The Development of Commercial Farming of Tilapia in Jamaica 1979-1983 Research and Development Series No. 31 International Center for Aquaculture Alabama Agricultural Experiment Station Gale A. Buchanan, Director October 1984 AID/ LA-C-I 166 Auburn University Auburn University, Alabama CONTENTS' Page BACKGROUND...................................................... PRINCIPAL GOAL AND INPUTS OF THE PROJECT.......................... 3 Goal.................................................... Inputs .................................................. FISH FARMING METHODS PROMOTED IN JAMAICA....................... 3 3 3 4 5 6 8 9 9 10 Brood Pond Management .................................. Nursery Pond Management................................. Management of Food-fish Production Ponds .................... Relative Size and Number of Brood, Nursery, and Production Ponds ..................................... Recent Introduction of a More Intensive Technology............. FISH PRODUCTION EXTENSION BRANCH.............................. Organizational Structure .................................. Extension Services Provided ............................... FOOD-FISH PRODUCTION ON PRIVATE AND PUBLIC FARMS............... Rates of Production....................................... Total Food-fish Production ................................ ECONOMICS OF SMALL-SCALE TILAPIA FARMING....................... 10 11 12 12 12 13 13 14 15 15 15 17 17 17 17 Costs and Returns for a Small-scale Fish Farm ................. Fish Farming and Competitive Land Uses.................... Capital Availability for Fish Farming......................... Fingerling Production in the Private Sector................... Harvesting of Private Ponds ............................... MARKETING FARM-RAISED TILAPIA................................... TREND TOWARD GREATER INDEPENDENCE OF PRIVATE PRODUCERS..15 Related Supply and Demand Considerations ................... Problems With Pre-existing Distribution Channels .............. Development of Distribution Channels for Tilapia .............. PUBLISHED 10/84-2M Information contained herein is available to all without regard to race, color, sex, or national origin. The Development of Commercial Farming of Tilapia in Jamaica 1979-1983 T.J. POPMA, F.E. ROSS, B.L. NERRIE, and J.R. BOWMAN' BACKGROUND PRINCIPAL GOAL AND INPUTS OF THE PROJECT Goal The basic goal of the project was to increase food-fish production throughout the country, primarily through extension support to private producers. Founded on a training program for participating farmers, extension agents, and professional aquaculturists, production targets for end-of-project (August 1984) were 600 participating farmers producing 545 metric tons annually from 232 hectares of ponds. Approximately 80 percent of the farmers and 20 percent of the production should be from small-scale farms, each with less than 4 hectares of arable land. ACOMMON REACTION to the idea of commercial fish farming in Jamaica is to question its appropriateness in a small country surrounded by the sea. This island, however, has long been a heavy importer of fish. The clear inshore waters so attractive to tourists do not support a large marine capture fishery, and many of the productive offshore fishing grounds near Central and South American countries have been lost because of expanded territorial claims by many of these countries. From 1973 to 1978, Jamaica, with a total population of approximately 2 million people, annually imported an average of 15,000 metric tons of fish. Imports would likely have been even higher if foreign exchange had not been limiting. Thus, an economically viable, local fish farming industry becomes more attractive. Tilapia (Oreochromis)mossambica, locally known as the African perch, was introduced to Jamaica in 1949. It flourished in irrigation canals, and a low-input subsistence level of management was attempted in a few small ponds. By 1977, a small hatchery and research facility were constructed, but the viability of commercial tilapia farming had not been demonstrated. In that year Auburn University was awarded a technical assistance contract (AID/1a-C-1166) for a USAID-sponsored Fisheries Development Project (532-0038), with Ken Randolph serving as Auburn's first resident advisor in Jamaica. Project objectives were to evaluate the economic potential of commercial fish culture in Jamaica and to increase Jamaica's institutional capacity to implement a fish production program. An existing 16-pond complex, with approximately 10 hectares of water, was acquired at the beginning of the project. Upgrading and expansion to 32 hectares were begun concurrently with the first production trials with monosex (male) Tilapia mossambica fed a commercial poultry ration. During the period of 1978 to 1979, more then 50 tons of food-fish were produced at this farm and marketed mainly through the government-owned Agricultural Marketing Corporation. Because of the high production and the promising economic analysis of the management system, a follow-up project (532-0059), jointly funded by the Government of Jamaica and the USAID, was begun in late 1979 to stimulate the development of warmwater fish culture in the private sector. The following sections describe the growth of commercial farming of tilapia during the first 4 years of that project. 'Assistant Professor, Department of Fisheries and Allied Aquacultures and International Center for Aquaculture, Fisheries Specialist, Jamaican Ministry of Agriculture, and Research Associates, Department of Fisheries and Allied Aquacultures and International Center for Aquaculture, respectively. Inputs Funding. Total financial obligations for the 5-year project were nearly US$9 million, 54 percent of which was funded directly by the Goverment of Jamaica, with the remainder being financed via loan and grant agreements with the USAID. Jamaican Staff. Emphasis was placed on the training of Jamaican personnel. More than US$600,000 has been committed for longand short-term training. Planned manpower levels at end-ofproject were 160 persons, including 24 professional staff members. Technical Assistance. Auburn University continued to provide technical assistance under the original USAID grant-funded contract. A total of 13 person-years of long-term and 14 person-months of short-term technical assistance were provided under this contract. The resident consultants included a senior technical advisor and two extension specialists. Support Hatcheries and Research Facilities. In the eastern region of the island, the Mitchell Town Support Facility, originally used to demonstrate the economic feasibility of tilapia farming, was upgraded and is now used principally for fingerling production. This farm now has 33 ponds with a combined surface area of 32 hectares. An office, a warehouse, feed storage bins, a seine room, a five-tank holding facility for live fish and other support equipment, and vehicles are now operational. Initial development efforts were concentrated in the eastern half of the island, but in 1982, the construction of an additional facility in the western region permitted expanded extension activity. The Meylersfield Support Facility has 18 earthen ponds with a combined surface area of 6 hectares. Support equipment and buildings, similar to those at the Mitchell Town Facility, are also provided. MITCHELL TOWN FISH FARM 33 ponds - 32 hectores 0.. A-SERIES Ponds I -II 0.1 to 0.4 hectare Ponds 12-16 2.8 to4.4 hectares Pond 17 I hectare Ponds 1-14 0.6 to 0.8 hectore Ponds 15.16 0.3 to 0.4 hectare Reservoir (R) 0.2 hectare ' '\ Adopted from aerial photograph I Practical research capabilities were increased by the addition of a 28-pond unit at project headquarters in Twickenham Park near Kingston. This facility is used to further refine the production technique currently practiced by farmers. Farmers' Commodities. Included in the project are loan funds for the importation of essential fish farming equipment. This provision was included to permit a more rapid decrease in dependency on government support. This equipment is made available at cost to participating farmers. Money from the sale is to be put into a revolving account for purchase of additional equipment through the Jamaica Agricultural Society. Vehicles. Mobility of extension and support personnel was enhanced by provisions to purchase 20 trail bikes and 24 four-wheel vehicles, including automobiles, pickup trucks, and vans. Three small bulldozers and two flatbed trucks for bulldozer transport were also purchased to facilitate the timely and more economical construction of farm ponds. FISH FARMING METHODS PROMOTED IN JAMAICA The following section is a technical description of the required inputs and the expected yields from the pond management practices currently promoted in Jamaica by project personnel. The rate of development in the private sector and the strategies used by the extension service to promote this management system will be described in the following two sections. The economic justification for semi-intensive farming of tilapia in Jamaica was provided in 1978-79 during the original project, and FIG. la, b, and c. The Mitchell Town Support Facility, above, is the largest of the three facilities, and is used primarily for fingerling production. The newest facility, Twickenham Park, below, is used to refine the production system used by farmers. The Meylersfield Support Facility, facing page, expanded the program to the western region of Jamaica. TWICKENHAM PARK FISH -CULTUREMAICA RESEARCH AND TRAINING CENTER 28 ponds - 1.8 hectares Kingston HHHHHHHH C-1 -5B-7 B-9 POND SIZES A ponds - 0.04 hectare each B ponds- 0.08 hectare each C ponds - 0.2 hectare each 'ki EY Li RSFEL D FISH FARM - 7 ponds 6 he~ctars D-4 04 D-3 0-3 Tilapia 'am g as usua.v paicticed in Jamaica, aI a 'le-a .ep procedure. Very small fish are 'irst harvested from brood ponds and are then transferred to a second pond !or rearing ic the intermediate size required -,' iiisually determine their seN. Males are -!inally rastocked into -!'lning ponds for grosw-cut 2o markelable size. cleset POND SIZES ponds- 0.04hiectore each ponds - 0.08 hectare each ponds - 0.2 hectore eatch ponds I hectare eoc., tptiiiiof, 11(11((IlSl Ntt"tied maiiat'iiiettt pr actices antd th iit eNl.lds obtain ed Iioti the 1)ro)d, (lilrset N alli 10(14lisi ;i o- dietol lilnderer pro eormmnal(er' rs,deflc. t.ced today . Tot ptc i t I t. jeell ti iall>l \ NIetepr utiott (tttl em le b\ii Nt u it o\ ti t.nd tIcowi Iit s>epartatt.'t frta the t al e.xatt ttlatittll of the speies Il, Ct.lttte A>is aitd iltt >tippletiiital fl'd Ii.Nt In eni lhtg> lit Ias adopte. ben produce tict all 1 i l1mixtdt. itiurscr\ potnds. Toaeeotoplish 'his. tanTiinu il size lfrom 40 to 2001)gtamtt ale sextt,(I itati e fisl, stockedi in e~a then pondiat eist' tAT .501)to 1(1 000 pet bec-Lar. The sex r atioi of t.' brood stock is apprtoxiniatekN 3 fenialc i per 1 ist ofitlet. ati a (aiN -te 1 2:3 kilogrili> per liectal e male. le a th ,reatert t t riate. ol 34 kilograms per mid, dotittg' the firt t i hectare. Fit iuunt patittal hatl-Netinig oF the It N hetim> .it) 7 Nweeks ifter dag stttekittg lrot 0111sh1. lIN\ ate etollected NN eklN. uim511 N\lotiit a seh(ie iIth 6-nilit etet Ntjal Le it e ii. III order.I t tobtain tit hllifo-I Brood ponds art mat taged 0o tdapi lot>uli~t~ju. ~ Thedamt subtiti ttit.' af iet \a h oritatt loittr i~t.> It h a s(1(0 itll\ iti talte teIitp i( (t l at in NN I 1>al. (1(11/ il Ni 1( 7 lt- o ca l fedtin Nf ~t euto'teineeteeae ~ ui i ler It5as o 12 irillhotttet S(~ac mteshi int'o a pat tialN subtmorgedl ea-e coN tcr.td 6 itiilimeter iteshi NN or plastic setreen. The as et age size of' ire the frN that ate sm itall ettoughi to pass thr ough the 12 tanillio eter 1 tmesh ittiti- and largec etiotilgh to he tretatned itt the( 0 millitnetet Niith bleaCl Itil dcak ihgtitott negatt a .fl111ads"llett-by prtistttt 197iotitl. fT >h~t ilttt 11i 198 1 itmha ((-1) aecen.too l laiers N\wa i tis la (17 lii alt- mes h eatie N aries lrottt 0. 6 to 1. 1 gt ,tt . Expcietuiee ias dlettonsltrated tiat It eqlit. lt patial hat Nesiitig a NN t.ek is esiential to the SLeet.Siiofthis tttaaettet utstalli tne (iddo slis:3 tirt.-lt.' m . ttl ttt baitudt.'tl t lii lt isof Moases actmi t 'itttd>.l airl ings ifipettl.' t ttll emtit a er an'l a t.o :3 ate r~ t rcatNest,apl.1 . .t.hugh e ot 1 ne hittl gt.' foue ati extenided't~ r te If' partial ltarx ests ale diljIhtitt perittd. tte a t age sitze tof the fingtetrtings increse.,,, tresudting it) inceatsed caniali ism toti tstmall fi atid agtreat retdtuttttion tle total Ofttt N . WXith att tilto tet rttpted tilt)grait of fre.qttettt partial ft tU till1(1 ti liars('its, somie brood tttids liai rtetined tirtodult c 11)1a pt.riodl of mtotrt than 8 tothst bitt the usauali trdtjise petriotd is 4 tt 6 pti mn th s. Ilkpretet iN teiar P ctiltoitn res ia 1981,iatc has rept.aictd NT. mtxd o . tila tte a ~s g sa re tetdttuIttre Yijeld statistics fot ItN ate base l resutmtlts frtot six brood ptottds w.ti oth k tomine art.a of l.5 htectares. Du.rtini ant averaue st irtct. tittle('t ol ill\\ eeki 3.4 miillioni frN, -,i ii alt cNrag. siz ot produtio 0.9 1 att teach, \ert. batN eited frott thtest. potitis. The as t.'tage rate ttl p~rtdtetmt N a 128,800 fr per betctar. pet Nit.t.k dturitig the pieriod ttf pa~rtial liarNemtig, tablt. 1. Thle ital Nieight ttlr btaN~ined hetare t.'tirting froim tile mix birood ptttois s was 1,983 kilogr auis per lt Of 6 tMonthis. Approximtately 1.7 ttCI the as etage pt odIICOet cC 1TABLE 1. PR(110 ', 01, FRN Il IA1) ILkAI N11H TIC I\ 1111 N Biiou I P01~ NIm 10 I Q CIs I PAFIIAAI 1U IAI,(; 11111-1A 6-NI \I AI EiIF,RNIIS Ill _SEI'sL 1O \IN NIlI I L 711 AN Stl (Ill Al I1 D 71 Nim 's,2 1 Im (I aOrcl Item I S.D. AkeekA . . .. . . . Annua HI Oppor C'ost,1/0 II isiAl ,e Iii lt\ (oI cost oil c icll 1 Ponid Illa itill 6 (11011th ic .. .. di( (IlCA use)................... I lp ( .1(,: 1 e . . . . . 50.00I CA 0le. million1s CI. iti r) II 65 -1 IA II ve HateiC oif Ii, Aprodu oC iiCI-I li I, sI lII pr i t & 6 (((iii 'fii 2 h Cs l1 IX .41.001( 251(0) 2 ti Ictor ii~u Ieiod CIaial Aofi Fee hAX 1 ((t (ii A tn(loLA7, i (11(1ii CA C ie . ha . .. . . . . . . . di IAX Ckeiihiiii Paulk SC1ppi aiities.1 1110produiciition CAcl bcII~,ail AX iI tile lioCkinof In-o h~!avest! lin of' Ar 10ga Lipp1 xii Alial\I 6 AXI''. lateli L Tim IA! Fe cI 2.5(16 kg ((o 8.604A ...... Pu 1 iint C)'716 ini .3(a c6.08 p 1) 1( 1( 9 ( 011 Ti.L C o II 1ici and Cl i ((tela dli ,H kl wk-I (a S1.25 kill . . . . . . . . . . . .. fiish.fl Par tial 1,512.00) 1,200.00( .520.00(1 37 72 00 Toi~ \I( 1 TO AL ttil 111 )11ge..........I .......... . . ..I AND ... I .. NAIABLE ALNI, FIXIED . . . . . ... kiloai 1)11 LI i 001 ilf'l AX IIAC was ((A IiA r each k(iluogram11 offr tr\ C Csed AI HBAsei COSTS1 8,81:3.00( I 1rali I of 0 TII(IIlI111 pCcditm illko cots tAI' thA aiXl A Ais C J84.8 (abToot US.5) perliiIAnOCI i wasA' pur111 por((( A IltiliA id c Ii inIOC'A1 JamaI xvili ica~, I T. tlitr Cf O nllf/i~tca Awill prclu Irakepricu 0, IfC)r Ar\ talC 110 raini wAi111 IIO tIN 2. N r nit co .1 ,00 1 fiO 'o ..... 411 0 . I' lIr pi anini a0il aitialf bro ixiNsiAC pro11perly ANNUA It 111t i19 Y ODuCI. I O N............. I tie("C ato til _ "''( 1 I' 1.00 .5 i i i Af fry Lit 01i an11(1l 111cC (If no( lc'AA than1 4 ut filli (,i.fiiilIA(Io IClrC~I ( 6 MCOitchAl To\ n(IfCilO tl!it II oAf. WXfil 111Ceflit ilialllg'11it iti A e 1))11CIiiih of coslt fa 0 -1111110 6 pcentCCIIiltill totali IcticeA fior IlLIrAC' An( 0111o~hic'tiion ponds~i. pra fi lrCi f'(i] wat aI rCAI (1ddcated tol tiilpial (111111e in Jamitca fAsT Ast BCcause tirequenit 1par1tiatl har%'ar AAf' e1'CsentialiII tothe Alicccs (lesci fll ui 011, is l0 17 (11 fAi l i limA k t l 1 iIi based5o3 tl' til workACmng liCiA (130 AId Con CCIC~or cI iiile I herc this mllldintit ilystem,11 the tcChnl1iu is nAolt .11111((111 Niursery Pond Managem an1 CIIatifafctory at enA tile gio\ Cr11 il1 dC'pcil dC'illanfl Ililit Aource perlet ie r i . AiliIolg1l ithIas prl\ ha~tcheICries Asmall-le of' fryIi foll If'll thiri saitce I'1rIICCCrA imi ht IIC~rAc'! A ist 1 10!t. thC' 6 The IxCiii IcC I IC.- A 1 HC ,lAnfcr! f-ool li brood ponds arei mC to i ii AC1 hlOX ('A cr AIIC ce hlA isL IC 111" 1i titx ai c' A i a1 1101111 I-C Iiing toi Asize o 20 to :30grl a fcll Ci t 1(010 111l11() (II0i I Al (lIlAci i Amall IALs-1i1s( hanil- ACCC and til'he o )11(1 p~rodction. tis heC CisCCL fiii FhA- lhas Seed fish for stocking nursery ponds are harvested weekl1y from brood ponds i' a size of only -1gram. Each year T early 5 million fry are harvested from each hectare of brood ponds. N orsdor p onds are etirrentIx stocked wxith I gram fix of' T. 1 niIotica it a tieisitx\ of' approxiioatex 180,000 pei let are &the solggest(i feedini, schedule begins at a daily irate of' 11 kilogi ainm per hiecto t duiii n the first week, and is gi nadoalx incirtaxed to 40 kiloifi ai peri hectare bx tlit fooriOxxeel aod 0f0 kilo,,rins per bieetart b\x the six~th xx eel. iotal ft ed tOI)LomptiOnl USoIalix rainges Froii 1,400 to 1,90(0 kilog-ramis pcr i eetai e per crop. flarx t ting of fingerinuos liegins ax earix as wxeeks 6. and is uixtallx Completed bx xweek 9 or 10. Although~ the hiai x st biegins hx sii mm a foil pond, the finial portion of the crop is reiiiox d Lx di amiiiiig. Before lilliiii thie pond xxitlh xxateI fior the folloxx imp Cx tie, tlhe pond bottoim is oxoualix allxxwed to drx fio a fews (laxs and aii iemaiining4 puddilex aire treated xxitli i otenone to i emoxe fish. Expected xjelds aire based on iresults frim 2 3 eairthen oorIISerxponds Nanx iie, ini sizc from 0.1 to) 0.4 hectare. A, total of' 1210,200 mnale fingerlings, xxith an axverage weight of' 27 gramns, xxax harvested froiii the 4.3 hectares of nurxerxy ponds during an average cy c of 9 xxeeks, table 3. Based on tin ax crage productioii iif34,20(1 male fingeirings per hectare (101inp, a 0 xxcek cirop and~ a ttirn-arountl timec of 2 xxeteks betx\ ccii ci ops the exptected ax tiape iinntial x ied froii 1 liectaie tof) ntiist's poods is 161,00t0 niale fiiip'erl uSs xxeiehfing 220 3i "irains to each. An atddit iional ann oil 1 prouc otioi of 4,(000 kilogri ms of femiial e fingerlings is a bx pi odtiet of ioai ginal x aloe. fTic animal fet d it qoirinits airc 9 to 10 imetici toins pci heetare. The hand-selection of male fingerlings for finaf grow-ou:- is most efficiently and comfortably done in specially constructed facilities (top photo), but most privately produced fingerlings are "handsexed" in the nursery ponds (middle photo). Many small-scale farmers simply work on the pond bank where female fish (foreground of bottom photo) are discarded or used as animal feed and the males are temporarily maintained in water-filled barrals. Iii RtiI \i; 05Mx i,,E xiii i~si N V ,iS o oPiuuuko Nit 1 i( IFMsx PONDS 5 Itc'.' Initial stocit I.. de i i .c ..sits fix .... Initial size iffn Doiration of i xsle, we eks.... Finial sizt of' onak fiii.eins i................. I) 1'-i Becox ii of i ud fiiierins, tiioosaod /lu/ct e ..... t t B,3 oui iui il I noi 23 points, 0. 1 to0) 1 liettaie eath. iuiitiaiix stotked 1sed xxitOl i eti sex 11 , At tht eotd of the xelse males xxere sepal attd Iiii the P femraics Lx x isu insopetioii Tht ftemales \xere a hx ptt)(i itt iiof cIriiia \ aitie. Fry collected from brood ponds are reared in nursery ponds for approximately 2 months to a weight of 20 to 30 grams at which size an experienced pond operator can separate the males from females by visual examination. At the end of a rearing period more than 30,000 male fingerlings are normally obtained from each hectare of jrserfp ondis. I; -ii ad 44 I"- TABLE 4. MAJOR PRODUCTION COSTS FOR MALE FINGERLINGS FROM NURSERY PONDS AT MITCHELL TOWN SUPPORT FACILITY DURING 19821 Item Annual fixed costs Opportunity costs on construction capital (8% ).............................. Pond maintenance........................ Depreciation (prorated): 76-m seine, 19-mm-square mesh (3 years life; 1/4 use).............................. ... 2 tractors, 1 wagon, 1 pick-up truck (7 yr. life; 1/40 use) .................. TOTAL............................. Cost, J$/0.4 ha (1 acre) 2 . 320.00 50.00 336.00 . 410.00 1,146.00 1,612.00 2,302.00 0..... 80000 520.00 5,472.00 1,187.00 12,893.00 14,083.00 21.70 8.04 64,7003 change substantially the economics of fish farming. At present prices, the additional revenue would offset no more than 15 percent of the total cost of producing male fingerlings. Nevertheless, in spite of the low utility of nearly one-half of the yield from nursery ponds, male fingerlings are currently produced at a cost that allows food-fish production of monosex tilapia to be economically feasible for Jamaican farmers. Management of Food-fish Production Ponds In Jamaica, T. nilotica are produced commercially in earthen ponds ranging in size from less than 0.1 to 0.4 hectare. Males, weighing 20 to 30 grams each, are stocked at a density of 15,000 per hectare. With supplemental feeding the grow-out period ranges from 10 to 15 weeks. Because of occasional errors made in the visual identification of the males or contamination of the pond with females through the water supply, some reproduction occurs in most ponds. Although reproduction is limited, the percentage of unmarketable fish usually becomes unacceptably high after 15 weeks. By this time, however, the originally stocked fish have attained an average weight of 180 to 260 grams, with total production of marketable fish varying from 1,800 to 2,900 kilograms per hectare for most well-managed ponds. Recovery of initially stocked fish is seldom more than 90 percent, and usually varies between 60 and 80 percent. Neither agitators nor aerators are used, but some farmers manage water quality by an occasional flushing with fresh water. Major fish kills from oxygen depletion have occurred but are infrequent. Small-scale commercial producers often fertilize their ponds with chicken manure, a practice which usually contributes to shorter production cycles (10 to 13 weeks) and decreased feed consumption. The amount of feed to produce a ton of marketable fish may be as low as 1.0 ton in manured ponds, but 1.5 tons is normal for well-managed unfertilized ponds. Depending on marketing strategy, ponds may be harvested in a single operation or over a 2- to 3-week period. All ponds are completely drained, and any remaining puddles are treated with rotenone to eliminate all fish before refilling and beginning the next production cycle. This generalized description of food-fish management practices is based on field data from private and government-owned ponds, table 5. Annual variable costs Fry (331,000 @ $4.87/1000) ............... . Feed (3,814 kg @ $0.60/kg)................ 3 3 Pumping (29,600 m @ 06.08 per 100m Tractor fuel and maintenance (8 km/wk . @ $1.25/km) .......................... Labor (1/30 of total labor budget)........... . Guard service (1/80 of total) ............... . TOTAL ............................. TOTAL ANNUAL FIXED AND VARIABLE COSTS ................................. Per unit cost ($/100 males) ................... ($/kg of males).................. Annual production of male fingerlings: ........ . . 'See footnote of table 3. 2 In 1983 an official devaluation set the exchange rate at J$3.15/US$. 3 The incidental production of approximately 1,600 kilograms of female fingerlings is assumed to have no economic value. The total cost of producing male fingerlings in nursery ponds at Mitchell Town in 1982 was J$21.70 (about US$7.25) per hundred, table 4. Sale of the female fingerlings would have lowered cost, but attempts to market them as "soup-fish" for human consumption have not been commercially successful at this farm. At present some females are kept as future brood stock, but most are either sold cheaply to a local small-scale manufacturer of fish feed, converted to silage as an experimental protein supplement for swine, donated to a local zoo, or discarded. By late 1983, one medium-scale farmer began selling small females from his nursery ponds for J$0.66 (about US$0.22) per kilogram to a food-fish distributor, who in turn received J$1.54 (about US$0.50) from consumers. If this trend continues, the sale of "soup-fish" will benefit lower income consumers, but will not TABLE 5. FOOD-FISH PRODUCTION FROM PRIVATE AND GOVERNMENT-OWNED PONDS STOCKED WITH MONOSEx TILAPIA FINGERLINGS (ABOUT 15,000 PER HECTARE) AND RECEIVING SUPPLEMENTAL FEED Public sector Item T.P. Final weight, g ........................................... Total production, kg/ha .................................... 184 2,568 2 M.T. Average 216 2,342 Private sector All ponds Top 80% combined 220 2,419 216 2,137 3 247 2,116 Duration of crop, weeks At first harvest....... ................................... At final harvest ........................................... Recovery of initial stock, pct............................... Feed conversion, kg feed/kg fish produced.................... Number of ponds ............ ............................. M anure..:............................................... 15 15 88 1.54 7 no 12 17 53 1.54 13 no 13 16 71 1.54 no 12 71 1.25 44 most ponds 12 65 1.59 55 most ponds Mitchell Town Support Facility is used primarily as a fingerling production farm, but afew 0.7-hectare ponds were kept in food-fish production to provide a buffer against fluctuations in available fish supplies from private producers. Data are from ponds stocked in 1982. 3 Data were taken from small-scale commercial ponds (with an average surface area of 0.3 hectare each) harvested between July and December 1982. It was the first crop for many farmers, and on occasions ponds were stocked at fish densities and/or harvested on dates that greatly deviated from suggested practices. The "top 80 percent' represents the average for reasonably well managed ponds and was calculated by eliminating the least productive 20 percent from the analysis. 'Twickenham ParkTraining and Research Center. Data are from experimental ponds (0.04 hectare) operated under carefully controlled conditions in 1982. 2 REG 0 L P NN ING -70%i 6% N~os a~a~sdrain and harvest t1heir 1ood-t'ish ponds a out 3 moulhs after sl'ocking. By 1hat time, the weight of an individual 'sh exceeds 200 gramns and the total crop is more than 2 metric tons per '~care oli pond. FIG. 2. On a regional or national scale, approximately 70 percent oi pond space is used in the final grow-out of hand-selected male tilapia. The remaining 30 percent must be utilized either as brood ponds for production of initial seed fish or as nursery ponds for subsequent rearing to a sexable size. Relative Size anid Number of Brood, Nursery, and Production Ponds tO' or ( Stal' l ti is'0 fo oi- ( p115'se of natio'nal iii ii -iiiial 1 planing. a total xwater ar ea ofl1(00 hiectaies xx oiid lutilic, 24 hcto ( s ill 110 rxers po0 I i id 6 lie Lii r'5 in hi 1111( ponuds to supiiport 1 M0 hetacx iiip1 i(Iictioo - oiitlx Ii''iic 2. Fori the I ii ('ir opieration s xxith sollhieut proidiietiiii pllitlx toi reqtie ii li 1(11 o ci ~ cii 11)1 al(1 l ie liii erliiigx thc elatix e size of' 0(11xi x pondics. as chililiiitI to p triloct ion pondils, ix no t citicaiilixxe fir siill-scaile ipci atioii w xithi fe\xxotui ct ion11 Doliiis the stic king of inalc tingei lgs is a per1iodic cx crit. lIn thix iase, i(t5 suext ii the si/c of the otrscrx ponds is sucih that the elitire %ieldfromii oile crop is juist suitfiiciint to colijii( tI lx xtock a 0o dulctionl piond. lFo accoinilisli tisx iuirserx ponds~i shuld~ lie aplirox1iiateix (0to 50 lperent islai as the pr oduictioni polucix. gt Sice the priodlictiiii cx ce in niii sel ponids is siloltui thain that ofi focichfish puilcs ahottI 1 xN -sos ai'ut t6 xxeeks) the uiiioer oif llorer xccxjid ithiuCdedl is less than thle iiiiiiil i i rocluctiioii poiilds Txx 0 usrx'in po1 ds, ech about halfi as I arige ax the prioducltio0n i ponsar c'l dcd to suporiiit thriee pt odulCtiii'i ponlds :3 1iiii xcix ponds are nechct for 4 produ ction lilits inch or 4 xxiii he needed 3 for a 5-prociiii poiid opecritionl. Sin11ce broodil pondits mo(st he' 1aiti al hai escxted onl a ncear xweeklx l bais toi remain p1 ldilctix c this txypc ii)ponld is iixtificcl oilx x lic'ii cemndc for fi : is also oii i near xx ('klx h~axis, Wihcl ii stified, hiowxecxer, a i illntlmr of txxo hirood pondcs is requ~lired( fir a clni'i 0011x supp)lx of' fix\ . Tis is chdic tol thec 2 mll tli ciiiii idiietix C' periodc (,eh bir1111 ponid 1hctxx ec the finaliization of one c\ cle andi iii thce hc'giniii( of' par tial hiarx cstiiug of' lix iii tlic followsing cxycl. Recent Introduction of~ a More Intensive Technology Flie l~-o i0( nrserx pIroduci(tioni pid nilnaiiemiiiii xstem descibcet aluii c, is piracticed byxiiost tilapiia farmers in Jamiiciia. The leveil of'si)opiis tieatiiin is a c'omprmis 0 e' o10 tile dies ire for high 1il productne and1 iiaxiiii m11 1 n econom11ic retun an( d onl tiieried to technicell s kills to lexe ls (0111patibl)1e wxith a xmall- O)1I e1di n) -seale faiie \\(1x little prlex 1115 expe i ence in aquacoltii e. itli B\ 1982, the haxie infrastriiictuiire and eritail i iai ke ting and pr odiut ion cihar acteistics iif apia farmiiing had I)eei suifficienutlx ii prmxen tou attract the intei ext of' prix ate comupaniesx hax ilg thle ixvextmet icapital and teclhnical exper tise needed hom in1oie intenl six e piroducetion sxstenux. Ini I98:3 txxo such coiaiiiilis lIeganl Collsti l1(t 11)1 (if farms deigi'lliii.'o a iiitii iif 100(hectares of xwatei. 1produtctioni at hoii of thexe fiarmx is h)axc ( oii tile 1)ilx enitni of the g'iant Ii eshxxater p1 axx Wlci obracli1111 r oxen/Ni ii) and a ni tilapia lix br1(T. i otico x~ T. acil producced froiii Strains selected in lIrael i exoitiiig in a neat I\ all ioaie lnix n pi iignx . T1he oither fLii isi alxo dexigned fiox ONCoit~ri C lt eiiiphaxis IS onl tihupia. It pi otiieed ncai Ix all hnale t'iapia lyix iii( i ii simiilar strains iii thn sanme two species but, in addtitionii ix mx i aluating the mnarket xah i oie a .eltilapia f of uncertain origain 1)iissilblN T. 111055111 /icaT i. i licaT'F. Ii )11(11ianii . Malde fi114('iliii g of thix red tilapia are o' utaicd lx xsex rex ci sal (oral adin~istratioi ioi a miale 0 1101-111011C tlff11(U gh fee'd t1 1 celt hatched Iil"xbeforie their. sx 01rgansx Ila\x e toiiii ('(1. Plans at tI is farmo are to groxx thle iiiajoiii of' the tilatuia to a xize oif .500 gi aix iin 0.1- to 0.2- hectare squairC eairthenu pondsi stocked ait fli ic(lesitles x eedhiji" 11)0),000 tper hectare. Each pond has two to four continuously operating electric paddle wheels arranged to keep the water flowing in a circular pattern. Approximately 10 to 15 percent of the pond water is removed by periodically opening the valve of a 150-millimeter drain pipe that extends to the center of the pond where the circular water flow has concentrated detrital material. Both of these farms began production in 1984. Their technology is not being promoted by project personnel, but they will be closely monitored to evaluate the applicability of the production system for other Jamacian fish farms, and as a potential source of male fingerlings for small-scale farmers. FISH PRODUCTION EXTENSION BRANCH Organizational Structure Prior to project implementation, some felt that General Agriculture Extension Agents having special training in fish farming might effectively transfer the new technology to farmers in the course of their traditional extension activities. Nevertheless, the project was organized with its own independent Extension Branch. This decision was based on two major considerations. First, since the technology was completely foreign to most farmers, new producers would require advice and reinforcement more frequently than for other agricultural activities. Second, because of the rapid evolution of fish farming and the relative inexperience of early extension agents, the opportunity for consultation with more experienced personnel should be increased by making an extension unit an integral part of the project. The organizational structure of the Extension Branch and its relationship with other technical sections of the project are diagrammed in figure 3. The extension staff includes the majority of the technical personnel assigned to the project. As of January 1983, 16 of the 26 Jamaican technical personnel were working full-time in extension. In addition to the Jamaican staff, each of the two regional extension officers served as counterparts to extension specialists from the technical assistance team, and 4 to 6 Peace Corp Volunteers were active as extension officers through most of 1982. ~~----1- -----0----- -- It r- T-lI oUIJI III--- .AI5IIHIlUll IIIUI" - -- uF 1 I West East I Marketing Specialist Parish Extension Officers Westmoreland (2) I Hanover / Parish Extension Officers Clarendon (2) St. Catherine (2) J Mitchell Town Support Facility Farm Manager " I I I---------I - St. Thomas (1) Portland/St.Mary Asst.FarmMgr. ASt.JamesM(1) Field Staff (31)i I St. Elizabeth (I) II ,.--.. L ..----------------- Stocking Specialist (1) Pt Pond Construction Spec. (2) - - - I . . I I . ----. . . ----.. ...-. m..,..,.. . FIG. 3. Organizational structure of the fish production Extension Branch in 1983 and its relationship with other technical personnel in the project. (Solid lines indicate vertical hierarchy of authority. Dashed lines indicate critical horizontal lines of communication for routine operations.) 10 Exes'lc Seirvflas Provided l> ci cid state l c It lt til cn I ' fiit i 'llm l ill"c xxl Tali tica. lici I' r! iltx '> iiilxi cod broadil suppttl \\xI' tx ased oil fouirt St'c'oi co il wrtohls ' 51,c c'sfci ioi t lllc'l's wacr ci c'lli lct t1 lixiia 111114 ( l. ciccli 11 i a0iti I ttcgX s 1m 1 andict el' ti'ttl 1 th S I tI 1p itsii~mnr m aqtf-d c5Illod Will Y.1Ci tij~ilC 111)11.Js Il Thrdisomettiof, thes 5(1 ski0 II ricc esalaticnical1( ntit. cl ~oc n i fttsh1 malt' 1l,,c1 i1s inCI 'd tO thll ' Seed Stuck. A4>t~lt tei l 16t1 Extension Branch'i ll('tinllt more colifiltx. IIIthc' earl\ role 01 tlit ere s51 toctked xxith finringos ( stagts- all smlsale~t fish poli 'tenio(0perIsonn0e1 ced at thie lPr111 ct suip port facilitics and ex1 pii0( ate xcicle. Tlthh mii stoekino( of api\ a flIOdcl dIIli\ Cr ill plo~ct't hax not \\ xxs\\ithIoot (hlage ao lllixtubsequent stockingsx wer Ile ilx xxitli 1Myinli0t oft ii ll i" at 011(1 1)1cI'1). stil 11i i fingtrhiiog pixate fa mei, miotixvated b\x Simos that eai lxperI iod Coplcted h\ shortac'>s aud tue phaised I 011 ixal of till stilbiit 1accepte do dret Iespoiosibilitx 01I thlt produIc tioln Id 198:3 1 a\c hong' 'i 01 Since late 1982, atstotkinig spel]ist coiltin 11(5 to make Idcl tin gclr ile 11( id ir-p iSil order ofh 1xei00 oiI' ticii\en 11 lins, buit project tiirulst has Li oil toxxard thie dexvelopmeint of the own s c d. 1 ioi iri is, illitdix eapacit\ iii suplx thi fariiii' cu Lx ixo ii xcXed" ol ix bx pi ojt e otm l, ait ' iixx also hiande 'i ox ad bx smid] gloi lp5itrained b\x somea il, the Lii ilili ' theol's t ilt contr act the1ir s(nic es tio >111111x do Proi proii1 (ct pel'soic Pcx onds. clillcls xxith nurui ih I' cd> ale lxxf & Feed! Supplies. Tlxxi iiaix iliamif'iie (iI 1111 distibiuiteci throtw ni'iorima l mnercia c'i11(1is. 'The( F stenxton local1(1 iiiltilosoxiinuiitiallx toicill l Br'anch \\xxas ilstrmna.l xxitl distribuitors to calx tihc produciit andc in coiii) clittiIppix Ilacc~ ilch ea tim lit'o tilt 151 proj~c'ct. T 1111ti1S s - ie 1)11. iame ioi oil li I i it'li fi IIt (1p ili i itii 11111' tefit - tbIf t jct e liclll ind It ll 11(1ix ittl ithe Ptrtod iuctixn i11 of111( a- lx osi Moe a\ailetnrlit f i'lx. ab refolinement pr\id e alitii Ilil p(' Il" of fii ill tillSr xaci \t stli -rac Is. R outine F'arm V~8is 'I'll c' I-i entx of' x isi tx bx all t'xteil 5001 ftt 0 sperili ce ofi ie( fo inllo \\it' xth aie 1> liidi'anc Ii Officct WXecklx or hi-xx ckix x txitx are iixiolxN requlired fill ioxx farmes 01111 lili ii ll 1' 01 Utles iii dinii and th ( ablex ti adjusit ett il i\ftei txxi i I 1114 reillic llo115 ilogI tol xx(1101 Ijiaiitx cioditions. Haisvest Schedules. As thet ex'p'ctecd LIIxtt cdaIte nearl Sod ip11 oifficeri max aiiitL' ii o 110 ti) estimaitte fisize/t. The textenioni I go'llx 1asis5 th it' 110(f a prolct thio fLi111(1 toi 011111aIt 1111a1 llt 11opertion,11 i (aillit a cosit that (1 l 01 amiortizaion xxte Itt' (t xx olfIl It ter aid cidetprec iatioil j 86.018 pet 1001cuic i tell aicci ~Iip at~((Il ~~~ ~Thec o\ ca eaiminate xz' r.ei creat stau'inc>ia "c". to u ta "e aaacaimn ips csor lc':""apeii' motivae cc "'eae Hll be mii'k Sit'C lso Id. oh i trmeiit ocoirc-a s shareeps and t coniton pr -sector parti itin t1111111 r Moxt small-]stilt' fai ili are IC's ilh to sthudlI 'c tlatix lix off iccr Sn fl i' Aiis tIi tlliillllllii 1 blt dxci stclxsio illiil a w1 ith o ltheri proues t'' T11 lari 111toti he iIs tillaq' selt' r "tt illS ch i b.\c iloltl> ixi del cait'i tidi ill mec] 111cr~~II, Ilr a-I0 1cidtlix labor.1 .-IS1(1t i]I stit of1' It tailt' t 1110 1 hes s oil t Illiltt' rlii xkis ii scillg allcilllo>, St \t1 f1111 ]ti~ siiickin"i. aid flai t iO tt 88le161 tiltn4'ix Wit th h o ('ow i Iilc F O.kcstociffcciiiac I)eae Pe ItO aucllit p-,Mt(' S l IM~l m thx ill 1 idx1 I O 1 il l nd time.tc '8,OWO 's ltialt Iil C1)s~ca17 P sls1'-" ''i"ntis!"djla eFil scale producers. Some of the earliest trainees were given a subsistence allowance during their 4-week period of training. Increasing numbers are now willing to forego a subsistence allowance to increase their employment opportunities. Extension Branch makes no employment commitments to these individuals, but most have obtained permanent positions on private farms or at least decreased their previous level of underemployment. Large-scale fish farms have also benefited from training received by project personnel. Private producers are not discouraged from hiring experienced technical personnel from the project. The formal and on-the-job training which makes these employees increasingly valuable as project staff also prepares them as competent managers for private fish farms. By December 1983, four of the technical staff had left the project to accept management positions at medium- and large-scale fish farms. FOOD-FISH PRODUCTION ON PRIVATE AND PUBLIC FARMS Rates of Production Rates of production from private commercial ponds are variable (CV = 37 percent in 1982). A producer has little control over some sources of variation, but other factors, such as feeding schedules, water quality, and theft, depend largely on his motivation and management skills. Reliable production data were obtained from 55 of 71 recorded private harvests in eastern Jamaica from July to December 1982. The median yield of marketable fish with average weights exceeding 200 grams was 2,100 kilograms per hectare. Approximately 20 percent of the harvests were less than 1,500 kilograms per hectare and more than 20 percent were greater than 2,600 kilograms per hectare, figure 4. Annual rates of production are greatly affected by the duration of the unproductive period between harvest and subsequent restocking, especially for this type of farming with its short grow-out period. The median turn-around time between crops for the above group was 6 weeks for all ponds combined and 5 weeks for "reasonably well managed ponds" (the top 80 percent). Based on actual yields, crop duration, and turnaround time between crops, the 10-C U) '4- O Actual A Targeted 9 Top 80% of small-scale commercial farmers a, 0 4w 0 '4-. _ o I, % o .Eo 8 U)2 "0c U - 0.CE C 71- ~ 61-P All small-scale ' 0 a 0 4 commercial farmers 5 I I I I I I I I I 2 3 4 5 6 7 8 Turnaround time between fish crops (weeks) FIG. 5. Average annual rate of fish production, as affected by actual and targeted turnaround times between fish crops. (Based on actual crop duration and yields from 55 small-scale private ponds.) annual rates of production were 6,100 kilograms per hectare for all ponds combined and 7,500 kilograms per hectare for reasonably well managed ponds. Annual yields from small-scale commercial ponds would have been 20 to 25 percent higher if the "down-time" between fish crops had been kept to 2 weeks, figure 5. The most common reasons for delayed restocking were fingerling shortages and delays in refilling ponds with water. 100 IA 90 -7 80 -Average . pond size O.3 ha Average duration of grow-out period 12 weeks Total Food-fish Production Total annual production of tilapia in Jamaica from 1979 through 1983 from both public and private sectors is presented in figure 6. During the first years of the project, 1979 and 1980, there was little net growth. Total production held at 16 to 18 tons per annum as the public sector phased out of food-fish production at approximately the same rate as private producers entered the industry. In 1981, however, total production doubled to about 36 tons, with growth occurring in both public and private sectors. In 1982 the harvest of 126 tons more than tripled levels of the previous year, and in 1983 output increased by an additional 27 percent to 160 tons. Pond construction activity during late 1983 indicated that growth will continue at least through 1984. 70 "oa 60 C 0C:, 5 0 ---- - I - --- 0 00 50 30 ', I I --- i.s . I I W1 0 . 20 --- i0 ,_ I -----.---. ° I0 U i &- 1,000 1,500 2,000 2,500 3,000 3,500 Yield per crop (kg/ha) FIG. 4. Frequency distribution of fish yields from small-scale commercial ponds. (Based on data from 55 private ponds harvested between July and December 1982.) 12 Small-scale Private Producers. During the first 3 years of the project, 1979 through 1981, all privately produced food-fish came from small-scale fish farms, typically consisting of one or two production ponds less than 0.5 hectare each. Growth was relatively constant during this period, gradually increasing from less than 2 tons in 1979 to 27 tons in 1981. The most rapid growth of small-scale 140- PRIVATE SECTOR 20Large-scale forms Small- scale farms 100 J 8013 so 40k ir.c" a L:o a ninua pe 7a~ _F is, C SUMptiOn E ve n wvi L s ,i r t~~ of fish products in Jamaica is nearly 10 kilograms. The economic feasibilIity of tilapia farming is enhanced by enth usilastic acceptance of this fish. The occasional larger tilapia weighing more than a kilogram, as above, may be taken first, but pan size, 200-gram fjsh are readily accepted. ish 11)01.1 6' 0(1 ctioll at tile 1))(1101 fcd ineted t N (( ftlll I \i)aN, to1 proN >o lm icic PUBLIC SECTOR I I 1979 1980 I 1981 Yeaor i i stabN at\ fish1 supplis t od) ohditiuos ac trl 1982 1983 ( annu1al I at( of' p ocloctioOl (lorillgth sicond11( olimc~plntlN . thE] 1 tols, semcster of 19,S dleclineod to IS101) bIlw lar~l\ I p ae b0(1I 0\Jpalc i tINr1odilctioll to O1ct t1he 1 (I I141>laisfon piIrnatc 1111 IN FIG. 6. Total production of marketable tilapia in public and private sectors from 1979 to 1983. fish fal rin~g N)s creased 1101)) toi1s, tilalpia. Mch r1 egistere 1 10111 odu111( 1)0is plill polld >. No flittii (INN-otl 1(10(1 1 p1(11 sector tlelilitioN ]] orPb ic lI 1981-1982 -hid 111(1 whn) 111 i 1(llctiolliIIproduced0( of existOig 27 to 66 tonls> 111 1983, produl~ctIin increased( to S5 oNf all p0lN atcix expa1nsion1 acoun~ltimlg tor 11001\ IN ECONOMICS OF SMALL-SCALE T1LAP FANVIN 1)f thl)is 111 lrtlcoite(1 roNNt Priv ate Producers. The] Costs and Returns !or a Small-scale 7isi Favm Large-scale 1c9S2 Dil-ll first 1(0 go >1410 fish tair o 110011) 1)110 11)1(1 (11(l to co)me) 111)11) that 1)111 hENc~ oil( incrase f1illO'elli IIo~iNtiol)1l1111> FItelii 1 tile last (1 111))1t1)N oftbat Na 0. 1 c~mliollof' 111111 tile i00r( >01110 Ocltelgry (0111111 oc Ic thle 01)01111(11 1N)(N\ec110101114 0011(1 (lot IliNc kept pace N)itil N)ithl ve w111(1 ha1( c1)1)111ct)ci a gIN)ot i1ll1 the p(IN Itc NOcto and N) PoIeiIa PIIN 01)1111 poio) to icchIce its u l cl eINc lOt ill tihe p111 tl\ I (fO prodution as11lpccts 1(111110111)11 Large-sc>al e product ion 111111 N exnded)(O 1 to 1111110thanl 8 1(ctare1 ( 1N f'N 11 I 1954, tile greatest 1)) At leas01tili the 011NN ill fish1>1 prodctio 1)00ur1ed i1i tilarge de1in Qi~ >sc1le category Nof' 1141 fliurc 11(1>wt Fof'tll-cc fish F11)l" wi 152th GoN ernnlent-mvned Farms. Tile earlIN dulcticon at tie govenm0 11citt facilities i -fs pro-1 1111)1 1from1 11 a1n1111l 1rIte of 'eao 2(1 depend1> NN 11 oilcte tlc1hn11i>cal Nill ilgt (11rtc~ their 11 1 cc 1and ioc eoilc poslilic\ of'11 a> 111)ilespo1 teste h 1980) Nwas ohcipii1InariN to I chaninl 011111 fctors nd11 to em phlasis, 11)11 fod fish1 p1> ro-11 i Mar1- 11)) Fot 11(11 >11,,c Nis oat )\ ll i ha b 11\ >011 finger ling produti~ction). NNNitc 1(1 that (101r1essed loo-cis to) re0(10 o ish £11 f i)l I cr l il01 Iciaa-llirntlNjl~ ste tSifed il till c tb a11106 111ultureO NjlOcl( xn)1 tile i((Xc duc11(1 1 ((jl ic \\lilcill1, il_ ct c ti pcte iclloili fo0mers had re4alizedo> o No ).c thei0r( 11(11>0 io Ictheir trl1)litionl chlils tile go)vernilliit-wnlcd Agicliturl Nlrsl p)01(> 1 1 . 5101 1(1NO 1 cos1ts0 mi1 d reunsfr11malscl tieaiONI 1~rl giwithI\( tok11 dilctiolln and ouItpu(t graual 1c14in lcreased tIN110 moc ta 1 :)31 1)1> 11111 toi m111982 11) 1111(198. The pilOn iI jutificaltion 1111 reb11111(111g i ponds, but he produces all the male fingerlings needed for his production ponds. Annual production of food-fish from this farmn is slightly less tban 10 metric tons (about 4 tons per hectare of land). Unit costs are J$3.50 (about US$1. 16) per kilogram of food fish. Approximately one-third of all costs is in fingerling production (about US$0.06 per male fingerling). Net annual returns to laud and managemnent are J$4,450 (about US$618/bectare of land). For family-operated farms, net returns to land, management, and labor (4.5 person-days per week) are J$10,388 (about US$1,440 per hectare). If ponds can be filled by gravity, instead of pumping, net annual returns are further increased by J$6, 257 (about US$870 per hectare of land). Jamaica, fish farming was originally emphasized as an alternate use for marginal or swampy crop land, but good returns and short crop times (12 to 15 weeks) enhanced the farm's cash flow, and made it a viable competitor with traditional farm crops. In table 7, the past and present economics of small-scale tilapia farming are compared with some common smTall-scale row crops and livestock enterprises. Early profit margins were very high for fish farming due to subsidized material and labor inputs from the-project to convince Fish Farming and Competitive Land Uses Land use is influenced by physical and social factors, but economic considerations, such as initial credit outlay, operational costs, labor requirements, and net profits, are the prime motivators. In 6. cautious farmers to enter tbe new industry. However, ev en after the gradual removal of most subsidies, tilapia farming continues to be competitive with other agricultural activities. For relatively small-scale animal husbandry, tilapia farming is less capital-intensive tban hog or broiler fattening. It requires more labor per unit of capital, but returns to labor and management are bigher. Tbus, tilapia farming has addressed one of the prime goals of tbe Government's agricultural development plan: a relatively low-capital but bigb-labor type of livestock enterprise. TABLE EXPECTED GOSTS AND RETURNS FOR SMALL-SCALE PRODUCTIUN OF TiLAPJA NILUTICA uN 2.4 HEC.TARES U! LAND CONTAINING THREE 0.4 HECTARE FOOD-FISH PONDS SUPPORTED By Two 0.2-HECTARE NURSERY PONDS' J$2 Item Nursery (20.2ha) Food-fish (3x0. 4-ha) 15, 000. 00 1,000.00 9,il000. 00 25,000.00 1,935.25 333.33 1,285.71 300.00 400.00 41,254.29 9,652.50 11,476.27 341.25 1,170.00 420.00 1,611.56 487.50 1,399.37 3,498.43 16,558.45 2,483.77 191,042.22 23,296.26 5,97.65 241,055.94 3,608.39 27,664.33 3l4,159.74 35 -35 Combined (2.4 ha of land) 21,000.00 2,500.00 121,000.00 35,000.00 2,748.05 833.33 11,714.28 400.00 800.00 6,495.66 1,L654.54 11,711.70 2,085.35 506.70 1,)560.00 600.00 INITIAL COSTS Pond construction............................ ...................... 6,000.00 Equipment (seines, holding . cages, . bucket, . scale, . dipnets)................... 1,500.00 3 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Water pump, 135 m /hr 3,000.00 Total............................................................. 10,000.00 ANNUAL FIXED COSTS Average interest on loan for initial costs amortized at 12% per annum over 5 years............................................. 812.80 Depreciation on equipnment . (3 . years' . life)................................. 500.00 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pump (7 years' life) 428.57 Pond maintenance ........... ,*.......................100.00 Use of existing storage structure........................................ 400.00 Total Fixed Costs.................................................... 2,241.37 ANNUAL VARIABLE COSTS Fry for nursery (@ J$5.00/1000)........................................ 1,654.54 Feed (@T J$600.OO/tnn)................................................... 2,059.20 Diesel fuel for pump (@ J$0.99/1)...................................... 609.08 Chemical fertilizer (@ J$3.50/ton)....................................... 165.45 Pump maintenance................................................... 390.00 Transportation for feed and fertilizer..................................... 180.00 Labor (@ J$25/man-day) Feeding (2.0 hours/day)............................................. 343.80 Harvesting (2 man-days/pond)........................................ 472.73 647.00 Hanid-sexing (J$1.00/100)............................................. Misc. (40% of total)................................................. 975.69 Subtotal, labor (237 man-days)....................................... 2,439.22 Subtotal Variable Costs ....... .............. ''... **..............7,479.49 Interest on production capital (15% per annum)........................... 1,124.62 Total Variable Costs.................................................. 8,622.11 TOTAL ANNUAL FIXED AND VARIABLE COSTS........................ 10,863.48 UNIT COST 16.79(J$/100 male fingerlings)............................................... (J$/kg food-fish).. . . . . . . . . . .*. . . . . . ...-4 NET ANNUAL RETURNS (@ pond-bank selling price of J$3.96/kg) To land and management5 . 5 .. . .. .. .. ... . .. . .. . . ... . ... . . . . . .. . . .. .. .. .. .. .. .. .. . . . . . . .. .. . . .. .. .. .. .. .. .. .. .. . . . . . . . . . 6 To land, labor management . . . . . .....- . . . . . . . . . . . . . - 10,387.91 4,450.26 'The econonmic analyses (late 1982) are based un a pruductiun systein which each 0. 2-hectare nlursery pund pruduces an average of 6,850 male fingerlings in an 11-week production cycle (including 2 weeks turnaround time); annual production is therefore approximately 64,700 male fingerlings. This provides a 10 percent margin of safety in stocking each of the 0.4-hectare food-fish ponds with 6, 000 male fingerlings every 16 weeks. Total annual production of 220 grams food -fish = 2500 kilograms per hectare x 3. 25 crops peryear x 1. 2 hectares = 9750okilograms. Based on actual production statistics from 55 private ponds harvested between July and December 1982, an average annual fish yield of four out of five small-scale fish farmers would equal or surpass this level of production provided average "down-time" between crops did not exceed 4 weeks (i.e. ponds without fish more than 25 percent of the year). 2 3 J$2.70 to 3.201US$ for most business transactions; by late 1983, an official devaluation set the exchange rate at J$3. 15/US$. Pump, costing $16,000.00, is used 75 percent in fish farming activities. Initial cost, depreciation, and maintenance are prorated accordingly. 'As of July 1983, most small-scale farmers were receiving J$4.40 per kilogram for fish at pond bank. Indicated net returns are therefore conservative. If. pumping is not required, net returns are increased by J$6,257. 00. ,'Approximately 1/4 - time manager is needed at this scale of operation. 6 An average of 4.5 man-days per week; 1/4 to 1/3 of the work is best accomplished by a team of 2 or 3 persons. Therefore, at this level, 2 or 3 part-time workers are needed to meet labor requirements. 14 TABLE 7. ANNUAL COSTS AND RETURNS FOR SMALL-SCALE PRODUCTION OF TILAPIA, SOME COMMON AGRICULTURAL CROPS, AND LIVESTOCK IN JAMAICA' Item Total costs, excluding labor (J$1000) .......... Net return (J$1000) to Land, labor, and management. .............. . . . . 6. . . . . . . . . . Labor and management Labor requirement (man-days) ................ Net return per unit of labor: (J$/man-day)' ... Crop duration (weeks) ....................... Tilapia (1 ha) Subsidized Independent 19812 19833 Sugar cane Vegetable and root crops (1 ha) Red Sweet Negro cassava yam beans Green corn Livestock 30-hog 1000-broiler unit' unit 4 6.6 . . 11.7 4.3 3.3 99 22 15 2.3 1.1 0 158 neg. 52 1.4 2.6 1.6 86 17 52 16.8 5.5 4.5 336 8 26 1.5 8.0 7.0 232 30 26 2.2 2.9 1.9 138 12 18 15.2 2.8 2.6 45 25 26 16.7 2.0 1.9 23 9 13 . 6.9 5.9 81 65 15 'Cost analyses of row crops, livestock, and "subsidized 1981" tilapia budget are adapted from F.E. Ross, 1983. An analysis of the economic viability of the Government of Jamaica/USAID Inland Fisheries Project. Ph.D. Dissertation, Auburn University, Alabama, 193 pp. 2 Based on a typical farm in 1981 with two 0.5-hectare production ponds that were stocked and harvested at a cost subsidized approximately 50 percent by the Government. 3 Based on management practices promoted in late 1982 for a 2.4-hectare plot with 1.6 hectares of ponds constructed to produce its own fingerlings and harvest its own food-fish (see table 6). 4 The scales of operation indicated are approximately the maximum if initial costs are to be kept in the range of eligibility for special low interest loans to small farmers. 5With purchase of 20-kilogram weaners and fattening to 90 kilograms. 6Assuming an annual return to land of 8 percent of equity value of J$12,500 per hectare. 'After subtracting a minimum return to management for risk of operating capital of 10 percent of total annual costs. Tilapia farming is more capital-intensive and equally or less labor-intensive than most common plant crops, but if investment capital can be acquired, returns are generally higher. The significant exception to this generality is cultivation of red beans, which would seem more appropriate for small-scale farmers with limited capital and excess labor potential. It should be pointed out that the two main economically competitive enterprises to fish farming, hog fattening and red beans production, have longer crop times, which can strain a small farmer's cash flow, and are very market-sensitive, with periodic gluts causing a fall in market prices of up to 30 percent. At current production levels tilapia has not been market sensitive. The low catch and high price of the inshore marine fishery, combined with the foreign exchange-related restriction on imported fish, have steadily driven up the price of pond-raised fish. Further, fish gluts are not expected in the near future. Fingerling Production in the Private Sector During the first half of 1980, male fingerlings were produced at the government support facilities at an annual rate of 342,000. With the exception of the second semester of 1981, when T. mossambica was being replaced by T. nilotica, their output has steadily increased, surpassing 800,000 in 1983. From the beginning, however, this growth rate was not sufficient to meet demands. Fingerling shortages soon became the principal limiting factor to further expansion in the private sector. In spite of this shortage, in 1981 farmers were not encouraged to produce their own fingerlings since little production and cost data were then available on the newly adopted brood and nursery pond management practices. By early 1982, sufficient data had been accumulated, and the subsidy on male fingerlings was reduced by doubling the charge from J$5.00 to $10.00 per 100. Where conditions permitted, farmers already in production were encouraged to include a nursery pond in their operation, and the policy was adopted that new farms with at least 0.4 hectare (1 acre) in food-fish production should be designed to produce their own male fingerlings. In 1983, the last of the fingerling subsidy was removed when the charge was again doubled to J$20.00 per hundred male fingerlings. During 1982, more than 1.2 million male fingerlings were produced in the public and private sectors. This was more than double the levels registered in 1980 and 1981. By the last 6 months of 1983, the rate of production had climbed to 2.0 million male fingerlings per annum. Although production at the government facilities had increased, the principal reason for the sudden growth was the participation of private producers, figure 7. Based on the total number of male fingerlings stocked in production ponds, the contribution of the private sector increased from zero in 1980 to 14 percent in 1981, 43 percent in 1982, and 54 percent in 1983. The present response of private producers suggests that they will continue to accept greater responsibility in the area of fingerling production. Capital Availability for Fish Farming Through the lobbying of the project, the Government of Jamaica has added small-scale fish farming to the list of enterprises eligible for low interest (9 percent) financing through an arm of the Agricultural Credit Bank. In these cases, a lien is taken on the crop, and the farmer has 5 years to repay the original loan. The demonstrated profitability of fish farming has also convinced the commercial banks to provide loans for development of medium-scale fish farms. These loans attract a higher interest rate of 15 percent, but have a 10-year loan repayment period. A signicant number of small-scale farmers have upgraded to the medium-scale category via this mechanism. TREND TOWARD GREATER INDEPENDENCE OF PRIVATE PRODUCERS Less than 5 years ago few Jamaican farmers seriously considered the prospect of semi-intensive fish farming. At that stage, special enticements were needed to convince serious farmers to transfer land from a traditional and proven agricultural use to a new highinvestment type of animal husbandry. Therefore, during the early stages many inputs were supplied by the Project. However, as production gradually expanded, the capacity to provide the required inputs increasingly became the limiting factor for continued growth. Fortunately, by this time the farmers had developed confidence in the economic viability of fish farming and a willingness to participate in all aspects of production. Harvesting of Private Ponds Fish harvesting equipment was not initially available to farmers since there had been no previous history of commercial fish farming in Jamaica. This equipment was to be purchased by the project and resold to producers at cost. All larger producers were to harvest their own ponds, and the small-scale producers, lacking the econ15 1983, most smaH-scale commercial producers of tilapia were rearing their osin male fingerlings. This farmn 3.2-hectare nursery ponds to providerhe required fingerlings for three 0.4-hecfare grow-out ponds. B as de signed wiTh Io omics ol scale to jostfx the purchase of harvesting gear, would 000 2,oo 1,800 5,00 0 -o L UM tq or "-0 hi ~00 o ~c ~0- &*N o cx 0 _ _ cx o Z C>. 00 i4 0IsC 3 % on go 0- 2 0 0 40 20 0 198 S 32 983 Yeacr RG.0. Production of male fingerlings by public and private sectors either share the cost with other larmers or arrane to have their ponds harvested on a contract basis. The first shipment of harxesting cquipment arrived in late 1982. Prior to this, the three largest producers were sold or !ent existing project gear on an emergency basis. Some small-scale producers harvested their ponds by draining, but most small commercial ponds xcre harvested with project equipment by project personnel. The capacity to provide this service was net seriousl 1axed until late 1982 when the weekly harvest schedules often included 3 to 5 production ponds. For promotional reasons, small commercial production ponds were initially harvested free of charge. However, with the availability of additional harvest equipment and the practical necessity for greater private sector involvement in harvesting, a harvesting charge was imposed to economically motivate he purchase of equipment. The harvesting charge was bxased on expected costs, plus a 28 percent return on fixed and variable costs, for a potential to insmall-scale processor-distributor xvho vertically 4ntegates clude contract harvesting in his operation. Accordingly, the harvesting charge was set at J$22.00 (about S$?.50) per 100 kilograms of fish. Labor and transportation costs accounted lr approximately 40 and 20 percent of this charge. Farmers who provide these inputs themselves paid a proportionally reduced charge. A farmer -with a well-constructed pond and an immediate demand for the entire fish crop can harvest at a considerably lower cost by simply draining his pond for a complete harvest. However, if marketing conditions require several partial harvests over a 2- or 3-week period, he must have a seine and minor related equipment. The purchase of this gear is justified only if the farmer has attained a minimum economy of scale. A farmer with at least 0.6 hectare of water can harvest using his own equipment more cheaply than by contracting for harvest. The initiation of the harvesting charge, and the rate at whIch it was set, were part of project strategy to motivate pr-vate sector from 1980 to 1983. (Fgures within graph indicate percentage of total production during a given year.) involvement. By late 1983 most of the equipment had been sold to fish farmers, and another shipment had been requested. Several producers now harvest their own fish, but a privatc contractharvesting operation has not been established. MARKETING FARM-RAISED TILAPIA Related Supply and Demand Considerations Jamaicans traditionally consume large quantities of fish. Per capita consumption in 1973 was 14 kilograms when 12,000 tons were harvested from local marine and freshwater sources (Ministry of Agriculture estimate) and an additional 19,050 tons of fresh and processed fish were imported (Jamaica National Planning Agency). By 1982, however, per capita consumption of fish had fallen to 9 kilos, as local catch fell to 8,000 tons and imports were held to 13,000 tons due to a deteriorating balance of payments situation, while the population increased by 9.2 percent over that 9-year period. Assuming that demand had been satisfied in 1973 and that it grew in proportion with the population, national demand for fish in 1982 was 34,000 tons, more than 12,000 tons greater than available supplies. The inshore marine fishing grounds of Jamaica are considered to be overfished. This, in addition to the high foreign exchange requirements for fisheries imports (e.g. boats, outboard motors, fuel, and replacements parts), offers little hope that the capture fishery will fill the existing unsatisfied demand. Similarly, greatly increased importation of fish is unlikely in the near future as prices are rising on the world market and the foreign exchange deficit continues. Increased annual production of tilapia, from 15 tons in 1980 to 160 tons in 1983, indicates that farm-raised tilapia larger than 150 grams have been well received by the Jamaican consumer. The freshness of the product has been an important factor since a high percentage is sold live with most of the remainder being marketed heads-on after gutting, degilling, and scaling. Local acceptance of tilapia was further reflected by a panel of local restauranteurs that rated it higher than imported snapper and grouper in odor, flavor, and texture, table 8. Problems with Pre-existing Distribution Channels for Fish During project planning, the distribution and marketing of tilapia were not expected to be a problem. Much of the marine catch had traditionally been moved through a network of smallscale middlemen, termed "higglers," and this system was expected to readily absorb the production from small farmers. In addition, arrangements had been made with the central government marketing body, the Agricultural Marketing Corporation (A. M.C.), to purchase and collect fish from the larger farmers at the pond bank. As agreed, quantities of fish in excess of 500 kilograms were initially collected by A. M.C. in refrigerated trucks and delivered to its fish processing plant in Kingston. Several factors had been overlooked or not anticipated in the early project planning. First, the capture, processing, and distribution of a high percentage of the inshore catch had been integrated into family operations, thus greatly reducing the risk of non-payment for fish taken on consignment. The families of most new fish farmers had no experience in marketing fish and no desire to do so, and, in addition, were often unwilling to extend credit to the highly mobile higglers. Second, the irregular harvest and widely scattered fish ponds made the organization of a timely and systematic distribution system difficult. Third, the A. M.C., which ran into a financial deficit in its general marketing operations, was closed down by the Government. Two other large wholesale/retail companies, Jamaica Frozen Foods and Grace Kennedy, although experienced in marketing fish, were geared to imported bulk fish purchased at a lower price. The initial low volume, inconsistent supply, and higher price of tilapia discouraged them from moving this product. It is probable that, as the fish farming industry grows and supplies are stabilized, these or similar large companies will begin marketing farm-raised fish. To date, however, most tilapia are processed and distributed by entrepreneurs who have begun operation only recently. TABLE 8. CONSUMER PREFERENCE TEST FOR THREE LOCALLY-AVAILABLE FISH AS JUDGED BY AN INDEPENDENT PANEL' Item Impression before testing Color.................... Appearance .............. Taste results Odor ................... Flavor ................... Texture .................. Tilapia 7.1 . 7.0 7.5 7.6 7.9 Grouper 7.4 7.2 Red snapper 7.6 7.5 Development of Distribution Channels for Tilapia The initial involvement of project personnel in the distribution process was substantial since semi-intensive fish farming requires heavy financial investment, and failure to reach a market, especially during the early stages of the industry, would have greatly retarded investment interest. The basic policy, however, has been to gradually withdraw support as confidence developed in the marketability of tilapia. Prior to 1981, most fish producers were reluctant or unable to locate market outlets. Although project personnel were seldom involved directly in the sale of privately produced fish, they usually coordinated the negotiations between producers and buyers. A small fraction of a harvest was often sold live at pond bank, but the major buyers were institutions such as schools, hospitals, and prisons. These institutional markets received the fish directly from the farmer or through a middleman with a previous history of distributing marine fish. Another occasional outlet was a small local market where fish were bought directly by consumers or by small-scale middlemen in lots of 5 to 20 kilograms. In both of the above cases, project personnel generally participated in the transportation and handling of the fish. The next stage of development was the establishment of a centralized distribution outlet. This was more practical than attempting to market at pond bank because production ponds were widely scattered and accessibility to public and private transport 17 7.2 7.5 7.1 7.4 7.1 6.6 'Taste test was conducted by Grace Kitchen and Consumer Center. The panel consisted of eight members, primarily restauranteurs. All samples were fried after adding usual quantities of salt and black pepper. Test characteristics were graded on a scale of 1 to 10, in ascending order of preference. The values indicated are the average from the eight panelists. Approximately one-third of all fish imported in 1982, nearly 5,000 tons, was chilled or frozen fish, which can be considered as a first-order competitor with tilapia. This quantity was more than 25 times greater than the rate of production of tilapia at that time and nearly 10 times greater than targeted levels at the end-of-project in 1984. With the stagnation of the fishing industry, continued import limitations, and an expanding population, the market potential for farm-raised tilapia is very favorable. Analyses of consumer response and the traditional sources and volume of fish products in Jamaica indicate the project goal of an annual production of 545 tons of tilapia (less than 3 percent of total national consumption of fish in 1982) is well below the level that would saturate national demand. LEFT: Fish tha-- are not sold I~ve at pond bank Effer harvest ae transported lve to a Central distrbu'lion point where they are unloaded fo immediate sale. BELOW: Fish are wesighed for customers at a cenctra distribution point. The obvious demand for these fish has motlvated argeT-soale was higly 1)variab le. Also, for reasons of seciritv and general farmn manage ment, manyx tarmers preferred that their planned harxvest date not he widely publicized. As a temporarx measure, limited marketirng facilities were made asvailable to smnall scale producers at the project's headquarters at Txwickenham Park. Farmers were allowsed to use holding facilities on the farm and to sell their fish on the compound. Tswo additional measures wvere taken to cnsurc the predictability of the supply of fish at this location. EFirst, prixvate sector lharsvests were schedoled so that fish would he asvailable for sale eveey Thursday. Second, fish from the gov erment production farms were used to boffer the market against periodic priv ate sector shortages. The majority of the fish wecrc sold by the producer or his designate directly to consomers or to small-scale distrihotors. The initiation of a central distribution point resulted in increased farmer confides cc and a gmowsing interest among potential processor/ditilutors. By mnid 1982, production wsas expanding beyood the capacity of this one-day market. Howesver, smlscale distributors had expanded their capacity to 2 or 3 tons of fish per week. To accelerate prixate sector involvement in marketing, a fee for the use of the distributimn outlet at project headquarters was suggested. Ensvate indiv iduals expressed interest in setting uip such outlets in two parishes, bout the project site wSill remain axvailable to farmers until such permanent central distrihution areas flaxve Leen desveloped. The project at this time is concentrating en developing tile local market. H-owexver, a pine-feasibility study on the export potential of tilapia, to be carried out by the Jamaica N ational fox estment Promotions Limited, has been proposed. Export of tilapia would be especially significant for Jamaica, where areouiction an foreign exchang-e earnings from its traditional exports has generain' strong interest in opening uip new oxverseas markets for non naditional export crops such as fish.