Ten years monitoring of fish populations

ONCHOCERCIASIS CONTROL PROGRA}I}IE IN WEST AFRICA: TEN YEARS MONITORING OF FISH POPULATIONS I +? [. r f, [,t I IF I I I C. L6v€que C.P. Fairhurst -K. Abban D. Paug), Il.S- Curtis X. Traore ORSTOII - ParisUniversity of Salford IAB - Kumasi ORSTOIi - Rams[sUniversity of SalfordIm - Abidjan tCONTENTS 1. Introduction 2. Aims and Protocol 3. Insecticide Treatments 4. River Hydrology and Biological Significance 5. Ecotoxicological Studies 6 . Resu I ts of Eco I ogi ca I l-toni tor i ng 6.1 - 6.r - 6. 3. 6-4- 6. 5. Species Composition Number of Fish Species Changes in Total Experimental Catches Changes in the Structure of the Fish Catches Coefficient of condition 7 - Conclusions a. Summary. Acl{nowl edgements Key to Figures Referencesfis FT rt o1 . I NTRODUCTION Human onchocerciasis is a dermal filariasis particularly seriousin Guinean and Sudanian African savannas where it causes irre- versible blindness among exposed populations. The filaria Onchocerca volvulus is transmitted to man by female blackfly ofthe Simulium damnosum complex (Philippon, L977). The larvae of these flies are aquatic and occur only in fast-flowing parts of rivers. Thus the disease is oost prevalent near water courses. In the absence of any effective cure suitable for large-scale use, vector control was the most effective way to prevent the spread of this disease. Adult control being difficult, chemical treatment of Iarval stages in the rivers was considered the onllfeasible method (Anon., 1985). The Onchocerciasis Control Programme (OCP) commenced in December 1974 under the auspices of the ldorld Health Organisation (tdHO) and was planned for a twenty year intensive implementation(Davies et al, 197a). The initial control area of 764,OOO km2 included Burkina Faso and parts of Ivory Coast, Ghana, Togo, Benin, Niger and }lal i (Figure 1). The f irst lnsecticide treat- Dents were in February 1975 in the central part of the OCp area, and have been progressively extended. Up to I8,OOO km of rivers have been monitored, and treated when necessary in the weekly spraying progremme. Prolonged and extensive use of insecticides coulcl have important environmental risks, and therefore it was necessary to evaluatethe possible short-, medium- and long-term effects of insecti- cides on the non-target fauna. Consequently, an aquatic nonito- ring programme hras devised before the start of OCp, to ensurethat insecticides would not seriously disturb the ecologicalfunctions of the water courses, and to provide information which would lead to the recognition of toxic effects (L6vCque et aI, 1979 ) . The monitoring is carried out by national teams of scientistsin the countries in the OCP area, aided by outside specialists.This support was essential because little infomation on the aquatic fauna uras avai lable. The survei I lance has been primari ly concerned with two major categories of organisms: the fish, by virtue of their economic iuportance, and the benthic inverte-brates, which nay utore quickly respond to insecticides. Animportant consideration for ocP is the demonstration to thelocal human population that care is being taken in reducing the risks of pol lution. An independent Ecological Group, consisting of experts ofinternational repute, leets every year and is in charge of the evaluation of the collected data which has been analysed by outside specialists (cummins, 1945). The Group advises ocp on safe insecticide use and new monitoring procedures. 2. A I I"I-S AND PROTOCOL All insecticides used b), OCP have to follow an intensive screening procedure in order to prove their high toxicity against the s. damnosum larvae, and their low toxicity for non-target fauna (L€v€que, 1986; Kurtak, ). The criteria arethat the pesticides should have neither any direct impact onfish nor any effect on their life cycles. Among hundreds ofinsecticide formulations tested by OCP, few are selected after operational field trials. It is therefore assumed that theseinsecticides have no acute effect on fish when employect at the recommended treatment rates. However, it is also necessary to evaluate the possible medium- and long-tern effects of appli- cat i ons. The fish monitoring programme was based on two fundamental i deas: - a) repeated Iong-term treatments could change the reproductive cycle of fishes, either by affecting the physiology or by clirect effect on eggs or juveniles. If so, there could be changes infish recruitment and, on a long-term'basis, a decrease in fish abundance. This would apply to the fish community as a whole, or to particular species which could be Eore sensitive to insec-ticides.b) Insecticides could affect the food chain leading to a serious reduction in diet. It should be noted that such investisations could hardly be conducted in the laboratory because of the number and diversity of fish species involved, and the difficulties of maintaining most of the species in rearing conditions to complete their life cyc I e. The monitoring stations (Figure 2) were chosen on the basis of accessibility at all seasons, suitability for sampling, availa-bility of hydrological data and abundance of fish stock. In establ ishing the monitoring programme, the terms of referenceincluded the introduction of simple standardized sampling tech- nlques for use by aifferent teams and under various environ- mental conditions (L6voque et al, t9z9). Experimental fishing istherefore carried out using sets of gill nets 25 metres long andtwo Detres deep with various mesh sizes (l5,zo,zs,3o and 40 nm)- Usually each collection is the result of two sets of gill netsfishing on two consecutive nights, but some protocol variationshave taken place, particularly in the early years of the prog- ramme. Data sheets for each sample include the number and total weight of individual species caught in the different mesh sizes.For comparison and standardization, results are expressed as catch per unit effort (CPUE) widely used by fishery biologists, which is the number or weight of fish caught in too m2 of netper night. llost of the monitoring stations were investigated elv'ery three months, but again there were soEe protocol varia-tions during the ten years, [rainly due to the accessibi I ity of stations and availability of teams. iFrom the results obtained in sampling the different stations, it was therefore possible to follow long-term changes in:- a) total catch for the set of gill nets with different mesh sizes or combinations of mesh sizes. b) the number of species caught. c) the quantity of each species caught. d) the structure of the fish catch i.e. relative abundance of species in each mesh size. coefficient of condition, used by fish biologists, ls a simpli-fied expression of the health of fishes. It allows an assess- ment of t-he abi I ity' of the f ish to f ind the food they require within their environment, and hrhether other ecological con-ditions remain favourable. Fish were individually measured and weighed, to estimate the coefficient of condition (l\) derivedfrom the formula:- K w x 1os iL3 where hl i sin mm. the weight in grarnes and L is the standard length Before the nonitoring progrepme there were very few detailecl studies on the biology and ecology of West African fishes. It was soon apparent that additional research was essential if the results of the monitoring programme q,ere to be correctly inter-preted. various studies have since provided a better knowledge of the biology of the main species; Alestes baremoze,Brvcinus nurse, B. imberi, B- racrolepidotus, B. longioinnis(Paugy, 1978, 1980a, 1ggob, tg,gaa, 1gg2b), petrocephalus bovei(Ilerona, 1980), Schilbe pystus and Eutropius mentalis (L€veque& Herbinet, lgaO, 1982>. A study of the Bandama basin (llerona, 1981) provided information on the ecology of the fish species and confirmed the represen-tative nature of nonitoring stations. Electro-fishing has alsobeen carried out in the rapids of some rivers to give a better understanding of fish populations oF these habitats which cannotbe samplecl by gill nets, and of their changes over time. In order to help the different teams in identification of species, a catalogue of fishes was produced (L€vEque & paugy, 19a5). Al Ithis information will be developecl further in other publications tICIDE REATYE Temephos ("Abate" @, is an organophosphorus Iarvicide which was used exclusively from 1975 to r98O. A 2OZ emulsion concentrate was applied at a dosage of o.O5 mgl-r per minute during the wet season, and at O.1 mgl per 10 mn in dry seasons. In December 1979 Temephos resistance developed in larvae of some cytospecies of the S. damnosum complex (Guillet et al, 198O; Kurtak, 1986) and spread rapidly to the southern forest zone andpart of the humid savanna zone. This situation led to a large- scale application of Bacillus thurinqiensis H14 (-Teknar" @ )during the dry season (dose rate 1.2 mgl-rper 10 mn) in these areas of resistance, together with "Chlorphoxim" R , another organophosphate, during the wet season (O.O25 mgl-r per 10 mn). However, a resistance to Chlorphoxim was discovered in JuIy 1981 in the forest species alread!' resistant to Temephos (Kurtak et al, 1982). This necessitated an acceleration of screening of other alternative insecticides, and "Permethrin" and only beenper 10 mn) di rect "Carbosul fan" @ appeared to be promising. These used in the fielcl during the rainy season (O.Or5 mgl-l where resistance to organophosphates was observed. No effect on fishes was apparent. By the end of 1985 the treatoent situation could be summarised as follotrrs- Tenephos was still used in those regions of the OCp area urhere no resistance had developed anong Sinulium popula-tionsi in the south-west, where strains resistant to this larvi- cide had appeared, Teknar was used where river discharge wasbelow 75 uns{. Above tnis level, the strategy was to alternate other larvicides such as Tenephos, chlorphoxin and, when neces- sary, PerDethrin (wHO, r9a6). As a result, insecticicle treatment varied between rivers. Figure 3 illustrates the treatment regimefor the nain fish monitoring stations since the start of obser- vations. llore details on insecticides and treatnent strategies wiII be found in (r9 )- It must be reoeobered that other insecticides ray also affectthe rivers in Ghana and Ivory coast. Large arounts of agricul-tural pesticides, which are difficult to evaluate, tay reach the water courses. According the Calamari (1995), 3oo t of DDT,600 t of Lindane, IOO t of lethylparathion and 30 t of other compounds uere used in 1976 in Ivory coast. since 1979 DDT hasbeen replaced by organophosphates, carbamates and pyrethroids.Similar values are given for Ivory coast b!'BaIk and Koeman(19a4) and are expected to increase two to three folct by the end of the century. The areas surrounding sone hlest African rivers are also treated with lnsecticides to control tsetse flies. The side effects ofhel icopter appl ications of dieldrin, endosulfan, permethrin, azamethiphos hrere monitored (Everts et al, r9B3a, r9a3b: Takken et al, 1978)- lio acute fish Dortality was observed, except inNigeria where a Eass mortality of fish occurred following endo- sul fan spra)'ing (Koeman et al , 1978) as wel I as in lvory coast(Everts et al, 19a3b). @had Another source of pollution is the Sugar and fruit factories often situated along the rivers. The residues from this organicpollution could cause fish mortalities, such as those reported from the Sassandra upstream of the Semien monitoring station in 1945- 1946. a , PIVtr R H\ DROLO (;\' A t) BIOLOGICA L SIGN T F I CAIiCtr a The ocP area co\.ers major river systems in hest Africa, such asthe Volta basin, part of the Niger basin, the northern parts ofthe sassandra, Bandama, como6, Mono and ou€m6 basins. I.lost ofthese rivers are savanna type, with a water regime characterisedby a flood period from July to November with a peak in September and a lengthy low h,ater period from January to June. For IvoryCoast, hydrological and Physicochemical characteristics for the main water courses are summarised in Iltis and Leveque (19a2).For the volta basin, Itoniod et al <t977) gave a synthesis ofhydrol ogi cal data. Many of the rivers in the central part of the OCP area are inter- mi ttent and may- dry' up compl etel y. For permanent rivers disch- arge is verl'lor+ during the dry season, and the upper course is sometimes reduced to a series of pools. There are therefore several seasonal changes in flow which result in major ecolog-ical changes for the fish species. However, the importance ofthe flood period is also directly related to the abundance of seasonal rains and as a result of climatic fluctuations, thehrater discharge of rivers exhibits large changes from year toyear (Figure 4). There nay be differences between basins, butit is clear that a poor hydrological situation prevailed in the whole ocP area from 1982 to 1984- In tropical rivers it is well known among fish biologists thathydrology plays a najor role in fish behaviour. Fish repro-duction tends to be highly seasonal and correlated prinarily with f lors (Helcomme, t9a5). This is the case for many tdestAfrican species which spawn cturing the earlier part of theflood: Alestes baremoze, Brvcinus nurse, petrocephalus bovei,I'larcusenius furcidens, Il. ussheri, Labeo senegalensis, L. coubie,schilbe mvstus, Eutropius mandibularis etc. (Albaret, tg8iz:L€veque & Herbinet, t9ao, 1982; paugy, 1979, 19ao). Hourever, sone species are known to breed throughout the year: B. irberi,B. macrolepidotus, Hvdrocvnus forskal i i, Ti lapia zi I I i,Hemichronis fasciatus (AIbaret, tg8,?; paugy, 19gO, 1982)- It is also assumed that breeding success and survival of fry of rany species could be related to the duration and water levelof the flood period (for review see hlelconme, lg7g, l9g5). Inyears when there is insufficient hrater, the young fish havefewer refuges, are Dore vulnerable to predators and have fetoer sources of food. Dansoko et al (1926) have shown that thereduction in commercial catches of Hvdrocynus brevis and H-forskal i i in 1972 and ].g73 was a consequence of inadequatelevels during the flood periods, and resulted in a poor condi-tion factor, limited growth and weak recruitnent to the fish stocks. In a detailed studt'of fish populations from the Logoue floodPlain Benech and Quensiere (19a3) were also able to demonstrateover ten years the e.xistence of a positive correlation betb.eenfish production and flood volume, as welI as changes in species composition and community structure related to drought periods. ,)tarrr fish species migrate lorrg ciistances upstream at the begirrnirtg of flood in order to spaun and find good conditions fat,ouring the development of fr!'. This is the case in the rivers considered here for A. baremoze, B. leuciscus, Mormvrus rume, I'tormvrops del iciosus, Distichodus rostratus, Eutropius ni loticus, E. mandibularis, Labeo seneqalensis etc. These migration pat- terns could be modified as a result of water management schemes such as dams and impoundments, whose numbers and surface area are expected to increase greatly by the end of the century(Clay, f9A4). These dams could act as barriers interrupting upstrean migrations, but could also favour species which clevelopln the lakes where they find good ecological conditions and migrate upstream during flood (See Bernaczek, 1994, for review).Such examples are the Volta and Xamji lakes (Kapetsky & petr, 1984) as well as the Kossou lake on the Bandama. Since thebeginning of the monitoring programme other dams have been built, such as the Taabo on the Bandama, Buyo on the sassandra and numerous smaller irrigation reservoirs on the upper reaches ofthe riru'ers. other projects are planned for the future in Ivorl' coast and Ghana. As a result of this management, changes in fish community structures are expected fn many rivers both upstream and downstream of the dams (Bernaczek, f9a4). In conclusion, the cooposition of experilental fish catches couldbe subject to: - a) seasonal changes as a result of nigrationsb) year to year changes as a result of clinatic fluctuations c) Iong-term changes following effects of impoundments.In the interpretation of data, attention should be paid to thesedifferent causes of influence with possible larvicicle impact. 5. ECOTOXICO ICAL STUDIES The effects of organophosphates (Temephos and Chlorphoxim) inlaboratory experiments showed that fish were able to accumulate Temephos ( I'li I es et al , 1976; Matthi essen & Johnson, 1978) ,but this accumulation seems to be limited and does not increaseindefinitely as was observed with DDT for instance. As an exam-ple, Sarotherodon mossambicus exposed weekly to operational doses (O.O5 mg l-r for 10 minutes) accumulated 3-4 mg kg-r by clirect absorption. They could also accumulate residues by eating contaminated food. An affinity of Temephos for fatty tissues has been observed, but in contrast to organochlorides there is no accumulation in the I iver. According to results obtained in field conditions (Quelennec et al, 1977) fish captured during the dry season just below the sprat,ing point e.xhibi ted traces of Temephos (between 1.3 and 14.3 mg I-:according to species) one da)'after treatment. Six days later contamination was lower (between I and 7 mg I-'). At a distance of 1 km below the spraying point, fish were weakly contaminated (between O and O.25 mg lr). In the rainy season, accumulation of Temephos is much lourer; O-O.4 rg I-r five hours after spraying, just below the spraying point, and O-O.O3 mg l-rfive days later. It should be mentioned that DDT residues u,ere also found (o.or-o.35 mg l-r ) in the fish stuctied, probably as a result of the use of this pesticide in agriculture. The inhibition of acetylcholinesterase activity due to organo-phosphates eras also studied in laboratory and field conditions.In the laboratory (Gras et al, 1982; Pelissier et al, tgBZ, 1983), where operational doses were tested (o.05 mg I for ro Dinutes)inhibition by Temephos was about Z5Z for Titaoia suineensis,but no fish intoxication was noted after repeated weekly expo- sures. When fish hrere exposed to the operational dose for 24hours, the inhibitory ef fect is much higher; 3€tt after one exposure and 69t after three weekly exposures- In the latter case the fish did not survive- The inhibition of acetylcholin- esterase activity appears more important with Chlorphoxin- In field conditions, the acetylcholinesterase activity in thefish brain does not seem to be significantly different in riverstreated with Temephos or untreated (Antwi, 1983, f98/r; Scheringa et al , 1981). tlhen Chlorphoxim is used, fish captured below spraying points exhibited a 2o,* reduction in enzynatic activity,but this inhibition was shown to be reversible (Antwi, f9a3, 19a5). 6 . RESI_'LTS C)F EC]()LOG I CAL }ION ] TC)R I NG 6. 1. Species Composition The monitoring stations are situated in different river basins and at different levels of the water course. It is therefore not surprising to observe differences between stations in the rela-tive abundance between species (Figure 5)- Alestes barenoze is the dominant species in the south-flowing Irror!, Coast rivers except in the Leraba-Comoe basin where Sch 1be m!,stus is co-dominant. The Sassandra demonstrates a spectrum similar to the Maraoue. The Bagoe (Niger basin) and VoIta exhibit a somewhat different fish fauna: Alestes leuciscus, Eutropius niloticus and Chrvsichthvs auratus instead of Brvcinus imberi,Eutropius mandibularis and Chrr-sichthvs maurus respectively. 6.2. Number of Fish Species Assuning that each sample is the catch of trrro sets of gill netsfishing on two consecutive nights, the number of fish species caught at different stations (Figure 6) exhibits a seasonal change urith a mElximum at low water. This represents the greatest efficiency of the gill nets- There are different long-term trendsin each river, but overal I there is no evidence of a reductionin species richness over the ten year period. For Ivory coast,there is a decrease in the n'rmber of fish species between 19gr and 1984 which could be related to the poor river dischargesduring that period (Figure 4> - The recovery observed in t9a5follows a good rainy season. The total number of species could mask osition, sone species replacing others not show any evidence of disappearance experimental catches. changes in species comp- However, the results clid of fish species in the 6.3- Chanses in Total D(perimental CatcLes t{hen considering changes in the fish catches, expressed as the mean CPUE for the standardized set of gill nets (Figure 6),a seasonal pattern is generalll' observed, with higher catches atthe end of the hisn t*ater period (November to Januar!,) and lower catches during the flood (August to september). This pattern ispartly the result of the fishing gear being Eore efficient inlour water conditions. The high cpuE values observed in the sassandra in 1980 appear to be due to an increase in Alestes/Brvcinus species (Alestes baremoze, Brycinus nurse, p. imberi)- A simi lar phenomenon was observed in rgao-tg8t in the Leraba clue toincreased catches of Alestes baremoze, Eutropius mandibularis,Schilbe mYstus and Lates niloticus. over the ten year period nolong-term reductions in mean catch can be discerned. wherr corlsi der r ng dr f f ererrt mestr si zes ( Fi gure 8) , and part i cu- Iarlv the smaller one efficient for juveniles, there is a more obvious correlation with year-to-year changes in hydrologl'. Rt Niaka on the Barrdama, for instance, catches in the 15 mm mesh size are lower in 1976-1977 and L9a2-19a4 which correspond topoor flood (Figure 4). The catch was better with more favourable hydrology between 1979 and 1981. A similar situation occurred on the Leraba (Figure Ab). 6.4. Chanqes in the Structure of the Fish Catches It is difficult to compare many graphs of species abundance against time and ol'erlat'these trith abiotic factors. Therefore multivariate anal!'ses are appropriate and the method used here is the factorial analy'sis of correspondence, the salient points of which are ampi!, described elseuhere (Benzecri, 1973: Lebar{ &Fenelon, 1973: Hi I I, 1971) - This method is particularll, inform- ative as it permits simultaneous graphical representation of species and samples. T\ro examples are given here, the first being the river Oti atSabari. The river has only received Abate treatment. Although treatments have been less frequent in recent years, the river IIOhas received heavy insecticide applications. This tributary joinsthe Oti before the Volta LaI<e so all migratory fish woulcl be exposed to Abate at some time. In the ordination presented in Figure 9a, the first axis separ- ated the samples into lotp water and wet season periods. Dry per-iod collections are characterised by Alestes, Labeo andChrvsichthvs species, as well as Svnodontis filamentosus and Schilbe mystus. The second a-riis has a group of high water 5ampleS in the middle, with rising and falling water periods being more diffuse. comparison with Figure ro gives some reasons for this pattern, asthe Petrocephalus spp- are not common after L979, while Eutropius niloticus continues to be abundant throughout the sampling period. Apparent reductions in abundance can also be correlated with a reduction in frequency of monitoring from monthly to quarterly-Therefore species such as E. mentalis may be under-represented asthe main migration period is nore likely to be nissed. In conclusion, the najor patterns of community change reflect seasonal influences rather than year-to-year differences- Changesin abundance of indit'iclual species over time have a partial e-xpla- nation in reduction of sampling frequency, but species such asPetrocephaus bovei and P- simus appear to become rarer, a trend which is difficult to explain in terms of hydrology- The secon6 gxampl€ is the river Bandama at Niakaranandougou. The river feeds into the lroussou Dam, and early treatments pere withAbate, other insecticides being employed from the end of r98o.The ordination (Figure 9b) appears to separate Abate treatmentsfrom the rest, but closer examination reveals a distinct community change at the end of L976, with A. nurse becoming more rare and spec i es such as _S. imberi , C. vel i f er and Hvdrocvnus f orslial i iincreasing in abundance. The second axis is dominated by Labeo Senegal ensl s, SSLrilbe fn st us arrd TiIania qaliIaea, indicative of arrottrer change at the begi rrrr i ng of 1982 when these spec i es became a smaller proportion of the catclr (Figure 16). In the centre of the ordinatton I ies Eutropius mental is which has maintained a relatively constant proportion of the catch throughout the sam-pling period. The reason for the changes could Iie in a combin- ation of a lag effect of the filling of the Koussou Lake and the wet years of 1979 ancl 1980. Therefore, effects of treatment are again not obvious, and this conclusion is reached for other sam-pl i ng stat. i ons. 6.5. Coefficient of Condi tion Long-term changes in the mean coefficient of conditon (X) were studied for the most abundant species at the clifferent nonitoring si tes. Sel ected e-\ampl es are gi ven i n Fi gure 10 f or spec i es vu'hose food is primari l-v- based on aquatic invertebrates. The rralues of the coefficient of condition are relatively random, fluctuating around a mean which does not seem appreciably altered over the ten year period of treatment. There is no evidence of a long-term decrease or irreversible modification in K. Nevertheless, when examined in detail, significant short-term decreases hrere obs- erved for a fetrt species- This is the case for Alestes baremoze and to a lesser extent for Eutrooius mandibularis in the Bandama river in 1976-1977 - A similar phenomenon occurred for Brvcinus nurse in the Sassandra and Leraba rivers between 1981 and 1983. Such decreases which apparently do not affect all species in the same river, or the same species at every site, seem difficult to explain. However, the drop in K for A. baremoze in t976-1977 was restricted to the course of the Bandama between the Kossou and Ferkessedougou dams (Paugy, 1978) and coincided with a period ofpoor floods (Figure 4). The fall in value of K for A. nurse between 1981-1983 coincided with severe drought in the Sassandra and Leraba- For sone species, therefore, there should be a rela- tionship between K and changes in hydrology. Itinor fluctuations in K relate to the life cycle and the seasons(Paugy, 1978, lgaO; L€v€que & Herbinet, 19BO). It was also obs- erved that for the s'ame species there could be differences in Kfor the different river basins- The relative stability, in coefficient of condition incticates thatfish are able to feed normally in treated rivers. In fact,stomach content anal)'sis carried out in 1975 just after the start oftreatment (vid)', 1976), and in ]-976-]-97i, did not reveal signi-ficant changes in the diet of fish species feeding entirely(Petrocephalus bovei) or partially (A. baremoze, B. imberi,B. nurse, E- nentalis) on the aquatic invertebrate fauna. It hasbeen observed that remephos treatment resulted in a temporary reduction of aquatic insects by only 30-40% (Dejoux, 19a3), andit does not seeu that treatment could lower food stocks to a critical threshold. Horeover, the adaptability of many fresh- water fishes to various types of food has been demonstrated a number of times (Lauzanne, t976; Reynolds, t973; tdelcomme, t9a5).Nevertheless, Corbet (1956) observed a change in the diet of some f ish, particularly, |lormyridae. Some species with a special izeddiet, such as ttastacembelus victoriae apparently suffered from nalnutrition or nigrated. Experimental fish sampling has been carried out for ten years inl{est African rivers treated weekly with insecticides by ocp using the standard protocol. The monitoring stations investi-gated di f f ered i n many uJays, wi th: - a) their Iocation on different river basins b) the relative abunclance of fish species, despite many speciesbeing common to different basins c) the insecticides used, some stations always being treated with remephos, others with Temephos up to 1980 and later with an alternation of insecticides (Figure 3). For the interpretation of data different sources of bias:- collected, we were faced with a) the human factor, in that data were collected by different teams and there u,ere changes in the personnel responsible forfish monitoring in Ivory Coast and Ghana. This could lead to some clifferences which on occasion could be recognised bydetai lecl statistical analysis. This source of bias dicl not appear to change the overal I picture. b) lack of knourledge concerning the changes in fish community structure as the result of changes in natural environmentalfactors. This is the case for instance for river discharge, which could vary greatly between years, and is known to affect the reproductive success of fishes. There is no direct evidence of the influence of hydrology, but using a compara-tive approach between stations, it was possible to assumethat some changes in fish abundance were the result of suchfactors. c) lack of knowledge concerning the uses and abuses of insect-icides other than those applied by OCp. It is knorpn (Calamari,1945; Balk & Koeoan, 1984) that large amounts of agriculturalpesticides are used in the ocp area, but the questions are; what reaches the river beds and urhat is the inpact on the aquatic fauna? A ferl direct observations also showed thatinsecticides are sometimes used as poison when fishing in rivers. Houever, the extent of this inf luence is not linown. 7 . CO]iCL,LISIO\.s 8. SL'YYAR\- The results provided in this paper do not show as a whole, any clear impact of ocP-applied pesticides on fish populations. The total catch, the number of species caught in each sample and coefficient of condition, appeared to fluctuate around a mean value, and no long-term drop was observed over the periodinvestigated. The seasonal pattern ls generally clear. In some cases longer term decl ines occur, general l.y being fol lotded by a rise correlatlng utith changing conditions. ACKOULEDGE}tE} TS The Ecological Group is recognised as the general instigator ofthis presentation. The authors would like to acknowledge the work of nany people in the collection, identification and data analysis phases. Particular mention should be made of J.J.Albaret, Antwi, R. Bigorne, Dankwa and B.de fterona. OCp ingeneral, and especially l. yameogo of the vector control Unit, harre given considerable assistance. REEF=EENCE=s AIbaret, J.-l . (198'). Reproductiorr et fecoundr te d'Ivoi re. des poissons d'eau douce de Cote 347 -37 t .Rev. Hvdrobiol - Trop. 15, Anon. ( 1945). Ten vears of Onchocerciasis control in West Africa Report OCP/GVA/ a5 - 18, WHO. 113 pp. Antwi , L- A. K. ( l9a3). The effect of abate and chlorpho-xim on the brain activitv of fishfrom some treated rivers in the VoIta basin area Report to OCP. 25 pp. mimeo. Antwi, L.A.l\. (198A) te on th s C of fish from tr' eaterl rivers in the L]pper Volt rit''ers tthiteVolta and Black Volta Report OCPi VCU,/HYBIOi 84. 13 Antwi, L.A-K. (19a5)- Effects of chlcDhoxim on the brain acetvlcholinesterase activitv of Tilaoia zillii Report te OCP. 5 pp. Balk, I-F. & Koeman, J-H. (19a4)- Future hazards from Pesticides Use. urith special reference toWest Africa and Southeast Asia Commission of Ecology Papers No.6. IUCN, Gland. IOO pp Benech, V. & Quensiere, J. (19a3).Migrations de poissons vers le lac Tchad d la decrue dedu Nord Cameroun. III. variations annuelles en functionI'hydrologie. Rev. Hvdrobiol. Trop- 16, 2,A7-3t6. la plaine de Bernaczek, c. (19a4)-Guidelines for dam design and operation to optimize fish pro-duction in impounded river basins (based on a review of the ecological effects of large dams in africa). CIFA Tech. Paper tl, 9a pp- Calamari, D. (1945). Si trr a tinn do la nnI I rr tion d2nc loc o2lr\- inf6rio ures de 'l'Afrinrrode l'Ouest et du Centre Document occasional du CPCA No. to. FAO. Rome. Clay, C.H. ( r9a4). New reservoirs in Africa. 198O-2OOO CIFA Occasional Paper No. tt. p3 pp. Corbet, P. S. ( f956 ). Some effects of Simul ium control by insecticide on the feedinghabits of insectivorous fishes Qon o nrl Qr rn nneirrn nh rfrinan Ltr Brazzavi I le. tgs6 CCTA : 81. rotr i alnnrr anrl rnlanA Fi chor i os Cummirrs, H.h- ( 196=r). The EcoI ogi caI Group. lrorld Health. October 1985 14-15. Dansoko, D., Brehman, N. & Daget, J. (f976). Influence de la sdcheresse sur les populations d'Hydrocynus dansle Delta Central du Niger. Cah. ORSTOII. r Hvdrobiol. lo, 7l-76. Davies, J.8., Le Berre, R., Onchocerciasis and Sinul ium llosqu i to Ne$s , 3a , 466- 472 . WaIsh, J.F. & Cliff, control in the Volta B. ( 197a) . River Basin t Dejoux, C. ( 1983). utilisation du Temephos en campagne de lutte contre Simulium damnosum en Afrique de I'ouest- Impact des premiers cycles detraitment sur le milieu aquatique. Rev. Hvdrobiol. Trop. 16, 165-179. Everts. J.td-, I'an Frankenhuyzen, K., Roman, 8., Cullen, J-Copplestone, J. & Koeman, J.H. (1983a). observations on side effects of Endosulfan used to controlin settlement areas in connection with s sampaign against sleeping sickness in Ivory Coast.Trop- Pest. ltanao. 29, 177-tAZ. tsetse human Gras, G., Pelissier, C. & Leung Tack, D. (fga2).Action du tom€phos sur I'activite ac6tylcholinesterasique du cerveau de .filapia ouineensis- lCre partie: Otudes exp€rimentales aux doses operationnel les.Toxicol. Europ. Res. 4, 301-3Oa. Gui I let, P. , H. Escaf fre, ouedraogo, lt. & Qui I l€ver€, D. ( lgao)ltise en €vidence d'une r€sistance au t€m€phos dans le complexe simuliun damnosum (s- sanctipauli et S. soubrense) en coted'Ivoire (Zone du programme de lutte contre I'Onchoceriose dansla region du Bassin de la volta).Cah. ORSTOI!. ser- Ent. m6d. et parasitol .Zgt-Zgg IItis, A- & L€v€que, C- (f9a2)-Charact€ristiques physico-chimiques des rivi€res cle COte d' Ivoire.Rev. Hvdrobiol- Trop. 15, tt5-f30. Xapetsky, J.t!. & petr, T. (r9a4).Status of African Reservoir Fishes. CIFA Tech. Paper. tO. 326 pp. Everts, J.w( 19a3b). van Frankenhuyzen, K Side effects of experimental pyret control erch. En of tsetse flies in a river vI ron. ntam. Toxicol - t2 Koeman, J . H. , Den Boer, W.llSpleithoff, p.C. (197a). Three years observation on tions of insecticictes used Ni ger i a. Environ. Pollut. 15, 31-59. Roman, B. & Koeman, J.H. hroid applications for the ine forest habitat (Africa) ,9l-97. .J-, Feith, A-F., de Jongh, H-H. & side effects of helicopter applica-to exterminate Glossina species in Xur t ak, D. ( Kurtak, D. (1986). Insecticide resistance in the Ochocerciasis Control Programme Parasitoloov Todav 2, L9-20. Kurtak, D., Ouedraogo, M( 79A2> . Prel iminarl' note on the Ocran, it Barro, T. & Guillet, P. aDDearance i n I vorv Coast of resi -s AN(.Pt to chl orpho-rim i n Simul i um soub rense./sanctipaul i larl'ae al readl, resistant to temephos (Abate R). trHO unpubl i shed document ldHO,;\'CB,/82, a5O Lauzanne, L. (1976). Regimes alimentaires et relations trophiques des poissons du lac Tchad. cah. ORSTOM. ser. Hvdrobiol. 1O, 267-31O- L6v€qe, C. (19a6). Tho rreo nf ineontiaidoe in tlro nhnr.orc i asi s (^rrht rrll prtrtrrammo and the aquatic ponitorino in west Africa(to be published in a scope volume). L€v6que, C. , fiei, Il. & Pugh Thomas, M. ( 1979). The Onchocerciasis Control Programme and the monitoring of its effects on the riverine biology of the Volta River Basin.In; Ecoloqical Effects of Pesticides- Eds. F.H. Perring & K. llel lanby. Linnean Society Symposium Series, No. 5, 133-143. LevCque, C. E Herbinet, P. (1940). Characteristiques Eeristiques et biologie des Schilbe nvstus(Pisces: Schilbeidae) en COte d'Ivoire. Cah. ORSTOH. s€r. Hvdrobiol. 13, 16r-r7(). L€v€que, C. & Herbinet, P. (1942).Charact€ristiques r€ristiques et biologie des Eutrooius mentalis(Pisces : Schilbeidae) en COte d'Ivoire. Rev. Zool. Afr. 96, 366-392. L€v€que, C. & Paugy, D. ( t9a4) -Guide des poissons d'eau douce de la zone de proeramme de lutte contre l'Onchocercose en Afrioue de l'Ouest Rapport cle conuent i on ORSTOI.I - OltS. OR-qTOM, paris. 393 pp. Matthiessen, P. & Johnson, J.S. (1978)- Accumulation of the organophosphate blackfly larvicide Abate(temephos) in Sarotherodon mossaabicus with reference to thelarvicidal control of Simulium damnosum.J. Fish Biol. 13, 575-546. Herona, B. de., ( l98O). Ecologie et biologie de Petrocephalus bovei (poissons|lormyridae) dans les rivieres de C6te d'Ivoire. cah. oRSTOI.i. ser. Hvdrobiol . 13, 177-t27- Merorla, B. de,. (1981 ). Zonat i on i chtht ol ogique du bassi n de Bandama ol. Tr 14, 63-75 Miles, J.w., Dales, ld.E. & Churchill, F.C Storage and analysis of samples of water, environments contaminated with Abate. Arch. Env. Cont. Toxicol. 5, 29-41. lloni od, F. , Pouyaud, B. & Sechet, Le bassin du fleuve Volta. Ilonosraphies Hvdrologioues No. 5. <1976). fish and mud from P. <7977>. ORSTOI',. Paris. 513 pp a Paugy, D. (1978). Ecologie et biologie des Alestes baremoze (Pisces des riviEres de COte d'Ivoire. cah. oRSToM. ser Hvdrobiol - 12, 2.45-275. Charac i dae ) Paugy', D. ( 198o a). Ecologie et biologie des Alestes imberi (Pisces : Characidae) des rivieres de COte d'Ivoire. Comparison m€ristique avec A.nioricauda. Cah. ORSOM. s6r. Hvdrobiol - 13, 129-tal. Paugy, D. (19ao b). Ecologie et biologie des Alestes nurse (Pisces des rivi€res de COte d'Ivoire. cah. oRSToI't. s6r. Hvdrobi ol . 13, 143- 159. Charac i dae ) Paugy, D. (19a2 a). Synonymie d'Alestes rutilus (Blgr- 1916) avec A. macrolepidotus(Valenciennes, 1849). (Pisces : Characidae). Biologie et variabi I it6 uorphologique. Rev. Zool. Afr. 96, 2a6-32A- Paugy, D. ( 19a2 b). Synonymie d'Alestes chaperi (Sauvage, 1882) avec A. longipinnis(Gunther, 1864) (Pisces : Characidae). Cvbium 6, 75-90. Pellisier, C-, Leung Tack, D. & Gras, G. (19a2). Action du t6r€phos sur I'activit6 acOtylcholinest€rasique du cerveau de Tilapia suineesis. 2 eme partie: 6tude exp€rinentalelors d'une exposition de 24 heures au toxique.Toxicol. Europ. Res. 4, 309-319. Pellissier, C., Leung Tack, D. & Gras, G- (19a3)-Action du temephos sur I'activitO acetylcholinest6raseique du cerveau de Tilapia guineensis. 3 eme partie: action compar€e du t6m6phos et de trois insecticides de renplacement.Toxicol. Euroo. Res. 5, 63-69. Phi I ippon, B- <t977> - Etude de la transmission d'Onchocerca volvulus (Leuckart. 1983) ( Nemat oda Onchocerc i dae )oarS imulium damnosum (Theobald. 19O3)(Diptera : Simul idae) en Afrique tropicale Travaux et Docunents ORSTOM. No. 63 3Oa pp. I Quelerrrrec, G. l'1iles, J.h., Dejour, C. & l,lerona. B. de. (1977). Chemiq_a_L__mp[itorinq for temephos in mud. ovsters and f ish from ui tlr in the on hoc erc i asi -s ContA rl Ve r C ol ProcIramm e in ther volta River' Ra-s i n ere.e, wHo, \'BC /77 .683. 6 pp Takken, td, Balk, F., Jansen, R.C. & Koeman, J.H. ( 1978). The experimental application of insecticides from a helicopterfor the control of riverine populations of Glossina tachinoidesin tlest Africa. VI - Observation on sicle-effects.Paris. 24, 455-466. Scher i nga, Fish brain E.J.F., Strik, J.J.T.td.A. & Antwi, L .A.X- (19a1). Abate appl ica-AC etvl chol i nesterase act i vi tv a terf imrr'l irrn rlamn rtfr i n f ha al + a i rror trac i n )OS v r t Unpubl i shed report to @P . 27 pp Vid1,, G. <t976)- Etude du reqime alimentaire de ouelques poissons insectivoresdans les rivier es de COte d'Ivoire. Recherch de l'influence deq t 12 ite mant e insectiridps cffpn t rr5c rl.ane lo n adre de La I rrt f o contre I'oneho ercosPC Rapport du centre ORSTOIi de Bouake No. 2, 36 pp. mimeo. Helcomme, R.L. (f 979).Fisheries ecoloqv of floodolain rivers Longman. London. 317 pp. lleleomme, R.L- ( r9a5).River fisheries. FAO. Fisheries technical paper No. 262. 33O pp hrHo. ( 19a6). Onchocerciasis Control Proaramme in hlest Africa-Expert Aclvisory committee - Report of the Seventh session, Bamako. 76-2U^ June 19a6. Jpc7.5 (ocp/EAC/a6.r)- 35 pp. nimeo. ADDITIONAL REFERENCES (Section 6.4. ) Benzecri , J. P. ( 1973 ). L'ana I vse des donnees. I . La ta-xonomi e: I I - L'AnaI vse des corres ndances Dunod, Paris. 631 & 6f9 pp HilI, lt.o- (1974). Correspondence analysis: a neglected Eultivariate nethod. App. Stat. 3, 34O-354. Lebard, L. & Fenelon, J.P. (1973) Qt at i ct i n ro of i n format i u rrnI i nrr5oc(2 €me edition). Dunod, Paris. 457 pp I aKE)' TO F I GURES FIGURE 1 llap of OCP area. FIGURE 2 Location of t-tajor Fish }lonitoring Stations in Ivory Coast and Ghana. l. Leraba - Pont Fronti€re. 2. Como6 - Gans6 3. Bandama - Niakaramandougou 4. Bandama - l,larabadiassa 5. Nzi - bridge of Dabakala road 6- Black Volta _ Bamboi 7- White Volta - Daboya a. Bagoe - Kouto 9. Oti - Sabari 10. I'laraoue - Itankono 11. Sassandra - Semien]-2. Pru - Asubende 13. wawa - Dodo Papasse FIGURE 3 Treatment at the ltajor Aquatic ltonitoring Stations FIGURE 4 FIGURE 5 FIGURE 6 FIGURE 7 FIGURE 8 Discharge(ms)perStations in the OCP Changes for the Ilonth for Some llonitoring area Total Numbers of the llost Conrnon Fish Caught in cill Nets per Unit Effort at the Major equatic ilonitoring Stations- l. Petrocephalus bovei Z. Hydrocvnus forskal i i 3. Alestes barenoze 4. Brycinus macrolepidotus 5. Brvcinus imberi 6. Brycinus nurse 7. Brvcinus leuciscus a. l-abeo seneqalensis 9. Schi lbe mystus 1O. Eutropius niloticus 11. Eutropius nanclibularis 12- Chrvsichthvs Eaurus Changes in Number of Fish Species per Sample at Various Stations in }tean C.P.U.E- (Catch per lOOm' tJhole Set of Gi I I Nets in Various per night) Stat i ons Changes in Catch (C.P.U.E. ) for Dif ferent Mesh Sizes at Two llonitoring Stations (Leraba and Bandama) Correspondence Analysis of Fish Numbers at: a) Oti at Sabarib) Bandama at Niakaramandougou FIGURE 9 FI GT'RE I T] Petrocephalus bovei Petrocephalus simus Svnodontis eupterus Svnodontis ocellifer Svnodontis sorex Svnodontis gambiensis Svnodontis schal I Synodonti s f i lamentosus Svnodontis bastiani Eutropius mandi bul ari s Eutropius ni loticus Hvdrocvnus forskal i i Charrges in CoefI)crt'rlt of Condittorr for Selected Spec i es Pb Ps Se So Sss Sg SS Sf Sb Em En Hf BM Bn BI Bi Ab Cv LC SM Ln Tg Gb Ls Brycr nus macrol epidotus Brycinus nurse Brvcinus leuciscus Brycinus imberi Alestes barenoze Chrvsichthvs vel i fer Labeo seneoalensis Labeo cubie Schi lbe mvstus Lates ni loticus Ti Iapia qal i lea Gnathonemus bruvere iI , I