Pupal Ecology and Role of Predators and Parastoids in Natural Population Regulation of Glossina pallidipes Austen (Diptera :Glossinidae) at Nguruman, Kenya

Abstract

A two-year progranme was carried out at Nguruman in the Kajiado District of the Rift Valley Province of Kenya, to identify the characteristics of larviposition sites, quantify pupal mortality rates, determine predators and parasitoids of Glossina. pallidipes Austen and to provide data on mechanisms of population regulation through the actions of predators and parasitoids. Pupal sampling was done using hand searching for two man-hour per site. The efficiency of this technique was measured and found to be 60%. The larviposition sites were usually found in dense shade under bushes. This is the first description of Q. pallidipes breeding sites in the Nguruman area. Though puparia were found in a wide range of soil types, they occurred more frequently in loamy-sand soils and showed a marked tendency to be aggregated in shade underneath big horizontal branches. In general , larvae pupate near the surface of the soil where and when it is very wet or covered with a thick carpet of leaf debris, tut were found deeper in the soil when it is very dry or has no leaf debris cover. There was evidence that seasonal variations in relat ive abundance and distribution of puparia resulted partly from a seasonal shift in breeding sites from low-lying sites to sites on hi lly slopes during the rainy season when most riverine habitats were flooded. Rainfall over 80mm per month increases pupal mortality through flooding and waterlogging of the larviposition sites. Adverse cl imatic factors caused 15.9% mortality through degeneration and/or decomposition of puparia. The survivorship curve from pooled age data indicated a mortality rate of up to 4% a day. An index of overall pupal loss rate (log number of pupae found - log nuwber of tenerals caught in biconical traps in the following month) was plotted against the log number of pupae found. The strong linear relationship indicates that the overall pupal loss rate is density dependent, and thus serve to regulate populati on size. Part of this mortality could be quantified by holding fieldcollected pupae in the laboratory till emergence. The causes of non-emergence were identified as developmental failures (mean = 18.0% ), emergence failures (mean= 24%); parasitism (mean= 12.3%) and fungal infections (mean = 33.7%). None of these was density dependent. Predation estimated from empty puparial cases showed that an average of 24.0% had arthropod-induced damage. Parasitism rates by Exhyalanthrax lugens Lw. and f· beckerianus Bezzi (also known as Thyridanthrax. argentifrons Austen) were usually below 10.0% and apparently inversely density dependent with a delayed density dependent component. The Agar gel double immune-diffusion technique developed to detect tsetse meal in guts of arthropods and to identify the natural predators of G. pallidipes, proved very sensitive and specific for Glossina. The length of time a tsetse meal remained detectable in gut of different predators varied .from a minimum of~ h for the gryllid Liogryllus bimaculatus to 48 h for another gryllid Phaeophillacris sp. Positive results were identified in 288 of 1,702 (16.9%) arthropod predators tested. Asiliidae, Gryllidae, Hymenoptera and Odonata were numerous and consistently had high proportions of positive results, indicating that they were the most important natural predators of G.pallidipes at Nguruman, Kenya. However, no numerical response could be demonstrated between these predators and tsetse numbers. Laboratory studies on the interactions of predators and puparia and adult G. pallidipes (prey) showed feeding responses that are best described by Holling 1s Type II and III functional response curves which have density dependent population regulation possibilities. Predation was assessed by burying known densities of puparia in larvipos ition sites and scoring them for predation two weeks later. Although predation levels were quite high (about 35%), there was no evidence that mortality was density dependent over the range of densities used (1 - 36 puparia/m2). Implications of the findings in this study in relation to tsetse control are discussed.