Chapter 3. THE PRESENT STATE OF KNOWLEDGE ON THE ENVIRONMENTAL IMPACT OF THE CHEMICALS USED IN TSETSE CONTROL AND ERADICATION

Although chemical control methods have been used in anti-tsetse campaigns in at least 20 countries in Africa, side-effects on non-target organisms have been studied to a greater or lesser extent in only about half of these. These studies are listed in Table 3.1. In this chapter a general account will be given of the studies referred to in the table.

Table 3.1 Studies on side-effects of the chemical control of tsetse flies considered in the present report

Botswana

Botswana was the first country in Africa where side-effect studies in relation to chemical control of tsetse have been carried out. In 1964 riverine forests were sprayed against G. morsitans in an area near Maun. Dieldrin was applied by both Unimog and knapsack spraying equipment. The effects of the motorised spraying on non-target organisms were observed for a period of 10 days following spraying. Among the animals found dead were many birds, mammals, reptiles and fish (Graham 1964). The environmental side-effects of the knapsack sprayings have not been determined, but they were certainly not as disastrous as with the motorised sprayer. Casual observations showed no difference between the wildlife populations of the sprayed and unsprayed areas (Davies, 1979 a). More recent studies conducted in Botswana all concern large scale fixed-wing ULV applications of Endosulfan, which started in 1975 (see p. 3).

The spraying is carried out in the Okavango Delta where the grassland and Mopane woodlands of the swamp are treated with five sequential applications of Endosulfan at a dose rate of 6–12 g a.i./ha. Some other insecticides were also tested. Environmental studies were started in 1975.

During a trial with Crotoxyphos the side-effects were monitored. The mortality of fish and tadpoles in exposure cages could hardly be attributed to the insecticide. There was a slight reduction in insect numbers. However, the results are not important within the scope of tsetse control programs, as the insecticide failed to achieve satisfactory control of tsetse fly.

The fauna elements that were subjected to research in relation to the Endosulfan sprayings, are summarized in table 3.2.

It was concluded that no serious harm to the non-target invertebrates resulted from the normal dose rate applications.

Bird studies indicated some effects, but the authors judge those implications only temporary.

Because of the high toxicity of Endosulfan to fish, extensive attention was paid to the survival and sublethal effects on fish.

Although acute mortality resulting from normal applications was low, physiological studies showed that surviving fish became significantly debilitated although recovery followed cessation of spraying. However, the authors state that histopathological studies showed abnormalities in brain tissue one year after spraying, suggesting that full recovery is slow. The results of this study have not yet been published in full detail.

Table 3.2. STUDIES ON SIDE-EFFECTS OF ENDOSULFAN USED IN TSETSE CONTROL IN BOTSWANA

It is suggested that a significant increase in spray concentrations or number of applications would be likely to result in heavy fish mortality and severe long-term pathological changes in the survivors.

Consequently, the general recommendation was made to the Director of Veterinary Services that by applying the present application rates of a maximum of 12 g Endosulfan/ha with a maximum of five applications, the effect on the ecology of the Okavango Delta would be minimal. An attempt of fly eradication by spraying the whole area in one season will probably be less harmful to the fish population than repeated applications over several years to parts of the area to be cleared (Botswana Government 1979).

The Endosulfan monitoring project which terminated in 1979 was funded by ODA.

Cameroon (United Republic of)

In the 1978/79 spraying season observations on side-effects of a single Dieldrin application on the Adamaoua Highlands were reported by Müller et al. (1980). Dieldrin was applied from a helicopter on a gallery forest at a dose rate of 750 g/ha for the control of G. morsitans submorsitans. Periodic control studies were undertaken from six weeks before until one year after Dieldrin treatment. As the wet season started immediately after spraying, a comparison between preand post-spray data on animal numbers and diversity could not be made. The prespray data were compared with the results of an investigation carried out one year later in order to examine possible longterm effects of the Dieldrin application. In contrast to previous observations in Nigeria where Dieldrin was applied using the same techniques and dose rates (Koeman et al., 1978), no dead or dying vertebrates could be found in the investigated gallery forest directly after spraying.

The residue levels of Dieldrin in the livers of birds (see table 3.3) do not suggest that any of them carried lethal or near-lethal levels of the insecticide. High levels though were found in bats. It is stated that levels up to 175 ppm (dry wt) did occur in the liver; this would correspond to levels as high as 44 ppm on a wet weight basis. These levels correspond to the levels observed in mammals or other warmblooded vertebrates exposed to lethal amounts of the insecticide. After one year the maximum residue level of the bats was reduced to 0.20 ppm. Fructivorous bats were found to occur only very rarely after one year.

One year after treatment the arthropod fauna of the ground surface as a whole showed a significant reduction in numbers. According to the authors the reduction in diversity of phytophagous insects of the herbaceous and foliage layers of the gallery forest indicate that some communities became instable. No data are given on numbers of the aquatic fauna, but it is stated that apart from a decrease in numbers of Amphicorisae (water-striders), the aquatic fauna did not seem to be affected in terms of abundance.

Table 3.3. Dieldrin content (ppm/DW) directly after spraying (March 15, 1979 to April 7, 1979) (Müller et al. 1980)

	Species		Min	Max	n
1.	Mammalia (liver)
	Praomys tullbergi (Rat)	0,30	0,00	0,73	13
	cf. Micropteropus (Fruit Bat)	76,21	1,48	174,81	14
	cf. Epomophorus (Fruit Bat)	7,73	0,00	12,08	8
	cf. Pipistrellus (insectiv. Bat)	4,04	0,48	7,28	8
2.	Aves (liver)
*	Halcyon malimbicus (insectiv.)	3,83	2,21	4,30	8
*	Ispidina picta (insectiv.)	0,93	0,28	1,28	14
*	Platysteira cyanea (insectiv.)	0,04	0,00	0,05	7
*	Terpsiphone viridis (insectiv.)	0,00	0	0	4
*	Turdus pelios (polyphag.)	0,43	0,20	0,91	8
	Nectarinia verticalis (nectariv., insectiv.)	1,83	1,03	2,44	6
3.	Pisces
	Aphyosemion buelanum	1,24	0,00	214,34	(ca 120)
	Clarias walkeri	1,08	0,00	3,42	5
4.	Insecta
	Lep., Sphingidae	1,04	0,00	1,83	(15)
	Isoptera (winged)	0,02	0,00	1,44	(ca 200)
	Califera	6,53	0,00	13,16	2
	Formic., Dorylus-♂♂	0			(ca 10)

* Bird species found dead after similar Dieldrin applications in Nigeria (Koeman et al. 1978)

Chad (Republic of)

In 1972–1974 DDT was employed in groundspray operations to eradicate G. tachinoides from the Assalle-Serbewel area between N'Djamena and Lake Chad both in Chad and Cameroon.

The direct action of DDT to wildlife seemed to be negligible, owing to the restricted application to tsetse habitats. However, in this connection it should be remarked that the paper does not indicate which efforts were made to trace undesirable effects.

The invertebrate fauna of treated treebark was seriously attacked, but since not all of the area was treated, non-target organisms may survive in biotopes bordering the sprayed zones.

A fishkill had to be attributed to poachers who poisened ponds with DDT or organophosphates. However, secondary effects of the campaign with regard to land development should not be overlooked. The cleared area was overrun by the local population and their cattle. Two years after the end of the operation the once so abundant forest galleries had almost completely disappeared.

Forests were demolished to fulfill the needs of fire wood and to clear the area for agricultural use. Thus controlled chemical eradication of tsetse was unintendedly followed by complete habitat destruction, preventing any future invasions of tsetse flies (Tibayrenc and Gruvel, 1977).

Ivory Coast (Republic of)

Side-effect studies of the Endosulfan and Decamethrin applications around Bouaflé in 1978–1978 were conducted by Everts (1979) and Elouard et al. (1979). Endosulan was used in five sequential aerosol applications (10 g a.i./ha) on the villages and as a residual barrier (267 g a.i./ha) around the villages. The fringing forest along the Maharoué River and its tributaries were treated with residual Decamethrin applications (12.5 g a.i./ha).

The residual applications of Endosulfan for the barrier around the villages repeatedly knocked down non-target insect populations, especially Diptera and Hymenoptera. Populations seemed to recover after each application. The residual activity lasted for about two weeks. The Endosulfan treatments apparently did not affect house-dwelling lizards (Agama agama), chickens or the hatchability of chicken eggs.

Decamethrin, used on the gallery forest, did not result in adverse effects on flying non-target insects, ground beetles or spiders. Penetration amounted to 5% to the forest floor, and 15% to the watersurface. Decamethrin caused an increased drift of benthic invertebrates, sometimes lasting for several days, causing partial depletion of the populations concerned.

Ephemeroptera (maylfy larvae), Trichoptera (caddis larvae), Hemiptera (waterbugs), Coleoptera and Lepidoptera species were affected. Some populations recovered after some time. Simulidae (blackfly larvae) populations were decimated too, but repopulation was completed within two weeks. Chironomidae populations increased markedly.

The little shrimp Caridina africana was virtually eliminated by the Decamethrin treatment, whereas populations of a larger prawn, Macrobrachium vollenhovenii were not severely affected. The latter is an economically important species. No effects on fish could be observed after the Decamethrin applications. Human exposure to both insecticides was found to be toxicologically insignificant (Everts, 1979).

Kenya

In 1968 observations were made on side-effects of Dieldrin groundspray operations against G. fuscipes and G. pallidipes in Lambwe Valley (Nyanza Province). As recorded by Koeman and Hadden (1968), Koeman et al. (1969) and Koeman and Pennings (1970), the use of Dieldrin killed a variety of insectivorous bird species in appreciable numbers. In table 3.4 residue levels in liver and brain of some of the victims are listed.

Analyses were also made of rodents, birds and fish, collected alive in the areas sprayed. The Dieldrin levels in the livers of 18 out of 20 small rodents (Arvicanthis niloticus, Lemniscomys striatus, Mastomys natalensis and Aethomys kaiseri varied from 0.01 to 0.52 ppm. Only in two specimens the level exceeded 1 ppm (respectively 1.4 and 9.7). Although under normal conditions at least some pregnancy could be expected, none of the rodents caught one week after spraying was found pregnant and their uteri were shrivelled. Further observations showed that two months later 44% of the females was pregnant. However, it seems unlikely that the shrivelled uteri were caused by Dieldrin exposure. Moreover none of the many experimental studies with this insecticide shows this effect.

Table 3.4 Residues of Dieldrin in the livers and brain of birds, found dead or dying in the Wandere Valley. For comparision data of birds caught alive are added.

Residue in ppm

¹⁾ Caught at Sikri Point 70 days after treatment
²⁾ Speciment caught alive carrying near lethal body burden

Dieldrin levels in livers of 13 fish-eating birds (8 cormorants and 5 pied Kingfishers) collected near the Lake Victoria shoreline sprayed, varied from 0.03 - 1.81 ppm (all data on a wet weight basis). In 15 larger fish (Tilapia esculenta, Alestes jacksoni and Clarias mozambicus) Dieldrin residues (total body or lateral muscles) varied from 0.009 - 0.086 ppm (Koeman et al., 1969). From this study it may be concluded that Dieldrin used in ground spray operations is especially dangerous to insectivorous birds, and also accumulates in the tissue of fish and fish eating birds following shoreline operations. From 1970–1972 the possible short and long term hazards to game animals resulting from aerial Dieldrin trials were studied in the Lambwe Valley. The uptake of Dieldrin by game animals and its photo-isomerization product photodieldrin, as well as residues in grass and soil, was determined immediately after spraying and in a 15–24 month period after the last date of aerial Dieldrin applications. It is reported that for this reason various game animals up to a total of 70 individuals were shot and their livers and brain analysed for residues (Allsopp, 1978; Richardson, 1970).

Mean residual levels of Dieldrin and photodieldrin in brain, liver and kidney of the animals (Reedbuck, Oribi and Hyena) were below 1 ppm for post-spray samples. After 15 to 21 months, the highest level recorded was 0.04 ppm. Hyena's in the area feed mainly on domestic stock, so this species is not representative for the higher trophic level of the ecosystem concerned. A Civet that may have been feeding on birds and rodents affected by Dieldrin had a liver residue of 4.3 ppm Dieldrin.

As might have been expected, the results are comparable to those of the rodent analyses referred to above, and indicate that both the groundspray and aerial spray procedures concerned do not give rise to a dangerous contamination of the grass and shrubs on which the larger game animals feed.

Niger

Observations were made on side effects of helicopter spraying of Endosulfan at a dosage of 900 g a.i./ha on the fringing forest along the Tapoa and Mekron in the Parc National du W du Niger. Based on previous experience in Nigeria (see under Nigeria) mortality of non-target organisms was expected.

Of the non-target insects Diptera, Formicoidae, Coleoptera, Hemiptera and Lepidoptera showed mortality, but after the spraying malaise trap catches showed the same qualitative composition of insect groups as before spraying, but numbers were less. Two dead froggs were recorded. Only a few dead birds were found, Endosulfan residues in their livers being low. However, several specimen of insectivorous birds, e.g. smaller Green-backed Eremomela, Grey-backed Camaroptera and Gonolek showed unusual tameness immediately after spraying, indicating that a slightly higher dose could have resulted in a much higher mortality.

The Endosulfan spraying caused a high mortality of fish. The large number of species found dead (48) indicates that all or almost all species were affected. Residue levels in fish were low and probably did not represent a danger for human consumption.

Long term effects resulting from food shortage for insectivorous birds or fish-eating birds and crocodiles have not been determined (Dortland et al., 1977). After the following rainy season the fish population of the Mekrou River appeared to have recovered to its original species diversity and numbers almost completely. However, dwarf species in the Tapoa River were present only downstream of the rapids. Upstream this important trophic level of the foodchain was absent (Roman, 1978).

Nigeria

Most studies concerning the environmental impact of tsetse control by chemical means, are reported from Nigeria. Figure 3.1 shows the areas where the studies were carried out.

In 1968 fish-eating birds were collected along the shore of Lake Chad and the estuary of the Yobé River in the lake (area 1, figure 3.1). The aim was to measure residue levels of DDT and its metabolites (in particular DDT).

In the previous period from 1952 to 1966 tons of DDT had been applied on the Komadugu Gana/Yobé River system for the control of G. tachinoides and G. morsitans submorsitans (Davies, 1964, 1971). The DDE levels in the livers of 15 Cormorants and Pied Kingfishers varied form 0.012 to 0.17 ppm indicating that at that time the level of contamination of the estuary and the lake was relatively low (Koeman and Pennings, 1970).

Acute and long-term effects of discriminative groundspray applications of Dieldrin were studied in a fringing forest habitat in the Sudan Savanna near Potiskum (area 2, figure 3.1) in 1969 and 1970. The results showed that the application of Dieldrin was associated with acute deaths of many wild animal species including among others insects, fish, reptiles, amphibians, birds (especially insectivorous species) and certain mammals (e.g. bats, squirrels). The increased death rates could be correlated with a marked decline of the population density of certain species. Observations carried out one year after spraying showed that almost all species of mammals, birds, reptiles and amphibians included in the census before and at the time of spraying had survived and that many of the depleted species showed signs of recovery; the latter very probably being due to immigration of individuals from adjacent unaffected areas. However, some species of insectivorous birds characteristic for the fringing forest biome were still present in low density as compared to pre-spraying conditions and control areas, while a few apparently had reached the point of local extinction (e.g. the White-crowned Robin-Chat, Cossypha albicapilla (Vieillot) and the Blue-breasted Kingfisher, Halcyon malimbicus (Shaw)). Dieldrin levels discovered in the brains of birds found dead varied from 2.2 to 16 ppm (wet weight basis). In general in birds, amphibians and fish collected one year after spraying residue levels of Dieldrin were low and very probably did not represent a toxic hazard for the animals analysed (Koeman et al, 1971).

In the period from 1974 to 1976 side-effect studies were made of helicopter applications of Dieldrin and Endosulfan and groundspray applications of Endosulfan and DDT. These studies were carried out on three locations in the Northern Guinea Savanna Zone (figure 3.1., area 3, 4, 5). It was found that all applications concerned may give rise to mortality in many non-target species. After helicopter application the period in which mortality occurred in birds, mammals, reptiles, amphibians, fish and non-target insects amounts to 2 to 3 weeks in the case of Dieldrin and approximately 1 week in the case of Endosulfan. In aerial spraying Endosulfan appeared to cause mortality mainly through direct contact, whereas with Dieldrin secondary poisoning of predators (herons, birds of prey) is more likely to occur than with Endosulfan. Koeman et al. (1978) relate this to the faster biodegradability of Endosulfan. Endosulfan seems to cause slightly more damage to coldblooded vertebrates than Dieldrin, while on the other hand certain warmblooded species are far less susceptible for Endosulfan than for Dieldrin.

Fig. 3.1. Map showing study areas

As compared with former studies on side effects of Dieldrin groundsprays, the helicopter aerial applications affect a wider range of bird species than discriminative groundspray applications. Certain species appeared to be very vulnerable and either were not recorded anymore or became extremely rare in the areas treated. Certain mammalian species were also affected by the spraying operations and their numbers markedly reduced (e.g. Tantalus monkeys after aerial applications of Dieldrin and Fruit Bats after aerial applications of Endosulfan) (Koeman et al., 1978).

The environmental effects of ULV fixed-wing applications of Endosulfan were studied in the Doma Forest Reserve (area 6, figure 3.1) in 1976. No adverse effects on wildlife (including mammals, birds, reptiles, amphibians, fish and non-target insects) were observed during and after the spraying cycles. This method seems to be less deleterious to wildlife than the helicopter application of Endosulfan tried out experimentally in other parts of Nigeria (Takken et al., 1976). However, the eradication of G. palpalis was not successful.

In 1977 and 1978 field trials were carried out to test three synthetic pyrethroids. These compounds combine a high insecticidal activity with low oral mammalian and avian toxicity and with a good physio-chemical persistence. After spraying no dead birds, nor other vertebrates were found and no marked changes in bird species abundance were noted. However, some small fish showed a temporary abnormal behaviour. Acute mortality occurred in many terrestrial and aquatic arthropods, notably water beetles and crustaceans. The abundance of flying insects was significantly reduced for at least 10 days after spraying. Shrimps (Caridina africana) and mayfly larvae disappeared from river benthos samples after spraying, but were found again in samples taken one year later. The severe reduction in aquatic arthropod abundance may represent a hazard to fish populations through reduced food supply. At the small scale of these field trials, the potential for recovery of the affected populations cannot be assessed, as immigration from adjacent untreated areas may be able to restore population levels (Smies et al., 1980).

Tanzania

The only side-effect studies conducted in Tanzania concern a DNOC application. DNOC (2-methyl-4, 6-dinitro-ortho-oresol) was used as a defoliating agent when 1000 ha of tsetse bush were aerially sprayed with a 20% DNOC solution (5.9 kg DNOC/ha). No residues were found in animal samples (buffalo, water-buck and gazelle) or in stagnant water 30 and 60 days after the spraying. Only a small trace of less than one ppm was found in the vegetation. It was therefore concluded that the use of DNOC at this dosage could lead to no danger of poisoning (Tarimo and Pallotti, 1979).

Uganda

The environmental impact of 8 to 10 years Dieldrin application in Eastern Ankole (western Region) is reported by Sserunjoji and Tjell (1971) and Sserunjoji (1973). During the period from 1964 to 1972 a total amount of 131 600 kg of Dieldrin was applied in discriminative groundspraying operations on a total area of 8600 km². The environmental study was restricted to residual analyses of soil, grass and fish samples. Dieldrin residues in soil varied from .066 to .014 ppm, residues in grass varied from 0.93 to 0.09 ppm. In general, both in soil and grass residue levels were inversely proportional to the time of application, Dieldrin still being detectable in samples taken from areas sprayed ten years from the date of sampling. Residues in fish also were low. Moreover, there is evidence of degradation in some fish species (Protopterus aethiopicus and Clarias sp.). In addition, dilution factors and adsorption involving the muddy nature of water run-off and grassfires would tend to in activate the residual insecticide and protect aquatic systems. Thus the general findings have indicated less risk than anticipated of the environmental contamination from tsetse control by Dieldrin, according to the authors (Sserunjoji, 1973; 1976).

Upper Volta

A WHO research project for the evaluation of various insecticides for the control of riverine Glossina vectors of human trypanosomiasis was carried out in the fringing forests along the Comoe and Volta Noire. A variety of insecticides was tested in 1977, selecting Endosulfan, Decamethrin and Permethrin for future trials. Studies were made to assess the direct effects, like mortality and morbidity in non-target species and to register population changes in birds, fish and some invertebrates during and after treatment.

Table 3.5 TOTAL INSECT AND SPIDER MORTALITIES FOLLOWING INSECTICIDE APPLICATIONS

(Takken et al., 1978)

Table 3.6 NUMBERS OF MACROBRACHIUM RARIDUS AND CARIDINA AFRICANA CAUGHT BEFORE AND AFTER INSECTICIDE APPLICATION

(Takken et al., 1978)

* Paralysed

In 1977 the applications did not result in any observable changes in the condition of birds or bird species abundance, and no adverse effects were found in fish. The pyrethroids, however, did cause paralysis and mass mortality in two species of fresh water crustaceans, notably Caridina africana and Macrobrachium raridens. Fenthion, Decamethrin and Permethrin had some effect on non-target insects, whereas Decamethrin as a residual spray reduced Diptera populations considerably and completely exterminated the crustacean populations. In table 3.5 and 3.6 data are given on insect mortality and shrimp population levels following insecticide applications (Taken et al., 1978).

In 1978 Endosulfan and Decamethrin were tried in residual applications, while Permethrin and Endosulfan were tested as aerosols in serial applications. The side effects of these insecticide applications are summarized in table 3.7. Endosulfan gave only slight mortality of insects and no mortality of fish and crustaceans when applied as an aerosol, but resulted in mass mortality when applied as a residual spray. Decamethrin generally proved more toxic than Permethrin and at 12.5 g/ha virtually eliminated the two crustacean species (Everts et al., 1978).

After one year the shrimp Caridina africana had returned in great numbers with many females carrying eggs. However, the prawn Macrobrachium raridens had not yet returned to the former treatment zone. The area where fish had been severely affected by Endosulfan was repopulated during the rainy season (Everts et al., 1979).

Table 3.7. Summary of side effects of spraying in the Volta Noire and Comoe area (Everts et al., 1978)

N.O. = No Observation
+ = slight effect
++ = marked effect
+++ = virtual disappearance

¹⁾ observed mortality of Gastropoda
²⁾ observed mortality of juvenile fish and tadpoles

Zambia

In the Chipangali area, Eastern Zambia, side effects of Dieldrin applications using knapsack and Unimog mist-blower spraying devices were studied in the years 1962–1964. Potential tsetse resting sites along road verges and on all trees lining access tracks, including rot holes in trees, overhanging branches and fallen logs were sprayed. Holes in the ground, particularly those of the antbear Oryceteropus afa, proved to be an important resting and breeding site of tsetse.

Dead and dying animals were collected only incidentally, including mammals, birds, reptiles and fish. In Table 3.4 mammals found dead after the Dieldrin applications were listed. It is reported that the Hyenas very probably died due to their feeding on Elephant meat which was contaminated with Dieldrin by accident. A total of 38 birds representing 24 species were found dead or dying, among which 15 insectivorous species were counted. A large number of various reptiles was collected, 17 snakes and 101 lizards. Due to an incident (washing of empty Dieldrix T15 tins in a dam) a total of 223 fish were found dead within a week after the incident, all Tilapia species (Wilson, 1972).

Table 3.5. Mammals found dead after Dieldrin applications with groundspray (mistblower) devices (after Wilson, 1972)

Studies on the side effects of ULV fixed-wing Endosulfan applications in Barotse Province (south-west Zambia) were carried out in 1968. Emphasis was laid mainly on non-target insect species by means of ground observations immediately after spraying, measurement of the insect ‘fall-out’ of trees with a dense insect concentration and observations on a predator (Dragonfly) species population. No dead or dying animals were found, nor was a reduction observed in the abundance of species or population density. Insect collections made in sprayed and unsprayed areas revealed a total of subsequently 163 and 159 different species. Considerable quantities of fish were caught from a sprayed pool for three successive days (Magadza, 1969).

Zimbabwe

The method of sequential aerosol applications of Endosulfan is now widely used throughout Southern Africa (Botswana, Zambia), not presenting serious hazards to non-target organisms. In Zimbabwe small fish in shallow pools suffered from the application, and some immediate reductions in natural population levels of non-target organisms were recorded, but these were short-term and population levels soon recovered (Cockbill, 1979).

Conclusions

Most applications of pesticides, commonly used in tsetse control and eradication programmes, can kill non-target organisms. However, there is a marked difference in the types of organisms affected between the high dose rate residual applications of compounds such as Dieldrin, Endosulfan and DDT and the aerosol applications of Endosulfan and pyrethroid compounds.

The residual applications have caused mortality in both vertebrate and invertebrate organisms. In particular, populations of various insectivorous bird species were affected seriously. Certain of these recovered to their former population densities quite soon, but others did not recover quickly, or they even disappeared completely from certain areas (isolated habitats). Among the mammals, monkeys (Green Monkey), squirrels and bats appeared to be the most vulnerable groups. Very high mortality in fish resulting in temporary elimination of many species have occurred in some localized situations. In general, it has been found that helicopter applications of residual insecticides affect a wider range of species than more discriminative groundspray applications.

The aerial aerosol applications of Endosulfan and the pyrethroids have not so far caused any measurable effects on warm-blooded vertebrates. At the dose rates applied, the effects on fish have appeared to be slight, only a small proportion of the populations of a limited number of species being affected or none at all. Many groups of invertebrates, especially arthropods, are extremely susceptible to the pyrethroids. Comparing the studies in Upper Volta and Ivory Coast with respect to the use of Decamethrin, it can be concluded that the temperature of the river water may have a marked effect on the toxicity on fish. The latter also applies to certain shrimp species like Macrobrachium raridens and Caridina africana. From experimental studies pyrethroids as well as DDT are known to show a negative temperature coefficient, viz. these compounds are relatively more toxic at lower than at higher temperatures (Van Bercken, 1972; Manck et al., 1976; Harris and Kinoshita, 1977). Besides the crustaceans many other groups of invertebrates, especially anthropods, are extremely susceptible for the pyrethroids. The long term effects on the populations of the species concerned cannot yet be judged because of the insufficient baseline data on their population dynamics. Currently, studies are being carried out in a number of river systems in Africa with the objective of comparing the effects of spraying with the overall seasonal changes in certain populations of terrestrial and aquatic arthropods.