|DDT in the Tropics: The Impact on Wildlife (NRI, 1994)|
|Part 2 : Scientific report|
|Effects on wildfire|
Invertebrates of mopane woodland
The effects of DDT
Invertebrates in the food chain
Invertebrates are extremely numerous and highly successful animals, which make many important contributions to the functioning of the living world. Some are involved in decomposition processes and nutrient recycling and some with pollination, whilst others have a major impact on plant biomass through herbivory or play a part in regulating animal populations through predation or parasitism. Invertebrates, in turn, provide the food for many vertebrates.
DDT was developed specifically to kill insect pests and so may be expected to have its most severe non-target impact on other insects and invertebrates. However, despite its broad spectrum of activity, relatively few invertebrate taxa have been shown to suffer adverse population effects from agricultural use of DDT, though some groups, particularly predatory, mesostigmatic mite populations, have been shown to decline after treatment.) Other soil mites may also be affected, particularly in the short term and in response to high doses of the insecticide. This has knock-on effects on the normal prey of these mites, including some detritivorous mites [Cryptostigmata], nematode worms, and springtails [Collembola: Isotomidae].Some species within all these groups are tolerant of DDT and increase in number following treatment.
The majority of work on environmental impact of DDT has been done in farmland, but in a wide-ranging study of the effects of aerial spraying of DDT on forest invertebrates in the USA, a number of insect groups were found to be affected by the insecticide in the short term. Included in this were many bugs [Hemiptera], which were greatly reduced or completely eliminated, and several species of flies [Diptera] (especially Calypteratae). Wasps [Hymenoptera], from the superfamilies Prototrupoidea and Chalcidoidea, also declined in numbers after spraying, and several species of ant also seemed to be affected e.g. Tapinoma sessile and Aphaenogaster treatae.
The ecological implications of these changes are unknown. In agro-ecosystems, the disruption caused by DDT to the natural balances within the system are recognized and have sometimes led to increased pest problems.Undoubtedly, similar changes will occur in more complex, natural habitats, but are less easy to detect than in the smplified ecosystems occurring in agriculture and horticulture. Until the ecology of natural habitats is better understood, the long-term impact of pesticides used in these areas will remain unclear.
Invertebrates played an important role in the ecological impact of DDT in temperate areas, as a route for entry of the insecticide into the food chain. For example, earthworms surviving DDT applications tend to contain higher DDT residues than the surrounding environment, with the concentration factor averaging about 9-fold.Contaminated worms were implicated in deaths of robins (Turdus migratorius) feeding around elm trees sprayed with DDT against vectors of Dutch Elm disease in Urbana, USA.There is also evidence that invertebrates continue to pick up residues for considerable periods after spraying has ceased and even in areas that have been previously unsprayed, so providing a potential reserve for accumulation through food chains.
The few attempts to monitor environmental impact of ground-spraying operations against tsetse fly have paid little attention to invertebrates. There are reports of effects on non-target insects, but no further details are given. The only specific example noted is the elimination of hersiliid spiders from the bark of sprayed trees in Chad.
Before the ecological effects of DDT use in the Siabuwa area could be assessed, baseline data on the composition of the invertebrate fauna were required. However, the sheer diversity of invertebrates precluded studying all of them and only the groups likely to be at greatest risk were investigated.
Study was restricted to the three groups at direct risk from DDT sprayed onto tree trunks, under fallen logs, under overhanging rocks and on the soil under thicket vegetation. These were epigeal invertebrates (those living on or above the soil surface); arboreal arthropods (specifically those living on and visiting tree trunks) and soil dwellers (also at risk from indirect contamination from wash-off).
There were a number of woodland habitat types in the study area, but invertebrate sampling was confined to Colophospermum mopane-dominated woodland. This is an important habitat for tsetse in the Zambezi valley and for ground-spraying operations.20
The use of widely speed sample sites allowed for any variation within treatment areas, but presented a number of problems in interpretation of the collected data, including lack of pre-spray information on sprayed sites and a variety of uncontrolled environmental factors within and between sprayed and unsprayed sites.
A spray monitoring area (SMA) was thus set up for more detailed study of the impacts of spray operations, to allow for pre- and post-spray sampling in a previously unsprayed habitat, something which could not be done over the rest of the study area (Box 3.1). This aided interpretation of the data from more widespread sites, where effects of the pesticide on invertebrate populations may otherwise have been overlaid by changes due to other environmental variables.
Sample points were selected, where possible, to coincide with those used by other members of the study team, particularly the ornithologist and soil microbiologist. This allowed any changes in bird populations or in the functioning of soil processes to be related to invertebrates found at the same sites. Sampling was carried out during the dry seasons (June and July) of 1988, 1989 and 1990.
Box 3.1 Invertebrate monitoring
Invertebrates were monitored at 20 sites spread widely over the study area, in both sprayed and unsprayed habitat, using pitfall traps, termite baits, soil cores and ant nest surveys. Only a small proportion of a 'sprayed area' is actually treated with DDT during ground-spraying operations and targetting of the insecticide is very specific. Sample sites within the sprayed area (S1-S10) were restricted to the zone treated in consecutive years since 1984, and were selected by examining trees for visible DDT deposits remaining from the previous year's operation. This ensured that only fauna directly exposed to the insecticide were monitored, whereas random selection of sites in the sprayed area could have resulted in sites being some distance from the nearest sprayed tree. For many sedentary invertebrates, such sites would be, to all intents and purposes, unsprayed. Study sites in the sprayed area thus represent the worst case, reflecting the conditions directly within spray swaths, (approx. 20% of the total area). Selected sites in the sprayed area were then matched with 10 similar sites in the unsprayed area (U1-U10). A study site consisted of one tree within an area of mopane woodland.
The 'spray monitoring area' (SMA) was set up in an area of mopane woodland stretching across the 1987 spray boundary into the unsprayed zone, at a point where the projected 1988 spray boundary coincided with the 1987 boundary (Figure 3.1). The area was divided into four sections:
'87 area': sprayed in 1987 (with DDT deposits still visible on a number of trees in June 1988) and again in 1988 and 1989;
'88 area': previously unsprayed, sprayed for the first time in 1988 and again in 1989; '89 area': previously unsprayed, sprayed for the first time in 1989;
'C area': unsprayed before and throughout the study.
Each area measured 50 m by approximately 40-60 m (the swath width of the spray team) and contained at least 25 trees of a size suitable as tsetse resting sites. Five of these were used as invertebrate sampling points for pitfall trapping, soil coring, termite baiting and litter bag studies; another five were used for trunk trapping. Ant baits were set up in the central part of each sector.
Invertebrates of mopane woodland
A total of 72 400 epigeal invertebrates from 400 pitfall traps (trapping effort of 2800 trap days) were sorted, counted and identified. Over 575 species were represented (Box 3.2), with many new to science. Springtails (Collembola) were most abundant within the catch (Figure 3.2a), but the spider fauna showed the highest diversity. Insects were caught in large numbers, made up the highest proportion of the biomass (Figure 3.2b) and were dominated numerically by the ants. The faunal composition shows characteristics typical of arid environments.The fauna represent a wide range of trophic levels including primary consumers, detritivores and scavengers, predators and parasites, all important in the functioning of the woodland ecosystem. In turn,they provide the major food source for a variety of reptiles, birds and mammals.
Box 3.2 Epigeal invertebrates
Epigeal invertebrates were sampled using pitfall traps (Figure 3.3), modified from the design described by Grant. The invertebrates caught span 7 classes, 26 orders, 183 families and over 575 species. The general composition of this fauna is given in Figure 3.2. The invertebrate fauna
trapped is of considerable taxonomic interest, with 27 species new to science in the collection.There were probably many other undescribed species present as well, but specialist taxonomic assistance was not found to identify all groups.
The Isopoda, Chilopoda and Diplopoda are poorly represented (Figure 3.2). The Isopoda are omnivorous detritivores, the Diplopoda feed on a variety of dead plant material and all the Chilopoda are predators, although some also take decaying plant material.
Amongst the Arachnida, the Acarina were caught in the largest numbers (Figure 3.2). The Prostigmata predominate and include fungivores, bacterivores and predators (Figure 3.4). The detritivorous
or fungivorous Cryptostigmata are next most numerous, with relatively few of the predatory Mesostigmata, the bacterivorous or fungivorous Astigmata and the ectoparasitic Metastigmata. Identification of mites was only attempted to family level. In general, it appears that the fauna was relatively limited consisting of only a handful of species per family.
The 1078 spiders caught come from 25 families and show the highest diversity of all the epigeal invertebrates (Figure 3 5). The Gnaphosidae and Salticidae are the most numerous and species-rich families. Classification of spiders to genus and species is tentative, as the majority caught in pitfall traps were immature and therefore difficult (or impossible) to identify past family level.All are predators.
Within the Insecta, Hymenoptera and Coleoptera predominate. The ants are most numerous, and scavengers, seed harvesters and predators are all represented. The Myrmicinae are most abundant and species-rich, dominated numerically by various Pheidole species, which were common across the whole study area.
The Coleoptera were numerically dominated by two families: the Tenebrionidae and the Histeridae. The tenebrionids were largely from the sub-family Tentyriinae and are probably detritivorous as larvae and adults, whilst the histerids may be predators of small invertebrates.
Few Isoptera were caught in the pitfall traps, despite their importance in this habitat. 3" The harvester termite Hodotermes mossambicus (Hager) IHodotermitidael was caught in the largest numbers, with small numbers of moundbuilding Macrotermes Spp. and Odontotermes spp.[Termitidael].
The composition of the pitfall trap catches indicate a number of faunal characteristics typical of arid environments, with Collembola outnumbering Acarina; larger numbers of Prostigmata than of Cryptostigmata; relatively low species richness of the ant fauna; presence of Hodotermes mossambicus and numerical dominance of Tenebrionidae within the Coleoptera.
Invertebrates walking up mopane tree trunks were sampled using trunk traps. The catch contained fewer species than the pitfall traps and comprised a mixture of specialist trunk dwellers, predominantly epigeal species and a number of species of flies (Diptera) (Figure 3.7). A total of 6811 invertebrates were trapped over 420 trap days, suggesting that there were fewer invertebrates on the tree trunks than walking on the soil surface (assuming equal efficiency of the traps). As with the epigeal fauna, the springtails and ants were the most numerous (Box 3.3).
Box 3.3 Arboreal arthropods
The traps used for sampling trunk fauna were based on a design developed in New Zealand. The traps were made of locally available materialsa plastic sandwich box trap and a thick grade polythene collar around the tree trunk (Figure 3.6). The trap was generally set at about 1 m above the ground. The gap between the tree trunk and the plastic collar was sealed with mud to block crevices in the bark.
The faunal composition was limited by comparison with epigeal fauna caught in pitfall traps. Only three classes were represented, with 13 orders, 52 families and 75 species. As with the pitfall trap catches, Collembola were numerically dominant, comprising 90% of the catch (Figure 3.7).
The Arachnida were dominated numerically by mites, which showed a similar composition to the
ground-dwelling fauna. The detritivorous Cryptostigmata were grossly outnumbered by Prostigmata, and predatory Mesostigmata occurred only in small numbers. The spider fauna of the tree trunks also shared many families with the grounddwellers, although the species composition within families was somewhat different. Numerically, spiders made up an even smaller percentage of trunk trap catches than they did of pitfall trap catches.
Within the Insecta, the proportion of Diptera and Thysanoptera in the catch was notably greater in the trunk traps than the pitfalls, whilst the Isoptera were completely absent from trunk traps and Coleoptera reduced to very low numbers (Figure 3.7). Again the ants were numerically dominant, but the fauna was species poor by comparison with the pitfall trap catches.
The sample of invertebrates extracted from soil cores indicates a very low density of soil invertebrates by comparison with temperate forest (Table 3.1). Nematodes occurred in the largest numbers and there were relatively few micro- and macroarthropods. No earthworms were recorded at all from soil samples taken in the dry season, nor using a standard formalin drench technique. The absence of earthworms has been noted in previous studies in the drier tropics, where their role in nutrient cycling is frequently taken by termites.
Box 3.4 Soil invertebrates
Invertebrates were extracted from soil cores taken to a depth of 5 cm, around the bases of sample trees, using a simple flotation technique. In general, the density of micro-arthropods was lower (Table 3.1) than in either temperate or African woodland, but was similar to those in a banana plantation in Uganda. In that case, no figures were given for nematodes, but Collembolan density was comparable (1.8 + 0.37 x 103/m), whilst mite populations appeared somewhat higher than in this study at 6.87 + 0.88 x 1 03/m.
Fauna were extracted from litter bags containing dry mopane leaves buried in the soil around sample trees in the SMA (Chapter 2). More microarthropods were recovered from litter bags than from the soil cores and nematodes were also well representated. Over 10 000 individuals were identified, representing 6 classes, 23 orders, 97 families and about 150 species. The mites were the most abundant group and were the best represented, with some 100 species. Nematodes were less numerous, but 22 species were identified. Relatively few springtails were found, mostly from the family Isotomidae and insect numbers varied greatly, but were generally low. Pot worms (Enchytraeidae) were also found in litter bags, though in small numbers.
The effects of DDT
The biomass, diversity, faunal composition and relative abundance of individual taxa within pitfall trap catches were compared between sprayed and unsprayed sites.
Biomass was very variable within sectors of the SMA (Box 3.5) and there were no significant differences detected between sprayed and unsprayed areas (Table 3.2). There is thus no evidence that DDT affected the biomass of surface-active invertebrates in the period immediately following spraying, or in the long term.
Box 3.5 Biomass
During sorting of post-spray pitfall trap catches from the spray monitoring area for 1988, invertebrates were measured using an eye-piece graticule on a binocular microscope. The biomass of invertebrates was then estimated using a weight versus length relationship.
For the majority of invertebrates weight was calculated using the formula:
W= 0.0305 L
However, for Diplopoda and Chilopoda a linear equation was used.
W= -0.792 + 0.571 L
where W= weight and L = length.
The results in Table 3.2 show that biomass estimated in this way was similar in all four sectors of the SMA, whether sprayed or unsprayed. Variance of biomass estimates for catches in the two sprayed sectors was higher, but there was no evidence to suggest that DDT had any adverse impact on the biomass of surface-active invertebrates in the area.
Abundance and Diversity
Fewer invertebrates were caught in the '87' area than in any of the other sectors of the SMA throughout the period of the study (Figure 3.8). This was the most heavily sprayed sector of the SMA, but as there is no pre-spray data for this sector, the differences cannot be attributed conclusively to DDT. They may relate directly to inherent environmental differences in the area. However, there are indications that DDT may have been affecting invertebrates in the SMA, with fewer invertebrates caught in the '88' sector a year after it was sprayed for the first time, whilst numbers rose in both the unsprayed sectors.
Species richness of the pitfall trap catches shows a similar pattern (Figure 3.9). Whilst there was an overall decline in 'species' richness as the study progressed, this was most marked in the area sprayed for the first time during the study. The results do not constitute proof of a detrimental impact of DDT, although this seems likely. Certainly, other environmental variables are more important than spraying in determining the abundance and diversity of surface-active invertebrates, as neither the total number of invertebrates nor the number of taxa caught in pitfall traps showed differences between unsprayed (U) and sprayed (S) sites within the whole study area.
A variety of diversity indices were also calculated for the pitfall trap catch data and these show similar results with no evidence of differences in diversity between sprayed and unsprayed areas in the extensive survey, but indications of minor impacts of DDT from the intensive study in the SMA.
Similarity in Faunal Composition Different sites can be grouped using cluster analysis to show the similarity of their fauna (Box 3.6). Figure 3.10 shows clear separation of the sprayed (S) from unsprayed (U) sites in the extensive survey area. However, the lack of pre-spray information on these sites and their spatial separation leave the possibility that factors other than DDT may be responsible for the differences in faunal composition between the areas. Within the SMA there was no simple division into sprayed and unsprayed sites. Faunal similarity was governed principally by natural seasonal and annual variation, with some indications of an underlying longterm effect of DDT. There are indications that sites sprayed for the first time during the first year of the study (1988) became less similar to the unsprayed sites, a year after spraying. Thus, natural variation has a greater impact on faunal composition of sites than DDT, but there is evidence that DDT may have some effect.
Box 3.6 Faunal composition
Similarity of sites based on the invertebrate faunal composition of pitfall trap catches was quantified using Sorensen's Quotient of Similarity:
Q/S = (2j/a+b) x 100
where j = number of species common to both sites;
a = number of species at site A;
b= number of species at site B.
The similarity quotients were then subjected to hierarchical, average link, cluster analysis. This procedure involves construction of a coincidence table comparing each site with every other one and selecting that pair of sites with the highest quotient of similarity. These are then grouped and quotients of similarity calculated between this group and the remaining sites. The procedure continues until the sites are classified into two groupings within which the extent of similarity between the individual sites is shown by the cluster arrangement. The results are then drawn up as a dendrogram.
Figure 3.10 a shows that, with the exception of two sites, faunal composition in the sprayed area is different from that in the unsprayed area. However, this pattern is not repeated in the SMA, where factors other than DDT are more important in determining faunal composition. Figures 3.10 b and c show very different patterns from each other, indicating that short-term seasonal variation in faunal composition is outweighed by natural annual fluctuations. Certainly, no consistent effect of DDT is noticeable. Figure 3.10 d shows that sites were primarily separated on the basis of sampling time, with all sites sampled in 1988 being more similar to each other in terms of their faunal composition (regardless of treatment) than the same sites sampled in 1989. In both years, the '87' area sites stand out as being different and form their own cluster. This may be due to an effect of DDT, as this is the most heavily sprayed part of the SMA, but may represent inherent differences in the fauna of this site. However, there were indications of a secondary effect of DDT in the longer term, underlying the primary grouping of sites on the basis of sample date (Figure 3.10 d). Thus the '87' sites are grouped together and are less similar to any of the other sites before spraying in 1988, but in 1989 there is some sign that the now sprayed '88' sites have become less similar to other sites and more similar to '87' sites. Thus DDT may have some influence on species composition of sites, although sampling date has a greater effect on faunal similarity of sites than does the insecticide.
Relative abundance of individual taxa
Relative abundance of the 95 most numerous taxa caught was compared between sprayed and unsprayed areas and within the spray monitoring area over the two years of the study. There were significant differences detected between sprayed and unsprayed areas for many taxa, but few showed a consistent change in relative abundance which could be attributed directly to DDT. The mites and springtails did show some signs of effects (Box 3.7), but temporal and other uncontrolled environmental factors were the greatest cause of variation.
Box 3.7 Micro-invertebrate abundance
Few of the micro-invertebrates sampled show consistent evidence of an adverse impact of DDT on their relative abundance. Surface-active Cryptostigmata were more abundant in the sprayed (S) area than in the unsprayed (U), but trap catches in the spray monitoring area showed no evidence that this difference could be attributed to DDT, as fewer Crytostigmata were caught in the most heavily sprayed '87' area than in any of the other sectors, throughout the study (Figure 3.11). Thus no clear pattern is apparent for epigeal Cryptostigmata. However, a single species of soil-dwelling Cryptostigmata, Hypozetes sp. [Ceratozetidae] (Plate 14), occurred in significantly higher numbers in litter bags from repeatedly sprayed sectors of the SMA than from sectors sprayed once or not at all. None of the other 37 species showed any differences in relative abundance between sprayed and unsprayed sectors.
Epigeal Mesostigmata were caught in small numbers throughout the study and there were too few to show statistically significant differences between sites. However, the data indicate that fewer Mesostigmata were caught in pitfall traps at sprayed sites (Figure 3.11), which is consistent with previous findings on DDT impacts.This trend is confirmed by results for soil-dwelling Mesostigmata recovered from litter bags, with Protogamasellus sp. mica species-group [Ascidae] occurring in smaller numbers in sprayed sectors of the SMA.
The epigeal Prostigmata showed enormous variability in numbers caught and few taxa showed consistent effects in response to spraying. The predatory Bdellidae, which have been shown to be very sensitive to DDT in temperate agricultural soils, showed no evidence of a decline in abun
Collembola showed no consistent differences in abundance between sprayed and unsprayed areas.
More Isotomidae were recorded in the sprayed (S) area than the unsprayed (U) area, which would be expected if DDT were affecting predators, as occurred in response to agricultural use of the insecticide. However, results from the SMA did not support any evidence of an effect, either from pitfall trap catches, trunk trap catches, soil cores or litter bag samples.
The Collembola and other micro-invertebrates were less abundant in the '87' sector of the spray monitoring area, supporting evidence that the fauna in this sector was impoverished, though not necessarily as a result of DDT use.
Several taxa were not caught at all in sprayed area pitfall traps, but were found in the unsprayed areas, including a previously undescribed woodlouse Aphiloscia sp. [Isopoda: Philosciidae]; and the ponerine ant, Ophthalmopane berthoudi. There was insufficient data to attribute this to the use of DDT, but these taxa may have been affected. Extensive surveying within the study area also uncovered only one individual O. berthoudi within the sprayed area and this was not in a tract of woodland which had been sprayed. This ant is often predatory on termites and may be susceptible to accumulation of DDT residues from its food. Another ponerine ant, Platythyrea cribrinodis, which was shown to accumulate high residue levels in sprayed areas (Box 3.8), also showed indications of population decline coincident with DDT spraying (Plate 15). It became increasingly scarce in the SMA and the sprayed area as a whole. There is no proof that this observed decline was due to DDT, but circumstantial evidence points strongly in that direction.
Pitfall trap catches of several other species of ant also showed significant differences between sprayed and unsprayed areas, although few were consistent in relation to DDT treatment. Pitfall trapping can give misleading results in attempting to estimate populations of ants,43 44 so other sampling techniques, including nest surveys and the use of food baits, were also employed. An extensive survey of nests of the common species showed no differences in numbers or distribution around trees in the sprayed (S) and unsprayed (U) areas. Similarly, the use of food baits in 1990 failed to find evidence of effects of DDT on the abundance and species-richness of the groundforaging ant community in the SMA. However, P. cribrinodis was absent from the catches, irrespective of sector, although it had been caught in pitfall traps there in 1988 and 1989. P. cribrinodis was attracted to similar baits elsewhere and its absence from the SMA in 1990 is further evidence of a DDT-associated decline.
No consistent adverse effects of DDT on arboreal arthropods were apparent from the trunk trap catches. Only the Collembola showed significant differences in abundance between sprayed and unsprayed sectors of the SMA, with more caught in the sprayed sectors. However, the dominant family, the Entomobryidae, showed no differences.
The Diptera and the Hemiptera showed short-term changes in abundance on trees following spraying, but as individual families within these orders showed varying patterns of change, it is unclear whether there was a real effect of DDT. Individual families and species differ biologically and with classification at order level, there is insufficient evidence to attribute differences in relative abundance to an impact of DDT. Data were too sparse to allow analysis of differences in numbers of individual families or species for these groups.
Only nematodes occurred in large enough numbers in soil cores to warrant statistical analysis and there was no evidence that DDT affected their relative abundance (Table 3.3). However, fewer species were extracted from litter bags in the sprayed sectors than in unsprayed (Table 3.4). This is consistent with previous findings of the effects of DDT on nematodes.
There was also evidence of effects of DDT on soil mites. Fewer species of Prostigmata were extracted from litter bags from the sprayed sectors than from the unsprayed sector of the SMA (Table 3.4).
Similar numbers of Astigmata, Cryptostigmata and Mesostigmata were found in the sprayed sectors of the SMA and the unsprayed and species richness was also similar. However, Hypozetes sp. [Cryptostigmata: Ceratozetidae] occurred in larger numbers and Protogamasellus sp. [Mesostigmata: Ascidae] in smaller numbers in sprayed sectors than in the unsprayed (Box 3.7).
Very few termites were found in soil cores or in litter bags and thus wooden baits were used in order to assess relative abundance and activity of these important insects in the woodland environment. Termites were more active around baits in sprayed woodland than unsprayed and in some cases attacked more baits around sprayed trees. The only termite species found on baits when they were collected, were workers and soldiers of Ancistrotermes latinotus (Holmgren). However, it is possible that other genera and species may also have been involved in attacks on other baits. A. Iatinotus is the most common termite found in Zimbabwe and favours mopane woodland (amongst other habitats), feeding on a variety of woody material, leaf litter and living vegetation (Figure 3.12).3°
The results contrast with those from experiments investigating the effects of DDT treatment to cleared forest soil in Nigeria, where fewer softwood baits were attacked by Ancistrotennes sp. and Pseuacanthotermes sp. in treated plots.
Invertebrates in the food chain
Insects and other invertebrates play an important part in the ecology of many higher animals and are the major food source for a wide range of reptiles, amphibians, birds and mammals. The contamination of invertebrates with DDT residues therefore has implications for the status of other animals in the ecosystem. These were considered through studies of diet in lizards (Box 4.1) and birds (Box 5.4) and by measurement of residue concentrations in samples of prey species (Box 3.8).
Residue analysis revealed that all the invertebrates sampled in the sprayed area contained DDT (Table 3.5). Many showed traces of the insecticide, even when taken in unsprayed areas, but residue burdens in insects taken from sprayed sites were always greater and in some cases reached very high levels. Ants showed the highest residue burdens of the fauna sampled, with most of the insecticide being in the form of p,p'DDT, the most acutely toxic form. Termites and beetles on the other hand had generally lower DDT burdens and a higher proportion of this appeared as the metabolite DDE, which is less acutely toxic to invertebrates. This may have been a result of sampling, as ants were taken directly from sprayed tree trunks, whilst termites and beetles were generally collected from the ground in the immediate vicinity of the same trees. Despite the high residue burdens, none of the ants sampled showed any unco-ordinated behaviour or tremors, characteristic of DDT poisoning.
Box 3.8 DDT residues in invertebrates
Individual invertebrates were collected alive from sample sites, using clean forceps, and placed in small aluminium canisters or glass sample vials with screw tops lined with aluminium foil. Samples were either frozen or kept in 10% formalin solution, from time of collection until residue analyses were carried out. Forceps were washed in xylene and rinsed in alcohol before changing sample insect or sampling site, but not between individuals of the same species taken at the same sampling site.
In every case, except that of one cricket, the DDT burdens were considerably higher in samples from sprayed areas than from unsprayed. DDT residues from invertebrates in the unsprayed area (except for the one cricket mentioned above) were mainly in the form of p,p'-DDE, whereas invertebrates from the sprayed areas contained a high proportion of unchanged DDT [mainly the p,p'-DDT isomer).
The ant species Camponotus spp. and Platythyrea cribrinodis showed the highest residue levels. This parallels the findings of a survey of carnivorous insects in forest and agricultural areas in Belgium.It was also notable that the highest proportion of DDT detected in these insects was in the form of p,p'-DDT, whereas the termites and Adesmini beetles collected had similar proportions of DDT and DDE, or a higher proportion of DDE. DDE is recognized as one of the first metabolites of DDT produced by a variety of invertebrates, and is much less acutely toxic to insects than DDT itself. The high level of unchanged DDT in the ants suggests that the species collected are tolerant to the insecticide without metabolizing it. Indeed one Pcribrinodis contained 2.9 1lg p,p'-DDT and was still apparently healthy, showing no lack of coordination or tremors characteristic of DDT poisoning. This fact could be significant for predators feeding on these ants (Plate 16).
DDT can persist in the biotic environment for long periods of time as evidenced by the 3.3 ppm S DDT recorded in Platythyrea cribrinodis, 2 years after the area in which it was caught had last been sprayed. Although levels were considerably lower than in individuals collected within 6 months of spraying, they were still much higher than in unsprayed areas.
In general, natural spatial and temporal variation in relative abundance of invertebrates was high within the study area and any impact of DDT would have to be fairly severe to show up against this background noise'. The differences in abundance of soil-dwelling mites parallel previous findings and indicate that despite discriminative application, sensitive fauna do suffer direct effects of the DDT. No such major impact was detected in terms of relative abundance of any epigeal taxon, nor changes in diversity or community structure of the surface-active assemblages sampled. However, there were indications of both primary and secondary effects of DDT both at the community and individual level.
The mopane woodland ecosystem is a poorly studied one and the ecology of this habitat remains largely unknown. This is reflected in the large number of previously undescribed species recorded from pitfall traps. There is little information on the biology and habits of many of the invertebrates found, even those to which names can be assigned. As a result, it is difficult to do more than guess at the ecological implications of any changes in the invertebrate fauna of the mopane woodland occurring as a result of DDT use. The results suggest that DDT-induced change in invertebrate abundance, if it occurs at all, is secondary to natural fluctuation in populations and activity. This is caused by microhabitat differences within different tracts of the mopane habitat and temporal and seasonal changes in temperature, relative humidity and other environmental factors. Nonetheless, evidence that DDT is having an influence on the invertebrate faunal composition within the woodland cannot be dismissed as insignificant. Until the ecology of this type of woodland is understood, any change in the fauna must be regarded as potentially leading to change in the functioning of the system.
Despite the fact that there is little evidence of population decline in the invertebrate groups sampled, their ability to carry high residue levels of DDT in their bodies and for it to persist over relatively long periods of time indicates that invertebrates are important components of ecological effects of DDT ground-spraying. The food chain link from invertebrates, to woodland birds (Chapter 5) and the skink Mabuya striata (Chapter 4) which suffered population decline in sprayed areas is clear.
1. Brown, A.W.A. (1978) Ecology of Pesticides. New York: John Wiley and Sons.
2. COPR (1978) Pesticide residues research, project. Final report and recommendations 1975-1978. Ibadan, Nigeria (ODM research scheme R2730): Ministry of Overseas Development/International Institute of Tropical Agriculture.
3. Spain, A.V. (1974) The effects of carbaryl and DDT on the litter fauna of a Corsican pine (Pinus nigra var. maritima) forest: A multivariate comparison. Journal of Applied Ecology, 11: 467-481.
4. Dindal, D.L., Folts, D. and Norton, R.A. (1975) Effects of DDT on community structure of soil microarthropods in an old field. In: Progress in Soil Zoology. (Vanek, J., ed.) pp. 505-513. The Hague: Junk Publishers.
5. French, N., Lichtenstein, E.P. and Thorne, G. (1959) Effects of some chlorinated hydrocarbon insecticides on nematode populations in soils. Jourrnal of Economic Entomology, 52: 861-865.
6. Sheals, J.G. (1956) Soil population studies.]. The effects of cultivation and treatment with insecticides. Bulletin of Entomological Research, 47: 803-822.
7. Edwards, C.A., Dennis, E.B. and Empson, D.W. (1967) Pesticides and the soil fauna: Effects of aldrin and DDT in an arable field. Annals of Applied Biology, 60: 11-22.
8. Dempster, J.F. (1967) A study of the effects of DDT applications against Pieris rapae on the crop fauna. Procceedings of the 4th British Insecticides and Fungicides Conference, 1: 19-25.
9. Butcher, J.W. and Snider, R.M. (1975) The effect of DDT on the life history of Folsomia candidn (Collembola:Isotomidae). Pedobiologia, 15: 53-59.
10. Hoffmann, C.H., Townes, H.K., Swift, H.H. and Sailer, R.J. (1949) Field studies on the effects of airplane applications of DDT on forest invertebrates. Ecological Monographs, 19: 16.
11. DeBach, P. (1974) Biological Control by Natural Enemies. London: Cambridge University Press.
12. Davis, B.N.K. (1968) The soil macrofauna and organochlorine insecticide residues at twelve agricultural sites near Huntingdon. Annals of Applied Biology, 61: 29-45.
13. Gish, C.D. (1970) Pesticides in soil. Organochlorine insecticide residues in soils and soil invertebrates from agricultural lands. Pesticides Monitoring Journal, 3(4): 241-252.
14. Edwards, C.A. (1973) Persistent Pesticides in the Environment. Ohio: CRC Press.
15. Barker, R.J. (1958) Notes on some ecological effects of DDT sprayed on elms. Journal of Wildlife Management, 22(3): 269-274.
16. Thome, J.P., Debouge, M.H. and Louvet, M. (1987) Carnivorous insects as bioindicators of environmental contamination: Organochlorine insecticide residues related to insect distribution in terrestrial ecosystems. International Journal of Environmental and Analytical Chemistry, 30: 219-232.
17. Koeman, J.H., Balk, F. and Takken, W. (1980) The Environmental Impact of Tsetse Control Operations. A Report on Present Knowledge. FAO Animal Production and Health Paper 7 rev.1. Rome: FAO.
18. Koeman, J.H. (1977) Effects of tsetse fly control measures on non-target organisms. Mededelingen van de Faculteit Landbouwwetenschappen Rijksunivsiteit, Gent, 42(2): 889-896.
19. Tibayrenc, R. and Gruvel, J. (1977) La campagne de lutte contre les glossines dans le basin du lac Tchad. II Controle de l'assainissement glossinaire. Critique technique et financiere de l'ensemble de la campagne. Conclusion generals. Revue d'élevage et de medicine vétérinaire des pays tropicaux, 30(1): 31-39.
20. Pollock, J.N. (1982) Training Manual for Tsetse Control Personnel. Vol. II Ecology and Behaviour of Tsetse. Rome: FAO.
21. Tingle, C.C.D., Lauer, S. and Armstrong, G. (1992) Dry season, epigeal invertebrate fauna of mopane woodland in northwestern Zimbabwe. Journal of Arid Environments, 23: 397-414.
22. Douthwaite, R.J. and Tingle, C.C.D. (1992) Effects of DDT treatments applied for tsetse fly control on White-headed Black Chat (Thamnolaea arnoti) populations in Zimbabwe. Part II: Cause of decline. Ecotoxicology, 1: 101-115.
23. Tingle, C.C.D. and Grant, I.F. (in press) The effect of DDT on litter decomposition and soil fauna in semi-arid woodland. Acta Zollogica Fennica.
24. Grant, I.F. (1989) Monitoring insecticide side-effects in large-scale treatment programmes: Tsetse spraying in Africa. In: Pesticides and Non-target Invertebrates. (Jepson, P.C., ed.) Wimborne, Dorset, U.K: Intercept, pp. 43-69.
25. Disney, R.H.L. (1990) Revision of the Alamirinae (Diptera: Phoridae). Systematic Entomology, 15: 305-320.
26. Disney, R.H.L. (1991) Scuttle flies from Zimbabwe (Diptera: Phoridae) with the description of five new species. Journal of African Zoology, 105: 28-42.
27. Disney, R.H.L. (1992) The 'missing' males of the Thaumatoxeninae (Diptera: Phoridae). Systematic Entomology, 17: 55-58.
28. MacFarlane, D. (Personal Communication), International Institute of Entomology, c/o Department of Zoology, The Natural History Museum, Cromwell Road, London SW7 5BD, UK.
29. Murphy, J.A. (Personal Communication), 323 Hansworth Road, Hampton, Middlesex.
30. Mitchell, B.L. (1980) Report on a survey of the termites of Zimbabwe. Occasional Papers. National Museum Rhodesia Ser. B, Natural Scince, 6(5): 187:323.
31. Moeed, A. and Meads, M.J. (1983) Invertebrate fauna of 4 tree species in the Orongorongo Valley, New Zealand, as revealed by trunk traps. New Zealand Journal of Ecology, 6: 39-53.
32. Wallwork, J.A. (1976) The Distribution and Diversity of Soil Fauna. London: Academic Press.
33. Raw, F. (1959) Estimating earthworm populations by using formalin. Nature, 184: 1661-1662.
34. Lavelle, P. (1988) Assessing the abundance and role of invertebrate communities in tropical soils: Aims and methods. In: Proceedings of a seminar on resources of soil fauna in Egypt and Africa. Cairo,16-17 April 1986. (Bhabbour, S.I. and Davis, R.C., eds) Journal of African Zoology, 102: 275-283.
35. Lasebikan, B.A. (1988) Studies in soil fauna in Africa: current status and prospects. In: Proceedings of a seminar on resources of soil fauna in Egypt and Africa. Cairo, 16-17 April 1986. (Ghabbour, S.I. and Davis, R.C., eds.) Journal of African? Zoology, 102: 301-311.
36. Block, W. (1970) Micro-arthropods in some Uganda soils. In: Methods of Study in Soil Ecology (Phillipson, J., ed.) Proceedings of the Paris symposium organized by UNESCO and the Internationai Biological programme. pp 195-202. Paris: UNESCO.
37. Rogers, L.E., Hinds, W.T. and Buschbom, R.L. (1976) A general weight vs. length relationship for insects. Annals of the Entholnological Society of Amnerica, 69: 387-389.
38. Rogers, L.E., Buschbom, R.L. and Watson, C.R. (1977) Length-weight relationships of shrub-steppe invertebrates. Annals of the Entomological Society of America, 70: 51-53.
39. Tingle, C.C.D. (unpublished) The effects of DDT on diversity and faunal composition of invertebrates in mopane woodland in northwestern Zimbabwe.
40. Perfect, T.J., Cook, A.G., Critchley, B.R. and Russell-Smith, A. (1981) The effect of crop protection with DDT on the micro-arthropod population of a cultivated forest soil in the sub-humid tropics. Pedobiologia, 21: 7-18.
41. Wallace, M.M.H. (1954) The effect of DDT and BHC on the population of the lucerne flea, Sminthurus viridis (L.) (Collembola) and its control by predatory mites, Biscirus spp. (Bdellidae). Australian Journal of Agricultural Research, 5: 148-155.
42. Tingle, C.C.D. (1993) Bait location by ground-foraging ants (Hymenoptera: Formicidae) in mopane woodland selectively sprayed to control tsetse fly (Diptera: Glossinidae) in Zimbabwe. Bulletin of Entomological Researh, 83: 259-265.
43. Douthwaite, R.J., Fox, P.J., Matthiessen, P. and Russell-Smith, A. (1981) The Environmental Impact of Aerosols of Endosulfan, Applied for Tsetse Fly Control in the Okavango Delta, Botswana. Final report of the endosulfan monitoring project, London: Overseas Development Administration.
44. Marsh, A.C. (1984) The efficacy of pitfall traps for determining the structure of a desert ant community. Journal of the Entomological Society of South Africa, 47(1): 115-120.