The Biology Of Ceranisus Menes (Walker) (Hym., Eulophidae), A Parasitoid of The Bean Flower Thrips Megalurothrips Sjostedti (Trybom) (Thys., Thripidae):A Comparison Between African and Asian Populations

Abstract

Cowpea is an important food crop in Africa, but suffers from a variety of insect pests. Among those is Megalurothrips sjostedti (Trybom), a major pest which is not successfully controlled by its local natural enemies. One of them, Ceranisus menes (Walker), a cosmopolitan endoparasitoid of thrips larvae, provides an efficient control of thrips populations in Southeast Asia. This has necessitated a study of this insect which involves a comparison of the biology of the African and Asian populations of C. menes. To achieve this, three (3) objectives were set: 1) to identify potential candidate for the adoption of the Novel Association Approach of biological control against M. sjostedti, using the Southeast Asian populations of this parasitoid; 2) to diagnose the problem of the local population in failing to provide good control of M. sjostedti, 3) and to obtain more information on the biological diversity of the different geographical strains of C. menes. Three (3) strains selected from Southeast Asia, i.e. one from Malaysia and two from India, and the local strain from Benin were studied. Quantitative evaluation of their biological characteristics, when M. sjostedti was offered as host, were obtained and compared. Based on these data, the compatibility of the different strains of C. menes with M. sjostedti for successful parasitism were assessed. On the question of the problem of the local strain, this study used as control the most compatible hostparasitoid pairings: M. usitatus-Malaysian, or Frankliniella schultzei-lndian strains, to evaluate the compatibility of the pairing M. sjostedti-local strain. A series of 3 laboratory experiments simulating the natural chronological sequence of events during the parasitization process were conducted. The first experiment studied the behaviour of host thrips larvae and C. menes parasitoid at the host selection phase. For this, a model of their behaviour was developed. Parameters were defined, and determined for host resistance, host acceptance, and host handling time, from the various reactions of the host larva, and the wasp. Host resistance refered to the defensive behaviour of the thrips, and host acceptance, to that of the parasitoid accepting the thrips larva for oviposition. Each of five (5) female wasps of the Malaysian strain, and ten (10) females of the other strains of the parasitoid, was observed, interacting with the thrips larvae during a total period of 300 minutes per strain. For each species of thrips, three different age-groups of larvae, 0-24 hours, 2448 hours, and 48-72 hours, were used as host, in order to determine any effect of host age on the behaviour of the parasitoid. Each female parasitoid was offered a unique category of host at a time. In the second experiment, host suitability for the local and the two Indian strains was studied. It was evaluated from the immature survival rate of the parasitoid on the host. In the last experiment, the developmental time and the demographic parameters of the parasitoid were studied, to compare the potential of each strain to maintain and build up its population on the host. All data were analyzed at the 5% level of significance. Host handling times were compared between treatments using Kruskal-Wallis test. Host resistance, and acceptance parameters, the immature survival rate, and the population sex ratio were compared using the logit loglinear model analysis. The jacknife method was used to analyze and compare the intrinsic rate of natural increase (rra) of the different populations of C. menes. During the first series of tests, it was found that the one and two days old larvae were slightly faster to handle, and also significantly less resistant, and more likely to be accepted by all strains of C. menes. Thus larvae from these two age groups were used to maximize parasitization success in the subsequent second and third tests. The results of this investigation indicated that there were in feet biological differences between the four strains of C. menes, on M sjostedti host. There were significant differences between strains in the sub-components of the host handling time, but no difference in the overall host handling time. The average host handling time for the Malaysian, the two Indian strains, and the local strains were respectively 27.4, 29.0, 25.8, and 31.6 seconds. It was also found that the four strains of C. menes were slightly different in their ability to overcome the resistance of M. sjostedti larvae. The results showed that 40.6 and 46.4 percent of larvae encountered by the Malaysian and local strains respectively, were resistant to the parasitoid. With the two Indian strains they were 53.2, and 53.5 percent. The analysis indicated no difference between the Malaysian and local strains; or between the two Indian strains. The first two strains were however slightly better to overcome M. sjostedti host resistance, than the two Indian strains. The acceptance of M. sjostedti was definitely more successful with the local than with the exotic strains. More than twenty five percent (25.5%) of the non-resistant larva were accepted by the local strains for oviposition. Acceptance of this thrips species decreased significantly in the order: the two Indian strains (4.7 and 3.3 percent), Malaysian strain (0.4 %). Thus, at the host selection phase of the parasitization process, the behaviour of the local strain was found to be the most compatible for a successful parasitization of M. sjostedti. However, in the physiological test, the suitability of M. sjostedti was found to be very poor for the local strain, as well as for the second Indian strain (Indian2). Their survival rates (proportion of eggs that survive up to adult stage) in this species of thrips was estimated at 0.06 for both strains. Unexpectedly, it was significantly higher (0.36) for the first Indian strain (Indian1).The possibility that M .sjostedti could have co-evolved with the Asian population of C. menes in the past was discussed. The results of the life history notes in the last test confirmed those of the first and second tests. The Indian2 strain was not able to reproduce in M. sjostedti, in contrast to the local and Indian1 strain. The sex ratio of the offspring population was 0 for the local, as compared to 0.67 for the Indian1 strain. On the other hand, the total fecundity of the Indian1Bstrain was very low, as compared to that of the local strains: Ten (10) female wasps of Indian1 strain produced an average of two (2) offspring, whereas the same number of females of the local strain produced an average of up to 867 offspring. As a consequence, the population growth parameters Ro and Rm were null for all the 3 strains of C. menes. The mean developmental time of the Indian1 and local strains were respectively 23.0 and 21.0 days for males. In fact, each of the 4 populations of C. menes was unique in the biological features it exhibited when M. sjostedti was offered as host. This can be seen as the onset of bio types differentiation among the 4 populations of this species of parasitoid. The difference on the two Indian strains, both originating from the same geographical area are discussed. When comparing the local strain on M. sjostedti against the Asian strains on their respective check hosts F. schultzei and M usitatus, the results of the first test showed that the host handling time by the local strain (31.6 seconds) was not significantly different from those of the two Indian strains (24.1, and 30.0 seconds). Host resistance against the local strain (40.6%) was also not significantly different from that against the Malaysian (47.2%). Host acceptance by the local strain (25.5%) was also not significantly different from that of the Malaysian strain (27.9%). However there were significant differences in the suitability of their respective check hosts, suggesting that physiological incompatibility must have been the only reason why the local strain could not perform as well on M. sjostedti as the Asian strains were doing on their check hosts. The survival rate of the local strain (0.06) was significantly lower than that of the first and second Indian strains, which were 0.20 and 0.46 respectively. The mean developmental time for males of the local strain (2 1 .0 days) was higher than that of the Indian2 strain (18.7 days), but it was lower than that of the Indian1 strain (22.5 days). The possible explanations for the lack of suitability of M. sjostedti to the local strain of C. menes are also discussed. In conclusion, this study has brought some insight into the biological diversity among the different geographical populations of C. menes. It has demonstrated that none of the 4 selected strains of C. menes from Southeast Asia and Africa would be able to parasitize successfully M. sjostedti. Each of them was hampered by behavioural and/or physiological incompatibility with this thrips species. The possibility of manipulating the Asian population to obtain better success with M sjostedti is discussed. The validity of the model and the critiques of the methodology used for the behavior study are also discussed.