
(20.12) (-4.15) 4.69) 3.51) 0.26
F 64.601 -0.00107' -0.0012 16.4 11.6 0.92 0.343 15.56
19.10 (-2.51) -1.69 4.19) 2.98 0.23
G 67.7 -0.00112 -0.000579 -0.00102 14.7 9.24 0.356 13.57
-0.07
18.17) (-2.63) (-0.68) (-1.53) 3.68) (2.28) (-0.02)
H 71.76 -0.00120' -0.000554 -0.000282 -0.00120 12.5 6.43 -2.89 0.371 12.24
17.44) (-2.27) -0.92 -0.46 -2.26) (3.08) 1.53 (-0.69)
Data sources:Chai-I FisheriesMarket (1999).
Year Books of Taiwan Fishery Bureau (1988-1999).
The prices used are defalted.
Number in parenthesesare t values.

80
3.7.2 Long-run variation of price and production

The degree of annual fluctuation or instability in the price and production of milkfish

can be evaluatedby using the Michaely index (f), where

f= y-it
's2 x'-'
x 100
n-1

The higher the value off, the more instability is implied. If f is above 20%, it means

extreme instability and slight instability if f is less than 10% (Lee 1983).The f value

of the milkfish price from 1986 to 1999 was 22.04% and the value of milkfish

production from 1987 to 1998 is 60.23%. This shows that the price of milkfish in

Taiwan was highly unstablein Taiwan during theseperiods, and production was even

more unstable.

3.7.3 Annual trend of yield and price of milkfish

The trend of production of milkfish in Taiwan from 1987 to 1999 can be expressedas:

Y= 35537+2094t R2= 0.176 F= 2.36 DW= 2.41

(3.28)** (1.54)

Where

Y= annual total production of milkfish in MT

t= number of yearsfrom 1987 to 1999

R2= coefficient of determination

DW= Durbin Watson d test

81
The trend of price of milkfish in Taiwan from 1986to 1999can be expressedas:

P= 100.669-3.6234t R2= 0.585 F= 16.91** DW= 2.16

(13.42)** (-4.41)**

Where

P= deflated averageprice of milkfish in NT$

t= number of yearsfrom 1986to 1999

R2= coefficient of determination

DW= Durbin Watson d test

Although DW value shows there is no positive serial correlation, the regressionshows

the long term trend of increasingproduction and decreasingprice.

3.7.4 Actual apparentconsumption

The real consumption level by domestic market can be expressedin kg/capita/year

and can be evaluatedby using actual apparentconsumption (AAC), where

AAC = (Production + Import - Export)/ Number of people

The result of AAC were shown in Table 3.29, from which it can be shown that the

highest AAC was in 1990,followed in 1994,reaching 4.4 and 3.2 kg. However, since

1994,the value of AAC decreasedgradually and was only 1.4 in 2000. The highest

value of AAC might be becauseof the massproduction in 1990.Table 3.26 shows

that since 1996,Taiwan developedthe export market, accounting for - 5-9% of total

milkfish production.

82
Table 3.37 The actual apparent consumption (AAC) of milkfish.

Year Production Export Import Population AAC*

(mt) (mt) (mt) (1,000 people) (kg/people/year)
1990 90,673 0.12 0.11 20,401 4.4
1991 41,232 2.93 0 20,606 2.0
1992 25,114 0.98 0 20,803 1.2
1993 45,513 2.79 0 20,995 2.2
1994 66,778 5.08 0 21,178 3.2
1995 63,254 26.71 0 21,357 3.0
1996 58,453 6,531.26 0.19 21,525 2.4
1997 62,749 10,765.00 0.01 21,743 2.4
1998 58,349 9,581.67 7.83 21,929 2.2
1999 50,824 9,640.69 0.97 22,092 1.9
2000 39,731 7,509.10 77.06 22,277 1.4
Data source:FisheriesAdministration, Taiwan (www. fa. og
v_tw).
Directorate Generalof Budget Accounting and Statistic, Executive
Yuan, R.O.C. (www. dgbas.gov. tw).
v_tw).
*AAC refers to actual apparentconsumption.

3.7.5 Consumerperspective

To understandbetter the attitudes of consumerstowards milkfish, 132 individuals

were surveyedin 1999.3 areaswere surveyed;Taipei (52 consumers),Taichung (45

consumers)and Tainan (35 consumers),representingnorthern, central and southern

part of Taiwan respectively.

As shown in Table 3.38, the majority of respondents(85%) prefer fresh milkfish.

Though consumersmay not have known that a lot of milkfish (dorsal part) is usedfor

producing fish meat ball or fish powder, fresh fish is clearly their favourite. As canned

tuna is - 10 NT$ cheaperthan cannedmilkfish per can (-230 g), it is difficult for

milkfish to competein this sector.

83
Table 3.38 Preferencesfor milkfish product forms.
Products Numbers Percentage Chi-square value P value
Freshfish 130 86.1% 74.42 1.89*10" **
Cannedfish 6 4.0% 8.82
Fish meatball 12 7.9% 5.78
Others 3 2.0% 10.58
Total 151 100% 99.6

Most of the respondentspreferred the belly part of milkfish (Table 3.39) becauseit is

the softest and most oily part. A very small number (5.7%) preferred the milkfish

head, while scaled and gutted or whole fish were moderately popular.

Table 3.39 Preferencesfor fresh milkfish product.

Products Numbers Percentage Chi-square value P value
Whole fish 34 21.7% 0.307692 0.000264**
Scaledand gutted whole fish 39 24.8% 0
Head 9 5.7% 7.692308
Bell part 75 47.8% 11.07692
Total 157 100% 19.07692

The survey suggestedthat consumptionof milkfish was not correlated with seasons

(Table 3.40), as respondents(>75%) did not buy milkfish in a specific season,though

they would buy a little bit more milkfish in summer than in winter. It might be

becausethe price in summer is cheaper.

Table 3.40 Seasonalpreferencefor purchasingmilkfish.

Seasons Numbers Percentage Chi-square value P value
Spring 6 4.4% 5.44 2.82*10" **
Summer 18 13.3% 1
Autumn 6 4.4% 5.44
winter 3 2.2% 7.11
Uncertain 102 75.6% 69.44
Total 135 100% 88.44

84
Most respondentspreferredthe size of milkfish at about 600g (Table 3.41), though the

market size of milkfish is usually about 200-500g. However if the size of milkfish is

too big (over 600g), its meat might be too firm and it would be over pan size.

Table 3.41 Preferredpurchasingsize of milkfish.

Sizes Numbers Percentage Chi-squarevalue P value
300 18 15.38% 1.4423086 5.62*10" **
600g 78 66.67% 27.08333
900 12 10.26% 3.39102
1200 9 7.69% 4.673077
Total 117 100% 36.58974

Most respondentsexpressedthat their frequency of buying milkfish was uncertain

(48.9%) (Table 3.42), suggesting that they would purchase at will, without pattern.

Somerespondentslike milkfish and will buy it regularly. A good number of

respondents(26.3%) like milkfish and will buy it every week.

Table 3.42 The frequenciesof buying milkfish.

Frequencies Numbers Percentage Chi-square value P value
Every week 35 26.3% 1.163636 4.2161*10' **
Every two weeks 24 18.0% 0.072727
Every three weeks 3 2.3% 6.913636
Every month 6 4.5% 5.254545
Uncertain 65 48.9% 16.91364
Total 133 100% 30.31818

In Taiwan, the unit of weight is 600g, and therefore 600g, 1200g and 1800gwere

chosenas purchasequantities. The most common amount of milkfish that respondents

report purchasing is about 600g (44.2%) (Table 3.43). Consumers might consider this

amount is suitable for a family. If they buy more they may not finish it in one meal.

85
Table 3.43 The quantity of milkfish purchasedeach time.
Amount Numbers Percentage Chi-square value P value
About 600 57 44.2% 6.331395 0.02929*
About 1200 21 16.3% 1.30814
About 1800 20 15.5% 1.40715
Uncertain 31 24.0% 0.052326
Total 129 1100.01% 9

When askedwhy they do not buy milkfish, more than 90% of respondentsanswered

that milkfish was too bony (Table 3.44). That may also be the reasonwhy some

consumers just buy the belly part, as there are no tiny bones in this part. The boniness

of milkfish appearsto be the biggest problem restricting market developmentof

milkfish in Taiwan.

Table 3.44 The reasonsthat consumersdo not buy milkfish.

Reasons Numbers Percentage Chi-square value P value
Bony 120 90.91% 43.7361213 5.2587*10" **
Too expensive 3 2.27% 12.7381663
Others 9 6.82% 9.28349693
Total 132 100% 65.7577846

More than 77% of respondentsconsideredthe price of milkfish to be acceptable

(Table 3.45), neither too expensivenor very cheap.Comparedto other aquaculture

products (Table 3.4) or other protein products, the price of milkfish is not high.

However, its boninessis a disadvantagein competition.

Table 3.45 Opinions of respondentsconcerningthe price of milkfish.

Opinions Numbers Percentage Chi-square value P value
Very expensive 4 3.0% 6.913636364 7.2238*10" **
Expensive 6 4.55% 5.254545455
Acceptable 102 77.27% 72.16363636
Cheap 14 10.6% 1.640909091
Very cheap 6 4.55% 5.254545455
Total 132 100% 91.22727273

86
Regardingthe quality of milkfish, most respondentshave a positive opinion. More

than 55% consider the quality to be good and only - 9% consider it to be below

acceptability, it might be becauseit is too bony (Table 3.46).

Table 3.46 Opinions of respondentsconcerningthe quality of milkfish.

Opinions Numbers Percentage Chi-squarevalue P value
Excellent 9 6.82% 3.822727273 3.06203* 10" **
Good 75 56.82% 29.82272727
Acceptable 36 27.27% 1.163636364
Bad 9 6.82% 3.822727273
Very Bad 3 2.27% 6.913636364
Total 132 100% 45.54545455

When combining quality and price, most respondentshave a positive opinion about

milkfish. More than 90% of consumersconsideredmilkfish are acceptableor more

than acceptable(Table 3.47). This suggestthat consumersmight be willing to pay

more.

Table 3.47 Evaluations of respondentsconcerningthe price and quality of milkfish.

Opinions Numbers Percentage Chi-square value P value
Excellent 12 9.09% 2.618181818 2.5983*10' **
Good 57 43.18% 11.82272727
Acceptable 54 40.91% 9.618181818
Bad 6 4.55% 5.254545455
Very Bad 3 2.27% 6.913636364
Total 132 100% 36.22727273

Most of consumersconsideredthat they would buy more milkfish if the price were

lower (Table 3.48). Milkfish must competewith other speciesand other food products

would also have to be taken into account.

87
Table 3.48 Situations in which consumers would buy more milkfish.

Situations Numbers Percentage Chi-square value P value

Price is cheaper 72 52.9% 5.4 0.01713*
Quality is better 33 24.3% 1.066667
Others 31 22.8% 1.666667
Total 136 99.99% 8.133333

3.8 Discussion
In Taiwan, milkfish is a traditional and very important sectorin aquaculture,

accounting for about 10-20% of total production (Table 3.3). Although fry can be

producedin hatcheries,and potential supplies are more than enough to provide for

domesticdemand,they are still imported from the Philippines becauseof the seasonal

First
shortage. spawning usually occurs in early April when the water temperature

rises to 26°C and so some farmers import fry to stock before May, to allow harvest

before that winter. Therefore, to make broodstock spawn from February to March, and

to cope with lower temperaturelarval rearing outdoors are big challengesfor

researchers.

Overall, milkfish rank first in fry production among finfish and the potential of supply

is more than enough to meet demandin the domestic market. It would thereforebe

more useful to balancethe demandand supply, to try to develop the fry market more

widely in Asia.
Southeast

Milkfish culture has beendevelopedover more than 300 yearsin Taiwan, and in

generalterms, its technology is quite mature, but a number of problems are to be

solved.The cold weather is a significant problem in the production of milkfish, as

water temperaturedrops below 10 °C, it causesmassmortality (Ding 1994).

Therefore, the accurateweatherforecasting is very important in allowing farmers to

88
set up overwintering canals or harvest before the cold weather comes. If possible, a

strategyshould be developedto set up production areasin the warmer part of Taiwan

to avoid cold-kill.

Being concernedabout the low profits from monoculture, a number of farmers used

polyculture in milkfish ponds with other species,such as tiger prawn, sandprawn,

fresh water prawn mullet and other economically valuable species to spread risks and

increaserevenues(Table 3.2 and Table 3.8). It also contributed to a smoother

harvestingpattern, and consequentlycash flow, throughout the year. Capacity can also

be utilized more evenly. However, this survey showedthat in polyculture, only farm

sizes in the category of 4-<5 ha had significantly higher profit than monoculture.

This survey suggestedthat the milkfish sectoris not economically sound.This might

be the reasonthat more than 75 % of milkfish farmers have other family income

(Table 3.5), and in the age category of 30 < 40 yearsold, the ratio of other family

income was even as high as 92% (Table 3.6). However most milkfish farmers have

their own land and have already made investmentsin facilities for cultivating. As a

result, it is difficult for them to changeto anotherusagethough the sunk costs also

meanthat their effective return may be greaterthan those suggestedby net profit

calculations.Even in years when the price of milkfish is not high enough to cover

costs, farmers still rear milkfish and hope the fish will fetch a better price the next

year.

Further understandingthe ratio of their labour input on milkfish farming and other

jobs, and the earning from milkfish and other job might be helpful in understanding

89
the involvement of farmers in milkfish farming. Farm size in the categoriesof 4- <5

ha for monoculture could appearto be more profitable. The lowest cost producer is

efficient as a result of economiesof scalefor family labour. However, when farm size

exceeded5 ha, it was difficult for a family to manage.In polyculture, the range of

profitability was wider than monoculture. Farmerswith better performancecould

attain higher profits than in monoculture. In the future, farms might be adjustedto 4-

<5 ha or replacedby better managedpolyculture farms.

The marketing channel is very complex. With at least five routes for selling fish and a

number of intermediary this

stages, suggestsinefficiency, and the need over the longer

term to shortenthe marketing channel.Even though production and marketing groups

have been set up, their function was not obvious (Wu 1998), though if thesewere

strengthened, this group might shorten the marketing channel. Production and

marketing groups will be discussedfurther in Chapter6.

The price of milkfish is easily influenced by production quantity. The fluctuation of

production might also causethe variation of price. Overproduction could make the

market prices drop dramatically. For example, in 1990, when the tiger shrimp sector

collapsed,some shrimp farmers adaptedtheir ponds to cultivate milkfish, which

induced overproduction.This overproduction made the price of milkfish drop from 60

to below 45 NT$ (Liao 1993) and causedeconomic distressto producers.Although

this extreme situation is unlikely to happenagain in the near future, the high values of

Michaely index show that the production and price of milkfish were unstable.Such an

unstablesituation might make this industry more risky. Therefore, proper information

about the production areasand predicted production amount might be helpful to avoid

90
this disaster.

To balancethe demandand supply more effectively, better planning of production is

required, and market conditions monitored by the government or Fishermen

Association to advise farmers if significant unbalances are likely to occur. As price is

also strongly correlated to seasonal

variation, farmers may be able to do more to

adjust the harvesttime to match the higher price period in April and harvest size. In

the long-term analysis,price was in a decreasingtrend. It might be becauseof the

long term increasingtrend of production causedby the introduction of deep-water

culture.

With respectto the potential for expandingmarkets, most consumershad positive

opinions on milkfish. Generally, consumerspreferred the belly part of milkfish and a

size of about 600g. The low price of milkfish might be an advantagefor market

competition with other products, however their boninessis a disadvantage.For

expanding market and competing with other products, the technique of boneless

preparationfrom the Philippines might be helpful. Although one processorhas learned

this techniquefrom the Philippines (China Times, December22,2001), it is still a

secrettechnique in Taiwan.

There has also been somerisk due to negative pressreporting. Thus when on

December25,2000, a newspaperreported researchresults by Chin-Hwa University

that milkfish were contaminatedby chlorine, the price declined to below 40 NT$ the

next day.Therefore, proper in

management fish farms to avoid any contamination and

appropriatequality control through FishermenAssociation and production and

91
marketing groups might improve the image of milkfish.

Since 1996,certain amount of export marketshave been developedand about 5-9 %

of total production was exported (Table 3.26), it might be benefit for the development

of milkfish industry in the long term.

92
 Chapter4

Eel Culture

4.1. Introduction

4.1.1. Background

Eel culture startedin Japanin 1879 (Matsui 1952), and during a similar period in Italy

and France (Gousset 1990). The Japanese started to rear glass eels in 1919, and

artificial feeds were introduced to the market in 1965. The first experiment of the

feasibility of eel culture in Taiwan beganin 1952, and in 1958 small-scale

commercial eel farming started.The first large-scaleexpansionof eel farming began

in 1964 with the raising of glasseels to stocking size fingerlings for Japaneseeel

farms. The first export of market size eels to Japantook place in 1970 (Chen1990).

Eel culture hasbeen one of the most important aquaculturesectorsin Taiwan. The

total value of eel production is the highest among all aquaculturesectors,with almost

90% of production exported to Japan.In 1988,its value had reached43.2% of total

aquaculturevalue (Table 4.1). Eel culture can be carried out in both of fresh and salt

water. However, most farmers use fresh water becauseof faster growth. Eel is a high

value product and farmers usually use intensive monoculture. Since 1988, the ratio of

eel culture areato total national aquacultureareahas ranged from 6.2 to 2.4% and that

of eel to total aquacultureoutput has rangedfrom 19.5 to 6.3%. The productivity of

aquaculturehas ranged from 3.5 to 4.5 mt/ha, while that of eel has ranged from 8.4 to

14.8 mt/ha (Table 4.1 and 4.2), illustrating its highly intensive nature. The ratio of eel

to total aquacultureproduction value has ranged from 43.2 to 18.7% since 1988

93
(Table 4.1). The averageprice (V/Q) of eel rangedfrom 200 to 289 NT$/kg from

1987to 1993.However, the averageprices increased,since 1994 to 1998,the prices

were higher than 300NT$ and reachedthe highest level of 457 NT$/kg in 1995. In

1999,the averageprice returnedto 266 NT$ (Table 4.1). This changemight be

influenced by the prices of eel seed(Table 4.4). Since 1988,the averageprices of

aquaculture products have ranged from about 90 to 130 NT$ and those of eel have

ranged from 200 to 460 NT$ (about 2 times the average for aquaculture) (Table 4.1),

illustrating its high price.

However, becauseof limited land and water resources,a shortageof eel seedand

competition with China for the Japanese

market, the sector has been declining (Fig.

4.1 and Table 4.1) and almost 1/3 (33.1%) of culture areawas suspendedin 1997

(Table 4.2). From 1987 to 1999, the averagechangeof eel output and value hasbeen -

61.01% and -64.07% respectively (Table 4.1). To reducethe use of land and ground

water, someproducershave recently developedindoor super intensive culture. In this

chapter,traditional and the super intensive eel culture will be describedand

compared.The averageyield of eel production was highest in 1991,reaching 14.8

mt/ha (Table 4.2). However, it decreasedto 8.44 mt/ha in 1999, implying that more

intensive culture was not applied widely.

94
Table 4.1 Output and value of eel culture in Taiwan.
Unit: Quantity: ThousandM. T.
Value: Million NT$
Year Total aquaculture Eel culture R R
Quantity Value V/Q Quantity Value V/Q
1987 305.4 35.23 115.4 42.5 12.23 287.9 13.9 34.7
1988 301.0 34.48 114.6 51.6 14.90 288.8 17.1 43.2
1989 250.0 26.52 106.2 48.0 10.61 220.9 19.2 40.0
1990 344.3 31.53 91.6 55.8 12.36 221.5 16.2 39.2
1991 291.9 30.26 103.7 55.6 11.11 199.6 19.1 36.7
1992 261.6 29.29 112.0 51.0 11.73 229.8 19.5 40.0
1993 285.3 29.82 104.5 40.0 11.15 279.1 14.0 37.4
1994 288.0 33.57 116.6 33.4 12.98 389.1 11.6 38.7
1995 286.6 36.51 127.4 25.5 11.67 456.9 8.9 32.0
1996 272.5 32.73 120.1 25.1 10.52 419.9 9.2 32.2
1997 270.1 27.10 100.3 22.3 8.55 382.8 8.3 31.6
1998 255.2 27.39 107.3 17.2 6.02 349.5 6.8 22.0
1999 263.1 23.51 89.4 16.5 4.39 265.6 6.3 18.7
Average -1.24% -33.28% -61.01% -64.07%
growth rate
(1987-1999)
Data source:FisheriesYear Book, Taiwan Area.
Rl representsthe ratios of total output of eel culture to total output of aquaculture.
R2representsthe ratios of total production value of eel culture to total production
value of aquaculture.
The figures are undeflated

95
Table 4.2 The area for aquaculture of Anguilla sp. Unit: Thousand ha

Year Total Eel culture R R2

A uaculture
Area MT/ha Mono- Poly- Suspended Total MT/ha
culture culture
1988 67.41 4.47 3.51 0.03 ------- 3.53 14.60 5.24 5.24
(99.28) (0.71)
1989 71.08 3.51 3.92 0.10 ------- 4.01 11.97 5.64 5.64
(97.58) (2.41)
1990 76.42 4.50 3.95 0.02 ------- 3.97 14.08 5.19 5.19
(99.55) (0.45)
1991 74.08 3.94 3.72 0.03 ------- 3.75 14.83 5.07 5.07
(99.22) (0.66)
1992 72.29 3.62 4.43 0.02 ------- 4.46 11.44 6.17 6.17
(99.48) (0.52)
1993 70.97 4.02 3.20 0.02 0.70 3.92 12.41 5.52 4.54
(81.68) (0.48) (17.84)
1994 69.60 4.14 2.90 0.07 1.23 4.20 11.22 6.03 4.27
(69.13) (1.66) (29.20)
1995 70.08 4.09 2.50 0.04 1.08 3.62 10.05 5.17 3.63
(68.96) (1.14) (29.90)
1996 67.61 4.03 2.17 0.07 1.07 3.31 11.19 4.89 3.31
(65.56) (2.13) (32.31)
1997 63.16 4.28 1.80 0.08 0.93 2.81 11.86 4.46 2.98
(64.02) (2.88) (33.10)
1998 63.19 4.039 1.46 0.07 0.59 2.13 11.23 3.36 2.43
(68.79) 3.40 (27.81)
1999 63.21 4.16 1.91 0.05 0.48 2.44 8.44 3.86 3.10
78.26 (1.93) (19.80)
Data source: Fisheries Year Book, Taiwan Area.
Figures in the parenthesis are the ratios to total area for eel culture.
R' representsthe ratios of total areasof eel culture to total areasof aquaculture.
R2representsthe ratios of real areasof eel culture (excluding suspendedarea)to total
areasof aquaculture.

96
 200000

160000

120000

I-f-Japan
80000
v --*-China

40000 --h-Taiwan

0
1988 1989 1990 1991 1992 1993 1994 1995 1996 1997
Years

Fig 4.1 The production amount of eels in China, Japan and Taiwan.

Data source:FAO (1999).

4.1.2 Eel seed

In Taiwan, the major speciesof culture is the Japaneseeel (Anguilla japonica), a

temperatecatadromousfish which is widely distributed in the rivers of NE Asia, i. e.

China, Taiwan, Japanand Korea (Tesch 1977). The catch levels of glasseel of

different countries are shown in Table 4.3 and Table 4.4.

The spawning ground of Japaneseeel was discoveredin the North Equatorial Current

west of the Mariana Islands, 15° N, 1400E (Tsukamoto, 1992). The leptocephali drift

with the North Equatorial Current to the continental shelf of the Philippines, then turn

northward into the Kuroshio Current conveyedby the mechanismof Ekman transport

(Kimura et al., 1994).

97
Table 4.3 Catchesof glasseel of different countries.*
Unit: mt
Year Japan Ch ina Korea
Million MT Million MT Million MT
seed seed seed
1991 255.8 46.5 220 40 49.5 9
1992 225.5 41 198 36 49.5 9
1993 236.5 43 ---- ---- ----
1994 155.1 28.2 192.5 35 ---- ----
1995 191.4 34.8 242 44 44 8
1996 160.6 29.2 82.5-99 15-18 38.5 7
1997 137.5 25 82.5-110 15-20 33 6
1998 68.8 12.5 41.3-48.4 7.5-8.8 9.9 1.8
1999 352.0 64 330 60 27.5 5
2000 93.5 17 275 50 27.5 5
Data source: Japan Aquculture News (2001).
* The figures of million seed are estimated from 5.5 million
glass eels per ton of eel seed.

In Taiwan, the fishery seasonof Japaneseelvers is from October to March, peaking in

Decemberand January.The glasseel are collected in estuarinewaters with traps,

seinesor scoop nets from boats or by wading (Chen 1990). Quantities caught have

beenvery unstable.From 1987 to 1999,the highest level was 155.1 million piecesin

1991 and the lowest level 8.0 million in 1998, an almost 20-fold variation (Table 4.4).

The averageprice was similarly unstable,ranging from a high of 37.5 NT$ per eel in

1994 to a lowest averageprice of 5.2 NT$ in 1990, a more than 7-fold difference

(Table 4.4). The average prices were influenced by the quantity of capture. Since

1994, averageprices were more than 25 NT$ per eel, though, in1999, this dropped to

14.3 NT$ becauseof the high quantity of capture, and the suspensionof production by

some farms. According to Chen et al (1994), the fluctuation of Japaneseelver catches

has a positive relationship with rainfall and a negative relationship with seawater

temperature.With total dependenceon the natural sources,the supply of seedis

limited, unpredictable(Table 4.4), and is one of the bottlenecksin the developmentof

eel culture.

98
Basedon Table 4.4, the relationship betweenaverageprice and quantities of glasseel

caught can be calculatedby a simple linear model and presentedbelow.

P= 26.5 -0.13Q F=3.25

(5.87)** (-1.80)

R2 = 22.8%

Where

P= Deflated averageprice of glasseel (NT$)

Q= Quantity of glass eel caught (106)

Although this is not a significant relationship it showedthat the averageprices had a

negativecorrelation with capturequantities.The coefficient for Q showedthat if the

capturequantity increasesby 106,the price would decreaseby 0.13 NT$ per eel.

Table 4.4 The quantity and value of caught glasseel in Taiwan.

Year Quantity Value V/Q
(106) (106NT$)
1987 21.0 182.8 8.7
1988 38.2 436.8 11.4
1989 137.6 852.8 6.2
1990 24.3 125.1 5.2
1991 155.1 920.1 5.9
1992 40.1 450.8 11.2
1993 12.0 217.3 18.1
1994 30.9 1,158.6 37.5
1995 35.5 1,195.5 33.7
1996 49.2 1,660.1 33.8
1997 12.6 348.1 27.6
1998 8.0 291.6 36.4
1999 47.0 673.3 14.3
Data source:Year Book of Taiwan Fisheries Bureau.
V/Q is the ratios of values to quantities and implied the averagenominal prices.

99
Basedon the culture areaand capacity in Taiwan, the demandof glasseel is around

250 million (- 50 t), while the size of the catch is only 50 million (- 10 t) (Tzeng,

1986). Tzeng (1986) noted that it was effectively impossible to increase local catches

of glass eel significantly, as exploitation in coastal Taiwan was 45-75% of the natural

population. The shortage must be made up by import from Korea, or from Mainland

China through Hong Kong. With China's recent development of a domestic eel

culture industry and thus control of its glasseel exportation, the shortageof glasseels

hasbecomemore serious.Before the introduction of superintensive culture, farmers

tried to use Europeaneel (Anguilla anguilla) to replaceJapaneseeel but suffered

numerousfailures. Comparedwith Japanese

eel, the Europeaneel was rejected by

farmers, consideredto suffer from a range of deficiencies, particularly that they:

" cannot stand high temperature,

0 are easily infected by parasites,

" develop large differencesin size betweenfast and slow growers,

9 have weak feeding behaviour and slow growth, and

9 require sedimenton the bottom of their ponds to be cleanedthoroughly.

Most farmers and researchersbelieve that in general(Usui 1991), the growth of eels is

faster in females than in malesbut they believe that higher densities support the

developmentof males as an adaptiveresponseto less than ideal living condition.

However, Holmgren and Mosegaard(1996) suggestthat males display, on average,a

higher weight increasethan females.More researchis neededto explore density and

growth interrelationship.

100
At market size, the Japaneseeel has a longer and slimmer body shape and is easy to

tell from the Europeaneel. However, they are difficult to distinguish morphologically

at glass eel stage.Usually, Japanese

glasseel are - 5.8 cm in length and - 5000 to

6000 individuals to a kilogram; while Europeanglasseel are - 7.8 cm and are -2500

to 3500 to a kilogram. To distinguish the two species,eel farmers can use a1 mg 1-1

solution of Giodrin (a pesticide), in which the Europeaneel die in 20 minutes whereas

the Japaneseeel survive (Chen 1990). However, because of the shortage and supply

fluctuation of Japaneseeel seed,Europeaneel is now the major speciesfor

industrialized superintensive eel culture.

4.2. Traditional eel culture

4.2.1. Introduction

In Taiwan, eel culture can be separatedinto 2 stages.The first stageis to rear glasseel

to fingerling (5-10g), while the secondis to rear the fingerling to market size (150-

200g) (Fig. 4.2). This two-stage approach,and the subsequentexport and marketing

will be discussedin the following paragraphs.

101
 Capture of natural glass Importation of glass eel

Rearing glass eel to fingerling

Rearing fingerling to market size

Export of fresh eel I Processingof eel

Exportation

Fig.4.2 The processof eel culture in Taiwan.

4.2.2. Facilities

Traditionally, there are two kinds of eel ponds in Taiwan i. e. "hard" and "soft" ponds.

Traditional hard ponds are normally rectangularin shapefrom 20 m2to 3,000 m2in

surfacearea,with vertical concreteor brick walls surrounding a clay bottom covered

with coarsesand.Usually, soft ponds are converted from milkfish and tilapia ponds,

in which fish farmers changeinlet and outlet systemsto develop eel ponds. Both the

walls and bottoms of soft ponds are clay and the arearangesfrom 0.5 to 0.8 ha. Eel

ponds normally also have a wooden feeding platform of up to 10m2,which provides a

shadedareafor eel to congregate.At the center of the feeding platform and under its

cover, there are one or more feeding cages.These are normally made of plastic-coated

wire mesh, with a mesh size, which allow eels to come into the cage. Paddle wheel

aeratorsare usually typically

necessary, one per 0.1ha of water surface (= 7.46 KW

installed per ha). The paddle wheels are installed along and parallel to the side of the

to
pond create a circular flow in the system.

102
4.2.3 Manipulation of glasseel

During the upstreamprocessfrom seawaterto fresh water, glass eels acclimatize to

the fresh-waterenvironment. If the glasseels were caught in an areawith high

salinity, this acclimatization must be done carefully and gradually, as dramatic

changesin temperatureand salinity should be avoided. Osmotic imbalance, shown by

the body of glass eels becoming opaque, can be adjusted by using a water bath in the

salt water with 0.5-0.7% salinity (Yu, 1994).

Before stocking the glasseels, ponds are usually sterilized; after clearing out the silt

on the bottom of the ponds by flushing and suction pumping, the ponds are filled with

water to a depth of 20 to 30cm and around 25 kg of powderedbleach is scatteredper

100 m2 of pond. The ponds are then stirred by a paddle wheel or pump, and drained

after 2-3 days exposure.The pH value is then adjustedwithin a range of 6-9 by

liming, the amount dependingon alkalinity. One week before stocking with glasseels,

the ponds start to be filled with water.

Glasseels acquire a dark pigmentation after severaldays' rearing, and reduceslightly

in size. Those that are freshly collected from the wild only take live food and must be

trained to adaptto artificial feed in a designatedfeeding area.Traditionally, tubifex

worms are first used as feed, then after lg weight, increasing proportions of

formulated feed are mixed in, with full weaning on to formulated feed when elvers

reach 2g. The tubifex worms are bought from tubifex farmers or collected from the

wild and must be stockedin clean water to dischargeany contaminating materials

from their outer surfacesor at stomachsbefore being used. If there are not enough

103
tubifex worms, minced oyster and boned fish are used as replacement.At the

beginning of feeding, feed is scatteredall over rearing pond, and over a few days this

is gradually concentratedtoward the feeding platform. Initially, feeding must be

conductedseveraltimes a day in the daylight hours and evening, after which evening

feeds are gradually eliminated. The weaning programme starts by placing the feeding

basket on the bottom and gradually raising it to the surface. This encourages the eel to

feed on the surfaceor out of the water, to reduce submergenceof the feed under water

and, thus, minimize loss of feed in suspension or through dissolution. Daily rations of

tubifex can be up to 30% of the body weight, divided into three feeds,in a quantity

that can be consumedin one hour. After feeding, the basket is lifted out of water.

Leftover feed in the pond should be avoided, as it can seriously affect water quality.

The stocking density of glasseels is 0.2-0.3 kg/m2 and thinning is required

periodically. The eels reach a weight of 2g in two months and 5-10g in another two

months. At 5-10g they are referred to as stocking size fingerlings.

4.2.4 Managementof on-growing eels

The initial stocking density for on-growing eels is 0.6-1.0 kg m"2.With a water

exchangerate of 20% daily, and after two or three thinnings, faster growing

fingerlings can reach the minimum market size of 150-200g(5-6 kg-1)in six months.

The slow growers need 18 months to reach the market size. With continuous 20%

water exchangedaily, the carrying capacity can be 3 kg m2 at harvest and at 40%

daily; this can increaseto 9 kg m 2. Usually, the aeratorsare turned on twice

exchange

every day, once in the afternoon to drive oxygen from the super-saturated surface

layers to the lower depthsand again from the evening to the next morning, to increase

atmosphericoxygen transfer into the pond water.

104
The survival rate from glasseels to market-sizeis usually about 60-70%, requires

100,000glasseels per hectarefor a yield of 10 t ha 1,or 10 glass eels per 1kg of

market size eels. If 5-10g fingerlings are used,survival to market size is 70-85% and

about 70,000 fingerlings are neededfor one hectare.

4.2.5 Feeding on-growing eels

Currently, formulated feed is the dominant diet, and trash fish are only added in small

quantities as a nutrient supplement. Formulated eel feeds are in powder or floating

pellet form. For the former, feeds are usually mixed with fish paste, a small amount of

fish oil and an equal amount of water, to form a dough-like consistency before being

put into feeding baskets. Feeding baskets are normally made of plastic-coated wire

Eels
screen. can go through the mesh of feeding basket or climb on to the basketto

get the feed. If the eels in the pond are not uniform in size, severalfeeding basketsof

different size are necessaryto ensurethat small eels can be adequatelyfed. One hour

after feeding, the uneatenpasteremaining in the basketsis lifted out of the pond, thus

avoiding fouling the pond. By contrast,pelletised feed requires little effort in

preparation,doesnot need feeding basketsand can prevent feeds from losing too

much nutrient due to suspensionand dissolution. However, it is difficult to add

medicine or other additive nutrients into the feed, though some farmers will dissolve

additives into fish oil or water and soak the pelletised feed in it before feeding.

4.2.6 Characteristicsof traditional eel farmers

In 1998,63 traditional farmers who also managedtheir farms were surveyedby

questionnaire. On average, producers were 45.59 years old. Regarding education

attainment,there were two main groups, those who had completed elementaryschool
105
(34.92%) only, and those who had completed senior high school (38.10%). Average

yearsof schooling were 9.1 years.Most had 1-20 yearsof experiencein eel farming,

averaging 17.1 years.More than half (57.1%) statedthat eel farming was not their

only sourceof family income. The averagehousehold size was 7.0 people; more than

60% of respondentshad a householdsize larger than 5 people (Table 4.5).

Table 4.5 Socioeconomic characteristics of eel farmers.

Socioeconomiccharacteristics Number Percentage
Age (years)
31-40 14 22.22%
41-50 32 50.79%
51-60 15 23.81%
61 and above 2 3.17%
Average 45.59 years

Education attainment
None 5 7.94%
Elementary 22 34.92%
Junior high school 6 9.52%
Senior high school 24 38.10%
College 6 9.52%
Average yearsof schooling 9.05 years

Experiencein eel culture

1-10 years 20 31.75%
11-20 years 28 44.44%
21-30 years 10 15.87%
31-40 years 3 4.76%
41 years and above 2 3.17%
Average yearsof experience 17.14 years

Sourceof family income

Eel production only 27 42.86%
Eel production and other source 36 57.14%

Household size
1-5 25 39.68%
6-10 28 44.44%
11 and above 10 15.87%
Average household size 7.02 peoples

106
The education attainment is correlated to the age, the younger groups having higher

educationattainment.The averageeducationattainmentsin different age categories

were 10.0,9.5 and 5.4 years in the categoriesof 30-<40,40-<50 and 50+ yearsold,

respectively.The averageyears of experiencein eel culture of different age groups

were 9.1,15.2 and 27.1 yearsin the categoriesof 30-<40,40-<50 and 50+ yearsold,

respectively.The averagehouseholdsizes were similar in different age groups,being

7.8,6.8 and 6.8 in the categories of 30-<40,40-<50 and 50+ years old, respectively.

The percentageof each group with outside income was correlated to the age.The

older group had a higher percentagewith outside income, accounting for 42.9%,

53.1% and 76.5% in the categoriesof 30-<40,40-<50 and 50+ yearsold, respectively.

The averageyield levels in different age groups were 9,832,15,063 and 13,035kg/ha

in the categoriesof 30-<40,40-<50 and 50+ yearsold, respectively.This implies that

older groups had more experiencein culture and higher yield level. However, when at

agesof greaterthan 50, more farmers engagedin other business(higher outside

income) and the yield level reduced.

Table 4.6 Averagesof education attainment,experience,householdsize, percentages

of farmers with outside income and yield levels in different age categories.
Age category 30-<40 40-<50 50->50
Education attainment 10.0 years 9.5 years 5.4 years
Experience 9.1 years 15.2 years 27.1 years
Household size 7.8 people 6.8 people 6.8 people
Outside income 42.9% 53.1% 76.5%
Yield level 9,832 kg/ha 15,063kg/ha 13,035kg/ha

4.3 Super intensive eel culture

Becauseof the shortageof Japaneseglasseel, fish farmers and the FisheriesResearch

Institute tried to import Europeanglasseel as a substitute.At the sametime, super

intensive fish culture systemswere imported. In these,a central control systemis

107
installed, which monitors temperature,pH value and dissolved oxygen, and controls

the automaticfeeder,emergencyoxygen-supply system,rotating net brushesand

pumps.

4.3.1 Culture tanks

Theseare madeof fibreglass with water inlets submergedat the bottom, with water

led in tangentially to create a rotating current. In this way, faeces and other particles

are forced to concentrate in the centre and on the bottom of the tanks by gravity. A

pipe from the centre of the tank baseis connectedto a bowl. The bowl is connectedto

and surroundsthe outlet. Through the pipe and the bowl, faecesand feed waste are

removed from the tank to the outlet pipe. The outlets are fitted with wire netting to

prevent the escapeof eels and are continuously cleanedby a brush rotating around the

net to keep the net from blocking (Fig. 4.3).

Usually, only pelletised feeds are used in superintensive systemsbecausethe

powdered form of formulated eel feeds easily fouls the water quality. The feed is

loaded in the feeding silo and leavesthe feed through a gap onto a plate. There, a

rotating scraperscrapesthe feed on the plate into the tank (Fig. 4.3). The size of the

gap can be controlled and therefore also, the feeding amount in certain time.

108
 Feeding Stainless wire, net
Scraping silo
blade
Brush

Feeding
plate
Outlet pipe

pipe

Outlet (only for grading) tank

Fig. 4.3 Culturing tank for superintensive eel culture.

4.3.2 The processof water flow

The overflow water is led into a drum filter, which retains particles on the filter

The
screen. retained particles would eventually block the flow of water and so, after

the water level reachesa certain point, a high-pressurewater jet is turned on

automatically to flush out the particles from the net. In the drum filter, part of the

faecesand uneatenfeed are eluded, leaving the partially cleanedwater to flow

through a biological filter tank packedwith a biofilter medium with specific surface

of 150m2ni-3,on which nitrifying bacteria are encouragedto grow, converting

ammoniato nitrite and nitrate. After flowing through the biological filter, an UV

systemis used to control bacteria levels, though typically penetration of UV light is

only 0.7-1.0cm. The useful life of an UV lamp is about 8000 hours. Finally, the water

is oxygenatedby using an oxygenation cone. At 25 °C, the dissolved oxygen can

109
reach 20 mg 1'' dropping to 16 mg 1-1at 30 °C. The system flow chart is as shown in

Fig.4.4.

Drum filter

Central control
FRP fish-rearing system Biological filter
tank

Oxygen cone UV light

Direction of water flow

Controlled by central control system """"'-""10.
Fig. 4.4 Flow chart of superintensive eel culture system.

4.3.3 Managementof water quality

In traditional pond culture, environmental conditions are maintained by balancing the

inputs of feed with the assimilative capacity of the ponds. A key to successful

recirculating production is the use of cost effective water treatment systems.Water

quality maintenancefor a recirculating system includes temperatureand pH value

control, reduced suspendedand dissolved waste fraction, oxidized ammonia and

nitrite-nitrogen, oxygenatedwater and effective sterilization etc. The required

of
management water quality control is summarizedin Fig. 4.5.

110
 Suspendedand Biological
dissolvedwaste Filtration
removal (Nitrification)
Fish culture tanks
Aeration and Disinfection
oxygenation (UV light)

pH value adjustment
(Liming)

Fig. 4.5 The required managementof water quality control.

Temperature

Temperatureis critical in growth and survival, with optimum temperaturerangesfor

growth, FCR and disease In

resistance. Taiwan, few farmers need to resort to

temperatureadjustment,but if necessary,this can be done by using submersible

heatersor coolers.

Dissolved oxygen (DO)

Oxygen supply is necessaryfor the eels in the tanks and nitrifying bacteria in the

biological filter systems.For maintaining optimum growth, D0 in rearing tanks must

not be lower than 5 mg 1"1,

addedto which, nitrifying bacterianeed at least 2 mg 1"1

DO to permit adequateconversion of NH3 to NO2 and NO3. To maintain adequateDO

level, oxygen must be fed. In recirculating systemsa better location to aeratewater is

in the recycled flow-stream just prior to re-entry into the tanks (Losordo et al., 2001).

The water is usually over saturatedin oxygen using a specifically designedoxygen

cone. This is better led into the bottom of the tanks where it can quickly mix with the

main tank water. If the mixing is too fierce, the over saturatedoxygen can easily be

lost to the air, while if mixing is inadequate,patchesof highly over saturatedwater

111
may causegas bubble disease,
which may hinder the normal physiological functions

and causedeath. Glass eels and fingerling are more sensitiveto this phenomenon.In

this system,eachfish tank is equippedwith an emergencyoxygen supply system

starting, when the central control systemdetectsthe DO as too low.

pH value

pH is a measureof hydrogen ion concentrationin water, indicating the degreeto

which water is acidic or basic. The pH of water affects the stateof many water quality

parametersand the rates of many biological and chemical processes.The acceptable

range for pH is usually from 6 to 9.5, though if the pH value changesby 2 units or

more very quickly, it may be harmful, especially, to small fish. The optimum pH

value for nitrifying bacteriato mineralize the waste in the biological filter systemis

about 7 to 8 (Gousset, 1990). However, pH reduces in recirculating systems due to

acids producedby nitrifiers in oxidizing NH3 to NO3 and due to accumulatedcarbon

dioxide releasedby the fish. For maintaining the appropriatepH value, lime water is

usually added.This is not addedinto the tank directly as it may causespatchy areasof

high pH and damagethe eels. The pH value in the systemmust be monitored and

adjusteddaily.

Dissolved nitrogenous waste

Ammonia is the major nitrogenouswaste in the eel culture systemand is convertedto

N02 and NO3. It exists in two forms in water, un-ionized ammonia (NH3) and ionized

ammonia (NH4'). At a pH of 7.0, most of the total ammonia nitrogen (TAN) is in

NHI form, while at a pH of 8.0, the majority is in NH3 form (Losordo et al., 2001).

NH3 and NO2 are toxic to eels and elimination is very important. If the concentration

112
of NH3 is higher than 0.035 mg 1"1,
gills will be damagedand growth retarded.

However, aquatic speciescan tolerate extremely high levels (>200 mg 1'1)of NO3

(Losordo et al., 2001). NO2 is a product of oxidation of NH3 while NO3 that of

oxidation of NO2. Nitrifying bacteria (eg Nitrosomonas spp) utilize NH3 as an energy

source and produce NO2, while Nitrobacter spp utilize NO2 as an energy source and

produce NO3. Both types of bacteria are present, and as culture develops, large

amountsof bacteria adhereto the biological filters and must be cleanedto maintain

their efficiency, getting rid of aging bacteria and supplying spacefor new bacteria.

However, the filters must not be too clean, leaving no bacteria on them. When there

are problems in the biological filters, feeding must be stoppedor reduced.

Suspendedsolids

Most of this arisesfrom faecesand uneatenfeed. Pelleted feeds usedin intensive eel

culture consist of protein, fat,

carbohydrates, minerals etc. The portion not eatenor

not assimilatedby the fish becomehighly organic wastes.When those wastesare

broken down by bacteria within the system,they will consumedissolved oxygen and

generateammonia.Their rapid removal is very important in the superintensive

system,as they will use oxygen and produce ammonia and other toxic gases.

Suspendedsolids can be removed by drum filters, though smaller particles are still

in
suspended the system.However, although they add to oxygen and ammonia loads,

they can also produce a substrateon which nitrifiers can attach, and so a certain

amount may be beneficial for the system.However, if levels are excessive,water

exchangemust be increasedto prevent deterioration of the water quality.

113
4.3.4 Husbandry

Feeds for super intensive system are pelletized, applied at - 3-4% per day for

fingerlings and 1-2% for on growing. When near harvesting, the total feeding amount

will reach its maximum, at which time, feeding continuously is better than 2 or 3

times daily. As the respiration rate increases significantly during feeding, feeding

small amounts regularly can prevent the DO from dropping abruptly below 5 mg 1"1.

Separatingthe fast growing eels from the slow growers is very important since it can

increasefeeding efficiency. In the processof grading, eels are drawn to the upper

layer above the rearing tanks by vacuum suction, go through the automatic grading

machine with a table of rotating bars and are separatedinto 3 sizes in plastic bags or

baskets.After that, the gradedeels are put in the tanks with pipes connectedto the

different rearing tanks for different sizes.The eels go down to the rearing tanks by

gravity. Usually, the eels are gradedevery 45-60 days.The processesof grading are

shown on Fig. 4.6.

114
 Grading
Pump machine Plasticbagswith
Tank different sizesof eel

Plastic bags with

different sizesof eel

Tank

Pipe

UU4
Fish Tank

Fig. 4.6 The procedure of grading in super intensive eel culture system.

4.4 Financial analysis of eel culture

4.4.1 Introduction

The eel farming industry has had experiencewith rising production costs (such as

increasingcost of glasseel) in Taiwan, and therefore, superintensive cultivation is

applied, using Europeaneel to reducethe cost of eel seed.Here, the feasibility of

traditional and superintensive eel culture are investigated and compared,using

financial surveysbasedon questionnaires.The surveyedareasfor traditional eel farms

were on Jang-Hwa, Yun-Lin, Cha-I, Tainan, Kaoshung and Ping-Tong Counties (Fig

2.4), where 63 farms are surveyed.The survey areasfor intensive eel culture were on

Taipei, Tao-Yen and Tainan County where only 5 farms were surveyedbecausethere

were only a few intensive farms in Taiwan.

115
Financial viability is analysed by using cost and benefit analysis, pay-back period

analysisand discountedcash flow investment appraisal.Additionally, the sensitivity

of the price of glasseel and the social costs are examined.

4.4.2 Cost analysis

Traditional eel culture

Two components,capital cost and operating cost can be established.The capital costs

comprisethe costs of a work shedand storagehouse,pond construction, preparation

and maintenanceof the ponds,repair and maintenancefacilities, power generator,

paddlewheeland pump. The operating costsconsist of the cost of eel seed,feed,

electricity, chemicals, wage, land

miscellaneous, rent, depreciation and interest. Three

important assumptionsare that:

1) investment costs are coveredby a loan at an annual interest rate of 8%;

2) the facilities and equipment are subject to straight line depreciation over the

useful lifetime;

3) the useful lifetime of buildings and pond construction is 20 years, of a

power generatoris 10 years and of a paddle wheel and pump is 5 years

respectively.

Becausespendingon operating costs is spreadthroughout the culture period the

interest is chargedon 50% of the outlay. Total expenditure on variable costs is not

paid out at the beginning and therefore, it does not incur the full interest charge for the

entire period. However, the interest is charged on 100% of the capital costs

(Christensen,1993).

116
Among the capital costsof traditional eel farm, pond construction is the highest,

followed by the work shedand storagehouse,paddlewheel,and power generator;

their ratios to total capital cost are 54.6%, 15.9%, 9.2% and 7.5% respectively. The

averagecapital cost per t of production was 108.74thousandNT$, the lowest and

highest capital costs per t being 45.91 and 195.72 thousand NT$, accounting for 42.2

and 180.0% of averagecapital cost, respectively.The rangesof capital costs were

very large, some items ranged from below 20% to more than 200% of the mean. The

items with higher cost variabilities were work shedand storagehouse,pond

construction and pump. This might be related to the sourceswhich the farmers bought

them from, their sizes and the yearsthey have beenused.

Table 4.7 The averageannual cost of eel production of per traditional eel farm.
Unit: Thousand NT$
Item Cost Useful life Percentage

Capital cost
Workshed and storagehouse 463.66 20 years 15.9%
Pond construction 1591.30 20 years 54.6%
Preparationand maintenanceof ponds 101.70 3.5%
Repair and maintenancefacilities 136.46 4.7%
Power generator 218.65 10 years 7.5%
Paddlewheel 268.13 5 years 9.2%
Pump 136.58 5 years 4.7%
Total 2916.48

Operating cost
Eel seed 6527.37 60.5%
Feed 1944.80 18.0%
Electricity 314.39 2.9%
Chemicals 128.65 1.2%
Wage 756.57 7.0%
Miscellaneous 7.79 0.07%
Land rent 270.50 2.5%
Depreciation 205.54 1.9%
Interest 639.54 5.9%
Total 10795.15 100%

117
Table 4.8 Average capital cost of traditional eel farm for producing per t of eel.
Unit: ThousandNT$
Item Cost Percentage of average
Work shedand storagehouse 17.29 (2.14-35.71) 12.4-206.5%
Pond construction 59.33 (16.67-138.89) 28.1-234.1%
Preparationand maintenanceof ponds 3.79 (0.67-6.67) 17.7-176.0%
Repair and maintenance facilities 5.09 (1.25-12.78) 24.6-251.1%
Power generator 8.15 (3.43-13.33) 42.1-163.6%
Paddlewheel 10.00 (3.60-16.67) 36.0-166.7%
Pump 5.09 (2.14-11.36) 42.0-223.2%
Total 108.74(45.91-195.72) 42.2-180.0%

The figures in the parenthesesare the range of highest and lowest cost.

The averageoperating cost per kg of eels is detailed in Table 4.9 and is about 402.6

NT$. This shows the relatively high cost of eel seedat 243.4 NT$ per kg of eel,

accounting for 60.5% of total operating costs, followed by feed, at an averagecost per

kg of 72.5 NT$, accounting for18.0% of total, and labour at 28.2 NT$ per kg,

accounting for 7.0% of total. The variation among observationsis relatively small.

The highest and the lowest operating costs for producing 1 kg of eel were 336.9 and

452.8NT$, accountingfor 83.7 and 112.5%of averageoperating cost, respectively

(Table 4.9). Comparedto capital cost, the variation of total operating cost was

smaller. Although the `miscellaneous' category had the highest variation (33.3-

400.0%) among operating costs, its contribution to total cost was small, the key

factors being seed,feed, wagesand interest. This suggeststhe importance of higher

survival rate, lower FCR, proper administration and lower capital cost in reducing

operatingcost.

118
Table 4.9 Annual average operating cost of traditional eel farm for producing Ikg of
eel. Unit: NT$
Item Cost Percentage of average
Eel seed 243.4 (178.2-340.1) 73.2-139.7%
Feed 72.5 (51.4-119.7) 70.9-165.1%
Electricity 11.7 (4.0-30.2) 34.2-258.1%
Chemicals 4.8 (1.3-8.9) 27.1-185.4%
Wage 28.2 (9.8-56.7) 34.8-201.1%
Miscellaneous 0.3 (0.1-1.2) 33.3-400.0%
Feeof renting land 10.1 (5.0-16.3) 49.5-161.4%
Depreciation 7.7 (3.9-16.9) 50.6-219.5%
Interest 23.8 (14.5-55.3) 60.9-232.4%
Total 402.6 (336.9-452.8 83.7-112.5%
The figures in the parenthesesare the rangeof highest and lowest cost.

Cost analysis of super-intensiveeel culture in Taiwan

In this system,capital costscomprise cost of buildings and the recirculating system.

Operating costs consist of the cost of eel seed,feed, electricity, oxygen, chemicals,

wage, miscellaneous,land rent, depreciation and interest. The first and second

assumptionsare as for traditional The

systems. useful lifetimes of buildings and the

recirculating system are 10 years. As with traditional eel culture, interest on operating

is
costs chargedon 50%, but interest is chargedon 100% of the capital costs.

In superintensive eel farms, most of the capital costs were in the recirculating system,

which on averageaccountedfor more than 70% of the capital costs (Table 4.10).

Average capital cost of super-intensiveeel farm for producing per t of eel was 64.7

thousandNT$. The highest and the lowest capital costs for producing per t of eel were

55.4 and 73.9 thousandNT$, accountingfor 85.6 and 114.2 % of averagecapital cost,

respectively (Table 4.11). The lowest cost producer might set up some part of

facilities by themselves,insteadof purchasingthe culture system.

119
Table 4.10 The averageannual cost of eel production for a super-intensiveeel farm.
Unit: ThousandNT$
Item Cost Useful life Percentage

Capital cost
Building 2,000 10 years 28.6%
Recirculating system 5,000 10 years 71.4%
Total 7,000

Operating cost
Eel seed 5,066.7 20.4%
Feed 9,441.33 38.0%
Electricity 2,060 8.3%
Oxygen 956.7 3.8%
Chemicals 320 1.3%
Wage 3,048.8 12.3%
Miscellaneous 1,172 4.7%
Fee of renting land 600 2.4%
Depreciation 700 2.8%
Interest 1,494.62 6.0%
Total 24,860.15 100%

Table 4.11 Average capital cost of super-intensive eel farm for producing per t of eel.
Unit: Thousand NT$
Item Cost Percentage of average
Building 18.5 (13.9-24.1) 75.1-130.3%
Recirculating system 46.2 (41.6-49.8) 90.0-107.8%
Total 64.7 (55.4-73.9) 85.6-114.2%

The averageoperating cost of eel culture for a super-intensivefarm is shown in Table

4.10. The averagecost of producing 1 kg of eels is detailed in Table 4.12 and is

shown to be about 229.5NT$. Feed is the highest cost component,the averagecost

per kg of eels being as high as 87.2NT$ and accounting for 38% of the total operating

costs (Table 4.10). Next to feed is eel seed,where the averagecost per kg of eels is

46.8NT$ accounting for2O.4% of the total. The third highest cost is labor at 28.1 NT$

kg
per of eels, accounting for 12.3%.The highest and the lowest operating costs for

120
producing 1 kg of eel were 207.6 and 256.7 NT$, accountingfor 90.5 and 111.9%of

averagecapital cost, respectively (Table 4.12). The key factors, which influenced the

production cost were eel seed,feed, electricity and wages.Similar to traditional eel

culture, higher survival rate, lower FCR and proper administration were the important

factors to reduce the operating cost.

Table 4.12 Annual averageoperating cost for a super-intensivefarm to produce 1 kg

of eels. Unit: NT$
Item Cost Percentage
Eel seed 46.8 (20.8-66.7) 44.4-142.5%
Feed 87.2 (77.1-136.8) 88.4-156.9%
Electricity 19.0 (6.3-26.7) 33.2-140.5%
Oxygen 8.8 (7.7-10.8) 87.5-122.7%
Chemicals 3.0 (1.4-5.8) 46.7-193.3%
Wage 28.1 (13.9-37.8) 49.5-134.5%
Miscellaneous 10.8 (5.4-16.6) 50.0-153.7%
Fee of renting land 5.5 (3.2-7.1) 58.2-129.1%
Depreciation 6.5 (4.9-7.2) 75.4-110.8%
Interest 13.8 (10.6-15.8) 76.8-114.5%
Total 229.5 (207.6-256.7) 90.5-111.9%
The figures in the parenthesesare the range of highest and lowest cost

4.4.3 Benefit analysis

The profit (P) is equal to the revenue(MI) minus operation cost (C), profitability can

be estimatedby the benefit-cost ratio (BCR) and the income ratio (IR)(Chen 1994).

The respectiveformulas are as follows:

BCR=P/C

IR=P/MI

Where P= Profit

C= Production cost

MI = Revenue

121
The higher are thesevalues the more financially soundis the operation. This also

indicatesthat the operation is economically soundand further developmentmay be

The
considered. averageamount of production of traditional eel culture per farm is

26.82 t, the averagecost per farm is 10.80x106NT$ with an averagerevenueper farm

of 11.12x106NT$. The averageprofit per year is 0.33x106 NT$. Therefore, the

averageBCR is 3.05% and IR is 2.96% (Table 4.13). The averagecost per super-

intensive farm is 24.86x106 NT$ and the average revenue per farm is 25.64x106 NT$.

The averageincome per year is 0.78x106NT$. Therefore, the averageBCR is 3.13%

and IR is 3.04% (Table 4.13). This showedthat on averagesuper intensive eel farm is

a little bit more financially sound than traditional eel farm. The range of profit

betweenbest and poorestfarms of traditional eel farms (-3.7- 6.6 million NT$) was

wider than that of super-intensiveculture (-0.7-2.2 million NT$).

Table 4.13 The benefit analysis of traditional eel farm and super intensive eel farm.
Unit: million NT$
Average cost Average revenue Average profit BCR* IR*
Traditional eel 10.8 11.1 0.33 3.05% 2.96%
farm (2.6-44.6) (1.7-44.8) (-3.7-6.6)
Superintensive 24.9 25.6 0.78 3.13% 3.04%
eel farm (19.3-26.7) (20.6-28.1) (-0.7-2.2)
* BCR is benefit-cost ratio and IR is income ratio.

Although super-intensiveeel farms had higher averageprofit, the distribution of

profitability shows that it is still possible for traditional eel culture to have a higher

profit than super-intensiveeel culture (Table 4.14). When attaining economiesof

scaleor having better performance,traditional eel culture can still make better profit.

122
Table 4.14 The distribution of profitability for traditional and super-intensiveeel
farms. Unit: NT$
Profit Number
Traditional eel culture Super-intensive eel culture
< -3 million 1 ----
-2 million->-3 million 1 ----
-1 million->-2 million 2 ----
0- >-1 million 12 1
0- <1 million 26 2
1 million -<2 million 8 1
2 million -<3 million 6 1
3 million -<4 million 4
----
4 million -<5 million 1 ----
>5 million 2 ----

4.4.4 Pay back period

The pay back period defines the time required to recover the initial investment out of

the expectedearningsfrom the investment,before any allowance for depreciation (

Shang 1981).The method is as follows.

T= C/E, Where

T= the pay back period (number of year)

C= initial investment

E= averageannual profit expectedfrom the investment before depreciation

The pay back period is expectedafter 5.45 years for traditional eel farming and 4.73

years for intensive eel culture.

4.4.5 Cash-flow and discountedfinancial indicators

A 5-year discountedcash flow analysis at 10% discount rate reveals that investment

in both types of eel culture remains viable (Table 4.15). The pattern of cash flow

includes capital cost, operating cost (excluding interest and depreciation) and revenue.

At 4271.1 thousandNT$, the NPV is higher in intensive eel culture than in traditional

123
eel culture (1534.9 thousand NT$). If the discount rate is increased, the value of NPV

will For
decrease. those whose capital costs are higher than averageand/or have poor

initial performance,the value of NPV might be negative.

Table 4.15 Cash-flow projection for a traditional eel farm and an intensive eel farm.
The discount rate for NPV is 10%. Unit: Thousand NT$
Year0 Year1 Year2 Year3 Year4 Year5

Traditional eel
farm
Cashflow
Capital cost 2916.5 0 0 0 0 0
Operatingcost 0 9950.1 9950.1 9950.1 9950.1 9950.1
Revenue 0 11124.5 11124.5 11124.5 11124.5 11124.5
Net cash flow -2916.5 1174.4 1174.4 1174.4 1174.4 1174.4

Discounted

cash flow
Cashoutflow 2916.5 9045.5 8223.2 7475.6 6796.0 6178.2

Revenue 0 10113.1 9193.8 8358.0 7598.2 6907.4

Net cash flow -2916.5 1067.6 970.6 882.3 802.1 729.2

NPV 1534.9

Intensive eel
farm
Cashflow
Capital cost 7000.0 0 0 0 0 0

Operatingcost 0 22665.5 22665.5 22665.5 22665.5 22665.5

Revenue 0 25638.8 25638.8 25638.8 25638.8 25638.8

Net cash flow 2973.3 2973.3 2973.3 2973.3 2973.3

-7000.0

Discounted

cashflow
Cashoutflow 7000.0 20605.0 18731.8 17029.0 15480.9 14073.5

Revenue 0 23308.0 21189.1 19262.8 17511.7 15919.7

Net cash flow -7000.0 2703.0 2457.3 2233.9 2030.8 1846.2

NPV 4271.1

124
4.4.6 Sensitivity of the price of glasseel in traditional eel culture

Overall profitability is clearly sensitiveto relatively small changesin valuesof certain

costs and selling price. Among the operating costs, the cost for fingerling is the

highest and the most sensitiveitem. Here, the relationship between the different cost

of fingerling to price is compared.In this study, the averagefarm size is 2.67 ha, the

averagedensity of fingerling is 96970/ha and the averagecost of fingerling is

6527365 NT$. With an average price per fingerling of about 25.2 NT$. The average

price per kg of eels is 414.85NT$ and the break-evenprice is 402.6NT$. Prices of 10,

15,20,25,30,35 and 40 NT$ of per fingerling was usedto comparethe sensitivity of

price (Table 4.16).

Table 4.16 Sensitivity analysisof eel fry price.

Price of per eel fry (NT$) 10 15 20 25 30 35 40
Priceof eelfry perproducing1kgeel (NT$) 100 140 190 240 290 340 390
Break-even priceperkg eels(NT$) 260 300 350 400 450 500 550
Benefit for each farm (ThousandNT$) 4266.4 2971.9 1677.4 382.9 -911.6 -2206.1 -3500.6
of fry costto operatingcost(%)
Thepercentage 37.8 47.6 54.8 60.3 64.5 68.0 70.8
The benefit of eachfarm is basedon the averageselling price, that is 414.85NT$/kg.

From Table 4.16, if the price of eel fry is higher than 30 NT$ each,the price of eels

kg
per must be higher than 450 NT$/kg, to break even. If the price of eel fry is 35 or

40 NT$, the price of eels must be as high as 500 and 550 NT$/kg (Fig 4.7). Thus

traditional eel farmers must confront the pressureof a high price for eel fry. The lack

of suitably priced glass eel seedhas made many traditional eel farmers stop

production. In some cases,some farmersjust partially stock their farms and suspend

some ponds when the price of glass eel is too high.

125
 600

500
400

300

200

x
1a)

I)) 15 20 25 3)) 35 40

Price of eel fry (NT$)

Fig 4.7 The break-even prices in different prices of eel fry.

4.4.7 Social cost

One of the particular reasons for the past prosperity of eel culture in Taiwan may be

attributed to the use of pumped ground water, legally or illegally, for which the users

had not been required to pay. Huang (1990) evaluated the cost of ground water,

revealing that this may vary from region to region. Based on his results, and assuming

water use to be the major externality (as evidenced by subsequent impacts) the

adjusted social cost of producing 1 kg eels can be estimated.

126
Table 4.17 Social prices of undergroundwater, and shadowcost and profit of
traditional eel culture and intensive eel culture.
Region Social Cost to Average profit Cost to Average
prices of produce 1kg per farm produce 1 profit per
ground eels (Traditional kg eels farm
water (Traditional eel culture) (Intensive (Intensive
(NT$/m3) eel culture) (Thousand eel culture) eel culture)
(NT$/kg)a NT$) (NT$/kg)b (Thousand
NT$)
Tun -Kan 0.018 403.1 315.1 229.5 778.8
Linpien 48.372 1611.9 -32099.4 263.4 -2893.3
Shuilichum 10.036 653.5 -6399.5 236.5 20.6
Wunfon 65.887 2049.8 -43841.8 275.6 -4214.8
Tachung 9.987 652.3 -6367.3 236.5 20.6
Lichiachun 0.238 408.6 167.6 229.7 757.28
Tonhai 4.039 503.6 -2379.9 232.3 475.5
Pathliao 0.063 404.2 285.6 229.5 778.8
Shuidiliao 0.018 403.1 315.1 229.5 778.8
Sinlong 6.678 569.6 -4149.7 234.2 269.7
a: 25m' ground water is neededfor producing 1 kg eel in traditional eel culture.
b: 0.7m3ground water is neededfor producing 1 kg eel in intensive eel culture.

The average prices in 2000 for Japanese eels and European eel were 414.85 NT$ and

236.69 respectively. From Table 4.17, it can be seen that including full costs of water

the traditional culture of Japaneseeel is only feasible in Tung-Kang, Lichiachun,

Pathliao and Shuidiliao and the highest profit is only 11.7 NT$/kg. With intensive eel

culture, only in Wufon and Linpein are the adjustedcosts higher than the average

price of Europeaneel. Here, the highest profit can reach 7.2 NT$/kg.

4.5 Post-harvestingprocessand marketing of eel

4.5.1 Introduction:

127
Although, somen(Japanesevermicelli), hiyamugi (iced noodle) kakigori (shavedice)

and watermelon are popular foods in Japanduring the summer, eel is also considered

a delicacy and good for the health during this season(The Japanese
Times, July 1,

2001), and therefore,the greatestconsumptionis in the summerperiod.

Recently, production of eel in Japanhas beenreducing from 40,098 mt in 1991to

24,904 mt in 2000, (Table 4.18). However, the consumption and imports have

increased.Total consumptionrose from 114,212mt in 1991 to 158,049mt in 2000

(average3.68 % per year) and imports increasedfrom 74,114 mt in 1991 to 131,352

mt in 2000 (average6.57 % per year).

More than 90% of eels producedin Taiwan are exported to the Japanesemarket, in

either live or in frozen roastedform. However, in recent years, China and other Asian

countries have committed an increasing effort in developing their eel industry (Liao

1996). As cheapereel from China becameavailable in Japan,the market shareof the

eel from both Japanand Taiwan startedto diminish. The market shareof eel produced

in Japanreducedfrom 35.1% in 1991 to 15.9% in 2000 and the market shareof the

eel produced in Taiwan reducedfrom 51.5% in 1991 to 18.9% in 2000 (Yu, 2001).

The rapid growth in eel culture in China in particular increasedcompetition in the

Japanesemarket, which has had important consequencesfor the domesticTaiwanese

eel industry.

128
Table 4.18 The production, consumption and imports of eel in Japan. Unit: mt
Year Production Total I mported amount
amount of eel consumption Fresh Processed Total
in Japan in Japan
1991 40,098 114,212 17,687 56,427 74,114
1992 37,397 114,752 16,745 60,616 77,361
1993 34,830 113,867 15,137 63,900 79,037
1994 30,380 111,232 15,832 65,020 80,852
1995 30,030 102,264 11,969 60,265 72,234
1996 29,517 116,796 11,442 75,837 87,279
1997 25,031 130,793 13,635 92,127 105,762
1998 22,845 122,548 13,033 86,670 99,703
1999 23,637 129,794 11,628 94,529 106,157
2000 24,907 158,049 14,355 117,187 131,352
Growth rate -37.9% 38.4% -18.8% 107.7% 77.5%
Data source: Yu, 2001.

4.5.2 Post-harvestprocessing

Two kinds of frozen roastedforms of eels are exported; without seasoning

(Shirayaki); and preparedeels (Kabayaki). Both forms are presentedwhole or sliced

skewered.The processis as shown in Fig 4.8 and is discussedbelow.

129
 Eel stocking
I
Selection

Icing

Heading
II Unheaded

Blood drained Ventral part cut open

Dorsal part cut open

II Boning

Selection

Slicing

Grading
II Whole eel skewered

Skewering
I
Roast (Kabayaki) Roast (Shirayaki)

Seasoning

Remaining viscera removed

Steaming
Remaining viscera removed

Seasoning

Roasting Roasting other side

Seasoning
Prefreezing

Packaging

Fast freezing

Frozen storage

Export

Fig. 4.8 The flow chart of the processfor the frozen roastedeel processin Taiwan.

130
Before processing, eels must be held in stocking ponds with flowing water for at least

24 hours, as far as possible to allow the feed in the eels' stomach to be excreted. In the

stocking ponds, total bacterial levels should not be over 102 /ml, and during stocking,

dead eels must be removed at least twice daily. Selection must then be implemented

before processing.This not only size-gradesthe eels,but also gets rid of eels of

unsuitable quality. After selection, the eels are cooled in a water bath with crushed

ice, at a temperatureof less than 4°C to anaesthetizethem. Icing lasts for 30 minutes,

after which they are preparedfor dissection.The first step of dissection is beheading

and the draining off of blood. They are then gutted, viscera removed, and boned. The

next step is to put them into a machine to remove the remaining blood, as the

percentageof blood content in the body will affect the taste, the speedof rotation to

remove part of the blood is very important.

Slicing and skewering: there are two kinds of roastedeels, whole skeweredand sliced

skewered.Usually, longer eels are sliced by hand, under highest standardsof hygiene,

into 4 pieces,and shorter eels into 3 pieces.During the slicing processany eels that do

not reach the required standardmust be removed. After slicing, the next step is

skewering.The pieces of eel flesh must be skeweredtightly and cannot be skewered

through the skin. The sharp ends of the skewers(bamboo sticks) usually project 5-cm

out of the flesh.

Roasting: the skin side of the eels is roastedfirst. After the skin side is roasted,the

fins of the eels are clipped and the eels are turned to the flesh side to roast again. If the

processis for Kabayaki, the fins of eels are not clipped during the first roast, but the

131
eels are steamedand the fins are clipped after steaming.Then, the eels are put into a

saucetank and roastedagain.The seasoningand roasting must be carried out twice.

The products were packedin plastic bag by vacuum and stored below -20 °C before

export.

4.5.3 Marketing channels

Eel producers are either individual eel farmers or companies, and more than 90% of

eels produced in Taiwan are exported to the Japanesemarket. Companiessell via their

own exporting agencyor to processingplants, while individual farmers will sell to a

producer associationor export agency,or to a processingplant through wholesalers.

The marketing channelis very stableand shown as Fig.4.9.

Individual farmer Eel company

Exporting agency
Wholesaler
Retailers
I, I Processingplant
Production association

Taiwanese market
<10%
Japanesemarket

> 90%
Fig. 4.9 The marketing channels of eel.

Data source(Chu 1996).

132
4.5.4 Import quantity of the Japanesemarket

In the Japanesemarket, Taiwaneseeels still have the biggest import market sharefor

fresh eels at around 60 to 80%, reaching as high as 88.68% in 1992 (Table 4.19).

However, for roasted eels, Taiwan has been replaced by China since 1994. In 1988,

Taiwan's market shareof roastedeels was 92.85% and that of China only 6.96%,

while by 1998, China's sharewas 91.27% and that of Taiwan only 5.65% (Table

4.20).

Table 4.19 Market shareof Japaneseimports of fresh eels, by countries. Unit: mt

Taiwan China Malaysia Other countries
Year Quantity Market Quantity Market Quantity Market Quantity Market
mt share (mt) share (mt) share (mt) share
1988 12617 66.8% 5569 29.5% 0 0% 709 3.8%
1989 13365 73.7% 3953 21.8% 0 0% 827 4.6%
1990 16033 79.7% 4027 20.0% 0 0% 47 0.2%
1991 13635 77.1% 4043 22.9% 0 0% 9 0.1%
1992 14850 88.7% 1871 11.2% 16 0.1% 8 0.1%
1993 11963 79.0% 2557 16.9% 553 3.7% 64 0.4%
1994 9319 58.9% 5699 36.0% 158 1.0% 655 4.1%
1995 8464 70.7% 3156 26.4% 122 1.0% 227 1.9%
1996 8067 70.5% 3096 27.1% 265 2.3% 14 0.1%
1997 9446 70.7% 3472 26.0% 268 2.0% 179 1.3%
1998 8122 62.3% 4329 33.2% 534 4.1% 49 0.4%
Data source:Customsbureau,Ministry of Finance,Japan.

133
Table 4.20 Market share of Japan imports of roasted eels, by countries. Unit: mt
Taiwan China Malaysia Other coun tries
Year Quantity Market Quantity Market Quantity Market Quantity Market
(mt) share (mt) share (mt) share (mt) share
1988 34142 92.9% 2560 7.0% 0 0% 70 0.2%
(20485) (1536) 42)
1989 35775 90.1% 3782 9.5% 0 0% 132 0.3%
(21465) (2270) (79.)
1990 41977 87.3% 6058 12.6% 0 0% 27 0.1%
(25186) (3635) (16)
1991 45215 80.1% 10453 18.5% 737 1.3% 30 0.1%
(27129) (6272) (442) (18)
1992 44116 72.8% 14141 23.3% 2324 3.8% 39 0.1%
(26470) (8485) (1394) (23)
1993 37591 58.8% 22556 35.3% 3668 5.7% 86 0.1%
(22554) (13534) (2201) (52)
1994 22150 34.1% 39374 60.6% 3406 5.2% 87 0.1%
(13290) (23625) (2045) (52)
1995 11856 19.7% 45884 76.1% 2445 4.1% 81 0.1%
(7113) (27530) (1467) (49)
1996 10750 14.2% 63008 83.1% 2053 2.7% 27 0.0 %
(6450) (37805) (1232) (16)
1997 7886 8.6% 82442 89.5% 1745 1.9 53 0.1%
4731 (49465) (1047) (32)
1998 4894 5.7% 79104 91.3% 2147 2.5% 525 0.6%
(2937) (47463) 1288 (315)
Data source:Customsbureau,Ministry of Finance,Japan.
The figures in parenthesesare quantities of processedeel, where the ratio of processed
to fresh eel is estimatedat 0.6.

4.5.5 RCA index

Balassa(1965) first developedthe concept of revealedcomparative advantage(RCA).

The relative export performancefor a particular product in a particular country can be

quantified in the form of an index, which indicates the pattern of revealed

comparative advantagein its trade (Ling et al, 1996,Traesupapet al, 1999). An RCA

value for the eel sector can be obtained by dividing a country's sharein the exports of

the type of eel product by its sharein the combined exports of eels of the exporting

countries.The RCA index is expressedas follows.

134
 RCA=(Xij/Xbj)/(Xa/Xb)

Or

=(Xij/Xa)/(XbjIXb)

where:

X= export quantity of eel product

i= eel exporting country I

j= eel product j

a= all eel products of country I

b= all eel products of all exporting counties

If the RCA index of a country in a given export eel product is higher than 1, it means

that the country's sharein this product has the comparative advantage,and if is less

than 1, it reveals a comparativedisadvantage.A value of 1 signifies comparative

neutrality.

From 1988to 1991, China had the comparative advantagein fresh eels, while since

1992,this shifted to Taiwan (Table 4.21), and to Malaysia in 1997 and 1998.

Regardingroastedeels,Taiwan lost the comparative advantageto China from 1992

but Malaysia also had the comparative advantagesince 1991 (Table 4.22). From the

RCA index, Taiwan's comparative advantagecan be shown to have changedfrom

processedto fresh eel, while that of China's shifted in the opposite direction. Unstable

supply of eel might be the reasonthat someprocessingfactories stoppedtheir

businessand causedthe comparative advantagechangedfrom processedto fresh eel.

135
Table 4.21 RCA indices of fresh eel imports in the Japanesemarket from 1989 to
1998.
Year Taiwan China Malaysia Other countries
1988 0.794972 2.01828 ---- 2.680691
1989 0.866877 1.628893 ----- 2.749452
1990 0.937031 1.353703 ---- 2.165942
1991 0.970956 1.168893 0 0.950281
1992 1.163551 0.539859 0.031592 0.783996
1993 1.260511 0.531706 0.683945 2.230713
1994 1.512447 0.645744 0.225578 4.507391
1995 2.513836 0.388408 0.287654 4.443594
1996 3.270161 0.357261 0.872345 2.624151
1997 4.301904 0.318955 1.051292 6.079245
1998 4.773435 0.396883 1.523247 0.653498
(Developed from Table 4.19).

Table 4.22 RCA indices of processedeel imports in the Japanesemarket from 1989to
1998.
Year Taiwan China Malaysia Other countries
1988 1.105352 0.476768 ---- 0.136396
1989 1.06086 0.712489 ---- 0.200205
1990 1.026343 0.85203 ---- 0.512234
1991 1.009103 0.947066 4.19067 1.015582
1992 0.954822 1.127107 4.588525 1.059668
1993 0.938289 1.110932 4.537491 0.708461
1994 0.875238 1.086249 4.88181 0.146077
1995 0.699349 1.121464 5.747539 0.316095
1996 0.657475 1.096977 6.755383 0.754946
1997 0.521007 1.098796 6.842132 0.263175
1998 0.432564 1.090695 6.126725 1.052106
(Developed from Table 4.20).

4.5.6. Seasonalvariation

Detailed monthly records of exports are provided in Annex C1, from which indices of

seasonalvariation can be calculated.The major seasonfor fresh eel exports from

Taiwan to the Japanesemarket is concentratedin June, July and August, the index

being over 100, the highest being in August, when the index is 305.4. For fresh eel

136
imported from China, there are 3 months, June,July and Decemberwhen the seasonal

index is over 100. Unlike Taiwan where the market is concentratedin the summer

period, China has 2 different peaksfor exporting fresh eels to the Japanese
market

(Annex C2). Comparedwith that of Taiwan, the Chineseeel supply is also more

evenly distributed.

In the caseof roastedeels imported from Taiwan to Japan,there are 5 months, March,

April, May, June and July, where the seasonalindex is over 100 (Annex C3), with a

more even distribution for roastedeel than fresh eel. For roastedeels imported from

China, there are 4 months-April, May, June and July where the seasonalindex is over

100 (Annex C4).

The seasonalindexes show, in the summertime (especially July), the demandfor

fresh eel in Japanesemarket attaineda peak. However, the demandfor roastedeel was

more evenly distributed in the whole year. When comparing Taiwan and China, the

demandfor fresh eel from Taiwan is higher in the summer while that from China is

more evenly distributed.

4.5.7 Analysis of export market

A simple econometricmodel of the export market was developedto outline the

factors associatedwith the performanceof Taiwaneseprocessedeels in Japan.This

linear model was basedon time seriesannual undeflated data covering the period of

1988to 1998, and estimatedwith an ordinary least squareprocedure.Data sourcesare

from the CustomsBureau, Ministry of Finance,Japan.The empirical equation is

presentedin the equation below:

137
PIP = 419.5- 0.000014DI + 46.04EC-0.003229QP1T+ 0.8116PPC
(0.46) (-0.01) (0.52) (-0.67) (5.73)**

R2 = 90.0% F= 13.46**

Where

PIP = Undeflated import price of Taiwaneseprocessedeel (W/Kg)

DI = Disposable income of Japanesehousehold ()

EC = Currency exchange rate of NT$ to

QPIT = Quantity of processed eel imported from Taiwan

PPC = Undeflated import price of Chineseprocessedeel (WlKg)

The empirical equation estimatedhas an R2 value of 90.0%, indicating that this

specification explains 90% of the variation in the import price of Taiwaneseprocessed

eels in Japan.Only the import price of Chineseprocessedeels has significant effect

on import price of Taiwaneseprocessedeels in Japan.The coefficient for PPC

showedthat if the price of Chineseprocessedeels increasesby lyen per kg, the price

of Taiwanese processed eels would increase 0.81yen per kg in the Japanese market.

However, the effect of Chineseprocessedeel prices in driving that of Taiwanese

processedeel needsto be judged very carefully, as other factors might also be

involved.

An econometricmodel for fresh eel was also developedfor the Japanesemarket. The

empirical equation is presentedbelow:

PIP = 1801.6-0.001383DI - 130.85EC- 0.0012QFIT + 0.6585PFC

(3.37)* (-1.67) (-1.89) (-0.05) (2.81)*

R2 = 96.9% F= 46.87**

138
 Where

PIF = Undeflatedimport price of Taiwanesefresh eel (v/Kg)

DI = Disposableincome of Japanesehousehold(Y)

EC = Currency exchangerate of NT$ to Y

QFIT = Quantity of fresh eel imported from Taiwan

PFC = Undeflatedimport price of Chinesefresh eel (v/Kg)

The empirical equation has an R2 value of 96.9%, indicating this model explains

96.9% of the variation in the import price of Taiwanesefresh eels for Japan.The

import price of Chinesefresh eels has significant effect on import price of Taiwanese

fresh eels in Japan.The coefficient for PFC showedthat if the price of Chinesefresh

eel increaseslyen per kg, the price of Taiwanesefresh eels would increase0.66yen

per kilogram in Japanese

market. However, the impacts of Chinesefresh eel prices on

Taiwanesefresh eel prices needto be very carefully considered,as other factors might

also be involved.

4.5.8 Market survey

Although most eel products are exported to Japanesemarket, the developmentof local

marketsis also consideredto be important. Therefore, an understandingof the

requirements of domestic consumers is essential. In 1999, in 3 areas, 132 consumers

were surveyed.These were in Taipei (52 consumers),Taichung (45 consumers)and

Tainan (35 consumers), representing the northern, central and southern part of Taiwan

respectively.

139
As shown in Table 4.23, most respondents (52%) preferred frozen roasted eel, though

a significant number preferred fresh eel (36%). Traditionally, the Taiwanese had

bought fresh eel and stewedthem with Chineseherbs.However, more and more

people now preferredconveniencefoods, and frozen roastedeel are becoming more

popular.

Table 4.23 Preference for product forms.

Products Numbers Percentage Chi-squarevalue P value

Fresh fish 48 36.36% 0.120579 0.003652**
Frozen roasted 69 52.27% 4.729986
Others 15 11.36% 6.374158
Total 132 99.99% 11.22472

The majority of respondents(66%) preferred the size below 300g (Table 4.24), which

is similar to demandin Japanesemarkets,for which the size is usually between200g

to 500g. However, the smaller size (200g group) can fetch higher prices.

Table 4.24 Preferred size of eel for domestic consumers.

Sizes Numbers Percentage Chi-square value P value
200 33 26.83% 0.5215 0.629202
300g 48 39.02% 0.39714
600g 42 34.15% 0.007966
Total 123 100% 0.926606

Although most respondents(66%) statedthat the consumption of eel was not

correlatedwith seasons(Table 4.25), 25% preferred buying eel in winter season.

Traditionally, Taiwaneseconsider stewedeel with Chineseherbs in the winter to be

wholesome,which is very different from the Japanese,who prefer to eat eel during

the hot summer.

140
Table 4.25 Seasonalpreferencefor purchasing.
Seasons Numbers Percentage Chi-square value P value
Spring 3 2,27% 6.913636364 1.584*10 **
Summer 3 2.27% 6.913636364
Autumn 6 4.55% 5.254545455
winter 33 25% 0.55
Uncertain 87 65.91% 46.36818182
Total 132 100% 66

Most people either stew eel with Chineseherbs or eat them roasted(Table 4.26). To

cook eel by roasting is perhapsinfluenced by Japanese

preferences,and some

companiesroasting eel before export also now sell them in domestic market. Such

factors may changetraditional cooking habits and consumerpreferences.

Table 4.26 Preferred ways of cooking.

Cooking ways Numbers Percentage Chi-squarevalue P value

Stewedwith Chineseherbs 60 44.44% 1.6667 0.03813*
Roasted 54 40% 0.6
Others 21 15.56% 4.2667
Total 135 100% 6.53333

More than 50% of respondentsconsideredthat they would buy more eel if the price

were lower (Table4.27). However, as the domestic market has been limited becauseof

high prices, most eel were exported to the Japanesemarket, for which prices might be

higher. Another 25% of respondentsconsideredthat they would buy more eel if the

quality were improved. Therefore, to improve the quality and service, and createnew

products might be helpful to develop the potential of domestic market.

Table 4.27 Situations in which consumerswould buy more eel.

Situations Numbers Percentage Chi-square value P value
Price is cheaper 69 52.27% 4.729986367 0.028233*
Quality is better 33 25% 0.918125426
Others 30 22.73% 1.486632584
Total 132 100% 7.134744376

141
A great number of respondentsnoted that their frequency of buying eel was uncertain

(93%) (Table 4.28). Eel consumptionis not popular in Taiwan and most consumers

did not buy them regularly.

Table 4.28 The frequencies of buying eel.

Frequencies Numbers Percentage Chi-squarevalue P value

Every week 0 0% 8.8 6.6822*10" **
Every two weeks 0 0% 8.8
Every three weeks 3 2.27% 6.913636364
Every one month 6 4.55% 5.254545455
Uncertain 133 93.18% 117.8227273
Total 142 100% 147.5909091

Most respondentsalso noted that the quantity purchasedwas uncertain (Table 4.29).

As in the precedingquestion, it seemedthat they did not buy eel regularly. Note that

as the unit of weight scaleis 600 g in Taiwan, units of 600 g and 1200 g were chosen

in this survey.

Table 4.29 Purchasingquantities of eel at each purchase.

Amount Numbers Percentage Chi-squarevalue P value
About 600 g 39 29.55% 0.187967258 0.006279**
About 1200 21 15.91% 4.002428377
Uncertain 72 54.55% 5.95058663
Total 132 100.01% 10.14098226

When askedwhy they do not buy eel, expensewas the biggest reason(51%), after

which were their boniness(32%) and difficulty of cooking (23%) (Table 4.30). As

more than 70% of respondentsconsideredthe eel to be expensive(Table 4.31), its

high price was likely to be the biggest problem restricting market development.

However, there were still 6.8% of respondentsdo not buy eel becauseof other reasons

(Table 4.30). To improve the quality might be their requirement.

142
Table 4.30 The reasons that consumers do not buy eel.

Reasons Numbers Percentage Chi-squarevalue P value

Difficult to cook 30 22.73% 0.090909 0.018566**
Bony 42 31.82% 0.818182
Too expensive 51 38.64% 3.272727
Others 9 6.82% 5.818182
Total 132 1100.01% 10

Table 4.31 Opinions of respondents concerning the price of eel.

Opinions Numbers Percentage Chi-squarevalue P value

Very expensive 9 6.82% 3.822727273 8.27367*10" **
Expensive 84 63.64% 41.89090909
Acceptable 36 27.27% 1.163636364
Cheap 3 2.27% 6.913636364
Very cheap 0 0% 8.8
Total 132 100% 62.59090909

Regardingthe quality of eel, most respondentshave a positive opinion. More than

65% of respondentsconsideredthe quality to be good/excellent.However, almost

only 32% consideredthe quality to be acceptableand 2% consideredit to be below

acceptability (Table 4.32). It seemedthat quality still has some scopeto be improved.

Table 4.32 Opinions of respondentsconcerning the quality of eel.

Opinions Numbers Percentage Chi-square value P value
Excellent 15 11.36% 1.640909091 1.77627*10' **
Good 72 54.55% 26.25454545
Acceptable 42 31.82% 3.072727273
Bad 3 2.27% 6.913636364
Very Bad 0 0% 00
8.8
Total 132 100% 46.68181818

When combining quality and price, most respondentsconsideredeel to be acceptable

(61.4%). Less than 30% of respondentsconsideredeel are good/excellent (Table

4.33). It suggestedeel might be an averagerather than a good level of value for

money.

143
Table 4.33 Evaluation of respondentsconcerningthe price and quality of eel.
Opinions Numbers Percentage Chi-square value P value
Excellent 9 6.82% 3.822727273 1.15787*10" **
Good 30 22.73% 0.163636364
Acceptable 81 61.36% 37.64090909
Bad 9 6.82% 3.822727273
Very Bad 3 2.27% 6.913636364
Total 132 100% 52.36363636

The market survey suggestedthat eel is a luxury product in Taiwan, which people did

not buy regularly. The biggest problem is the price is too high, and this might be the

reasonwhy most of the eel products were exported to Japanesemarket. However,

there was still a certain level of respondents(25%) who consideredthat they would

buy more eel if the quality was improved. To improve the quality to satisfy consumers

requirement might be helpful to develop domestic markets.

4.6 Discussion

Although the history of eel culture in Taiwan is less than 50 years old, the technology

of eel culture is quite mature. However, problems still exist, including overuseof

fresh water resource,lack of eel seedand market competition. To overcome some of

these,Taiwan tried to apply new technology from other countries and usedEuropean

eel to replaceJapaneseeel.

Traditional eel culture in ponds requires large quantities of water, and in many areas

of Taiwan is not feasible becauseof limited water supplies or an absenceof land for

pond construction. Recirculating systemsmay offer an alternative, through water

treatmentand reuse,requiring only a fraction of the water required by ponds to

produce similar yields. These systemsusually also use tanks for production,

144
substantiallyreducing land requirements(Losordo et al., 2001). The level of water

renewal in a recirculating aquaculturesystemdependsfirst, on its efficiency in

removing toxic outputs from fish metabolism and secondly, on the amount of water

that is lost when removing the accumulated waste product (Chaves et al., 1999). On

average, the production of 1 kg eels requires 20 to 30 t of water in traditional eel

culture, but only 0.8 tin intensive recirculated culture. Intensive culture is thus also

less likely to create adverseenvironmental impact.

Superintensive recirculating systemswere imported during the 1990s,but are still not

widely developedin Taiwan. Up to 2000, only 12 farms used this system(Chen

personalcommunication, 2001). Most eel farmers still do not want to invest in

intensive culture, and they consider the density to be too high, growth rate too low

and the capital cost too high.

Until captive propagationcan be successfullyachieved,the supply of eel seed

dependstotally on wild capture, and the shortageof natural seedremains a huge

constraint to the developmentof eel culture, also burdening eel farmers with a higher

cost of production (Table 4.7 and Table 4.9). Therefore, any stepsto increasethe

fishery resourceof eel seedwould be valuable imperative to restore the development

of eel culture. The situation is similar to that describedby Houvenaghel(1989)

concerning the decline of Europeaneel, due to:

9 overfishing;

" environmental changes(land reclamation, river banking, dams,etc.);

" water pollution especially in the rivers usedby the eels during ascendingand

descendingmigration;

145
 9 decreasein recruitment.

This situation is even more seriousin Asia and for Japaneseeel. To overcomesuch

constraints,the Taiwanesegovernmenthas tried to increasethe natural resourceby

releasingadult eel into ocean,and importing Europeaneel. However, successhas

beenlimited.

As Table 4.34 showsthe percentagecost of eel seedin producing Japaneseeel has

increasedgreatly, from 8.2% in 1975to 60.5% in 1997.The increasing cost of eel

seedmakesthis industry difficult to develop on a sustainedbasis.

Table 4.34 The percentageof cost structurefor producing Japaneseeel in Taiwan.

Item 1975 1977 1979 1981 1983 1985 1987 1990 1997
Eel seed 8.2% 25.5% 10.1% 25.5% 25.2% 25.7% 29.9% 37.9% 60.5%
Feed 40.9% 38.3% 60.7% 37.4% 36.2% 36.1% 34.1% 34.4% 18.0%
Wage 10.6% 9.0% 5.4% 9.0% 9.2% 0.3% 1.7% 8.4% 7.0%
Electricity 5.3% 3.9% 4.5% 3.6% 4.3% 1.8% 1.2% 5.3% 2.9%
Renting land 11.2% 7.0% 5.6% 7.6% 7.9% 4.8% 4.5% 3.1% 2.5%
Interest and depreciation 13.0% 3.0% 9.4% 2.4% 12.4% 7.9% 5.6% 7.9% 7.8%
Miscellaneous 2.7% 3.3% 4.5% 4.5% 4.8% 3.3% 3.1% 3.1% 0.07%
Data source: 1975-1990:(Chen 1997).
1997: this study.

The various forms of averagefinancial appraisalhave shown that intensive eel culture

has a slight advantageover traditional eel culture. However, the distribution of

profitability shows that it is still possible for traditional eel culture to have higher

profits than super-intensiveeel culture (Table 4.14), the advantageof super-intensive

culture being primarily due to the cheapereel seed.The averageprices of eel seedof

Japaneseeel and European eel are 25.21NT$ and 4NT$ respectively, and as shown in

Table 4.9 and Table 4.12, eel seedtakesup 60.5% and 20.4% of the operation costs in

traditional eel culture and intensive eel culture, respectively. The highest operating

146
cost in intensive culture is feed instead of eel seed.Most eel farmers still do not want

to try intensive eel culture as it is difficult for many to invest more than 7,000,000

NT$ (= 218,750 US$) in the necessaryfacilities (Table 4.10). Farmers also consider

that the growth rate of eels in intensive culture is slower than in traditional culture.

Although the deficiency of eel seed makes operating costs of traditional systems rise,

eel farmers would rather suspendtheir farming than invest in the facilities for

intensive culture.

If the full costs of ground water are accounted,the real cost of traditional eel culture is

high. In different regions, the social cost of ground water is different and here, it

might be appropriatefor the governmentto assistby setting up special areasto

develop eel culture by reducing the social cost. This will be discussedin Chapter6.

The quality of frozen roastedeel producedin Taiwan is as good as that producedin

Japan(Yu, 2001). However, Table 4.21 and Table 4.22, show that Taiwan has lost the

comparative advantagein roastedeel for the Japanese

market. One reasonis that the

supply of raw material is unstable and this has causedsome developersto transfer

their investment to China or stop their business.

The import price of Chineseeels has a significant impact on the import price of

Taiwaneseeels in Japan.If Taiwaneseeels are to improve their potential, it will be

important to differentiate their quality of eels and to develop new products. To reduce

risks of contamination, aquacultureproducts needcareful processingand strict

adherenceto the safety guidelines. Unlike production processesbasedon raw material

147
from capture fisheries, the manufacturing process of aquaculture products begins at

the farms.

The massimport of eel from China and Taiwan has causeda crisis of eel fanning

industry in Japan.In 2001, the number of Japanesecompaniesengagedin eel

growing, around 500, is only one-eight the number of the industry's heyday in the

mid-1970s (The JapaneseTimes, May 9,2001). To protect this industry, Japaneseeel

growers have askedtheir governmentto impose an emergency"safeguard" import

curb. Recently, Japaneseeel growers negotiatedwith their Chinesecounterparts,but

there have beenno substantiveagreementsto date (K. Katsuyama,Deputy Director,

Administrative Division, JapanFishery Agency, 2001, personalcommunication by e-

mail).

If the situation of conditions in the Japaneseeel industry getting worse, the Japanese

government may restrict imports using the WTO rule that empowers member

countriesto impose import restrictions if the domestic industry can suffer serious

damage(The JapaneseTimes, May 9,2001). However, this might induce retaliation

from the Chinesegovernment.

Traditionally, Japaneseconsider the quality of domestic eel products to be better than

that of imported products, and domestically grown eels are retailed at prices some30

percenthigher than those of imported eel (The JapaneseTimes, May 9,2001).

However, not all people can tell the differencesbetweendomestic and imported

products and some importers

Japanese had representedeels from China or Taiwan as

domestic.To addressthis problem, the Japanesegovernmenthas now required that all

148
eels should be sold with the information provided on the place of origin and

processing.Although this is not a discrimination measure(Katsuyama,2001, personal

communication by e-mail), it appearsthat the Japanesegovernmenthopesthat the

patriotism of Japanese
citizens can help to reduceimports from other countries.

The mass production of eel from China may make a big impact on Taiwan's eel

industry, becauseboth areasdependon the Japanesemarket very strongly. In addition

to improving products and extending product range,Taiwan should develop its

domestic market and explore marketsin other countries to alleviate the possible

impact of the Japanesegovernmentin setting up trade curbs on eel import. However,

there appearto be few immediate prospectsfor other countries to import such large

quantities of eel. It may therefore be critical to negotiate with Japanesecounterpartsto

obtain a good import quota. Although, the domestic market appearsto be still difficult

to develop becauseof the high price of eel, there is still some spaceto improve. To

improve the quality or develop new products, which are more suitable for Taiwanese

consumersmight be useful for developing domestic markets.

149
 Chapter 5

Cage culture

5.1 Introduction

The origins of cageculture are somewhatunclear, though it is probable that the first

cage was the use of simple containment by fishermen as a convenient holding facility

for fish until they could be accumulated and ready for sale (Beveridge 1996). Some

documents show that the lower reaches of Mekong River was the place of origin of

cageculture (Hickling 1962; Pantulu 1979). On the basis of recorded documentsHu

(1994) proposedthat cageculture had beenpracticed in ancient times in China. In

Zhou Mi's work Bieji describedhow the fry were placed in cloth cagesin open water

with bamboo sticks supporting the four corners, allowing them to grow bigger for

marketing after one and half months. From this documentdescribing what in modern

aquaculturecould be called a `hapa', the origin of cage culture might be traced back

to at least 1243 in China.

Modern cagesutilize synthetic mesh or netting materials and commonly have floating

structural collars, usually fabricated from synthetic polymers and metals.Japanis

consideredan important sourceof inspiration for modern cage culture (Beveridge

1996). In Japan,the first experiment in cageculture was conductedin 1954, and the

commercial culture of yellowtail (Seriola quinqueradiata) startedin 1957 (Milne

1974). In Norway, cages were used to culture Atlantic salmon (Salmo salar) in the

early 1960sand in Scotland the White Fish Authority commencedsalmon cage

culture trials around 1965 (Beveridge 1996).

150
In Taiwan, the first trial of cageculture startedin Sun Moon Lake in 1970 and

thereafter,investigations in the techniquesof fresh water cageculture were conducted

in the Shyr-Men Reservoir, the U-Shan-Tour Reservoir and the Der-Ji Reservoir. By

1987, there were 300 nets in the inland reservoirs in Taiwan producing some 1800 t

annually. In inland cage culture, tilapia was the most important species by volume,

while the averageprices of perch and eel were higher (Annex D). In 1990,production

from inland cagesattainedits highest level at 2,314 t (Table 5.1). However, the

problem of eutrophication and deterioration of water quality forced the responsible

agenciesto prohibit the setting up of more cagesin reservoirs, and becauseof the

limited resources of water and land, the further development of this type of

aquaculturewas inevitably constrained(Fig 5.1). By 1990, production from inland

cagewas reducing, and by 1997, there has been no commercial inland cage

production (Table 5.1). However, Taiwan is an island, surroundedby open sea,and

therefore, has wide opportunities to develop seacageculture. The advantageof sea

cage culture would include availability of relatively unmodified natural environments,

increasedproductivity in limited space;maintenanceof good water quality becauseof

water currents, and the reduction of dependenceon fresh water and underground

water.

Sea cage culture in Taiwan commenced in the Pen-Hu area in 1977, though before

1989, there were no records on quantity and value of production from this sector. By

1989, according to the Year Book of the Taiwan Fisheries Bureau, the total quantity,

and value of production was 21 t, NT$ 3,218,000 respectively with an average value

of about 153 NT$/kg. Production then increaseddramatically, by 1998, reaching 884 t

151
and a total value of 179,642,000NT$ from a total areaof seacage culture of 1072,896

m2 (Table 5.1). From 1989 to 1998,the averageannual growth rates of production,

value and area are 51.5%, 56.3% and 69.2% respectively (Table 5.2). Of the species

from sea cage culture, red porgy (Pagrus major) was the most significant, with

quantity and value of 261 t and NT$ 75,817,000 by 1997. Red porgy (Pagrus major),

grouper (Epinephelus spp.) and cobia (Rachycentron canadus) had higher prices

levels (about 300 NT$/ kg) among the species used, though the price of cobia dropped

to 125 NT$/ kg in 1998 (Annex D). The average price of cobia dropped 350 NT$ in

1995 to 125 NT$ in 1998, and the production increased from 3 tin 1995 to 17 tin

1998.It implied this fish is price elastic (i. e. elasticity of demand>1.0). The other

speciescan be cultivated in land basedsystems,and so it is difficult to comparetheir

elasticity.

Taiwan cageculture
development

Inland cageculture Seacageculture

In Sun Moon Lake, Shyr-Men Started near shore in the bay of

Reservoir, U-Shan-Tour Pen-Hu District, and Dah-Perng-
Reservoir and Der-Ji Uan and Chu-Keng of Ping-Tung
Reservoir (1971) distrct (1977)

The developmentof inland cage Special areas of cage culture were

culture was limited becauseof set up. The production area, scale
eutrophication and deterioration and production amount are
of water quality (1994) increasing (1997)

Fig.5.1 The developmentprocessof cageculture in Taiwan (modified from Chang,

2000).

152
The prices of products from seacageculture are generally higher than those from

inland culture. Since 1992, averageprices from seacageculture were over 200 NT$

comparedwith averagesfrom inland cage culture which were usually lower than

100NT$. However, sea cage culture had lower productivity; at an average level of 1

kg m-2 in 1990 compared with levels as high as 177 kg m-2 from inland cage culture

(Table 5.1).

Table5.1 Production, value and areaof cageculture in Taiwan, 1989 - 1998.

Year Quantity Value Price Area Productivity

(metric ton) (1000 NT$) (NT$/Kg) 1000m2 (kg m2
1989 Seacage 21 3218 153 ---- ----
Inland cage 633 27480 43 16 40
1990 Seacage 103 13726 133 ---- ----
Inland cage 2314 253946 110 13 177
1991 Seacage 86 11541 134 76 1
Inland cage 508 15043 29 14 36
1992 Seacage 130 29327 225 113 1
Inland cage 330 10898 33 4 84
1993 Seacage 138 31529 228 181 1
Inland cage 368 13955 38 4 93
1994 Seacage 150 36278 241 181 1
Inland cage 21 802 38 1 21
1995 Seacage 357 94128 263 513 1
Inland cage ---- ---- ---- ---- ----
1996 Seacage 678 147500 218 534 1
Inland cage ---- ---- ---- ---- ----
1997 Seacage 837 186074 222 391 2
Inland cage 0 0 ---- 0 ----
1998 Seacage 884 179642 203 1073 1
Inland cage 0 0 ---- 0 ----
Data source:Year Books of Taiwan FisheriesBureau (1990-1999).

153
Table 5.2 The average annual growth rate of production, value and area of sea cage
culture in Taiwan, 1989 - 1998.
Quantity Value Area*
Annual growth rate 51.5% 56.3% 69.2%
* The average growth rate of area is from 1990-1998.

5.2 Development conditions

5.2.1 Culture area

The successfuldevelopmentof cageculture is very dependenton having accessto

suitable sites (Jensen1996). Beveridge (1996) recommendsto establishfacilities for

cageculture in areaswhere current speedsare normally less than 50 cm/sec.

Currently, the major prefecturesfor cageculture are in Pen-Hu and Pintung. In Pen-

Hu, there are about 900 single cages,360 single cagesin Pintung and 10 single cages

in I-Lan. In Miaoli, Hwalian and Taitung, there are 4 single cages respectively.

In Pen-Hu Prefecture,cage culture has beencarried out for more than 10 years.

Initially, cageswere set up by fishermen for keeping fish stock or as an appendix of a

set net. Later, fishermen startedfeeding the stocked fish and, thus, commencedcage

culture. Becauseof the strong monsoonsduring the winter period, most cageswere set

up in the inshore bays of Pen-Hu. However, although this can reducemonsoon

damage,problems of self-pollution may quickly arise. Owing to the monsoon,water

temperaturedecreasesduring the winter, thus reducing feeding and winter growth

rate.

154
In Pintung Prefecture,cageculture startedin Da-Pern-Wan.As Da-Pern-Wanis

surroundedby land on three sides and is connectedon one side to the open sea,water

circulation is very restricted, and owing to this poor circulation, the cage farms in this

area caused serious eutrophication. The local government therefore forced the farmers

in this area to move elsewhere.

In 1997,Pingtung prefecture governmentstartedpromoting cageculture and set up a

specific areafor cage culture in Fon-Kong and Shioa-Liu-Chou areas.In southern

Taiwan, in the areafrom Tunkung to Fon-Liao, there is no monsoon and there is no

typhoon from October to May of the following year. The typhoon period from Juneto

Septemberis the seasonfor the propagationof several speciesof marine fish. After

nursing, the fingerling can be transferredinto cages,by which time they can avoid

typhoon damage.Thesecircumstancesmake Pintung prefecture a very suitable area

for developing cage culture.

5.2.2 Legal rights for cage culture

Aquaculture is strongly affected by public law becauseit involves many social areas

which concern modem governments(Rieser 1996).To avoid conflicts betweencage

farmers with navigation rights, the capturefisheries sector and other related functions,

governmentmust provide farmers a special legal right to develop cage culture.

Among these,the establishmentand protection of property rights of private culturists

are paramount.Laws providing ownership of cageculture to the culturist have to be

promulgated.Otherwise, investors would have little incentive to invest and to

innovate. In Taiwan, cageculture is consideredas a form of fisheries, and cage

farmers must obtain fisheries rights to becomelegitimate, two types of which are

155
suitable.These are the Zoning Fishery Right (ZFR) and the Specification Fishery

Right (SFR). Both establishthe right for fishermen to fish in a certain water area.

However, applications for SFR are limited to the Fishermen's Association or Fisheries

Production Cooperatives,while applications for ZFR can also include individuals or

companies.The areaallocations for ZFR are modest: for individual farmers less than

3 hectares and groups less than 10 hectares. However, the SFR can be hundreds of

hectares(Table 5.3). Applicants can obtain licensesfrom the local Fishermen

Association, which has beendelegatedthe authority, apart from in Pen-Hu, where fish

farmers can only get the license directly from the local government.

Table 5.3 Features of Zoning Fishery Right (ZFR) and Specification Fishery Right
(SFR).
Item Zoning Fishery Right (ZFR) Specification Fishery Right (SFR)
Qualification of Individual, partnership, FishermenAssociation and
applicant company, institute of FisheriesProduction Cooperation
research,Fishermen
Association and Fisheries
Production Cooperation etc.
Fishing Fishermen Association and
FisheriesProduction Cooperation
set up the rules for fishermen
fishing in the areaof SFR
Area for applicant Less than 3 hectarefor Hundredsof hectares
individual and less than 10
hectarefor groups.
Duration of tenure 5 years 10 years
Current situation Pen-Hu prefectureusesZFR The other prefecturesexcept Pen-
for cageculture Hu prefecture use SFR for cage
culture
Coordination agency The FisheriesDepartment of FishermenAssociation and
local government FisheriesProduction Cooperation

5.3 Systemfeatures

5.3.1 Cage structure

Cagesfor fish culture have been constructedfrom a variety of materials and at a range

of sizes.Basic requirementsare that cagematerials are strong, durable and non-toxic.

156
In 1970s, in the Pen-Hu Islands, the initial net cages were mounted on a 5mx5mx3m

(m3) or 6mx6mx4m (m3) framework. After that, cages were gradually enlarged to

7mx7mx4m (m3), 8mx8mx4m (m3) and even to 10mx10mx5m (m3). These are either

in the form of a wooden square supported by floats made of a 1.OxO.6x0.5m3 pieces

of styrofoam wrapped in plastic sheet, or in the form of a series of five or six

connected styrofoam pieces on each side (Fig 5.2). Detachable top nets are used to

prevent fish from escaping. The bottom of the net cage is weighted with lead; four or

five net cages are connected to become a group. Cages should not be too close

together for avoiding the likelihood of low dissolved oxygen (Massser, 1997).

Currently, a range of cages is used in Taiwan.

1. Flexible frameless cage: the framework is made of Styrofoam buoys and

connected to the net. The shape is maintained by the buoys, sinkers and anchors.

Usually, the sizes are from 7mx7mx4m deep (196 m3) to 10mxlOmx5m deep

(500 m3).

2. Rigid steel net cage: the framework is madeof stainlesssteel. Nets are tied to the

framework of the cages.The size are usually 5mx5mx3m deep (75 m3).The major

purposeof this kind of cageis rearing fingerlings.

3. Wooden semi-rigid cage: the framework is made of wood and connectedto a

flexible net cage,the size is usually 7mx7mx4m deep (196 m3). When the weather

for
worsens, example, in a typhoon, the framework is disassembledand hauled to

the shore.In normal conditions, farmers can stand and work from the framework

for feeding, changing nets and other manipulations.

4. PVC rigid cage: the framework is made of PVC pipe and connectedto the net

cage with buoys made of Styrofoam. Similar to rigid steel frame cage, the size is

about 5mx5mx3m deep (75 m3), and it is usually used for rearing fingerling.

157
5. Submersiblecage:the framework is made of circular flexible plastic PVC. During

a period of wave attack, this kind of cage can be submergedunder the surfaceof

water to reducedamage.A hoseis attachedto the ring of the cage framework,

which when submerged, allows water to fill the framework and sink the cage.

Later the cagecan be raisedup by feeding compressedair down the hoseto empty

the water from the framework. Originally, the brand named Hvalpsund net cage

was imported from Denmark, though now, some factories in Taiwan can produce

systemsof this type. The size is 16mOx8m deep (1600 m3), and it can

accommodate16t of stock at 19 kg m-3.

Buoy
Il 111 1

1I11 1
Il I Il I Il I Il I1
Il I I! I1
111I 1
1
1
nnnn /ý nnn n
i
i
i
i
i
i
i

Sinker
Rope

Connectedto anchor

Fig 5.2 The traditional structureof cagein Taiwan.

Netting of different mesh size is usedin the cagesduring the production phase,using

a larger mesh as the fish get larger. Nets may also be replacedwhenever there is

extensive growth of algae.To do this, about two-thirds of the old net is disassembled

from the framework, on to which is then mounted the new net. The remaining part of

the new net is made to envelopethe old one and the fish in the old net are emptied

158
into the new one. After that the old net is removed and the new one is mounted

completely. When the old nets have been removed, fish farmers use high pressure

waterjets to clean them.

Most fish farmers use flexible insteadof rigid frames, which only use styrofoam

buoys instead of a rigid framework, while the framework of cages is usually mounted

when changing nets. During the typhoon period some fish farmers pound steelbars to

the seabottom to serveas a heavier sinker to which the net cage is mounted, with its

top submergedto about 2m below the seasurface.It is said that the flexible and

submergednet cagesare more weatherresistant.

5.3.2 Anchoring

Cagesmust be fixed in designatedplacesby anchoring. Initially, 4 or 5 cageswere

connectedtogether and the two ends of eachgroup were each attachedto a 70 kg steel

anchor in 5-8m of water (Chen 1992). Some fish farmers used cement blocks or bags

of gravel to replace the steel anchors.More recently, cement blocks of about 5-15 t

have becomethe most widely usedmaterial for anchoring. After the blocks are placed

into the water, the cagesare connectedusing ropes or steel chains. If the water bottom

type is sand,the anchorbecomesembeddedowing to the current flow. If the bottom

type is rock, some farmers will use pins to anchor this systemdirectly to the rock.

5.3.3 Net management

Biofouling, the growth of algae and other animals on the sidesof the cage,is a

common problem. In tropical inshore areas,nutrient-rich water acceleratesmarine

biofouling, which will reducemesh size and restrict the flow of water through cages,

159
thus reducing the rate of DO supply and waste metabolite removal. Biofouling will

also increaseresistanceto water flow and increasethe force of the current on the cage

structure and netting, thereby causing deformation of the bag; decreasingthe cage

volume for fish; reducing the useful life of the nets and potentially causing damageto

the cage structure.Although there are other methodsfor coping with biofouling, for

example, chemical or biological control agents,fouling resistant or rotating designs,

fish farmers in Taiwan have usedthe most basic approach,exchanging nets and

cleaning them with high pressurewater jets. The frequency for exchangingnets

dependson the weather and the speciesbeing reared.In the summer seasonthe

frequency of exchangingnets is higher than in the winter as biofouling is more rapid

and as fish grew faster in higher temperaturesand the feeding ratio is higher. The

specieswhich grow most rapidly, require more frequent net exchange.For example,

in the summer seasoncagesfor rearing cobia (Rachycentroncanadus)usually

required a net exchangeevery 25 days where those for silver bream (Sparussarba)

need to be exchangedevery 90 days (Chen, personalcommunication, 1999).

5.3.4 Feeding

Cagedfish in most caseswill receive no natural food and, therefore, must have

nutritionally complete feeds. These feeds should have adequateprotein and energy

levels, balancedin amino acids and in essentialfatty acids, and be supplementedwith

a complete array of vitamins and minerals. In Taiwan, feeds for cage aquaculture

include trash fish and moist or dry formulated feed. Their use varies with species.

Thus, groupers (Epinephelus spp.) are usually fed with trash fish and cobia with

formulated feeds.During the winter season,with reducedfeeding activity, feeding

frequency and ration is also reduced.In the summer season,the typical feeding

160
frequencyfor fingerling is 5-6 times per day and for harvestablefish is once per day.

Hand feeding are also important, as from the behavior of fish in feeding, farmers can

observeand understandtheir stateof health.

5.4 Cultured species

Many speciesof fish are suitable for cageculture. Becausecapital and operatingcosts

of cageculture are high in Taiwan, priority is given to farming of high value species

to offset the high investment,particularly in the caseof carnivorous fish which require

high cost diet basedon fishmeal. The major cagecultured speciesare extremely

diversified and include cobia (Rachycentroncanadus),red porgy (Pagrus major),

silver bream (Sparus sarba), Spangledemperor (Lethrinus nebulosus),Dumerils's

Amberjack (Seriola dumerili), Brown croaker (Atrobucca nibe), snappers(Lutjanus

spp.), grouper (Epinephelus spp. ) red drum (Sciaenops ocellatus) and others. The

fingerling size of most of the speciesis about 3 cm except for cobia, which is about 9-

10cm (Table 5.4). Among those species,cobia is the fastestgrowing and provides the

best output per fry supply, growing up to 10 kg after 18 months.

In many areasof the world, fry and fingerling production is the main technical

problem yet to be overcome.However, in Taiwan this issueis not serious,as almost

all speciesfor cage aquaculturecan be artificially propagated.Taiwan's aquaculturists

have sophisticatedskills in propagation and most fry and seedfor seacageculture can

be obtained from hatcheries(Liao 1995). According to the numbersand facilities of

hatchery farms, the TaiwaneseFish Breeding Association estimatedthat the potential

production of fingerling in Taiwan was far more than the current demand(Table 5.4).

161
Hatchery production seasonsare shown in Table 5.5. The size at harvest, stocking

density, and the culture period are shown in Table 5.6. The harvest sizesof most

speciesare between0.3-1 kg except for cobia and amberjack,which are 6-10 kg and

1.5-2 kg respectively.The stocking densitiesbefore harvest of most speciesare

between6-12 fish/m3 except for cobia and amberjack,which are 2-5 (about and 3-4

fish/ m3 respectively.The culture period of most speciesis about 1 to 1.5 years.

Currently, cobia is the most widely producedfish in Pen-Hu and accountsfor 38% of

its total production, followed by red porgy, which accountsfor 28% (Table 5.7).

Amongst the speciesgrown, cobia is consideredas having the greatestpotential in

Taiwan, attaining a weight of 6-8 kg in one year with food conversion ratios ranging

from 1.6 to 2 using dry pelleted food (Chang 2000). The survival rate of cobia can be

as high as 90% and about 150 t of whole fish, at an average6 kg, was exported to

Japanin 1999 at a wholesaleprice of around US$ 4.48 to $ 5.70 kg-1(Chen 2000).

Table 5.4 The size of fingerling, domestic demand,estimatedproduction potential and

average price of different species for cage culture in Taiwan.
Species Size Demand Estimated Average price
potential (NT )
Cobia 9-10 cm 1,390,000 1,500,000 22
(Rachycentron canadus)
Red porgy 3 cm 200,000 1,500,000 6.5
(Pagrus major)
Silver bream 3 cm 4,400,000 10,000,000 3
(Sparus sarba)
Snapper 3 cm 6,650,000 30,000,000 4.5
(Lutjanus spp. )
Grouper 3 cm 4,450,000 18,000,000 18
E ine helus spp.)
Red drum 3 cm 1,000,000 30,000,000 2.5
(Sciaenopsocellatus
Data source:Fish Breeding Association', 1999.

162
Tahle 5.5 The nronagation season of different species for cage culture in Taiwan.
Species Mo nth
Jan. Feb Mar Apr May Jun. Jul. Au Se Oct Nov Dec
Cobia (Rachycentron canadus) ---- ---- ----
Red porgy (Pa rus major ---- ---- ---- ----
Silver bream (S arus Barba) ---- ---- ---- ---- ----
Brown croaker (Atrobucca nibe ---- ---- ----
Grouper (E ine helus s. ---- ---- ---- ---- ---- ---- ---- ---- ----
Red drum (Sciaenops ocellatus) ---- ---- ---- ----
Spangled emperor (Lethrinus ----
nebulosus
Data source: Fish Breeding Association, 1999.

Table 5.6 Size at harvest,stocking density and culture period of main fishes cultured
" nn f"T
111 U11Jf UiG Vtl GJ 111 1aiwall.

Species Size at harvest stocking densit' before culture period

(kg) harvest (fish/m (months)
Cobia 6-10 2-5 12-18
(Rachycentron canadus)

Dumerils's Amberjack 1.5-2 3-4 12-14

(Seriola dumerili)

Pink snapper 0.6-1 6-12 10-12

(Lutjanus erythropterus)

Grey snapper 0.4-0.6 6-10 10-12

(L. argentimaculatus)

White-spotted snapper 0.6-1 6-10 10-12

(L. stellatus)

Grouper 0.6 6-10 10-12

(Epinepheluscoioides)

Pompano 0.3-0.6 6-10 10-12

(Trachinotus blochii)

Red porgy 0.3-0.6 8-12 12-14

(Pagrus major
Source: Su et al., 1999.

Fish Breeding Association did not mention how the demandsand potentials were estimated.

163
Table 5.7 The production amount and percentageof cage aquaculturefish in Pen-Hu,
1997.
Fish species amount Produced Percentage(%)
(tonnes/year)
Cobia (Rachycentron canadus) 660 38.0
Red porgy (Pagrus major) 483 28.0
Silver bream (Sparus sarba) 255 14.7
Span led emperor (Lethrinus nebulosus) 118 6.8
Dumerils's Amberjack (Seriola dumerili) 97 5.6
Brown croaker (Atrobucca nibe) 44 2.5
Others 78 4.5
Total 1735 100
Data resource: statistical data of Peu-Hu government (1998).

5.5 Markets

5.5.1 Marketing channels

In the Pen-Hu area,most cage farmers sell their fishes through dual distribution

channels,selling directly to consumersas well as through independentmarketing

intermediaries.Usually, farmers' wives or relatives will sell some fish directly to

consumersat the retail market in Pen-Hu (aout 20%). The other fishes are sold to

Taiwan Island by wholesalers.By contrast,cage farmers in Taiwan usually sell all

their fishes to wholesalers,who also collect products from fish farmers and deliver

them to auction markets.In auction markets,products were sold to retailers, who then

sell to consumersin retail markets.

164
Somecageculture farmers (such asFuw -Cheng Marine Biotec. Co. Ltd. in Pen-Hu)

also try to sell their fishes to Japanese

markets as materials for raw fish (sashimi).

Farmersconsider cobia is a good candidatespeciesfor the sashimi market in Japan

and have tried to develop this. According to Tsai's (1999, personal communication)

estimation, the marketing channel of cageculture could be summarizedas shown in

Fig. 5.3.

< 10%

Fig 5.3 The marketing channelsof products from cage aquaculture.

5.5.2 Market characteristics

In Taiwan, there is a market preferencefor live fish, which are transportedin live-haul

trucks. Usually, the weaker fish are separatedand sold locally, or slaughteredfor

further markets, as they are not likely to survive long distancetransportation,which

may be more than 300 Km and take more than 3 hours.

Prior to transport, fish are starvedfor a day, or longer for larger fish, to reducethe

contamination of water by excrement and to sedate the animal to make it less active.

Often, fish are also sedatedby gradually lowering the water temperatureduring

transport,gradually raising it again to ambient when arriving at the destination.

165
Lowering the temperaturehas also been shown to improve the texture of the fish flesh

(Subasinghe1996).

5.6 Economic evaluation

5.6.1 Introduction

Becauseof the breakthroughof the technology in rearing larvae and the diversified

demandof consumers,severalfish speciesmay be consideredas candidatesfor

marine cage aquaculture.Most of the candidatesare high priced species,typically

those sold to restaurantsor retail marketsdirectly and fetching premium prices.

Among those candidates,the dominant speciesinclude red porgy (Pagrus major),

silver bream (Sparus sarba), grouper (Epinephelusfario), Dumeril's amberjack

(Seriola dumerili), the spangled emperor (Lethrinus nebulosus), and cobia

(Rachycentroncanadus)etc.

However, it must be recognizedthat the developmentof a commercially viable

aquatic farming systemrequires much more researchand developmenteffort than the

experimental raising of a few animals under controlled conditions. It was therefore

consideredlogical to concentrateefforts on a smaller number of speciesand farming

systemsin order to have an adequatenumber of properly testedtechnologiesfor

commercial application (Pillay, 1994).

In 1998,Wang and Huang analysedthe cost structure of cage aquaculturein Taiwan.

However, they did not compareprofitability betweenthesecandidatespecies.The aim

of this section is therefore to comparethe costs and benefits of different cage

aquaculturecandidates.

166
For this purpose, 22 cage farmers were surveyed by questionnaire in 1999,10 in Pen-

Hu and 12 in Ping-Tong county. The averagefarm size is about 32.14 cages.The

selection of samplesis describedin Chapter 2.

5.6.2 Cost analysis

The average cost of cage aquaculture can be categorised into two components: capital

and operating cost. The capital costscomprise the costs of net, rope, buoys,

anchoring,installation, raft, building, and outboard motor. The operating costsconsist

of variable costs (costs of fingerlings, feeds, labour, fuel and miscellaneous)and

overheads(interest and depreciation).

In carrying out this exercise,three important assumptionshave been made:

1) investment costs are coveredby a loan at an annual interest rate of 8%;

2) facilities and equipment are subjectedto straight line depreciationover their

useful lifetime;

3) the useful lifetime of the net, ropes,buoys and anchoring is 10 years,while

those of the building, boat and outboard motor are 30,20 and 5years

respectively.

Becausespendingon variable costsis spreadthroughout the culture period the

interest is assumedto be chargedon an averageof 50%, as total expenditure on

variable cost is not paid out at the beginning. However, interest is chargedon 100%

of the fixed cost (Christensen.1993), as this is incurred at the outset of eachperiod

in operation.

167
To simplify the comparativeanalysis,cost were separatedinto those which were

similar regardlessof the speciesgrown, and those, such as fingerlings and feeds,

which differed with The

species. averagecommon costs of cage aquacultureacross

different speciesare shown in Table 5.8. The differential costs of fingerlings and

feeds are separatelyspecified in Table 5.10 and Table 5.11. Among the cage

aquaculture the
species, price of grouper fingerling is the highest, at 34.7 NT$ each,

the averagecost of fingerlings per farm is 3.40 million NT$ and the averagecost per

kg producedis 46.7 NT$ (Table 5.9). Next to grouper is Dumerils's Amberjack,

whose fingerling price is 34.6 NT$, averagecost of fingerlings per farm is 2.88

million NT$ and the averagecost per kg producedis 29.6 NT$. The averagecostsof

feeds for different speciesare listed in Table 5.11.The highest is cobia, at 5.2 million

NT$ per farm, followed by silver bream, at 5.03 million NT$. The highest average

cost of feeds for producing 1kg fish is silver bream, at 57.9 NT$, followed by grouper,

at 56.0 NT$. The averagesurvival rate and feed conversion ratios of different species

are shown on Table 5.9.

Operatingcosts are shown in the Table 5.12, which shows that the major costs are

feed, fingerling and labour. Apart from grouper and Dumerils's Amberjack, which

have higher fingerling costs, ratios of feed to total operating cost are more than 50%.

Next to feed, fingerling is the second highest cost, for grouper and Dumerils's

Amberjack at 33.0 and 26.8% of total, respectively. Ratios for others range from

12.43to 14.49%.The third highest cost is labour, at 12.2 to 16.1% of total operating

cost.

168
Table 5.8 Fixed and variable cost structure(exclude feed and fingering) per farm for
cage aquaculture in Taiwan. Unit: thousand NT$
Items Total cost Annual
depreciation
Capital cost
Net 2016.6 201.7
Rope 390.1 39.0
Buoy 317.6 31.8
Anchoring 250.0 25.0
Install 387.9 38.8
Building (including feed storage,office fridge etc.) 564.6 18.8
Boat 295.2 14.8
Engine of outside boat 124.5 24.9

Total capital cost 4346.5

Operating cost
Labour 1316.1
Fuel 202.2
Interest of fixed cost 347.7
Subtotal of variable cost 1866

Depreciation 394.8
Miscellaneous 197.6
Subtotal of overhead 592.4

Total operating cost 2458.4

Table 5.9 The averagesurvival rate and feed conversion ratio (FCR) of different
species.
Fish species Survival rate FCR
Cobia 48.9 2.0
Red porgy 41.7 1.7
Silver bream 57.5 1.9
Spangledemperor 48.6 2.0
Dumerils's Amberjack 51.2 2.1
Grouper 55.7 2.3
Data source: Wang and Huang (1998).

169
Table 5.10 The average cost of different species of fingerlings for cage aquaculture
per year.
Fish species Average Average Average Average Average
price of density cost/cage cost/farm cost/kg
fingerlings (fingerlings/ (NT$) (Million produced
(NT$) cage) NT$) (NT$)
Cobia 18.9 2,083 39,446 1.27 11.9
Red porgy 6.4 5,274 33,640 1.08 9.4
Silver bream 3.6 10,018 35,779 1.15 13.2
Spangledemperor 8.1 4,533 36,805 1.18 13.7
Dumerils's Ambe 'ack 34.6 2,588 89,598 2.88 29.6
Grouper 34.7 3,047 105,744 3.40 46.7

Table5.11 The averagecost of feed for different speciesof cage aquacultureper year.
Fish species Average Average Average Average Average
price of amount cost of feed cost of cost /kg
feed of feed (NT$/cage) feed produced
(NT$/kg) (kg/cage) (Million (NT$)
NT$/farm
Cobia 24.7 6614.9 162,775 5.23 49.2
Red porgy 24.1 6254.0 150,602 4.84 41.9
Silver bream 30.2 5198.1 156,639 5.03 57.9
Spangledemperor 25.1 5235.0 131,287 4.22 48.9
Dumerils's Amberjack 25.1 6243.5 156,236 5.02 51.5
Grouper 24.5 5182.1 126,760 4.07 56.0

170
 Table 5.12 The cost structureof the averageoperating cost of production of cage
aquaculture in Taiwan. Unit: thousandNT$
Cobia Red Silver Spangled Dumerils's Grouper
porgy bream emperor Amberjack
Variable cost
Fingerling 1267.9 1081.3 1150.0 1183.0 2879.9 3398.9
(13.6) (12.4) (12.8) (14.5) (26.8) (33.0)
Feed 5232.1 4840.8 5034.8 4219.9 5021.9 4074.4
(56.2) (55.6) (56.1) (51.7) (46.7) (39.5)
Labour 1316.1 1316.1 1316.1 1316.1 1316.1 1316.1
(14.2) (15.1) (14.7) (16.1) (12.2) (12.8)
Fuel 202.2 202.2 202.2 202.2 202.2 202.2
(2.2) (2.3) (2.3) (2.3) (1.9) (2.0)
Miscellaneous 197.6 197.6 197.6 197.6 197.6 197.6
(2.1) (2.3) (2.2) (2.4) (1.8) (1.9)
Subtotal 8,215.9 7,638 7,900.7 7,118.8 9,617.7 9,189.2
(88.3) (87.7) (88.1) (87.0) (89.4) (89.2)
Overheadcosts
Depreciation 394.8 394.8 394.8 394.8 394.8 394.8
(4.2) (4.5) (4.4) (4.8) (3.7) (3.8)
Intereston fixed costs 347.7 347.7 347.7 347.7 347.7 347.7
(3.7) (4.0) (3.9) (4.3) (3.2) (3.4)
Intereston variablecosts 344.4 321.3 331.8 300.5 400.5 383.4
(3.7) (3.7) (3.7) (3.7) (3.7) (3.7)
Subtotal 1086.9 1063.8 1074.3 1043 1143 1125.9
(11.6) (12.2) (12.0) (12.8) (10.6) 10.9
Total 9,302.8 8,701.8 8,975 8,161.8 10,760.7 10,315.1
Figures in bracketsare percentagesof total.

The averagecosts for producing 1 kg of products are shown in Table 5.13. This table

shows that grouper has the highest cost per kg at 141.8NT$ (4.4 US$), followed by

Dumerils's Amberjack and silver bream at 110.5 and 103.1 NT$, respectively. The

lowest is red porgy, at 75.3 NT$. The costsfor producing 1kg of product were 87.5,

75.3,103.1,94.6,110.5 and 141.8NT$/kg, the highest and lowest cost ranged from

78.2-95.1,65.6-83.5,93.2-110.8,86.4-109.7,98.6-119.3 and 129-149.2NT$/kg in the

speciesof cobia, red porgy, silver bream, spangledemperor, Dumerils's Amberjack

and grouper, respectively. The key contributory factors in the cost for producing 1 kg

of products were feed, fingerling and labour.

171
The cost of feed for producing 1kg of product were 49.2,41.9,57.9,48.9,51.5 and

52.0 NT$/kg, the highest and lowest cost of feed ranged from 46.7-53.1,37.1-43.6,

54.2-60.8,46.9-49.7,48.9-53.7 and 54.8-57.4NT$/kg in the speciesof cobia, red

porgy, silver bream, spangledemperor,Dumerils's Amberjack and grouper,

respectively.

The cost of fingerling for producing lkg of product were 11.9,9.4,13.2,13.7,29.6

and 46.7 NT$/kg, the highest and lowest cost of fingerling ranged from 7.8-14.2,7.9-

13.2,9.7-15.1,10.2-15.6,23.1-34.5 and 42.3-49.2 NT$/kg in the speciesof cobia, red

porgy, silver bream, spangledemperor,Dumerils's Amberjack and grouper,

respectively.

The cost of labour for producing 1kg of product were 12.4,11.4,15.1,15.3,13.5 and

18.1 NT$/kg, the highest and lowest cost of labour ranged from 9.7-14.5,8.6-13.6,

12.1-18.3,12.1-18.4,11.8-15.8 and 15.8-21.3NT$/kg in the speciesof cobia, red

porgy, silver bream, spangledemperor,Dumerils's Amberjack and grouper,

respectively.

Better FCR, higher survival rate and lower cost of labour are important in reducing

cost. Therefore, better bigger

management, farm to reducethe labour cost by

economiesof scale might be important in reducing production cost.

172
 Table 5.13 The cost structure of the averageoperating cost for producing 1 kg of cage
aauacultureproducts in Taiwan. Unit: NT$
Cobia Red porgy Silver Spangled Dumerils's Grouper
bream emperor Amber'ack
Variable cost
Fingerling 11.9 9.4 13.2 13.7 29.6 46.7
(7.8-14.2) (7.9-13.2) (9.7-15.1) (10.2-15.6) (23.1-34.5) (42.3-49.2)
Feed 49.2 41.9 57.9 48.9 51.5 56.0
(46.7-53.1) (37.1-43.6) (54.2-60.8) (46.9-49.7) (48.9-53.7) (54.8-57.4)
Labour 12.4 11.4 15.1 15.3 13.5 18.1
(9.7-14.5) (8.6-13.6) (12.1-18.3) (12.1-18.4) (11.8-15.8) (15.8-21.3)
Fuel 1.9 1.8 2.3 2.3 2.1 2.8
(1.5-2.3) (1.5-2.2) (1.9-2.6) (2.0-2.6) (1.6-2.4) (2.4-3.2)
Miscellaneous 1.9 1.7 2.3 2.3 2.0 2.7
(1.6-2.3) (1.4-2.1) (2.1-2.6) (2.0-2.6) (1.8-2.2) (2.4-3.0)
Subtotal 77.3 66.2 90.8 82.5 98.7 108.2
Overhead
costs
Depreciation 3.7 3.4 4.5 4.6 4.1 5.4
(3.2-4.6) (3.1-3.8) (4.1-4.9) (4.3-4.9) (3.7-4.4) (5.1-5.7)
Interest on 3.3 3.0 4.0 4.0 3.6 4.8
fixed cost (3.0-3.5) (2.8-3.3) (3.6-4.3) (3.7-4.2) (3.3-3.9) (4.4-5.1)
Interest on 3.2 2.8 3.8 3.5 4.1 5.3
variablecost (2.9-3.6) (2.4-3.2) (3.4-4.1) (3.2-3.8) (3.7-4.4) (4.8-5.6)
Subtotal 10.2 9.2 12.3 12.1 11.8 15.5
Total 87.5 75.3 103.1 94.6 110.5 141.8
(78.2-95.1) (65.6-83.5) (93.2-110.8) (86.4-109.7) (98.6-119.3) (129-149.2)
Fi 0inres in narentheses are the range of hig hest and lowes t cost.

5.6.3 Benefit analysis

The profit (P) is equal to the revenue(MI) minus production cost (C). Profitability can

be estimatedby the benefit-cost ratio (BCR) and the income ratio (IR)(Chen 1994).

The respective formulas are as follows:

BCR=P/C

IR=P/MI

Where P= Profit

C= Production cost

MI = Revenue

173
If the values of BCR and IR are specifically high compared with other options, it

indicates the operation is economically sound and further developmentis feasible.

Revenuesdependon market prices, which may differ widely from place to place.

Usually, the samespeciescan fetch higher price in Pen-Hu than elsewherein Taiwan,

as people in Pen-Hu traditionally like to eat fish and are willing to pay higher prices.

As Table5.14 shows, the highest revenue per farm unit of 97.42 tonnes annual

production is attained from Dumerils's Amberjack, with an average annual revenue of

29.23 million NT$ (820,000 US$) after which is red porgy, with an averageprofit of

23.11 million NT$ (722,000 US$). The lowest revenueis seenwith grouper, at only

16.73million NT$ (523,000 US$). Regarding averageprofit, benefit-cost ratio and

income ratio, Dumerils's Amberjack also offers the best performance,at 18.47 million

NT$, 1.72 and 0.63 respectively, following which is red porgy, with an averageprofit,

benefit-cost ratio and income ratio of 14.40 million NT$, 1.66 and 0.62 (Table 5.15).

Table 5.14 The averagecost of production, amount of production, price, revenueand

benefit for different cage aquaculturespecies.
Fish species Average Average Product- Average Average
cost of amount of ion cost price of revenue
production production (NT$/kg) fish (Million
(Million (tonnes (NT$/kg) NT$/farm)
NT$/farm) /farm)
Cobia 9.30 106.31 87.5 160 17.01
Red porgy 8.70 115.53 75.3 200 23.11
Silver bream 8.98 87.02 103.1 220 19.14
Spangledemperor 8.16 86.29 94.6 200 17.26
Dumerils's Amberjack 10.76 97.42 110.5 300 29.22
Grouper 10.32 72.74 141.8 230 16.73

174
Table 5.15 The averageprofit, benefit cost ratio and income ratio of different cage
aquaculture species. Unit: Million NT$
Fish species Average BCR IR
profit/
farm
Cobia 7.71 0.83 0.45
Red porgy 14.40 1.66 0.62
Silver bream 10.17 1.1 0.53
Spangledemperor 9.10 1.1 0.53
Dumerils's Amber'ack 18.47 1.72 0.63
Grouper 6.41 0.62 0.38

The ranges of highest and lowest profit shows that Dumerils's Amberjack had the

highest profit at 19.6-17.6.million NT$ (Table 5.16). However, cobia might have

higher profit than Spangledemperor and grouper might have higher profit than cobia

when there is a better performance.Better FCR, higher survival rate and lower cost of

labour are the important factors to increaseprofit.

Table 5.16 The range of profit, benefit cost ratio and income ratio of different cage
aquaculture species. Unit: Million NT$
Fish species Profit BCR IR
Highest Lowest Highest Lowest Highest Lowest
Cobia 8.70 6.90 1.05 0.68 0.51 0.41
Red porgy 15.53 13.46 2.05 1.40 0.67 0.58
Silver bream 11.03 9.50 1.36 0.99 0.58 0.50
Spangledemperor 9.80 7.79 1.32 0.82 0.57 0.45
Dumerils's Ambe 'ack 19.61 17.60 2.04 1.51 0.67 0.60
Grouper 7.35 5.88 0.78 0.54 0.44 0.35

5.6.4 Cash-flow and discountedfinancial indicators

The pattern of cashflow includes capital cost, operating cost (excluding interest and

depreciation)and revenue.A 5-year nominal and a discountedcash flow analysis (at

10% discount rate) of cageculture farm reveal that investment in Dumerils's

Amberjack can obtain the highest NPV at 67921.6 thousandNT$, followed by red

175
porgy at 52662.2 thousandNT$ (Table 5.17 and Table 5.18). The lowest NPV is seen

with grouper at 23446.6 thousandNT$.

Table 5.17 Nominal cash-flow projection for cageculture.

Unit: Thousand NT$
YearO Yearl Year2 Year3 Year4 Year5

Cash flow
Capital cost
Cobia 4346.5 0 0 0 0 0
Red porgy 4346.5 0 0 0 0 0
Silver bream 4346.5 0 0 0 0 0
Spangledemperor 4346.5 0 0 0 0 0
Dumerils's Amberjack 4346.5 0 0 0 0 0
Grouper 4346.5 0 0 0 0 0
Operating cost
Cobia 8215.9 8215.9 8215.9 8215.9 8215.9
Red porgy 7638.0 7638.0 7638.0 7638.0 7638.0
Silver bream 7900.7 7900.7 7900.7 7900.7 7900.7
Spangledemperor 7118.8 7118.8 7118.8 7118.8 7118.8
Dumerils's Amberjack 9617.7 9617.7 9617.7 9617.7 9617.7
Grouper 9189.2 9189.2 9189.2 9189.2 9189.2
Revenue
Cobia 17009.6 17009.6 17009.6 17009.6 17009.6
Red porgy 23106.0 23106.0 23106.0 23106.0 23106.0
Silver bream 19144.4 19144.4 19144.4 19144.4 19144.4
Spangledemperor 17258.0 17258.0 17258.0 17258.0 17258.0
Dumerils's Amberjack 29226.0 29226.0 29226.0 29226.0 29226.0
Grouper 16730.2 16730.2 16730.2 16730.2 16730.2
Net cashflow
Cobia -4346.5 8793.7 8793.7 8793.7 8793.7 8793.7
Red porgy -4346.5 15468.0 15468.0 15468.0 15468.0 15468.0
Silver bream -4346.5 11243.7 11243.7 11243.7 11243.7 11243.7
Spangledemperor -4346.5 10139.2 10139.2 10139.2 10139.2 10139.2
Dumerils's Amberjack -4346.5 19608.3 19608.3 19608.3 19608.3 19608.3
Grouper -4346.5 7541.0 7541.0 7541.0 7541.0 7541.0

176
Table 5.18 Discounted cash-flow projection for cageculture.
The discount rate for NPV is 10%. Unit: ThousandNT$
YearO Yearl Year2 Year3 Year4 Year5
Cashoutflow
Cobia 4346.5 7394.3 6654.9 5989.4 5390.5 4851.4
Red porgy 4346.5 6874.2 6186.8 5568.1 5011.3 4510.2
Silver bream 4346.5 7110.6 6399.6 5759.6 5183.6 4665.3
Spangledemperor 4346.5 6406.9 5766.2 5189.6 4670.6 4203.6
Dumerils's Amberjack 4346.5 8655.9 7790.3 7011.3 6310.2 5679.2
Grouper 4346.5 8270.3 7443.3 6698.9 6029.0 5426.1
Revenue
Cobia 0 15308.6 13777.8 12400.0 11160.0 10044.0
Red porgy 0 20795.4 18715.9 16844.3 15159.9 13643.9
Silver bream 0 17230.0 15507.0 13956.3 12560.6 11304.6
Spangledemperor 0 15532.2 13979.0 12581.1 11323.0 10190.7
Dumerils's Amberjack 0 26303.4 23673.1 21305.8 19175.2 17257.7
Grouper 0 15057.2 13551.5 12196.3 10976.7 9879.0
Net cashflow
Cobia -4346.5 7914.3 7122.9 6410.6 5769.5 5192.6
Red porgy -4346.5 13921.2 12529.1 11276.2 10148.6 9133.7
Silver bream -4346.5 10119.3 9107.4 8196.7 7377.0 6639.3
Spangledemperor -4346.5 9125.3 8212.8 7391.5 6652.3 5987.1
Dumerils's Amberjack -4346.5 17647.5 15882.7 14294.5 12865.0 11578.5
Grouper -4346.5 6786.9 6108.2 5497.4 4947.7 4452.9
NPV
Cobia 28063.4
Red porgy 52662.2
Silver bream 37093.2
Spangledemperor 33022.5
Dumerils's Amberjack 67921.6
Grouper 23446.6

5.6.5 Sensitivity

Overall profitability is clearly sensitiveto relatively small changesin values of certain

costs and selling price. Here, market prices of fishes, feed and fingerling prices, and

survival rateswere used to test the sensitivity of profit.

The break even price of cobia, red porgy, silver bream, spangledemperor,Dumerils's

Amberjack and grouper were 87.5,75.3,103.1,94.6,110.5 and 141.8 NT$/kg,

177
accounting for 54.7,37.7,46.9,47.3,36.8 and 61.7% of the current market prices,

respectively. The specieswith the highest profit sensitivity to market prices is

grouper,where profit increased131% when market price increased50%. This was

followed by cobia, profit from which increasedby 110% when market price increased

50%. The lowest candidate is Dumerils's Amberjack, where profit increased 79%

when market price increased50%. When prices decreased50%, the profits of grouper

and cobia becamenegative (Table 5.19).

Table 5.19 Sensitivity of profitability in changing market price of products.

Unit: Million NT$
Increase50% Increase25% Decrease25% Decrease50%
Profit BCR Profit BCR Profit BCR Profit BCR
Cobia 16.21 1.74 11.96 1.29 3.45 0.37 -0.80 -0.09
(110) (55) (-55) (-110)
Red porgy 25.96 2.98 20.18 2.32 8.63 0.99 2.85 0.33
(80) (40) -40 (-80)
Silver bream 19.74 2.20 14.96 1.67 5.38 0.60 0.60 0.07
(94) (47) (-47) (-94)
Spangledemperor 17.73 2.17 13.41 1.64 4.78 0.59 0.47 0.06
(95) 47 -47 (-95)
Dumerils's Amberjack 33.08 3.07 25.77 2.39 11.16 1.04 3.85 0.36
79 (40) (-40 (-79)
Grouper 14.78 1.43 10.60 1.03 2.23 0.22 -1.95 -0.19
131 (63) (-63) -131
The figures in parentheses are the percentages of increased profit.

The highest sensitivity of profits to feed prices is shown with cobia, where profit

increasedby 33.9% when feeds prices decreasedby 50%, followed by grouper, where

profit increasedby 31.8% when feed prices decreasedby 50%. The lowest is

Dumerils's Amberjack, where profit increasedby 13.5% when feed prices decreased

by 50% (Table 5.20).

178
Table 5.20 Sensitivity of profitability in changing feed prices.
Unit: Million NT$
Increase 50% Increase 25% Decrease 25% Decrease 50%
Profit BCR Profit BCR Profit BCR Profit BCR
Cobia 5.07 0.42 6.38 0.60 9.00 1.12 10.32 1.54
(-32.2) -17.3 16.7 (33.9)
Red porgy 11.96 1.07 13.18 1.33 15.60 2.08 16.82 2.67
(-16.9) (-8.5) (8.3) (16.8)
Silver bream 7.63 0.66 8.89 0.87 11.42 1.48 12.68 1.96
(-25.0 (-12.6) 12.3 (24.7)
Spangledemperor 6.97 0.68 8.03 0.87 10.14 1.43 11.20 1.85
(-23.4) (-11.8) (11.4) (23.1)
Dumerils's Amberjack 15.93 1.20 17.19 1.43 19.71 2.07 20.97 2.54
(-13.8 (-6.9) 6.7 13.5
Grouper 4.36 0.35 5.38 0.47 7.43 0.80 8.45 1.02
(-32.0) (-16.1) (15.9) (31.8)
The figures in parentheses are the percentages of increased profit.

The highest sensitivity of profits to fingerling prices occurs with grouper, where profit

increased26.5% when fingerling prices decreased50%, followed by cobia, where

profit increased8.3% when fingerling prices decreased50%. The lowest is red porgy,

where profit increased3.8% when feed prices decreased50% (Table 5.21).

Table 5.21 Sensitivity of profitability to changing price of fingerling.

Unit: Million NT$
Increase 50% Increase 25% Decrease 25% Decrease 50%
Profit BCR ProfitBCR Profit BCR Profit BCR
Cobia 7.07 0.71 7.39 0.77 8.02 0.89 8.34 0.96
(-8.3) (-4.2) (4.2) (8.3)
Red porgy 13.86 1.50 14.13 1.58 14.67 1.74 14.94 1.83
(-3.8) (-1.9 (1.9) (3.8)
Silver bream 9.59 1.00 9.88 1.07 10.46 1.20 10.74 1.285
-5.7 (-2.9) (2.9) (5.7)
Spangledemperor 8.50 0.97 8.80 1.04 9.39 1.19 9.69 1.28
(-6.6) (-3.3) (3.3) (6.6)
Dumerils's Amberjack 17.03 1.40 17.751.55 19.19 1.91 19.91 2.14
(-7.8 (-3.9) (3.9) (7.8)
Grouper 4.72 0.39 5.56 0.50 7.26 0.77 8.11 0.94
-26.4 -13.3 13.3 26.5
The figures in parenthesesare the percentagesof increasedprofit.

179
Different specieshave different averagesurvival rate (Table 5.9). The highest

sensitivity of profits to increasingsurvival rates is grouper, where profit increased

98.6% when survival rate increased50 %, followed by cobia, where profit increased

76.4% when survival rate increased 50%. The lowest is red porgy, where profit

increased 63.4% when survival rate increased 50%. (Table 5.22).

Table 5.22 Sensitivity of profitability in different survival rates.

Unit: Million NT$
Increase50% Increase25% Decrease25% Decrease50%
Profit BCR Profit BCR Profit BCR Profit BCR
Cobia 13.60 1.14 10.65 1.00 4.76 0.60 1.82 0.27
(76.4) (38.2 -38.2) -76.4
Red porgy 23.54 2.12 18.97 1.91 9.84 1.31 5.27 0.84
(63.4) (31.7 (-31.7 (-63.4)
Silver bream 17.22 1.50 13.70 1.34 6.64 0.86 3.11 0.48
(69.4) (34.7) (-34.7) (-69.4)
Spangledemperor 15.62 1.52 12.36 1.34 5.84 0.82 2.58 0.43
(71.7) (35.8) (-35.8) (-71.7)
Dumerils's Amberjack 30.57 2.30 24.52 2.04 12.41 1.30 6.36 0.77
(65.5) (32.8 -32.8) (-65.5
Grouper 12.74 1.03 9.58 0.85 3.25 0.35 0.87 0.01
98.6) 49.3 -49.3 -98.6
The figures in parentheses are the percentages of increased profit.

5.7 Constraints

5.7.1 Environment

Becausethere are strong currents, such as the Kuroshio, and typhoons on the eastern

coast of Taiwan, major sites for cageaquaculturehave been limited to the western

coast.However, this coast still suffers seriousdamagefrom typhoons and monsoons

and this has hampereddevelopment (Beveridge 1996). Apart from the southerncoast,

every coast of Taiwan suffers from monsoonwaves continuously during the winter

(Twu et al 1986), during which season,significant wave heights are estimatedto be

180
3m on average,with periods of around 10 seconds.Taiwan is also prone to large

stormsduring the annual typhoon season,which are particularly acuteon the coastal

fringe. In summer and autumn, almost all coastsare subject to the threat of typhoon-

generated waves, and immense waves exceeding 10 m have occurred quite often. In

1998, typhoon Zeb damaged a number of cage units in Taiwan, including breakage of

anchorropes, broken nets, deformed and destroyedcage frames, fish escapingfrom

nets, and mortality due to overcrowding and friction induced by deformation of the

nets. Usually, the typical lossesfrom typhoon are the damageof cage and fish die

becauseof friction. Whole cageswere broken and all the fish escapedare not

common. Almost every year typhoon will attack Taiwan. However, the level of

damagefrom typhoon dependson the frequency and strength of typhoon, and the

prevention that farmers did.

5.7.2 Diseases

Cageculture is one of the most intensive forms of aquaculture,and as a consequence

fish diseasescan be problematic. Most diseasesaffecting cage aquaculturea are

epizootics, including Benedeniasis(Benedenia ),

spp. sealice (Caligus ),
spp.

Dactylogyrus (Dactylogyrus spp), Ichthyophthirius (Ichthyophthirius spp.),

Trichodina (Tricodina spp.), Myxospora (Myxosporidia spp.) and Epistylis (Vorticella

spp) and ulcer on the skin of groupers.Among these,Benedeniasis,

the ulcer on the

skin of groupers and sealice are the most serious,potentially causing serious

mortality to the extent of only 20% survival rate and increasing the cost in feed, fry

and labour.

Although Benedeniaand sealice can be cured by freshwater baths, they are not

particularly effective for sealice, and this also dependson the tolerance of stock to

181
fresh water. Cagefish farmers usually dip fishes in freshwater for less than 10 minutes

every 10-12 days. The fish can tolerate a fresh water bath for 10minute but Benedenia

spp can only tolerate this for 3 minutes (Chen, personalcommunication, 1999).

However, this work is tedious and laborious. Benedeniasis infections usually occur in

the spring and autumn season,during the months of October, November, February,

March and April. Skin ulcer of groupershappensthroughout the whole year, though

autumn and winter are the most seriousseasons.It is postulated that this diseaseis

causedby bacteria and occurs most often after the transportation of small fish or after

infection by parasites.Small fish that cannot stand strong currents will rub against

cagenets and induce secondarybacterial infection. Sealice infections happenthe

whole year round though the most seriousperiod is summer.During the rainy period

when the salinity is lower and turbidity is higher, sea lice infections become less

serious.

5.7.3 Management

According to Wang et al (1998) and this investigation, the major costs of production

are feeds (about 50%), fingerlings (about 20%) and labour (about 15%). Becauseof

high feed cost, it is difficult to lower the price of cage aquacultureproducts, and

therefore,the market size may be limited. If cheaperformulated feeds can be

developed,this may greatly increasethe potential for development.Having personnel

on hand to make frequent inspectionsfor sign of the fatigue of structuresand nets to

ensurethat the cultivars are performing properly will help operatorsto avoid

catastrophicloss. However, this requires higher expenditurein the cost of labour. For

example, to prevent the mortality from Benedeniasis,

farmers need to spendplenty of

time in fresh water bath, which is crucial in management.In Taiwan, most farms are

182
run by families with limited additional staffing, and they did not spend so much

money in labour. As a result a number of serious losseshave occurred. A better

can
management improve the survival rate to 70%.

5.8 Discussion

The successof a cage aquaculturefarming facility dependson the combination and

fine tuning of cages,nets, and moorings utilized in responseto the local site-

conditions (Lisac . 1996).Masser (2000) points out that the superior characteristicsof

candidatesfor aquacultureinclude marketability, easeof breeding, rapid growth,

toleranceof poor water quality, diseaseresistance,tolerance of crowding and

handling, easeof harvest,easyto feeding formulated diets and omnivore. A number

of specific changesmay be consideredto improve opportunities for development:

(a) Improvement of the cage structurefor poor weather conditions

The expansion of cage aquaculture is restricted by the specific environmental

conditions. Constant stressfrom waves and currents,with intermittent exposureto

storms that greatly increasestresslevels and result in rapid fatigue of structures

and nets, requiring the use of strongermaterials and advancedengineering. In

addition, currents can causedeformation of net pens, which may reduce

productive capacity For example, submersiblecagescould be less impacted by

waves, permit avoidanceof surface storms without relocating the system,and be

less susceptibleto biofouling and corrosion than their floating counterparts.

(b) Feed improvement

183
 Feed is the main cost of cageculture, and technical problems in feed manufacture

are critical in its development.Currently, fish farmers use trash fish or formulated

diets to feed fish and the protein content of the formulated diets is as high as40%.

Reduction of the protein content would lower the cost of feed and potentially

widen the range of raw materialsbut there is still insufficient research.

(c) Industrialise production system

In Norway, the averageannual production of cage aquaculturesalmon reached

350 t per FTE* personemployed (Taiwan Fisheries Bureau, 1997). In contrast,the

averageannualproduction in cage aquaculturein Taiwan is only about 30 t per

FTE personemployed. Thus, while this is partly a function of high labor cost in

Norway, modernisation of the production systemcould increasecompetitive

ability and reduceproduction cost. For example, feeding is still manual from a

raft, when it is now possible to use an automatic feed delivery systemwhich also

improves feeding efficiency. The developmentof innovative technology to allow

appropriatelevels of feeding, long distancemonitoring and communication, and

carefully planned responsesto emergencysituations are important. Craft or barges

designedfor efficient servicing of the cages,harvesting fish and dealing with

routine tasks associatedwith the operation of cage farming should also be

investigated.However, all of thesemust demonstratetheir potential for cost

effective performancewith the wider range of speciesinvolved in this sector.

Full time equivalent

184
(d) Marketing promotion

Marketing is one of the main problems in developing cage aquaculture,and the

developmentand promotion of marketsin Taiwan, Japan,China and other

countries is a critical issue. If the target market is for restaurants,the preferred size

is usually 1-2 kg. Grouper is suitable to promote for this market becauseof its

harvesting size. Bigger fish, such as cobia, which is harvestedat 7-8 kg, may be

promoted for the raw fish and fillet market. To profit from aquaculture,adequate

information about potential marketsis required, including domestic and foreign

production, exports and imports, product forms (such as live, fresh, frozen, fillet

etc.), and processingand distribution cost.

Cageculture may releasesomepressureon the demandfor land and water. Seacage

culture could also avoid impacting the more environmentally sensitive coastalzone

and there might be some benefit in term of improved product quality (Stickney 1997).

In Taiwan, the developmentof seacageaquaculturehas some potential, as it is

surroundedby open ocean,there are many candidatespeciesfor seacage culture, and

farmers have experiencein other sectorsof aquaculture.Technology from Norway is

being imported, and governmentconsiderscage culture to be an important step in the

future developmentof aquaculture.

It is common that the cage farmers in Taiwan rear two, three, or even more, speciesto

reducethe risks in monoculture, aiming for both more speciesand more markets.

Diversified farms can cope better with market and climate fluctuation as they can

contribute to a smootherharvesting pattern, and consequentlycash flow, throughout

the year. Capacity can also be utilized more evenly. As overheadcosts would be

185
higher for sea cage aquaculture and there would be a limited coastal environment of

suitablequality (depth, current, and water quality etc.,) the production of luxury

species,rather than high volume, low value ones, would undoubtedly be given highest

priority.

The establishmentof a licensing or permit processcould provide orderly development

of the industry, and a sound legislative and leasing program is imperative. The legal

rights for cage culture were described in 5.22. However, the following features are

likely to be important.

1. Both the seabedand the water column should be included in the lease.It can avoid

the conflicts between farmers and other activities, such as diving and fishing.

2. The period of leasing should be long enoughfor farmers (e.g. longer than pay back

it
period, needsat least 2 years in this study) to start and establish a viable culture

operation.Within this period, farmers can make profits and would be preparedto

invest on system.

3. The leasedareasshould have adequatelegal protection againstdegradedwater

quality, theft of culture organism and facilities, and trespassing(DeVoe and Mount

1989), as it is essentialto reduceoverall risks in the production system.In Taiwan,

although the legal protections exist, sometimesit is difficult to find the criminals.

Somecoastalregions in Taiwan are increasingly polluted (Hsiao, 1994a).Cage

culture located in those nearshorewaters can be subjectedto such pollutants. The

capacity of a farming site to accept the waste from fish farm can also be exceededif

the intensity of culture becomesexcessive,though comparedwith municipal,

186
industrial, agricultural and nonagricultural river inputs, the inputs of phosphorusand

nitrogen from aquacultureis commonly very small (Stickney 1997).

Establishing systems of a large size with a relatively low biomass per unit area can

also help ensure the proper dispersal of released nutrient. To prevent excessive growth

of cageculture within a given location, careful attention to the sites and density of

cageculture within a given locale will ensurethat the sustainability of the activity is

maintained.If the structureof cagescan be improved to resist moderatecurrents,

cagescan be set up in areaswith somecurrents which will be useful to prevent

sedimentation,as they may assistin cleaning the seabed.Nutrient releasedfrom fish

farms could also be convertedto biomassin polyculture facilities. Thus, developing

seaweedculture in associationwith cageculture may be an option for recycling

nutrients releasedfrom cagesand incorporatedinto seaweed(Stickney 1997).

The dispersionof nutrients releasedand the uneatenfeeds from cagefarms might

increaselocal floral and faunal productions and attract wild fish in offshore areas.

Sport and commercial fishing could be enhancedbecausethe structure associatedwith

offshore mariculture facilities can serve as fish attracting devices and increased

nutrients levels can promote overall local productivity. Modest increasein offshore

nutrient levels might be actually a benefit (Stickney 1997). It may consequentlybe

appropriatefor governmentto assistfish farmers to set up some facilities for

recreational fisheries.

According to ADCP (1983), feed represents40-60% of the total operating costs in

intensive aquacultureand is very similar to the results of this research(Table 5.12 and

187
Table 5.13). It is therefore, only feasible if the fish being cultured can fetch a

sufficiently high to
price generatea profit when harvested(Beveridge, 1996), as is the

case in Taiwan. The sensitivity of profit to market prices and feed prices were

particular notable for grouper and cobia (Table 5.19 and Table 5.20). However, for an

ideal profitability target, it was suggested (ADCP 1985) that feed costs should not

exceed20% of farm gate value of the fish. In this survey, most specieshad feed costs

of more than 50%, suggestingthat reducing feed costs relative to market price is an

important issue.This investigation found that although Dumerils's Amberjack (Seriola

dumerili) and red porgy (Pagrus major) can make higher profits than other species,

fish farmers still have great expectation for cobia (Rachycentron canadus). Because

Taiwan's market is not big enough, this is based primarily on the hope that the

Japanese market can be developed, and cobia can become a candidate for sashimi

(raw fish).

The survey revealedthat mortality ratesin cageculture was very high (Table 5.10).

Whether this is becauseof the quality of feed, fingerlings or other factors is still not

clear. The survival rates definitely will influence the profit of cage culture. The

to
sensitivity of profits survival rates was higher in grouper and cobia (Table 5.22).

Marketing channelsfor seacage speciesare quite similar to those for other fisheries

products in Taiwan. Most products are sold to wholesalers and re-sold in auction

to
markets retailers. As wholesalerscollect the products and commonly control prices,

the establishmentof production and marketing groups might help farmers to get

higher profits. This will be discussedfurther in Chapter 6. Some farmers have

attemptedto open up Japanese

markets, but have still confronted difficulties. Another

188
threat is fisheries products from China, which competeat a number of levels. Some

fishermen do not catch fish, but trade fish, bought from Chinesefishermen in the open

oceanand re-sold to Taiwan. This has lowered prices of some marine fish products,

and threatened the profitability and development of sea cage culture.

Apart from efficiency and market issues, cage culture faces objections concerning

issuessuch as visual pollution; potential impact on non-targetbacteria from the use of

antibiotics in fish feed; infection of wild fish with diseasescarried by cultured fish;

pollution of the water column and destruction of the benthic community from waste

feed and fecal deposition.There are also concernsfor interference with navigation and

removal of accessto traditional commercial fishing. Successfulcage culture should

emphasizethe developmentof appropriatesite-specific systemsand public policy

strategiesthat integrate the various disciplines, including social, physical and

biological sciences.With improvementsin structure design, materials and farm

operation, cageculture is expectedto expand in the future. However, this will require

a critical coalition or team building approach(Jensen1996), and the important lessons

for investment can be learned from other countries such as Norway, Scotland, Canada

or Chile.

189
 Chapter 6

Sustainability

6.1 Introduction

Although the final goals of environmental protection and developmentof aquaculture

are not necessarily in conflict but may indeed be the same, namely to improve the

human quality of life or welfare for present and future generations, aquaculture, like

all farmed food production may have a large effect on the environment. Those effects

can be negative, such as reduction of the abundanceand diversity of wildlife; change

in soil, water and landscapequality and occupation and fragmentation of former

natural habitats (Pullin 1993).

The viability of aquaculturecan be viewed from two levels: farm and society.At the

level of the farm, sustainability dependson its productivity and on market factors

reflected in costs for production and revenuesfrom its products. However, the social

costs from aquaculturecannot be assessedin isolation from the rest of the economy

(Shangand Tisdell 1997). In Taiwan, the profits from aquaculturein the past two

decadeshave attractednumbersof new entrants.However, a range of potentially

negative environmental, economic and social factors may pose increasing challenges

in the future. It is important, therefore, to consider the issueswhich may allow the

aquaculturesector to remain viable and effective, i. e. sustainable,in delivering a

range of economic and social benefits, and also to consider those factors which may

threatenits future potential and would needto be effectively addressedto enablethe

sectorto prosper.
190
Sustainable development is an increasingly widely used concept in planning, thought

subject to different definitions. The UN World Commission on Environment and

Development defined sustainable development as:

"Meeting current needs without compromising the ability of future generations

to meet theirs" (Anon 2000).

The FAO (1991) has defined sustainabledevelopmentin a similar manner,extending

it more specifically to be:

"The management and conservation of the natural resource base, and the

orientation of technological and institutional change in such a manner as to

ensure the attainment and continued satisfaction of human needs for present

and future generations. Such sustainable development conserves land, water,

plant and animal genetic resources, is environmentally non-degrading,
technically appropriate, economically viable and socially acceptable".
Aquaculture is sustainableonly when it can cope with and recover from stressand

maintain or enhanceits capability and assetsboth now and in the future, while not

undermining the natural resourcebase.Unsustainablesystemswill depleteor run

down capital (financial or natural resource),leaving less for future generations.The

sustainabledevelopmentof aquaculturecan be judged in economic, environmental

and social systems,but invariably, the result is a seriesof trade-offs between all three

of thesecomponents,as summarizedin Fig. 6.1.

This implies that a sustainableaquatic farm must be economically viable, ecologically

sound and socially It

acceptable. would not specifically strive to maximize any single

result, but rather to achieve a long-term balance among outcomes.The processof

trade-offs among goals must also be adaptive,and the priority among goals should be

weighted by the three systems(Barbier 1987).The trade-off includes the tension

191
betweenobtaining better income and environmental sustainability, the tension

betweenmaximizing production in the short term and guarding againstvulnerability

to external shocksin the long term and the tension of conflict between fish farmers

and others.

The sustainability of aquaculturecan be describedas depending on two setsof

factors: intrinsic and extrinsic. Intrinsic factors are related to adequateon-farm

planning and such

management, as water quality, culture techniques,location and

operation of facilities, seedsupply, speciescharacteristicsand availability of artificial

and natural feed, etc. Extrinsic factors refer to off-farm factors, such as national

policy, natural hazard,pollution, the market, sociocultural conditions and legislative

control (Chua 1997).The following sectionsattempt to examine theseissuesas they

affect the aquaculturesector in Taiwan, and in particular, the focus areasdefined in

previous chapters.

Sustainableaquaculturesystem

Economically viable Environmentally sound I Socially acceptable

Fig.6.1 Sustainableaquaculturesystem (modified from Shangand Tisdell 1997).

192
6.2 Social and economic issues

6.2.1 Introduction

Most aquaculture in Taiwan is carried out intensively. Usually, intensive aquaculture

is not previously driven by food shortage objectives, but by market price (Folke

1997). Intensive fish farming relies on high stocking rates and formulated diets to

achievehigh yields of marketablefish within purpose-built rearing facilities by means

of close supervisionover the entire production cycle (Shepherdand Bromage 1992).

The logic of intensification is not related to a biological concept of efficiency, but to

resulting net gain of commercial benefits that can be attained in certain economic

conditions. If the revenueis lower than the cost, the industry will not be feasible.The

economic factors of aquacultureas outlined in earlier chaptersare shown in Fig 6.2.

Clearly, feasibility dependson the difference betweencost and revenueand hencethe

potential to decrease
cost and increaserevenuei. e. improve financial performanceis

critical. These issuesare discussedin the following sections.

193
 Feed

Fingerling

Labour

Production in
Investment Kg
cost Production
Revenue
cost
Fuel and
electricity Market price in
$/Kg

Depreciation
Net return on
Interest investment

Miscellaneous

Fig 6.2 Economic factors of aquaculture.(Modified from Klemetson, and Rogers

1985).

6.2.2 Improving financial performance

Strategiesfor improving financial performanceof the aquacultureindustry have been

discussedin specific casesin earlier chaptersand cover both supply and demand

issues.The attributes or variables that influence financial and economic performance,

and the nature of the links betweenthose attributes and end performanceis illustrated

schematically in Fig. 6.3. (Sherer 1982). From Figure 6.3, it can be seenthat the

situation of supply, demand and market structure will affect the conduct and

performanceof the sector as a whole. In attaining economiesof scale, increased

supplies from large-scaleaquacultureenterprisesmay lower prices and in turn further

increasethe pressuresfor efficiency.

194
 Basic condition

Supply Demand
Raw materials Price elasticity
Technology Substitutes
Unionization Rate of growth
Product durability Cyclical and seasonal
Value/ weight characteristics
Business attitude Purchasing method
___ Public policies Marketing type

Market structure

Number of sellersand buyers

Product differentiation
Barriers to entry
Cost structures
L--------- Vertical integration
Level of aggregation

Conduct

Pricing behavior
Product strategyand advertising
Research and innovation
Plant investment
Legal tactics -----------

Performance

Production and allocative efficiency

Progress
Employment
Equity

Fig 6.3A model of industrial organization analysis (the dotted lines indicate feedback
effects) (adaptedfrom Sherer 1982).

195
However, limited land resourcesand expensivelabour costs may favour family labor,

and without substantialinvestment in it

mechanisation, may be difficult to develop

large-scaleaquacultureenterprisesin Taiwan. Limited land resourcesforce producers

to apply intensive aquaculture,though, this increasesrisks (such as D. O. depletion or

water quality deterioration), and needshigh levels of experienceand close

supervision. Investment may also be constrained by the conditions in which wealthy

investors do not want to take the risk to invest while experienced farmers did not have

enough capital to expand their farm, nor does the risk allows them to obtain adequate

loans or insurance.This makesthe condition of aquaculturein Taiwan different from

that in Norway, where good accessto capital and high labour costs have resulted in a

highly mechanisedindustry.

National tradition has also restricted the developmentof larger-scaleenterprise:in

many sectors,there is a joke that if you throw a stone in the street of Taipei, you are

likely to hit a chairman of a board. There is one company for every 18 people in

Taiwan (the highest density in the world) (Anon 1998), people preferring not to work

for others "better the head of a chicken than the tail of an ox", according to an old

Chinesesaying, and most Taiwanesefeel most comfortable in a company financed

and run by their own family.

Small to medium-sizedenterprisestherefore make up 98.5% of Taiwan's companies,

75-80% of all employment and 47% of the total economy (Anon 1998) and about

60% of aquatic farms in Taiwan are less than a hectarein area(Liao et al 1995b).

Governmenteconomistsrefer to corporateTaiwan as an "army of ants". There is,

therefore, a bias againstlarge-scaleaquacultureenterprise.

196
Industry structure

In early years,the aquaculturesectorwas often characterizedby disorganized

gaps
marketing, with seasonal of supply. Since most aquatic farms are small and

independent, they have a disadvantage in selling their products, where prices are

decided by market condition and not by farmers. As with agriculture, markets are

completely competitive, and small producers were easily open to major price falls

during period of high availability. To obtain higher profits, farmers can organize

production and marketing groups, combining together to negotiate prices in

purchasing inputs and selling their product, taking the advantage of larger scale. An

equally important function of these bodies would be to establish and maintain

appropriate product quality standards to which members must adhere. This would

it
make easier to undertake sales promotion for the industries as a whole.

By "collusion", small farmers can unite and set up groups and agreementsto control

production to optimize profits, in effect developing cartels, of which 3 kinds can be

defined (Wu 1998).

Price cartels: most companiesagreein price levels, or define the lowest price
-.

to avoid price competition.

Production cartels: companiesagreeto limit production levels to control the

-.

price in the market.

Territory cartels: the goal is to separatedifferent areasfor different companies

-.

to avoid competition.

In general, a production cartel might be a more suitable approachas, price cartels are

illegal in Taiwan, and Taiwan is not big enoughto use a territory cartel. Moreover,

197
although some species have relatively high price elasticity of demand, the E. O. D. of

most fishery products is less than 1 in Taiwan (Wu, 1998), producers can obtain larger

profits by using production levels to control the price rather than using price to control

demand.However, if the production and marketing groups combine to control

production levels, they must confront the problem of quota; as each member would

compete for as high a quota as possible, and it may therefore be difficult to attain

Even
agreement. if a quota is agreed,it is difficult to guaranteethat some farmers,

`free-riders'- will not produceextra fish abovequota. If a lot of farmers "cheat" and

produce more than their quotas,the agreementwill collapse.Therefore, sound rules

with severepenaltiesfor infringement, a good marketing systemand reasonable

production levels are imperative. To the

assess potential for such action, in 1997,Wu

(1998) sent 636 questionnairesto fish farmers with 168 responses(26.3%). The major

difficulties of production and marketing groups for aquaculturein Taiwan were

summarized in Table 6.1, which suggests that considerable issues are to be overcome.

The potential organization of a production and marketing group is illustrated in

Fig.6.4, in which the cooperationof government,researchinstitutes and private

sectorsmight all be important in development.

Table. 6.1 The major difficulties of production and marketing groups for aquaculture
in Taiwan.
Major difficulties Percentage
1. Lack of financial support 25%
2. Members are not strongly combined together 17%
3. The knowledge of aquaculturetechnology is insufficient 15%
4. Difficult to control production amount 13%
5. Difficult to predict natural disaster 11%
6. Have not sound system of production and marketing 9%
7. Deteriorated water quality causes fish disease 6%
8. Lack of instruments for aquaculture 4%
Total 100%
Source:Wu, 1998.

198
 Agricultural consultantcommittee

Universities II Fishery Administration Fisheries Research

Institute

Local government

FishermenAssociation

Production and Marketing groups

Fig. 6.4 The organization of aquaculture production and marketing groups in Taiwan.

Marketing development

According to Porter (1980), the stateof competition dependson five basic forces;

rivalry among existing firms, threat of new entrants,suppliers, buyers and substitutes.

In coping with the five competitive forces, Porter categoriessuccessfulstrategiesas

involving one or more of three elements:

" Cost leadership

" Product differentiation

is Focusing on a particular market segment

In cost leadership,the aim is to reducecosts. Here, production efficiency and

economiesof scale are required. Economies of scalerefer to reductions in unit costs

199
of a product as the absolutevolume per period increases.
Such economiesof scalecan

also deter entry by forcing the entrant to come in at large scale and risk strong

reaction from existing firms or come in at a small scaleand accept a cost

disadvantage.If a large portion of the total supply is purchasedby a few given buyers,

this will raise the buyers' power, enabling them to force down prices and/or

bargaining for higher quality or more service.In contrast, when dealing with buyers,

greater power is associated with a large market share, and large volume makes it

possible to supply a greater number of market outlets on a cost- effective basis. In

aquaculture, economies of scale can be created by integration. This can be

implementedthrough a farm being bought out or merged,or by setting up the

production and marketing groups.

However, if production economiesof scaleare not large enough, or if the factors of

site capacity and legal constraintsrestrict the scaleto expand,the industry might be

fragmented.In Taiwan, e.g. traditional cage culture is a fragmentedindustry and

therefore faces a marketing disadvantage.As a consequence,the developmentof

small grower groups to market products collectively may be a practical option to deal

with this problem. Thus, if large modern cage farmers or grouped farmer associations

can control the production levels and henceinfluence price, they would have a

potential marketing advantage,

and so the structure of businesssectormay

increasingly be dominatedby them.

The secondgeneric strategyis to differentiate the products. In aquaculture,product

differentiation can be carried out by differing the culture species,harvesting season,

fish size, offering a variety of products and/or by setting up brand names.In a

200
fragmentedindustry, producerscanjoin together to becomean associationand set up

their own brand names.Differentiation can reduce shorter-term competitive rivalry

becauseof brand loyalty by customersand their resulting lower sensitivity to price.

Differentiation can yield higher margins, and can mitigate buyer power, since buyers

lack comparable alternatives, and are thereby less price sensitive. However, achieving

differentiation may reducethe market share,and is often incompatible with high

market share.

The final generic strategyis to focus on a particular buyer group. Unlike the low cost

and differentiation strategies,aimed at achieving objectives for the whole industry,

this strategybuilds around serving a particular target basedclearly on the premise that

firms are able to identify and service their narrow target most effectively. As a result,

the firm achievesbenefits of differentiation either by better meeting the needsof the

particular target, or by lower costs in serving this target, or both. In aquaculture,

applying this strategy requires a clear demographicinvestigation and also implies

some limitations on the overall market shareachievable.As such, this strategy it may

be
not very suitable for aquaculture.

Potential international competition

The factors influencing the potential for international competition may include those

of resource,technology and administration. The resourcescan be classified as natural

and non-natural. Natural resourcesfactors include water, land and their qualitative

features.Such factors are crucial in the growth and survival of aquatic animals and

will influence the carrying capacity. One of the reasonsthat the tiger prawn (Penaeus

monodon) industry collapsed in Taiwan was becauseof the lack of quantity and

201
quality of water. Non-natural resourcesrefer to factors such as capital and marketing

organisation.A farm needsa loan to cover not only construction costs but also

operatingcosts for the first few yearsuntil the first or secondharvest generatesa

positive cash flow. Therefore, capital is usually a problem during the start-up phaseof

a new company. In the competition of commercial sales,it is difficult to distribute

products far from the point of production without a strong salesorganisation.

Another important factor for international competition is technology. This is the basis

of any industry, and new technology must be developed as conditions changeif any

industry is to remain competitive (Ackefors, 1994). Supply of better seeds,feeds,

feeders,cages,therapeuticand prophylactic drugs, aeration devices and other

equipment is important for the successof aquaculture.Competent managementis also

critical to the operation of an aquaculture farm. A successful farm manager must be

able to recognise and correct biological and water chemical problems, and make

effective emergencyfield repairs to complex equipment

The third important factor is administration. Legislation must be favourable to

farmers. If the political will is very strong in opposing aquaculture,little significant

developmentcan be expectedAquaculture is dependenton institutional support,

including universities. The transfer of new technologiesfrom governmentalresearch

to the private industry and diagnosis of diseasesare also important.

Comparedto other Asia countries (such as China, the Philippines, Thailand and

Malysia etc.) Taiwan has a disadvantageof limited natural resourcesand higher

wages.In the longer term, if those countries develop better technology and obtain

202
governmentalsupport, they may be able to start supplying the market at lower prices

and Taiwan may lose long-term competitive advantage.For example, China has

replacedTaiwan as the biggest eel exporting country in the Japanesemarket, and

Thailand has overtaken Taiwan to become the biggest producer of tiger prawn

(Penaeus monodon) in Asia, and even the biggest producer in the world (FAO. 2001).

In the longer term, if Taiwan loses the advantage of cost leadership, focusing on

product differentiation might be another way forward in international competition.

Developing market demand

Though traditionally thought of as a simple post production issue, marketing is a

much broaderconcept underlying the entire basis in which production is established

and developed.The link betweenproducersand consumers,it consists of individual

and organizational activities that facilitate and expedite satisfying exchange

relationships in a dynamic environment through the creation, distribution, promotion

and pricing of goods (Dibb et al. 1997). In the processof getting the products from

producersto consumers,the product passesfrom one owner to anotherthrough a

network of marketing channels,from the place of production to the point of final

consumption.

203
 Buying power II Fish availability

Demand for fish

Development of knowledge to meet demand

Fish oriented Market oriented

. Reproduction "Consumption behavior
"Nutrition 'Processing development
. Health control
. Bio-Engineering

Marketing
Aquaculture Commercialization

Fish Supply

Fig 6.5 Featuresof a demandpull aquaculturemarket (Modified from Huisman,

1986).

A greaterimpetus for the developmentof aquaculturewill be market demandfor fish

and local pressuresfor new forms of livelihood and enterprise.There are two kinds of

market of aquaculturein Taiwan, describedas demand-pull and supply-push

respectively.The developmentof a demand-pull market (Fig 6.5) arisesfrom demand

from consumers(such as milkfish and eel etc.). By contrast, the developmentof a

supply push market (such as cage aquaculture)is not specifically becauseof demand

from customers,but becauseof increasing supply from producers (e.g. through a new

speciesand/or technology) (Fig 6.6). Thus in cage aquacultureproduct (fish)

expansion as a result of the collective consequences

of decisions to expand by

individual businessesmust createthe market for development.

204
 Limited Water and Land Resource

Development of knowledge of cage culture

Fish oriented
Market oriented
"Reproduction
*Consumption Behavior
*Nutrition
*ProcessingDevelopment
"Health control
. Bio-Engineering

Aquaculture Commercilization II Marketing

Fish Supply
Fig. 6.6 Features of a supply push aquaculture market (Modified from Huisman,

1986).

Potential marketing strategy for aquaculture

With regard to the main marketing strategiesin salmon farming, Shaw (1997)

categorizedfive major They

elements. are market targets,products, pricing, market

channelsand promotion.

Firstly, the market targets of production must be set up. Depending on whether they

are international (such as shrimp or eel) or domestic (such as milkfish) or both, it must

be known which market sectorsand which countries are emphasized.

Secondly,product strategiesmust be set up. This would include harvest timing,

harvest sizes and quality control. In principle, it would be desirableto be able to plan

supplies in line with forecast demand,although this has not always been achievedin

205
practice. When considering harvest sizes, preferred market sizes should be known.

Farmers with little experience of marketing can also underestimate the importance of

quality to consumers, being usually more concerned with finding a sale for the fish

rather than with how consumption ultimately occurs. However, a farmer who

producesgood quality fish will be interestedin consumersbeing satisfied with the

products. Quality factors of particular importance are accuratesize, grading,

appearance,lack of any off-flavours, professional packing and freshness.Service

issuesare also increasingly critical - physical quality alone may be insufficient. If

processorscan develop new popular products, it can also help the development of the

industry significantly. The attractivenessand convenienceof the packaging materials

are influential in selling the dressedfish, especially in supermarkets.Neat and

appealingcontainersincreasethe sale of fish in retail markets. If farmers can join

together and set up trade associations,education programmeson quality for members

can be incorporated into the activities of trade association(such as the eel production

and marketing associationin Taiwan).

The third element is pricing. In Taiwan, traditional aquacultureis fragmentedand

prices are usually decided by the market, i. e. producersare price takers. However, if

the fish farmers canjoin together and regulate the amount and quality of product (i. e.

under oligopoly condition), they might enjoy some degreeof market power and be

able to influence the price of their products.This can be further extendedwhere

consumersneedsare better targeted.

206
Mudie (1994) describes a typical pricing approach as:

P=UVC+(F/X)+(rK/X)

Where P is selling price,

UVC is unit variable cost,

F is fixed cost,

X is standard unit volume,

r is profit rate desired and

K is the capital employed.

But when targeting a price, demandelasticity must also be considered.It may be that

demandfor a certain type of fish or its product is highly sensitive to relatively small

changesin price. This price elasticity of demandi. e. the units of demandchangeas a

result of the units of price change,can be an important consideration for fish farmers

in developed markets who may be tempted to produce for premium prices before

trying to establishthe new potential market. Intensive farming is characterizedby low

fixed costsbecauseof high productivity per area,but high variable cost mainly for

feeds and water quality maintenance.If market prices are favorable, intensive farming

remains profitable. Once prices drop, so does profitability becauseof the high

production or variable cost (Primavera 1991). It may therefore, be more suitable to

choosehigh price speciesto cultivate in Taiwan due to that high variable cost

associatedwith the intensive forms of aquaculturewhich may be required due to other

factors.

The fourth strategyis to shortenthe marketing channel.This can reducethe margins

of the channel, maintain sufficient first-sale price, reduce the retail price, increase

207
consumer demandand benefit the producersand consumers.Usually, the highest

profit margins achievedby fish-farmers result from supplying live fish, though this

does not guaranteea high profit in absoluteterms as the quantity sold, and the cost of

be Although
selling must considered. some supermarkets(such as Welcome,

Carrefour and Macro etc.) can apply economiesof scale to fish purchasing and

handling, selling fish to supermarketsis relatively unpopular among fish farmers in

Taiwan, as payment terms are usually delayed, with buyers routinely paying by check

or asking for credit of 1-2 months. Small farm owners need cash for operating costs

and may find it difficult to work with the credit terms. There is no simple solution for

this, but good local representationand contracts are essentialfor mitigating this

problem. This can be achievede.g. by fish farming trade associationsetting up an

extensive network and selling products directly to retailers or supermarkets,

or by fish

farmers making contractsdirectly with retailers or supermarkets.

The final strategyis product promotion. Becausefish are common commodities, the

role played by product differentiation may be small and there will be a lot of free

riders who take the advantagefrom advertising by others.To solve this problem, large

farmers can set up their own brandsand small farmers can set up brandsin the names

of trade Not
associations. only do the trade associationspromote the products, they

can control its quality as for generic advertising to be successfulthe actual quality of

the products must match the image and expectationcreated.

6.3 Environmental factors

6.3.1 Introduction

Sustainingfood supply requires protecting the environment as the basis for

208
production. Environmental sustainability is achievedwhen the productivity of a

natural resourceis conservedor enhancedfor use by future generations(Chua 1997).

Increasing production through intensifying culture createsthe risk of taxing the

carrying capacity of the environment. Concentrating many animals in a small space

createshigh oxygen demands,increasesthe concentration of waste products (such as

nitrate and phosphateetc.) and increasesthe transmission of diseases.The application

of various forms of chemical or antibiotic treatment against diseasescan also pose

public health problems.As aquaculturedevelops, it is possible that someexternal

effects will happen.When evaluating the feasibility of aquaculture,the social costs of

waste treatment,pollution prevention and taxes on discharging effluents are usually

neglected.However, environmental impact should be consideredand should be fully

integrated into developmentdiscussion.

6.3.2 Key issues

Eutrophication or hypernutrification

Becauseland is limited in Taiwan, almost all its aquacultureis intensive, in which the

aquatic animals gain energy from allochthonous (external) rather than autochthonous

(internal) sources.That meansthat extra feed is imperative. However, feed lossesare

inevitable. Intensive farms, including cage systemshave a high production per unit

areaand a correspondingly large amount of particulate organic waste,as well as

soluble-inorganic excretory waste. Fish farms/cagesproduce orthophosphateand

nitrogenousnutrients, and high concentrationshave been observedin adjacentsurface

waters (Hansen et al., 2001). Since the raw material for fabricating the feed pellets

often originates from other water bodies, the result is also a net addition of nutrients to

the receiving environments (Folke 1992).Depending on the quantity and composition

209
of the effluents and the susceptibility of receiving environment, these emissions may

have varying and sometimessevere,ecological impacts. Nutrient releasesfrom

intensive farms representa net addition to environmental loading. Excessive nutrient

enrichment or eutrophication can causeimpact in coastal areasand may affect

socioeconomicactivities that are dependenton the quality of these areas(Folke and

Kautsky, 1989), including negative feedbackon the aquacultureactivity itself.

However farmers might arguethat the severity of pollution causedby aquacultureis

far less seriousthan causedby other industries and domestic sewagewastes.Further

researchis neededto clarify who is the most serious sourceof pollution.

Chemical waste

A wide range of chemicals is usedby the aquacultureindustry. Most of these are

biocides, usedto control bacterial, fungal, protozoan and other diseases(Beveridge et

al., 1994). If thesechemicals and materials such as those used for anti-fouling

treatment of cagesare releasedoutside their targetedpurpose,there may be adverse

effect on the local environment.The high incidence of infections followed by

medication through feeds and water baths has been a burden to the aquaculture

industry. Feed basedtreatmentsare used in land-basedfish farms and seacages.

However, chemical treatmentsusing water baths are usually used in land-basedfarm

rather than in seacagesbecauseof the effect of dilution in the open areaand because

the techniquehas not been well developedin seacagesyet. In Taiwan, currently, there

is a lack of transparencyin the industry regarding the use of chemical inputs and there

has been little public debateon theseregards.The absenceof a monitoring

mechanismcomplicates the introduction of industry standards.This has lead to the

feeling that fish from aquacultureare not as clean and healthy as their wild

210
counterpart, and to adverseeffects for aquaculture in attempting to develop and/or

improve markets.However, somefish farmers may arguethat the pollution from

industry and domestic sewagemight be more seriousthan from aquaculture(Stickney

1997). Therefore, the severity of the pollution from aquacultureand other industry

need further investigation.

6.3.3 Key biological impacts

Changes in aquatic fauna and flora

The establishmentof aquacultureusually increaseshuman activity in the immediate

vicinity, which in turn can have impact on wildlife, especially in remote areas,

through disturbanceof breeding or feeding. Increasedconflicts between man and

wildlife, especially with piscivorous (fish-eating) animals are occurring in Taiwan.

The end result may be the death of animals, either deliberately (shooting, trapping) or

accidentally (entanglement),or the loss of stock for fish farmers. For example, a

number of aquaculturedevelopmentshad occupied the habitat of the spoonbill

(Platalea minor) - an endangeredspecies,resulting in conflicts betweenfish farmers

and animal protection groups.Finally, the government had to negotiate between these

intereststo set up special protection areasto solve the conflicts.

High densitiesof farmed fish and food also attract predatorsand scavengers,which

may in turn displace local species(Beveridge et al., 1993). Releasednutrient from

fish farms can also changethe composition of speciesflora. In 1998, there was a

serious `red tide crisis' in Hongkong, which was causedby aquacultureand nearly

decimatedall investmentsof the fish farmers in a short period of 2 days (Lai and

Lam, 1999).The precipitation of uneatenfeed and faecal materials can also cause

211
severe disturbance of the macrobenthic community (Brown et al., 1987). Some

benthic fauna and flora may be replacedby bacteria becauseof environmental

deterioration. The ecological changemight interfere the sustainabledevelopmentof

aquaculture.

Hybridization between wild and farmed strains

Whether deliberateor not, it is inevitable that some fish will escapefrom aquaculture

facilities, and cagescan be particularly risky. During storms and through other

incidents, large numbersof fish can escape.As an aquacultureindustry develops,the

advantageoffered by strains that are superior to those from the wild in terms of

growth, diseaseresistance,color, shapeetc. will becomeincreasingly apparentand

selectedfor. The selectedorganismsare increasingly domesticatedand would

potentially exhibit lower fitness in the wild (Donaldson 1997). When those selected

strains escape,they may possibly interbreed with wild strains and thus reducethe

variability of the wild population. There are fears that such interactions will adversely

affect the gene pool through the introduction of nonadaptivegenes,though this

dependson the potential of aquaculturestock to breed in the wild (Beveridge et al.,

1994). Except for eel, whose fingerlings are collected from wild, most aquaculture

speciesin Taiwan are artificially selectedand hatchery produced. However there have

beenno investigations on genetic interactions carried out in Taiwan. According to

Primavera (1991), such impacts on aquatic biodiversity are rarely positive,

occasionally neutral but usually negative. However according to observationof cage

farmers in Pen-Hu, most feral fish will stay near the original cages(Chen personal

communication 1999), so the developmentof recreational fishing near the cagemay

perhapsreduce the genetic interaction between wild and feral stock.

212
Spreadof diseasesfrom farmed to wild fish

Although diseasesin wild fish appearto be uncommon, these may be a seriousthreat

in intensive culture, due to behavioral stressand limited environmental conditions.

The development of the fish culture industry had led to an associated increase in the

number and severity of diseases of farmed fish and it is understandable that concern

has been expressed about the possible transfer of these diseases to the wild (Saunders,

1991). Such transfers may occur through the discharge of infectious waste water to

the wild, through contact of wild fish and farmed fish on each side of the cages,

through feral farmed fish and through contact with contaminatedgear etc (Hastein and

Lindstad 1991). However, potential negative interactions are not just one-way, since

wild fish may in some casesbe an important reservoir for pathogens,which can create

problems of diseaseeradication at culture sites.Although it is difficult to demonstrate

transmissionfrom farmed to wild fish, the high frequency of diseasesand on farms

and the concentration of pathogensmay have the potential to causeoutbreaksof

diseasesin wild fish population.

Developmentof antibacterial resistant bacteria

In Taiwan, different kinds of antibacterialsare usedin aquaculture,with application

typically by addition to the water or by incorporation into the feed. Long-term

antibacterial treatment of cultured fish may result in increasedlevels of resistancein

bacteria in the surrounding environment. Antibacterial resistancemay also be

genetically transferredfrom harmlessand normal sedimentarybacteria to bacterial

fish pathogens(Herwig et al., 1997).The three well establishedmechanismsof gene

transfer,transduction,transformation, and conjugation, are all believed to occur in the

213
aquatic environment (Saye and Millers !989). Genetic transfer of antibacterial

resistancewas demonstratedwith bacterial isolates obtained from sediment samples

collected beneathNorwegian fish farms. Such antibacterial resistant strains of fish

pathogenmay make it more difficult to treat fish (Sandaaet al., 1992). Heavy usage

of antibacterials,such as OTC (Oxytetracycline) may also result in their persistencein

the environment for months to years (Coyne et al., 1994). These increased residues

may representa significant threat in maintaining the health of the cultured fish and the

continued viability of the farm itself, and the product might even threatenhuman

health.

6.3.4 Physicochemicalimpacts

Sedimentation

Wastesfrom fish farms include organic solids and dissolved organic and inorganic

nutrients. Most organic solids are from uneatenfeed and faecal materials. Even the

high-energy or low-pollution commercial fish feeds were introduced, there are still

10-25% of the dry weight of the feed consumed,or 100-250 kg dry weight per tonne

fish production, is voided as faeces(Cho et al. 1994, Chen et al. 1999). Increased

loads of organic materials to the sedimentshift decomposition processesfrom aerobic

to anaerobic.The featuresof such sedimentsare substantially lowered redox

potentials and the presenceof hydrogen sulphide in the pore water, mats of sulphide-

oxidizing bacteria and severedisturbanceof the macrobenthiccommunity (Brown et

al., 1987).The azoic (devoid of oxygen) zone is enriched with carbon, nitrogen and

phosphorus,and may be completely devoid of macrobenthos(Beveridge, 1994). In

sedimentswith severeorganic enrichment, methanogenicbacteria will produce gas

and lower pH value. The releasedgas has been shown to consist of methanewith up

214
to 1800 mg1"1of hydrogen sulphide (Samuelsenet al., 1988). The impact of

is
sedimentation more seriousin seacage culture becausethe rearing area in land-

basedareawill be clearedafter harvest and is easierto be ignored by fish farmers in

the open sea.However, in Taiwan investigations of sedimentationbeneathseacages

have still not beenreported.

Land subsidenceand salinisation

In Taiwan, the fresh water usedfor aquacultureis usually groundwater pumped from

aquifers, rather than river water, which may be contaminatedwith domestic,

agricultural and industrial pollutants. Where this is acceptableat a modest level,

massive extraction of freshwater from undergroundaquifers posesa seriousthreat to

the environment, with depletedaquifers are subjectedto physical destabilisation

and/or salt-water intrusion. The uncontrolled enthusiasmfor intensive prawn

cultivation required very large amountsof pumped ground water that causedwater

level decline and attendantcompaction of aquifers, which eventually led to land

subsidenceand vulnerability to floods. In addition to land surface depression,

salinisation of surrounding areasmay decreaseagricultural production and affect the

soil, to preclude conversion to agriculture or even other aquatic crops. According to

Primavera (1991), roughly 6600 m3 of fresh water are neededto dilute full seawaterin

a one-hectarepond at one-meterwater depth over a cropping period of 4 months. In

Taiwan, land subsidencedue to excessivepumping of undergroundfreshwaterby

prawn farmers has causedtwo-story housesto become one-story bungalows.

215
6.3.5 Managementapproaches

Zoning of production area

For reasonableuse of limited water and land resources,zoning of aquaculture

production areais one approachsuggestedto addressissuesof environmental

deterioration. Zoning means dividing an area into definable parts, and regulating the

use of land or waters within these (Corbin and Young 1997). In areaszoned for

production, fish farmers can obtain the right of tenure and water use- legitimately,

also providing protection againstother usersetc. Well-controlled zoning of

aquacultureproduction can avoid conflict between fish farmers and other interested

users and can avoid the use of unsuitable types of aquaculture.It can also provide

opportunities for the governmentand private sectorsto systematically plan and

develop the allocated areas,such as grouping together to construct proper engineering

systems.In such areas,aquacultureand aquaculture-friendly activities may be

allowed, protecting them from adverseeffects of other activities, and allowing a more

focused approachto environmental management.

Managing aquaculturalzonesneedsthe cooperationof the fish farmers, other related

industries, non-governmentalorganizations(such as fishermen associations;and

production and marketing groups), and agenciesof local governments.When

designing an aquaculturalzone, expert advice and cautious evaluation are imperative

e.g. knowledge of the carrying capacity is essentialto determine the type and size of

an aquacultural zone. The planning processneedsto consider the possible

environmental impacts, and the preventive and mitigating measuresagainstpossible

216
adverseeffects of other activities within or outside the zones on its potential

development.

In aquaculturalzones,governmentcan also offer supporting infrastructure. By 1993,

42 aquaculture production areas with a total area of 12,713 ha had been set up in

Taiwan. These were in the counties of I-Lan (7), Chung-Hwa (3), Yu-Lin (6), Chia-I

County (8), Tainan (6), Koashung (4), Ping-Tung (7) and Hwa-Lian (2). In each of

production areas,10 km of. road, 7 km of inflow and effluent canals,7 km of dikes

and 40 water gateshave beenconstructed.Of the 42 production areas,6 areasin Hwa-

Lian, Ping-tung and Chia-I are for fresh water speciesand the other 36 areasfor

brackish water species.There are 7 fresh water speciescultivated in the 6 fresh water

areasand 19 speciescultivated in the 36 brackish water areas(Table 6.2). The

production quantity of the 42 areasis about 98,000 t, about 38% of total aquaculture

output in Taiwan. The value of production from theseareasis about 10 billion NT$,

(-300 million US$) about 36% of national value in Taiwan.

217
Table 6.2 cultivated in a uaculture areas.
mber of Species
areas Common name Scientific name
Tilapia Oreochromis sp
Japaneseeel Anguilla iaponi
Japanese sea perch
water 6 Barramundi Lates calcarifer
Large mouth bass Micropterus salmoides
Corbiculas Corbiculas formosana
Giant river prawn Macrobrachium rosenl ii
Milkfish Chanos chanos
Orange dotted groupc Epinephelus coioides
Garrupa Epinephelus fario
Yellow fin sea bream Acanthopagrus latus
Black porgy Acanthopagrus schlege
Gray snapper Lutjanus nebulosus
Striped threadfin Polynemus plebelus
Silver bream Sparus sarba
Red porgy Pa,erus major
sh 36 Flathead mullet
Jacks Caranx

Siganusfuscescens
Grass prawn Penaeus monodon
Kuruma prawn Penaeusjaponicus
Sand shrimp Metapenaeus ensis
Mud crab Scylla serrata
Poker chip venus Meretrix lusoria
Gracilaria Gracilaria spp.

Improving feed efficiency

Intensive aquaculture of carnivores requires concentrated protein and fish oil and has

led to increasing dependency on wild fish. In 1999, almost one quarter of the total

fishery production was utilised as raw material for the production of animal feed,

reaching 29 million tonnes (FAO, 2001). It was estimated that about 17% of fishmeal

was used in aquaculture in 1994 and it that this would reach 23% in 2010 (Pike,

1997). Moves to wider the resource base for feeds, with a range of ingredient options

are important for the future. The determination of nutritional requirements at different

life-history stages is of key importance in maximizing the efficiency of nutrition and

218
in the developmentof diets for potential aquaculturespecies(Donaldson 1997). Three

aspectsare important. Firstly, alternative protein resourcesfor fish feeds need to be

developed.Secondly,the developmentof low-pollution diets, improving feed

conversion efficiency and reducing phosphorus and nitrogen excretion is important in

environmental protection. Thirdly, using least-cost formulation that integrates the

knowledge of nutritional requirementswith ingredient cost can be valuable in

reducing production cost. When extraneousfeeding of fish cultivars is necessary,it

will be better, if this can be achievedin an ecologically sustainablemanner.

Reducing genetic impacts from feral stock

Even if great care is taken to prevent the escapeof fish from aquaculture,it might

occur, and becomedetrimental to one or more native species.To avoid genetic

impacts on wild stocks,techniques,such as chemical sterilization and polyploidy,

could be used to produce cultivars for stocking that are unable to reproduce (Rogne,

1995, Stickney 1997).Also, it should be prudent to avoid the culture of exotic species

except for making sure that the escapeeswill not be able to establish reproducing

populations.

Reducing use of chemicals

To control diseasesand reducefouling of cages,a wide range of chemicals is used by

the aquacultureindustry. As mentionedin the previous section, those chemicals may

have adverseeffects on the local environment. For sustainabledevelopment,effective

regulation is important to control the use of chemicals,with proper standardsfor

monitoring and administrating their use.However, it is costly to implement these

works. If possible, national agenciesshould establish dialog and collaborate with the

219
private sector (particularly local farmers), non-governmental organizations and

international institutions (Corbin and Young 1997).

6.4 Other issuesof sustainability

6.4.1 Risk

Aquaculture is a relatively new technology and hencerisky for new entrants.

Producersface a variety of production related risks, and risks such as the fluctuation

of market demandare also particularly important. Price volatility can be an important

constraint and if prices paid to producersare subject to wide fluctuations, the viability

of aquaculturecould be Natural
threatened. disastersmay also be important; Taiwan is

a typhoon and monsoon-pronecountry, with typhoons typically occurring almost 3-4

times each year.The developmentof some forms of aquaculture(such as cageculture)

might be an impediment to some types of fishing activities, and therefore, vandalism

or poaching may occur. Farms located in accessiblepublic water bodies may be

especially vulnerable. Aquaculture is in turn often impacted by other activities. Water

pollutants (e.g. heavy metals, chemical wastes, and pesticides from industries,

agriculture and domestic sewage) can threaten its environments. The contamination of

products that are harmful to the health of customers might also threaten development.

In Taiwan, oyster had been found to be contaminated by the discharge of copper

compound from recycling factories (Han and Huang 1990), and although the

incidence was just local, the oyster farmers of whole island suffered.

6.4.2 Role of the government and other institutions

The sustainability of aquaculturein Taiwan could be consideredto have reacheda

Further
crossroads. developmentneedsa reorientation not only in operation and

220
managementat the farm level but also greater control, integrated planning, and

managementof the industry by the state(Chua 1997). National policies are important

to the developmentof aquacultureand can have major impact on the distribution of

benefits. A national policy not only contributes to the avoidance of use conflicts but

also createsinvestment opportunities for the new economic activities, such as cage

culture. The development of aquaculture requires certain interventions from

government,including appropriatepolicies and planning, and the developmentand

adoption of new technologies. The appropriate roles of government would include

national planning and legislation, infrastructure support, research,extension and

information service.

Governmentcan also promote the establishmentof cooperativesfor fish farmers,

though which would be easierand more effective for extension workers to

disseminateinformation and training. Currently, the lack of appropriateextension

serviceshas resultedin farmers engagedin aquaculturerelying heavily on their own

perception and their neighbors' or friends' experience.Extension servicesshould be

gearedtoward promotion of technologiesfor and adoption of aquacultureand

economic analysis.These advisory efforts may include site selection and construction,

feed composition and management,water quality control, diseasemanagementand

marketing. Encouraging researchinstitutes to link to industry needsand participation

can also help the developmentof extension work. Interventions should be targetedto

promote the long-term environmental and economic sustainability.

When planning an aquaculturedevelopmentproject, both positive and negative

aspectshave to be identified and valued, and comparedwith alternative opportunities

221
in using limited resources.For developing aquaculture,government needsto set up

the agenda,which includes the necessaryadministrative, promotional, regulatory and

enforcementframeworks. The processof policy formation includes problem

definition, criteria for evaluation, generationof alternatives, proceduresfor

implementation, and identification for next steps, which include procedures for

monitoring, evaluation and reassessment (Corbin and Young 1997).

Governmentcould take the lead in cooperationwith the industry, providing the

guidelines and managementmeasuresto prevent, control and mitigate adverse

impacts. To guide sustainabledevelopmentfor aquacultureas for other sectors,

externalities and socioeconomicfeasibility should be assessedmore comprehensively.

The misuse of scarceresourcesoften results in real social lossesin the long term and

will undermine the basic requirementsfor sustainabledevelopment.If the farms aim

to maximize the profits rather than cover the social costs, selective economic

intervention from governmentmay be required providing good practice, but to

bringing private costs of farms into line with their social costs, thereby,internalizing

the environmental externality (Shangand Tisdell 1997).To encourageinternalization,

incentives, such as tax reductions,or controls such as heavy penaltiescan be used.

Information for timely policy and managementintervention, such as farming systems,

financial investment, operators,products market, socioeconomicbenefits and

constrains must be gathered, updated, and analyzed by government.

When performing the plan, the cooperationof line agencies,non-governmental

organizationsand other stakeholderstogether with government should be sought to

utilize power, influence, and resourcesto assurethat industry embracesthe desired

222
characteristicthrough direct and indirect intervention (Corbin and Young 1997).The

local administrative structure (in Taiwan, such as fishery agenciesin local

government) has the greaterability to discern potential problems and can be more

promptly and effectively engagedin preventive, rather than reactive management

(Chua 1997). For the long-term developmentof aquaculture,local government must

understandthe economic values and social benefits and commit to development.

However, potential issuesof the costs and effectivenessof enforcement activities need

to be realized, and self-regulation or voluntary compliance could be encouraged

where feasible to reducethe necessityand costs of governmental oversight

6.5 Discussion

No matter how ecologically sound,the industry of aquaculturecannot attain

sustainability if it is not profitable. Neither will it be sustainable,if it is not

ecologically sound,no matter how productive or profitable it may be in the short run.

Improvements in aquaculturewill not be sustainableunlessthey are met by adequate

policies, socio-economiccriteria and an environmentally sound regulatory framework

(Anon 2000). Successfulsustainableaquaculturehas to maintain an aim of meeting

requirementsfor production at socially acceptableeconomic and environmental costs.

The most sustainabledevelopmentwill be one that attains the best possible

relationship of the forces active in the local and regional dynamic of cultural and

economic systemsas well as in larger dynamic, but normally slower-changing,

ecological system (Bardach 1997).

In the Philippines and other developing countries, the developmentof intensive

aquaculturehad confronted a range of social impacts.These include the displacement

223
of labour, credit monopoly by big businessmen and the transfer of natural resources

into privately-owned single purposeresource.Intensive aquaculture with its high

capital cost has a poor employment-to-investmentratio, and the benefits have

remained with farm owners, entrepreneursand traders without trickling down to

community residents(Primavera 1991).The development of aquaculturehas not

improved living standardsnor village welfare, but instead,brought about social

displacement and marginalization of fishermen on the top of ecological cost. In

contrast, the situation of aquaculturein Taiwan is different. High risk, lack of capital

and the characteristicsof the Taiwanesepeople and society have causedmost aquatic

farms to be small to medium sized, insteadof being monopolized by corporate

interests.However, they then could not get the advantageof economiesof scale.To

commercialize, producersrequire high capital input. Well-organized production and

marketing groups, fishermen associationsor big businessmenin joint ventureswith

their own collateral, together with the credit offered by banks and financial institutes,

might help to overcomethis problem.

To develop aquaculture,the first question is whether a suitable market exists,

as
expressed potential salesvolume, price and harvest pattern. For example, cage

aquacultureis a new industry in Taiwan and the commercial distribution of its

production is problematic. Aquaculture products can enjoy distinct advantagesover

those from capture fisheries, offering to remove the uncertainty of supply associated

with traditional fishing, enabling food retailers to place contracts and plan forward

sales,as is customary for conventional livestock products. Harvest can be timed to get

maximum price benefit. Contract growing also offers advantageto the processor,with

uniform supply and higher levels of plant utilization (Lee 1981).

224
The demandfor fish will be influenced not only by its price, but also by the price of

potential substitutes(such as white and red meat), as well as by habits, health

consciousnessand income levels of consumers(Shepherdand Bromage 1992).An

important factor for Asia's dominancein aquacultureis that demand for fish is higher

than that for poultry and red meat when compared with Europe or North America, due

in part to traditional eating habits (Shepherd and Bromage 1992). To compete with

other meat product, uniform, palatableproducts are essentialand products with off-

flavor or undesirablesizes should be rejected.

Often, quality and uniformity of the product can improve the marketability of

aquacultureproducts, as comparedto those from capture fisheries. In order to achieve

maximum profits, it is clearly important to organize harvesting, packing, processing

and distribution of the products.The increasing awarenessof quality among consumer

and the high quality standardsof many importing countries has necessitatedthe

processorof aquacultureproducts to adopt guaranteedmethodsof assuringproduct

quality and safety. The implementation of quality management

systems,fulfilling the

requirementsof internationally acceptedstandard,for example, the HazardAnalysis

Critical Control Point (HACCP) (Subasinghe1996) and the ISO 9000 series

(Jakobsen1993), is an effective way to meet and even profit from the increasing

demandsraised by customers.

The past emphasison maximizing the efficiency of producing a single food

commodity has led to very high harvest levels per unit surface area.However, the

environmental cost of such monoculturesis also very high. When comparedto low-

225
density culture, intensive farms are more vulnerable to diseasesbecausethe crowding

and build-up of wastesfavors the growth and transmission of pathogens(Primavera,

1991). As a consequencechemicals or antibiotics may be overused.The releaseof

byproducts (such as excesslime, organic wastes,pesticide, and diseasemicro-

organisms) may directly or indirectly affect estuarineand marine organismsand

of Withdrawal
produce resistant strains pathogens. of groundwater and pumping of

harmful byproducts into coastalwaters may produce largely negative results.The

dilution of pollution by and of aquacultureis being less and less acceptedpublicly.

Good farm managementpractices,such as using good site selection and design;

adding effluent treatmentponds; using various modes of water recycling and aeration;

controlling stock rates; using biofilter and sedimentmanagement;maintaining

adequatedistancesbetween groupsof cagesmight all reduce the pollution and

internalize the externalities (Shang and Tisdell 1997). However, these are big

for
challenges engineersand biologists. To further simplify the control and regulation

of environmental impacts related with aquaculture,the development of aquacultural

zones and a sound licensing system might be useful. In Taiwan, the monitoring of

environmental quality and the enforcementof regulations on licensing are still not

well grounded.

For reducing the overuseof limited resource,aquaculturein Taiwan can try to change

towards greaterspecialization in particular aspects,such as fry or fingerling

production for export to other countries.Ancillary products, such as feeds, chemicals,

aerator,pumps and the technology can also be exported to other countries. However,

this requires particular skills, which will needto be provided.

226
The health and potential of a national aquaculture industry depends on economic

factors, legal policies (legislative, judicial and enforcement), government/ private-

sector cooperation, import-export policy, and customers' practice (Bardach 1997). The

role of government in guiding, directing, and monitoring the development of

aquaculture is important to achieve expansion in a way most beneficial for society

(Corbin and Young 1997). It is important to realize that natural resources should be

sharedwith all potential usersin a way that will benefit the society while not harming

the ecosystem.The problems of aquaculturecan be solved by learning from

experienceand finding ways to tackle them as they arise, not only in theory but by

combining this with practical application. Even though Taiwan has experienceda

prosperousperiod in aquaculture,a growing industry in itself is not a measureof

success.It is the content of growth that matters (Arrow et al., 1995).The successof

aquacultureshould be measurednot by fish production alone but by a range of fish

and other crop products, and environmental and cultural benefits.

227
 Chapter 7

Conclusions

7.1 Introduction

In Taiwan, the history of aquaculturespansover 300 hundred years.However the

problems of environmental deterioration, diseases, and export market competition

with other countries have causedTaiwan to readjust the direction of its aquaculture

development.Here, three main strategiesfor readjustment- adjusting existing

production (e.g. diversifying productive and marketing strategies),improving existing

systems(e.g. intensification of eel culture to reduce use of ground water) and

developing new systems(e.g. cageculture) were examined, with pond basedmilkfish

culture, pond and tank culture of eel and cage culture of higher value marine fish the

focus of study.

7.2 Milkfish culture

The successof artificial propagation of milkfish causedtraditional fry collection from

the wild to decline, with the potential supply of hatchery fry being more than enough

for domestic demand.However, the seasonalshortageof fry required some farmers to

import fry to stock before May. The developmentof deepwater culture has increased

the density and productivity of milkfish culture. However, cold weather is a

significant problem, as massmotalities arise as water temperaturedrops below 10 °C,

it cause.

This survey suggestedthat milkfish culture is not economically sound and showed

that more than 75% of milkfish farmers had other sourceof family income. A number
228
of farmers usedpolyculture with other speciesto spreadrisk and increase revenue.

However only farm sizesin the 4-< 5 ha category showed higher profit levels than

monoculture. When farm size exceeded5 ha, it might be difficult for a family to

manage.In the future, milkfish farm could be adjusted to 4-<5 ha, or replacedby

better managed,industrialised, larger farms.

The marketing channelfor milkfish is very complex. It might be possible to shorten

this by strengtheningthe functions of production and marketing groups.The price of

milkfish is easily influenced by production quantity and seasonalvariation, and

therefore, better information on expectedproduction levels and better harvesting

strategiescould be important in improving the market power. Although the low price

be
of milkfish can consideredas an advantagein competing in wider food markets,

however, their boninessis a disadvantagefor consumerpreference.Here, new

techniquessuch as those of bonelesspreparation may offer future advantages.Proper

in
management fish farms and appropriatequality control through Fishermen

Association, and production and marketing groups might improve the image of

milkfish to fetch a higher price.

7.3 Eel culture

Eel culture is one of the most important aquaculturesectorsin Taiwan with total value

the highest among all aquaculturesectors,and almost 90% of production exported to

Japan.However, the limited land and water resource,a shortageof eel seed,and the

competition with China for Japanese

market has causedthis sector to decline. To

overcomethese problems, super-intensivesystemsbasedon the culture of European

glasseel have been introduced.

229
Although super-intensivesystemshow higher averageprofit, the distribution of

profitability showsthat it is still possible for traditional eel culture to have a higher

profit than super-intensiveculture. Most farmers still do not want to try more

intensive eel culture as it is difficult for many to invest amount of 7,000,000 NT$

(218,750 US$) or more in the necessary facilities. Farmers also consider that the

growth rate of eel in super-intensivesystemis slower than in traditional system.

However, when the effect of social cost, such as the cost of ground water extraction,

was considered,superintensive eel culture was more economically sound than

traditional eel culture.

The massproduction of eel from China may make a big impact on Taiwan's eel

industry, as both areasdependon the Japanesemarket very strongly. It might be

important to differentiate the quality of eel and to develop new products. In addition

to improving products and extending product range, it might also be useful to develop

domestic market. To develop a domestic market, it is important to createnew products

which are more suitable for Taiwaneseconsumers,and reduce the production cost,

such as increasing survival rate and reducing feed cost.

In the future, Taiwan might lose the advantage of cost if environmental costs are

brought into consideration,and lose competitive advantageto China to the Japanese

market. To negotiate with Japanesecounterpartsto obtain a better import quota might

be critical in the near future.

230
7.4 Cageculture

Cage culture is a new sectorof aquaculturein Taiwan, with typical unit being25-75

cagesrun by family from hatchery fry. This sector was found to be broadly profitable

as most speciesgrown can fetch a sufficiently high price to generateprofit. Feed is the

highest operating cost and represent40-56% of the total operating cost, suggesting

that reducing feed cost is an important issue.

Although this survey found Dumerils's Amberjack (Seriola dumerili) and red porgy

(Pagrus major) can make higher profit than other species,producershave great

expectation for cobia (Rachycentroncanadus).A notable variation in profitability was

observedfor the speciescurrently farmed, with 18.47 million NT$ profit in

Dumerils's Amberjack and 6.42 million NT$ in grouper (Epinephelusspp.). Market

sizesin all caseswere relatively small, though there is a potential market for cobia in

the Japanesemarket as raw fish (sashimi). The likely impacts of expandedproduction

might be an environmental impact and a limited market. However, increasing farm

size might result in economiesof scale,and so reducethe production cost and selling

price. This should increasemarket size.

The sectoris restricted by specific environmental condition with strong currents and

typhoon exposureas seriouspotential constraints.The improvement of cage structures

and systemsto enablethem to reliably withstand poor weather conditions is therefore

crucial for development,as though theoretical returns currently appearto be

the
acceptable, risks may be perceivedto be too high for significant investment.A

survey of current farm units revealedthat mortality rates were very high, either

becauseof diseasespoor nutrition or unhealthy fry. Further researchin diseases,

231
nutrition, hatcheryproduction and fry quality would be helpful.

As Taiwan's market is not big enough,this is primarily on the hope that cobia can

becomea candidatefor sashimi in Japanesemarket. For the further developmentof

cage culture, marketing promotion is required.

7.5 Sustainability

A sustainableaquacutureindustry must be economically viable, ecologically sound

and socially acceptable.To improve financial performance,it is suggestedto set up a

production and marketing group through which producers might be able to control

production level, enlargeeconomiesof scale, shortenmarketing channel, control the

quality of products, and promote products.

To protecting the environment, zoning of production areasis suggested.Well-

controlled zoning of aquacultureproduction can avoid conflict between fish farmers

and other interesteduser,provide opportunities for the government and private sector

to systematicallyplan and develop the allocated areas,and protect environment from

adverseeffects of aquacultureor other activities.

The developmentof aquaculturerequires appropriatepolicy approachesand certain

interventions from government,including appropriatenational planning and

legislation, infrastructure support, research,extension and information service. In

carrying out such initiatives, governmentagentsshould use power, influence and

resourcesto assurethat industry embracesthe desired characteristic through direct or

indirect intervention. However, self -regulation or voluntary compliance should be

232
encouragedto reducethe necessityand costs of governmental oversight.

7.6 Further research

7.6.1 Economies of scale and industry aggregation

In this research,although there were apparentrelationship in specific cases,the

evidence of economiesof scale was ambiguous,and it is not really clear what effect

increasing the scaleof production has on efficiency and unit cost. The lack of

evidenceof economiesof scale was in part due to the limited sample sizes,as there

were only 64,5 and 22 samplesin traditional eel culture, super intensive eel culture

and cage culture, respectively.Although there were larger sample sizesin milkfish

culture and farm sizesin the categoriesof 4-<5 ha appearedto be more profitable than

other categories, the economies of scale were not still very clear. This may have been

due to the relatively poor financial returns in the sector,or deficiencies in data quality.

In all casesthere may have beento high level of site or operator-specificvariability in

performanceto identify clear scalar relationships.Related to this is the question of

whether such economieswere leading to changesin farm size and possible

concentrationof production into smaller number of larger, more efficient units.

7.6.2 Market research

This is very important for new product development,since the successof product

innovation dependson being able to identify and measuredemandcharacteristicsof

In
potential purchasers. this researchsome market features and consumerattitudes can

be further investigated,e.g. most eel were exported to the Japanesemarket, wherein,

the attitude of Japaneseconsumerstowards different eel products and products from

different countries would be very valuable in understandingpotential competitiveness.

233
Another areaof researchconcernsthe attitude of consumerstowards new products,

such as bonelesspreparationin milkfish, as the cost of this preparation and the prices

that consumerswould be willing to pay would be crucial for the wider adoption and

developmentof this technique.More broadly, consider future potential for expansion,

demandelasticity must also be consideredin pricing, and further researchin

measuringthe (expand)E.O.D. for aquacultureproducts and specieswouls be

important.

7.6.3 External costs of aquaculture

Intensification of aquaculturein Taiwan has stimulated the use of more chemicals, and

with great nutrient input, createdmore seriouseutrophication. The overuseof ground

water had also causedserious land Although

subsidence. the social cost of ground

water was estimatedin this the

research, costs of pollution and other impacts of

aquacultureis as yet relatively unexplored, and the methodsused to estimatesocial

costs of ground water are themselvescontroversial. Further is

researches required

acrossall the issuesof externalities.

234
 References:

Ackefors, H., J. V. Huner and M. Konikoff. 1994. Introduction to the General

Principles of Aquaculture. Food Products Press. 172 pp.

ADCP (Aquculture Development and Coordination Programme). 1983. Fish Feed and

feeding in developing country- an interim report on the ADCP Feed

Development Programme.Rome, UNDP/FAO, Aquaculture Development and

Coordination Programme.ADCP/REP/83: 97pp

Anonymous. 2000. Taiwan's offshore potential- first symposium outlines country's

cageculture. Fish Farming International. Vol. 27, No. 1, p26-27.

Arrow, K., B. Bolin, R. Costanza,P. Dasgupta,C. Folke, C. S. Holling, B-O. Jansson,

S. Levin, K. G. Maler, C. Perrings and D. Pimentel. 1995. Economic growth,

carrying capacity, and the environment. Science.No. 268, p 520-521

Balassa,B. A. 1969.The theory of demandfor products distinguished by place of

production. IMF. Staff papers 16, p159-178.

Bardach, J. E. 1997.Fish as food and the casefor aquaculture.In Sustainable

Aquaculture. (ed. by J. E. Bardach).John Wiley & Sons,Inc. p1-14.

Bargarinao,T. 1994. Systematics,distribution, geneticsand life history of milkfish,

Chanoschanos.Environmental Biology of Fishes.Vol. 39, p23-41.

Beveridge, M. 1996. CageAquaculture secondedition. Fishing News Books, Oxford.

Beveridge, M. 1984. Cageand Pen Fish Farming -Carrying Capacity Models and

Environmental Impact. FAO FisheriesTechnical Paper255.131pp.

Beveridge, M., L. CxRoss and L. A. Kelly. 1994.Aquaculture and biodiversity. Ambio

Vol. 23. No. 8. p497-502.

Beveridge, M. C. M., M. J. Philips and R. M. Clarke. 1991. A quantitative and

235
 of
qualitative assessment wastesfrom aquatic animal production. In: Advances

in World Aquaculture. D. E. Brune and J. R. Tomasso(eds). World

Aquaculture Society,Baton Rouge, Vol. 3, p506-533.

Boonyaratpalin, M. 1997.Nutrient requirement of marine food fish cultured in

SoutheastAsia. Aquaculture. 151, p283-313.

Brown, J. R., R. J. Gowen and D. S. McLusky. 1987.The effect of salmon farming on

the benthosof a Scottish sealoach. J. Exp. Mar. Biol. Ecol. Vol-109. p39-51.

Chang, C. Y. 2000. The Economic Analysis of Cage Culture Industry in Taiwan. Msc

dissertationof Institute of Aquaculture, National OceanUniversity, Taiwan. (in

Chinese).89 pp.

Chang, S. L., M. S. Su. and I. C. Liao. 1993.Milkfish fry production in Taiwan. In

Finfish Hatchery in Asia

Chaves,P. A., R. M. Sutherlandand L. M. Laird. 1999. An economic and technical

evaluation of integrating hydroponics in a recirculation fish production system.

Aquaculture Economics & ManagementVol. 3, No. 1, p83-91.

Crear, D. 1980. Observationof the reproductive stateof milkfish populations Chanos

chanosfrom hypersalineponds on ChristmasIsland (Pacific Ocean).In J. W.

Avault Jr. (Editor). Proceedingsof World Mariculture Society. Vol. 21.

Louisiana State University, Baton Rouge, 548pp.

Chen, C.-C. 1994 An economic analysis of major species in brackishwater

aquaculture. In Socioeconomics of Aquaculture-Proceedingsof International

Symposium '93 ed. by Yung C. Shang, PingSun Leung, Cheng-ShengLee,

Mao-Sen Su, and I Chiu Liao. p173-191.

Chen, L. C. 1990 Aquaculture in Taiwan. Fishing News Books. 273pp

236
Chen, M. J. 1997.The production Economics of Japaneseeel and American eel

aquaculturein Taiwan. Msc thesis in Institute of Fisheries Economics,

National Taiwan OceanUniversity, Taiwan.

Chen, Y. L. L., H. Y. Chen, and W. N. Tzeng. 1994. Reappraisalof the importance of

rainfall in affecting catchesof Anguilla japonica elvers in Taiwan. Australian

Journal of Marine and FreshwaterResearch.45(2), p185-190.

Chen, Y. S., M. C. M. Beveridge and T. C. Telfer. 1999. Setting rate characteristics

and nutrient content of the faecesof Atlantic salmon, Salmo salar L., and the

implications for modelling of solid waste dispersion. Aquaculture Research.

Vol. 30, p395-398.

Chen, Y. S. 2000. Taiwan's offshore potential. Fish Farming International. Vol 27, No.

1, p26-27.

Cho, C. Y., J.D. Hynes, K. R. Wood and H. K. Yoshida. 1994. Development of high-

nutrient-dense,low-pollution diets and prediction of aquaculturewastes using

biological approaches.Aquaculture. Vol. 124, p293-305.

Christensen M. S. 1993. An economics analysis of floating cage culture of tinfoil

barb, Puntius schwanenfeldii, in East Kalimantan, Indonesia, using chicken

manure and other fresh feeds. Asian Fisheries ScienceNo. 6, p271-281.

Chu, I. 1996. The difficulties and solutions of Taiwaneseeel industry. Taiwan Bank

Quarterly. Vol. 42, No. 3, p260-273. (in Chinese)

Corbin, J. S., and L. G. L. Young. 1997.Planning, regulation, and administration of

sustainableaquaculture.In SustainableAquaculture. (ed. by J. E. Bardach).

John Wiley & Sons,Inc. p201-233.

Cyone, R., M. Hiney, B. O'connor, J. Kerry, D. Cazabon and P. Smith. 1994.

Concentration and persistencein marine sedimentsand invertebrate following

237
 chemotherapyin aquaculture.Aquaculture, Vol. 145, p55-75.

DaVoe, M. R., and A. S. Mount. 1989. An analysis of ten state aquaculture leasing

system:issuesand J.
strategies. Shellfish Res. Vol. 8, p233-239.

Didd, S., L. Simkin, W. M. Pride and 0. C. Ferrell. 1997. Marketing- Concepts and

Strategies.Third EuropeanEdition. Houghton Mifflin, New York. 842pp.

Ding, Y. Y. 1994.Milkfish culture. In The Encyclopedia of Agriculture inTaiwan,

Fisheriesedition. (in Chinese).p185-189.

Donaldson,E. M. 1997.The role of biotechnology in sustainableaquaculture.In

SustainableAquaculture. (ed. by J. E. Bardach).John Wiley & Sons,Inc.

p101-126.

El-gayer, 0. F. 1997 The use of information technology in Aquaculture management.

Aquaculture Economics and Management. Vol. 1, No. 1, p109-128.

FAO. 1991.Environment and sustainability in fisheries. COFI/91/3. Document

presentedat the 19`hSessionof the committee on Fisheries,April 1-12. Food

and Agriculture Organization of the United Nation, Rome. 23 pp.

FAO. 1997. Aquaculture production 19886-1995.FAO FisheriesCircular No. 815

Revision 9. FAO, Rome.

FAO. 1999.Aquculture Production Statistics. FAO Fisheries Circular. No. 815,

Revision 11.

FAO. 2001 World review of fisheries and aquaculture.www.

htr
fao.org/DOCREP/003/X8002E/x8002eO4.

FisheriesAdministration. 2000. FisheriesStatistical Yearbook, Taiwan Area, 1999.

400pp.

Folke, C. and N. Kautsky. 1989.The role of ecosystemsfor a sustainabledevelopment

of aquaculture.Ambio. Vol. 18. No.4, p234-243.

238
Folke, C., N. Kautsky and M. Troell. 1992. The cost of eutrophication from salmon

farming, implication for policy. Journal of Environmental Management.No.

40, p173-182.

Folke, C., N. Kautsky and M. Troell. 1997. Salmon farming in context: responseto

Black et al. Journal of Environmental Management.No. 50, p59-103.

Gousset,B. 1990.Europeaneel (Anguilla angullia L. ) farming technology in Europe

and in Japan:application of a comparative analysis. Aquaculture. Vol. 87,

p209-235.

Han, B. and T. Huang. 1990. Green oysters caused by pollution on the Taiwan coast.

Environmental Pollution. Vol. 65, p347-362.

Hansen,P. K., A. Ervil, M. Schaanning,P. Johannessen,J. Aure, T. Jahnsenand A.

Stigebrandt.2001. Regulating the local environmental impact of intensive,

marine fish farming H. The monitoring programme of the MOM system

(Modelling- Ongrowing fish farms- Monitoring). Aquaculture. Vol. 194, No. 1,

p75-92.

Hastein, T. and T. Lindstad. 1991.Diseasesin wild and culture salmon: pssible

interaction. Aquaculture. Vol.98, p277-288.

Herwig, R. P., J. P. Gray and D. P. Weston. 1997. Antibacterial resistantbacteria in

surficial sedimentsnear salmon net-cagefarms in Puget-Sound,

Washington.

Aquaculture, Vol. 149, No.2, p263-283.

Hickling, C. F. 1962.Fish culture. Faber and Faber.London. 295pp.

Holmgren, K. and H. Mosegaard.1996.Implication of individual growth statuson the

future sex of the Europeaneel. Journal of Fish Biology. Vol. 49, p910-925.

Houvenaghel,CýT. 1989.Assessmentof the needsto culture the eel Anguilla anguilla

in Europe. In Aquaculture-A Biotechnology in Progress.Edited by N. Depauw,

239
 E. Jaspers, H. Ackefors and N. Wilkins. European Aquaculture Society,

BredeneBelgium.

Hsiao, G. K. 1994a.Water resource.In The Encyclopedia of Agriculture inTaiwan,

Generaledition. (in Chinese)p19-22.

Hsiao, G. K. 1994b.Climate. In The Encyclopedia of Agriculture inTaiwan, General

edition. (in Chinese)p31-35.

Hsiao, S. M. and L. C. Tseng. 1979. Induced spawning of pond-rearedmilkfish,

Chanoschanos (Forskal). China Fisheries Monthly No. 330, p7-13. (In

Chinese,English abstract).

Hu, B. T. 1994.Cageculture developmentand its role in aquaculturein China.

Aquaculture and Fisheries Management. Vol. 25, No. 3, p305-310.

Huang, C. H. 1990.Economic valuation of undergroundwater and man-inducedland

subsidencein aquaculture.Applied Economics, No. 22, p31-43.

Huang, C. H. 1997 Aquaculture and the endogenousdamagecost of water pollution

:the caseof Taiwan. Aquaculture Economics and Management.Vol.!, No. 1,

p99-108.

Huisman, E. A. 1986, Current statusand role of aquaculturewith special referenceto

the African region. In Aquaculture Researchin the African Region. (ed. by E.

A. Huisman). p11-22.

Jensen,G. L. 1996. Coalition building to support U. S. marine aquaculture.In: Open

OceanAquaculture. Proceedingof an International Conference,p513-522.

May 8-10, Portland, Maine.

Jakobsen,M. 1993. Quality management-ISO 9000. INFORFISH International.

No.2, p31-36.

Kannicki, Z. S. 1995Market forces, responsiblefishing and resourcesustainability.

240
 INFOFISH International. No.5, p22-28.

Kelley, C. D. and C. S. Lee. 1991.Milkfish culture in the U. S. In J. P. McVey (Ed.)

CRC Handbook of mariculture, Vol. II, Finfish Aquaculture. CRC Press,Boca

Raton, Florida. p211-226.

Kent, R. 1999.Marketing Research.International Thomson BusinessPress.334 pp.

Kimura, S., K. Tsukamotoand T. Sugimoto. 1994.A model for the larval migration of

the Japaneseeel: rolesof the trade winds and salinity front. Marine Biology.

Vol. 119,p185-190.

Klemetson, S. L. and Cl L. Rogers. 1985. Engineering and economic consideration for

aquaculture development. Aquacultural Engineering. No. 4, p1-19.

Kuo, C. M. and C. E. Nash. 1979. Annual reproductive cycle of milkfish Chanos

chanos Forskal, in Hawaii water. Aquaculture 16, p247-251.

Lai, L. W. C. and K. H. Lam. 1999.The evolution and future of pond and marine fish

culture in Hong Kong. Aquaculture Economics & Management.Vol. 3, No. 3.

p254-266.

Lee, C.S. 1983.Production and Marketing of Milkfish in Taiwan: An Economic

Analysis. ICLARM Contribution No. 91.41pp.

Lee, C. S. 1985.Environmental factors in the reproduction of milkfish. In Cheng-

ShengLee and I-Chiu Liao (Eds) Reproduction and Culture of Milkfish,

Tungkang Marine Laboratory, Tungkang, Taiwan. p99-144.

Lee, C. S. 1995.Aquaculture of Milkfish (Chanos chanos).Tungkang Marine

Laboratory, TFRI, Taiwan and The Oceanic Institute, Hawaii, USA, 141pp.

Lee, C. S. 1997 Constraintsand governmentintervention for the developmentof

aquaculturein developing countries. Aquaculture Economics and

Management.Vol. 1, No. 1, p65-71.

241
Lee, C. S., C. S. Tamaru, J. E. Banno, C. D. Kelly, A. Bocek and J. A. Wyban. 1986.

Inducedmaturation and spawning of milkfish, ChanoschanosForsskal, by

hormone implantation. Aquaculture 52, p199-205.

Lee, C. S., P. S. Leung and M. S. Su. 1997. Bioeconomic evaluation of different fry

production systemfor milkfish (Chanoschanos).Aquaculture, No. 155, p367-

376.

Lee, J. S. 1981.Commercial Catfish Farming. The Interstate Printers & Publishers,

Inc. 310 pp.

Liao, I. C. 1991 Aquaculture: The Taiwanese Experience. Bull. Inst. Zool., Acad. Sin.

Monoger. 16:1-36.

Liao, I. C. 1993.Finfish hatcheriesin Taiwan, recent advance.In Finfish Hatchery in

Asia. p65-73.

Liao, I. C. 1992. Aquaculture-theTaiwaneseexperience.INFOFIFH International,

No.3, p49-54.

Liao, I. C. and T. I. Chen. 1984. Gonadal developmentand induced breeddingof

captive milkfish in Taiwan. In J. V. Juario, R. P. Ferrari and L. V. Benitez

(Eds.) Advancesin Milkfish Biology and Culture, Island Publishing House

Inc., Manila, Philippines. p41-52.

Liao, I. C., W. C. Lee and Y. K. Hsu. 1995. Aqua-culture in Taiwan: Toward a

SustainableIndustry. Proceedingof thr 7a' Biennial Conferenceof the

International Institute of FisheriesEconomics and Trade Vol. 2, p1-13.

Liao, I. C., W. C. Lee and Y. K. Hsu. 1995b Aquaculture in Taiwan: a reassessment.

Proceeding of the 7`h Biennial Conference of the International Institute of

FisheriesEconomics and Trade. Edited by David S. Liao. Vol.2, p297-312.

242
Lin, L. T. and H. H. Lin. 1987. Studies on the induced breeding of milkfish (Chanos

chanosForskal) rearedin ponds-Natural spawning of pond-rearedbroodstock

in 1986.In J. Y. L. Yu, C. M. Kuo, S. H. Shih and S. J. Chen (Eds.) Fish

Reproduction and Its Endocrine Control: Basic and Practical Aspects.Institute

of Zoology, Academia Sinica and Department of Fisheries, Council of

Agriculture, Taipei. COA Fish. Ser. 7, p199-210. (In Chinese,English

abstract.)

Lin, L. T. 1984. Studies on the induced breeding of milkfish (Chanos chanos Forskal)

in
reared ponds. China Fisheries Monthly No. 378, p3-29. (In Chinese, English

abstract).

Lin, L. T. 1982.Further successin induced spawning of pond-rearedmilkfish. China

FisheriesMonthly No.357, p17-19. (In Chinese,English abstract).

Ling, S. W. 1977. Aquaculture in South Asia-A Historical Overview. University of

Washington Press.Seattle, 108pp.

Ling, B. H., P. S. Leung and Y. C. Shang. 1996.Export performanceof major

cultured shrimp producersin the Japanese

and US markets. Aquaculture

Research,27, p775-786.

Lisac, D. 1996.Recent developmentin open-seacage:practical experiencewith the

Tension Leg Cage.In: Open OceanAquaculture. Proceedingof an

International Conference,p513-522. May 8-10, Portland, Maine.

Losordo, T. M., M. P. Masser and J. Rakocy. 2001. Recirculating aquaculturetank

production systems-anoverview of critical consideration.World Aquculture.

Vol. 32, No. 1, p18-21.

Lovel, T. 1989 Nutrition and Feeding of Fish. Van Nostrand Rheinhold. New York.

243
Marte, C. L., N. M. Sherwood,L. W. Crim and B. Harvey. 1987. Induced spawning of

maturing milkfish (Chanos chanos,Forsskal) with gonadotropin-releasing

hormone (GnRH) analoguesadministeredin various ways. Aquaculture. Vol.

60, p303-310.

Masser, M. P., 1997. Cageculture- cageconstruction, placement, and aeration. SRAC

(SouthernRegional Aquaculture Center) Publication. No. 162, pl-4.

Masser, M. P. 2000. The statua and future of inland aquaculture. World Aquaculture.

Vol. 31,. No. 4, p34-39.

Matsui, I. 1952. Studies on the morphology, ecology and pond culture of the Japanese

eel (Angullia japonica Temminck & Schlegel ). J. Shimonoseki College of

Fisheries,245 pp. (in Japanese,with English abstract).

Milne, P. H. 1974. A visit to Japan'sfish farming industry. Fish Farming International.

1(2), p38-55.

Mudie, P. 1997.Marketing: An Analytical Perspective.Prentice Hall. London. 368 pp.

New, M. B. 1997 Aquaculture and the capturefisheries. World Aquaculture. Vol

28(2), p11-30

Pantulu, V. R. 1979.Floating cageculture of fish in the lower Mekong Basin. In:

Advance in Aquaculture (ed. by T.V.R. Pillay and W. A. Dill), p423-427.

Fishing News Books, Blackwell Scientific Publication, Oxford.

Pike, I. H. 1997.Future suppliesof fish meal and fish oil: quality requirement for

aquaculturewith particular referenceto shrimp. 26p. In Feed Ingredient Asia

'97. Uxbridge Turret-RAI.

Pillay, T. V. R. 1990. Aquaculture. Principles and Practices.Fishing news books,

1990.

Pillay, T. V. R. 1994. Aquaculture Development. Fishing News Books. 182pp.

244
Pullin, R. S. V. 1993. An overview of environmental issues in developing-country

aquaculture.In: Environment and Aquaculture in Developing Countries. (ed.

by H. R. Pullin and J. L. Maclean), p. 1-19. ICLARM Conf. Proc. 31.

Porter, M. E. 1980ComparativeStrategy.The Free Press.New York. 396 pp

Primavera, J. H. 1991.Intensive prawn Farming in the Philippines: ecological, social,

and economic implication. Ambio. Vol. 20, No. 1, p28-33.

Ridler, N. B. and M. Kabir. 1988.Economic aspectof salmon aquaculturein Atlantic

Canada. Canadian Industry Report of Fisheries and Aquatic Science 188.

Rieser,A. 1996.Defining the federal role in offshore aquaculture:should it feature

delagation to the state? In: Open Ocean Aquaculture. Proceeding of an

International Conference, p19-40. May 8-10, Portland, Maine.

Rogne, S. 1995. Genetics and breeding in aquaculture.In SustainableFish Farming.

(ed. by H. Reinertsen and H. Haaland). Proceedingof the First International

Sympsium on SustainableFish Farming./Oslo/Norway/28-31 August.

Roquillo, I A. 1975. Biology studies on bangos(Chanos chanos).Philip. J. Fish. No.9,

p18-37.

Samuelsen,O. B., A. Ervik, and E. Solheim. 1988. A qualitative and quantitative

analysis of the sediment gas and diethylether extract of the sediment from

Aquaculture. Vol.74, p277-285.

salmon farms.

Sandaa,R. A., V. L. Torsvik and J. Goksoyr. 1992. Transferable drug resistance in

fish-farm
bacteria from sediments.Can. J. Microbiol. Vol.38, p1061-1065.

Saundaers,R. L. 1991. Potential interactions between cultured and wild Atlantic

Aquaculture. Vol. 98, p51-60.

salmon.

Saye, D. J. and R' V Miller. 1989. The aquatic environment: consideration of

horizontal gene transmission in a diversified habitat. In: Gene Transfer in the

245
 Environment. (ed. by Levy, S.B. and R. V. Miller), McGraw Hill, NY, p223-

259.

Schuster,W. H. 1952.Fish culture in brackishwater ponds in Java. Special

Publication. Indo-Pacific Fish Council. Vol. 1,140pp.

Shang,Y. C. 1981.Aquaculture Economics: Basic Concept and Method of Analysis.

Westview Press,Boulder, CO, USA, 154pp.

Shang, Y. C. and C. A. Tisdell. 1997. Economic decision making in sustainable

aquaculturaldevelopment.In SustainableAquaculture. (ed. by J. E. Bardach).

John Wiley & Sons,Inc. p127-148.

Shaw, S. A. 1994 Marketing. In Salmon Aquaculture. (ed. by Heen, K., Monahan, R.L.

and Utter, F.) p255-273.

Sheeks,R. B. 1989.Taiwan's aquaculture-at the cross road. INFOFISH International.

No. 6. p38-43.

Shepherd, C. J. and N. R. Bromage. 1992. Intensive Fish Farming. Hartnolls Ltd,

Bodmin, Cornwall, 402pp.

Smith, J. G. 1987.BusinessStrategy,2°aed. Basil Blackwell, Inc. Oxford, 222pp

Stickney, R. R. 1997. Offshore Mariculture. In Sustainable Aquaculture. (ed. by

Bardach,J. I. ). John Wiley & Sons,Inc. p255-273.

Su, M. S., Y. H. Chien and I. C. Liao 1999. Potential of marine cageculture in Taiwan.

The First International Symposiumon CageAquaculture in Asia. (Abstract)

p24.

Subasinghe,S., 1996.Handling and marketing aquacultureproducts. INFOFISH

International. No. 3, p44-51.

Taiwan Fisheries Bureau 1987-1998Fisheries Year Book, Taiwan Area, 1988-1997.

Taiwan Fisheries Bureau, Department of Agricultural and Forestry, Provincial

246
 Governmentof Taiwan, Taiwan, R.O.C. (in Chinese).

Taiwan Fisheries Bureau 1997. Development of cage culture. Fisheries Weekly. No.

558, p1-3. (in Chinese).

Tesch,F. W., 1977.The eel-biology and managementof anguillid eels. Chapman&

Hall, London.

Traesupap,S., Y. Matsuda and H. Shima. 1999. Diversisication of shrimp products in

the Japaneseand US markets during the 1990s.Aquacultre Economics &

Management. Vol. 3, No. 2, p167-175.

Tseng, L. C. and S. M. Hsiao. 1979. First successful case of artificial propagation of

pond-rearedmilkfish, Chanoschanos. China Fisheries Monthly No. 320, p9-

10. (In Chinese,English abstract).

Tsukamoto,K., 1922.Discovery of the spawning areafor Japaneseeel. Nature, Vol.

356, p789-791.

Twu, S. W., R. C. Kao and H. H. Hwang 1986. On a wave dissipation method for

offshore aquaculturearea.Aquacultural Engineering Vo1.5,p271-186.

Tzeng, W. N., 1986.Resourceand ecology of the Japaneseeel Anguilla japonica

elvers in the coastal waters of Taiwan. (in Chineses)China Fish. Monthly, No.

404, p19-24.

Usui, A. 1991.Eel Culture. Fishing News Books. 148pp.

Wang, K. L. and K. M. Hung 1998.The Analysis of Production Cost of Cage

Aquaculture in Taiwan. Fish World (in Chinese)No.7. p14-22.

Williams, M. J., 1997.Aquaculture and sustainablefood security in the developing

world. In SustainableAquaculture. (ed. by J. E. Bardach).John Wiley & Sons,

Inc. p15-51.

Windsor M. L. and P. Hutchinson, 1995.Minimising the impacts of salmon

247
 aquacultureon the wild salmon stock. In: SustainableFish Farming. (ed. by

Reinertsen,H. and H. Haaland). Proceedingsof the First International

Symposiumon SustainableFish Farming/Oslo/Norway/28-31 August 1994.

Wu. Y. L., 1998.A casestudy on ManagementAspects of Fishery Production and

Marketing Groups in Taiwan. Msc. dissertation in National Taiwan Ocean

University. (in Chinese).

Yu, K. M., 2001. The implication of Japanese government's defending policies on the

eel importation. Fisheries Extension. No. 179, p45-50.

Yu, T. G., 1994. Eel culture. In The Encyclopedia of Agriculture inTaiwan, Fisheries

edition. (in Chinese) p139-144.

248
Annex A. 1 The questionnaire for milkfish farmers

1. Farm environment
1.1 The ownership of the land for fish farm
Q Owned by yourself Q Rent from government Q Rent from private
Q Others

1.2 Farm size

Pond size Number

1.3 Water source

O Underground water Q Stream or rive Q Reservoir Q Sea water
Q Others

1.4 Type of culture Q Monoculture Q Polyculture

1.5 If polyculture, what kind of fishes are cultivated with milkfish?

2. Production costs and revenues

2.1 Canital cost
Items Number Price Useful life Cost of maintenance
Land Cost
Workshed and storage house
Pond construction
Power generator
Paddle wheel
Feeder
Pump

2.2 Cost of fingerlings

Amount Price

2.3 Cost of labor

Number Wage
Full time labors
Part time labor

249
2.4 Cost of feed
Amount Price
Fingerling
Juvenile
Marketing size

2.5 Other costs

Items Amount Price
Chemicals
Electricity
Miscellaneous

2.6 Revenues
Harvest amout Size Price

3. Personalinformation
3.1 Age:

3.2 Education attainment

Q None Q Elementary Q Junior high school Q Senior high school
QCollege or above

3.3 Experiencein milkfish culture

Q 1-10 years Q 11-20 years Q 21-30 years Q 31-40 years
Q 40 years and above

3.4 Is milkfish culture your only sourceof family income

QYes QNo

250
Annex A. 2 The questionnairefor traditional eel farmers

1. Farm environment
I. Me ownership of the land for fish farm
Q Owned by yourself Q Rent from government Q Rent from private
Q Others

1.2 Farm size

Pond size Number

1.1 Water source

Q Underground water Q Stream or rive Q Reservoir Q Others

1.2 Type of culture Q Monoculture Q Polyculture

1.3 If polyculture, what kind of fishes are cultivated with eel?

2. Production costs and revenues

2.1 Canital cost
Items Number Price Useful life Cost of maintenance
Land Cost
Workshedand storagehouse
Pond construction
Power generator
Paddle wheel
Pump

2.2 Cost of eel seed

mount Price

2.3 Cost of labor

Number Wage
gull time labors
'art time labor

2.4 Cost of feed

Amount Price
: el seed
ruvenile
vlarketin size

251
2.5 Other costs
Items Amount Price
Chemicals
Electricity
Miscellaneous

2.6 Revenues
Harvest amout Size Price

3. Personalinformation
3.1 Age:

3.2 Education attainment

Q None Q Elementary Q Junior high school Q Senior high school
QCollege or above

3.3 Experience in eel culture

Q 1-10 years Q 11-20 years Q 21-30 years Q 31-40 years
Q 40 years and above

3.4 Is eel culture your only sourceof family income

QYes QNo

252
Annex A. 3 The questionnairefor super-intensiveeel farmers

1. Farm environment
1.1 The ownership of the land for fish farm
Q Owned by yourself Q Rent from government Q Rent from private
Q Others

1.2 How many intensive systemsin your farm.

1.3 Farm size

Tank size Number

1.4 Water source

Q Underground water Q Stream or rive Q Reservoir Q Others

2. Production costs and revenues

2.1 Canital cost
Items Number Price Useful life Cost of maintenance
Land Cost
Workshed and storagehouse
Cost of systems

2.2 Cost of eel seed

Amount Price

2.3 Cost of labor

Number Wage
Full time labors
Part time labor

2.4 Cost of feed

Amount Price
Eel seed
Juvenile
Marketing size

2.5 Other costs

Items Amount Price
Chemicals
Electricity
Oxygen
Miscellaneous

253
2.6 Revenues
Harvest amount Size Price

3. Personalinformation
3.1 Age:

3.2 Education attainment

Q None Q Elementary Q Junior high school Q Senior high school
QCollege or above

3.3 Experiencein eel culture

Q 1-10 years Q 11-20 years Q 21-30 years Q 31-40 years
Q 40 years and above

3.4 Is eel culture your only sourceof income

QYes QNo

254
Annex A. 4 The questionnaire for cage farmers

1. Farm environment

1.1 Farm size

Kinds of cages Number Size

1.2 Culture species

2. Production costs and revenues

2.1 Capital cost
Items Number Price Useful life Cost of maintenance
Cage (include net)
Rope
Buoy
Anchoring
Install
Workshed and storagehouse
Boat
Engine of outside boat

2.2 Cost of
Amount

2.3 Cost of labor

Number Wage
Full time labors
Part time labor

2.4 Cost of feed

Price

255
2.5 Other costs
Items Amount Price
Fuel
Miscellaneous

2.6 Revenues
Species Harvest amout Size Price

3. Personalinformation
3.1 Age:

3.2 Education attainment

Q None Q Elementary Q Junior high school Q Senior high school
QCollege or above

3.3 Experiencein cage culture

Q 1-10 years Q 11-20 years Q 21-30 years Q 30 years and above

3.4 Is cageculture your only sourceof family income

Q Yes Q No

256
Annex A. 5 The questionnairefor consumersof milkfish

1. What kind of products of mikfish do you prefer?

Q Fresh fish Q Cannedfish Q Fish meat ball Q Others

2. What kinds of fresh part of milkfish do you prefer?

Q Whole fish Q Scaled and gutted whole fish Q Head Q Belly part

Q Others

3. Which season do you prefer for purchasing milfish?

Q Spring Q Summer Q Autumn Q Winter Q Uncertain

4. What are your preferred purchasing sizes of milkfish?

Q 300g Q 600g Q 900g Q 1200g

5. What are the frequenciesof buying milkfish?

Q Every week Q Every two weeks Q Every three weeks Q One month

Q Uncertain

6. What are the amountsof buying milkfish every time?

Q About 600g Q About 1200g Q About 1800g Q Uncertain

7. What are the reasonsthat you do not buy milkfish?

Q Bony Q Too expensive Q Others

8. What is your opinion concerning the price of milkfish?

Q Very expensive Q Expensive Q Acceptable Q Cheap Q Very cheap

9. What is your opinion concerning the quality of milkfish?

Q Excellent Q Good Q Acceptable Q Bad Q Very bad

10. What is your evaluation concerningthe price and quality of milkfish?

Q Excellent Q Good Q Acceptable Q Bad Q Very bad

257
11. What kinds of situations will you buy more milkfish?

Q Price is cheaper Q Quality is better Q Others

258
Annex A. 6 The questionnaire for consumers of eel

1. What kind of products of eel do you prefer?

Q Fresh fish Q Frozen roasted Q Others

4. Which season do you prefer for purchasing eel?

Q Spring Q Summer Q Autumn Q Winter Q Uncertain

5. What are your preferred purchasing sizes of eel?

Q 200g Q 300g Q 600g

6. What are the ways that you cook eel?

Q Stewedwith Chineseherbs Q Roasted Q Others

7. What are the frequenciesof buying eel?

Q Every week Q Every two weeks Q Every three weeks Q One month

Q Uncertain

S. What are the amountsof buying eel every time?

Q About 600g Q About 1200g Q Uncertain

9. What are the reasonsthat you do not buy eel?

Q Bony Q Too expensive Q Difficult to cook Q Others

10. What is your opinion concerning the price of eel?

Q Very expensive Q Expensive Q Acceptable Q Cheap Q Very cheap

11. What is your opinion concerning the quality of eel?

Q Excellent Q Good Q Acceptable Q Bad[] Very bad

12. What is your evaluation concerning the price and quality of eel?

Q Excellent Q Good Q Acceptable Q Bad[] Very bad

259
13. What kinds of situations will you buy more eel?

Q Price is cheaper Q Quality is better Q Others

260
Annex B. 1 The culture area for milkfish in Taiwan from 1987 to 1997.
Unit: ha
Year Total Area
aquaculture R1* R2**
area
Brackish Fresh water Subtotal
and and
1987 66,302.12 monoculture 6117.00 763.20 6,880.20 12.53 12.53
polyculture 841.84 588.40 1,430.24
Subtotal 6,958.84 1,351.60 8,310.44
1988 67,406.10 monoculture 4659.60 1092.11 5,751.71 11.19 11.19
polyculture 1068.24 720.57 1,788.81
Subtotal 5,727.84 1,812.68 7,540.52
1989 71,082.53 monoculture 4986.62 1204.30 6,190.92 13.23 13.23
polyculture 1663.53 1552.47 3,216.00
Subtotal 6,650.15 2,756.77 9,406.92
1990 76,421.90 monoculture 6781.23 2460.12 9,241.35 16.81 16.81
polyculture 2207.92 1395.73 3,603.65
Subtotal 8,989.15 3,855.85 12,845.00
1991 74,078.76 monoculture 6546.21 1802.63 8,348.84 16.97 16.97
polyculture 2225.84 1993.67 4,219.51
Subtotal 8,772.05 3,796.30 12,568.35
1992 72,293.00 monoculture 6508.13 1685.93 8,194.06 17.30 17.30
polyculture 2833.05 1476.61 4,309.66
Subtotal 9,341.18 3,162.54 12,503.72
1993 70,965.01 monoculture 5403.14 1624.11 7,027.25 15.05 14.38
polyculture 1801.31 1375.74 3,177.05
suspended 363.62 113.12 476.74
Subtotal 7,568.07 3,112.97 10,681.04
1994 69,602.78 monoculture 5655.66 1533.29 7,188.95 16.17 14.97
polyculture 1834.48 1398.01 3,232.49
suspended 702.97 130.53 833.50
Subtotal 8,193.11 3,061.83 11,254.94
1995 70,075.31 monoculture 6166.61 1436.86 7,603.47 17.33 16.13
polyculture 1887.45 1872.29 3,759.74
suspended 713.43 70.50 783.93
Subtotal 8,767.49 3,379.65 12,147.14
1996 67,613.37 monoculture 4062.60 1667.28 5,729.88 17.43 14.30
polyculture 2408.22 1529.08 3,937.30
suspended 2071.89 44.60 2,116.49
Subtotal 8,542.71 3,240.96 11,783.67
1997 63,155.51 monoculture 3247.87 1291.54 4,539.41 17.16 14.75
polyculture 2719.01 2059.85 4,778.86
suspended 1420.82 97.90 1,518.72
Subtotal 7,387.70 3,449.29 10,836.99

261
1998 63,188.84 monoculture5278.57 1609.95 6888.52 19.24 17.47
polyculture 1932.83 2219.50 4152.33
suspended 814.59 300.87 1115.46
Subtotal 8,025.99 4,130.32 12,156.31
1999 63,214.74 monoculture 6146.93 1158.38 7305.31 19.19 17.95
polyculture 1758.32 2281.66 4039.98
suspended 451.24 334.79 786.03
Subtotal 8,456.49 3,774.83 12,131.32
2000 62,567.10 monoculture 6464.61 1363.42 7828.03 22.61 21.35
polyculture 3337.22 2193.53 5530.75
suspended 602.29 184.23 786.52
Subtotal 10,404.12 3,741.18 14,145.30
Data source:Year Book of Taiwan FisheriesBureau
* R1 is the ratio of milkfish culture areato total aquacultureareain Taiwan.
** R2 is the the ratio of real milkfish culture areato total aquaculturearea in Taiwan. The real milkfish
is
culture area referred to that the milkfish culture areadeductsthe suspendedmilkfish culture area.

262
Annex B. 2 The production quantity and value of milkfish in Taiwan from
1987to 1997. Unit: mt/1000NT$
Produc tion qua ntity Production value
Year Total Brack- Fresh Total R1* Total Brackish Fresh Total R2**
Aqua- ish water milk- Aquaculture water water milk-fish
culture water pond fish pond pond
and
1987 305,428 19476 9351 28,827, 9.44 35,232,460 1058348 509924 1,568,272 4.45
1988 300,974 23161 16511 39,672 13.18 34,478,389 1168196 776820 1,945,016 5.64
1989 249,755 12601 8481 21,082 8.44 26,524,516 577103 417722 994,825 3.75
1990 344,263 75244 15429 90,673 26.34 31,530,574 2331927 481759 2,813,686A
1991 291,885 27106 14126 41,232 14.13 30,256,203 980181 427909 1,408,090
1992 261,648 15580 9534 25,114 9.60 29,292,039 946343 517888 1,464,231
1993 285,275 16844 28669 45,513 15.95 29,815,944 760266, 1157966 1,918,232 6.43
1994 287,965 26188 40590 66,778 23.19 33,566,439 1148663 1778478 2,927,141 8.72
1995 286,634 28058 35196 63,254 22.07 36,514,231 1455262 1900564 3,355,826 9.19
1996 272,525 27806 30647 58,453 21.45 32,727,444 1437008 2133046 3,570,054 10.91
1997 270,139 31259 31490 62,749 23.23 27,100,002 1321106 1335728 2,656,834 9.80
1998 253,339 28359 29990 58,349 23.03 27,043,476 1287480 1214694 2,502,174 9.25
1999 263,069 22649 28175 50,824 19.32 23,780,415 1104451 1320822 2,425,273 10.20
,
2000 256,399 16267 23463 39,731 15.50 25,912,938 880497 1138709 2,019,207 7.79
Data source:Year Book of Taiwan FisheriesBureau
* RI is the ratio of total milkfish production quantity to total aquacultureproduction quantity.
** R2 iS the ratio of total milkfish production value to total aquacultureproduction value.

263
Annex C. 1 The seasonalvariation of fresh eels imported from Taiwan for
the Japanesemarket from 1994 to 1998. Unit: mt
Year Jan. Feb. Mar. Apr. May Jun. Jul. Aug Sep. Oct. Nov. Dec. Ave
348.69 555.02 557.84 731.91 1060.42 3087.21 1180.37 347.32 179.66 181.84 495.85 776.57
1994 592.73
490.60 732.82 801.41 781.01 820.84 2063.50 1061.02 142.30 154.20 228.42 519.82 705.29
1995 667.59
1996 463.13 481.50 562.11 557.29 545.50 783.01 1869.81 1189.13 485.20 316.39 287.01 527.07 672.26

1997 453.29 488.61 503.24 515.07 601.49 978.08 2463.32 1123.66 644.01 472.71 482.04 720.07 787.13

757.08 663.65 646.79 602.01 763.82 1565.90 636.78 337.41 294.03 401.20 646.56 676.82
1998 806.64
Ave. 596.67 513.30 603.37 615.68 652.38 881.23 2209.95 1038.19 391.25 283.40 316.10 581.87 723.62

Index 82.46 70.93 83.38 85.08 90.16 121.78 305.40 143.47 54.07 39.16 43.68 80.41 100.00

Data source: Customs Bureau, Ministry of Finance, Japan.

Annex C.2 The seasonalvariation of fresh eels imported from China for
the Japanesemarket from 1994 to 1998. Unit: mt
Year Jan. Feb. Mar. Apr. May Jun. Jul. Aug Sep. Oct. Nov. Dec. Ave
589.25 432.80 403.96 419.55 334.31 739.53 542.18 488.11 350.45 411.25 400.41 474.90
1994 586.94
172.37 174.93 172.16 150.26 439.45 661.45 537.40 203.63 131.17 118.77 107.96 263.00
1995 286.49
155.92 118.83 110.85 155.38 237.42 458.19 778.41 157.06 138.49 123.51 248.60 413.49 258.01
1996
1997 383.74 279.12 256.24 296.16 323.40 520.58 728.73 146.42 112.78 167.00 247.14 280.24 311.80

166.69 157.36 319.18 400.05 431.05 509.24 948.48 204.35 123.99 152.46 394.46 521.20 360.71
1998
Ave. 315.96 263.39 258.80 285.54 312.34 452.35 771.32 317.48 213.40 184.92 284.04 344.66 333.68

Index 94.69 78.93 77.56 85.57 93.60 135.56 231.15 95.14 63.95 55.42 85.12 103.30 100.00

Data source: CustomsBureau, Ministry of Finance,Japan.

264
Annex C. 3 The seasonal variation of processed eels imported from

Taiwan for the Japanesemarket from 1994 to 1998. Unit: mt

Year Jan. Feb. Mar. Apr. May Jun. Jul. Aug Sep. Oct. Nov. Dec. Ave
1.44 1039.20 1270.78 1567.14 1435.97 1244.64 1974.09 878.66 1096.62 583.96 707.38 495.15 1107.50
1994
1995 265.88 431.54 787.16 1308.97 1192.38 1069.87 636.50 240.60 274.86 296.86 248.76 359.98 592.78

279.36 448.33 942.94 970.88 1084.30 694.04 279.65 503.72 323.86 327.35 300.38 537.49
1996 295.01
1997 332.56 357.67 644.82 510.37 548.59 513.76 407.72 336.24 463.44 240.51 169.34 206.46 394.29

199.35 428.57 426.49 470.83 477.99 261.17 31.77 53.42 53.76 125.44 148.85 244.72
1998 258.98
Ave. 429.77 461.42 715.93 951.18 923.73 878.11 794.71 353.38 478.41 299.80 315.65 302.16 575.35

74.70 80.20 124.43 165.32 160.55 152.62 138.12 61.42 83.15 52.11 54.86 52.52 100""00
Index

Data source:CustomsBureau, Ministry of Finance,Japan.

Annex C.4 The seasonalvariation of processedeels imported from China

for the Japanesemarket from 1994 to 1998. Unit: mt
Year Jan. Feb. Mar. Apr. May Jun. Jul. Aug Sep. Oct. Nov. Dec. Ave
1332.29 1601.87 2700.15 2737.76 2325.88 2130.83 1186.54 854.38 2170.42 2144.22 2416.60 1968.72
1994 2023.65
1885.78 1250.62 2000.37 2595.63 3410.50 3780.80 3044.15 1138.45 823.70 2169.09 2996.18 2434.88 2294.18
1995
1913.24 1461.38 3919.93 5883.35 5747.87 3632.94 1701.51 1767.18 2564.78 2801.07 3541.58 3150.41
1996 2870.11
4922.67 3117.71 4439.24 5181.08 6679.40 5403.38 4864.33 2657.37 3014.66 3297.17 3086.33 2801.67 4122.08
1997
2722.66 2463.45 4353.19 6268.52 5969.87 4736.54 5851.05 3190.35 1586.44 2487.96 4176.04 3656.49 3955.21
1998
Ave. 2884.97 2015.46 2771.21 4133.06 4936.18 4398.89 3904.66 1974.84 1609.27 2537.89 3040.77 2970.25 3098.12

65.05 89.45 133.41 159.33 141.99 126.03 63.74 51.94 81.92 98.15 95.87 100.00
Index 93.12

Data source: CustomsBureau, Ministry of Finance, Japan.

265
Annex D. 1 Production and value of different speciesused in cage culture
in Taiwan. Unit: ton, thousandNT$
Offshore cageculture Inland cage culture
Year Species Quantity Value V/ Species uantit Value V/
1989 Black seabream 5 1000 200 Tilapia 475 17527 37
Misc. seabream 16 2218 139 Common carp 118 5310 45
Crucian carp 8 416 52
Big head carp 1 47 47
Perch 26 3930 151
Milk fish 5 250 50
1990 Red porgy 1 200 200 Tilapia 534 17070 32
Black seabream 3 518 173 Common carp 110 4730 43
Misc. seabream 99 13009 131 Crucian carp 19 969 51
Eel 1651 231177 140
1991 Misc. seabream 86 11541 134 Tilapia 366 9742 27
Common carp 121 4356 36
Crucian ca 21 945 45
1992 Red porgy 69 20859 302 Tilapia 218 6322 29
Misc, seabream 53 8289 156 Common carp 96 3840 40
Malabar cavalla 8 179 22 Crucian carp 16 736 46
1993 Red porgy 58 18241 315 Tilapia 209 6509 31
Misc. seabream 80 13288 166 Common carp 120 5400 45
Crucian carp 36 1872 52
Other fishes 3 174 58
1994 Red porgy 59 20644 350 Tilapia 11 281 26
Misc, seabream 91 15634 172 Common carp 3 99 33
Crucian carp 5 350 70
Other fishes 2 72 36
1995 Red porgy 137 47692 348
Misc. seabream 108 18686 173
Brown croakeer 31 3320 107
Groupers 68 22130 325
Cobia 3 1050 350
Other scads 2 350 175
Other fishes 8 858 107
1996 Seaperch 16 1034 65
Red porgy 200 56913 285
Black seabream 18 4186 233
Misc, seabream 141 24870 176
Brown croakeer 27 2949 109
Groupers 141 38253 271
Cobia 13 4680 360
Other scads 20 1463 73
Other fishes 101 13154 130
1997 Seaperch 10 756 76
Red porgy 261 75817 290
Black seabream 16 3586 224
Misc. seabream 137 24126 176

266
 Brown croakeer 28 3027 108
Groupers 150 44559 297
Cobia 9 3330 370
Other scads 69 12471 181
Other fishes 156 18402 118
1998 Seaperch 12 856 71
Red porgy 201 62133 309
Black seabream 13 1820 140
Misc. seabream 156 28698 184
Brown croakeer 29 3120 108
Groupers 151 38579 255
Cobia 17 2133 125
Other scads 98 16528 169
Other fishes 209 25775 123
Data source: Year Book of Taiwan Fisheries Bureau

267