Abstract:
Grain legumes are among the key economical crops widely grown in western and eastern Africa as important sources of food and animal fodder. However, the production of grain legumes in Kenya is seriously affected by a complex of insect pests particularly thrips. Yield losses of 20 to 100% have often been reported in some areas. The bean flower thrips (BFT), Megalurothrips sjostedti is considered to be the most important thrips pest of grain legumes. Chemical control is still the main management strategy, with detrimental consequences on the environment, users and consumers. Entomopathogenic fungi (EPF) are among the most promising alternatives to chemical pesticides. Inundative sprays are the most common application techniques for EPF.Although efficient and environmentally safe, the performance of entomopathogenic fungi is affected by several environmental parameters such as UV light, temperature, drought and rain. In order to improve the efficacy of EPF, an autodissemination system has been developed for the management of thrips in greenhouses. In this system, thrips are attracted to an autoinoculator where they are infected with an EPF before returning to the environment to disseminate the EPF to conspecifics. It therefore provides promising prospects, but for effective control, the conidial persistence and thrips attraction need to be optimized, while the EPF and the semiochemical should be compatible. The objective of this study was therefore to optimize the autodissemination system for thrips management on grain legumes in Kenya.The semiochemical Lurem-TR, has been found to inhibit conidia of EPF when put together in an autoinoculation device. The effect of spatial separation of Lurem-TR on the persistence of conidia of EPF, Metarhizium brunneum and Metarhizium anisopliae was therefore evaluated to develop an autodissemination strategy for the management of M. sjostedti. Influence of spatial separation of the semiochemical on thrips attraction and conidial acquisition by thrips from the autoinoculation device was also investigated in the field. This study showed that conidia persistence of both fungal species increased with distance of separation from Lurem-TR. Attraction of thrips to the device also varied significantly according to distance between the device and semiochemical. More thrips were attracted when Lurem-TR was placed in a container below the device and at 10 cm distance from the device. Conidial acquisition by thrips was not significantly different between spatial separation treatments of conidia and Lurem-TR. Seven alternative thrips attractants, namely 4-anisaldehyde, ethyl benzoate, cis-jasmone, linalool,methyl anthranilate, trans-caryophyllene and phenylethanol were also screened for their compatibility with M. anisopliae ICIPE 69 in utodissemination devices and for their attraction to M. sjostedti in the field. Methyl anthranilate (MA) was found to be the attractant most compatible with M. anisopliae and its attractiveness to M. sjostedti was similar to that of LuremTR.The performance of the attractant, methyl anthranilate, was compared to the commercial attractant Lurem-TR in autoinoculation devices treated with M. anisopliae under field conditions for two seasons. Densities of M. sjostedti in plots with the two semiochemical-baited autoinoculation devices were less than in the control plots during both experimental seasons.Plots with MA-baited and Lurem-TR-baited devices had similar densities of M. sjostedti during both seasons. However in the second season thrips densities in plots with the Lurem-TR-baited devices did not differ significantly from the control plots. Conidial viability of M. anisopliae was significantly higher in semiochemical-free baited devices (control) than in semiochemical-baited devices in both seasons. Conidial germination decreased over time in all the treatments but remained above 45%, 12-15 days post-exposure. The average number of conidia acquired by a single M. sjostedti ranged between 2.0 and 10.0 x 103 conidia in both semiochemical-baited device treatments during both seasons. Significantly more conidia were acquired by single thrips in MA-baited devices compared to Lurem-TR baited devices during the podding stage of the crop during the second season. Significantly higher mortality of M. sjostedti was caused in field plots by Lurem-TR baited and MA-baited autoinoculation devices compared to mortality of M.sjostedti collected from the control plots in both seasons. Cowpea yield also differed significantly between the treatment plots. The highest yield was recorded in plots where MAbaited devices were placed. From this study, it could therefore be recommended that methyl anthranilate be used in autoinoculation devices for the management of M. sjostedti on grain legumes. The success achieved with MA in these trials resulted in the evaluation of this EPF for possible use in a spot spray strategy.The efficacy of spot spray and cover spray applications of M. anisopliae in combination with the thrips attractant Lurem-TR was compared in field experiments for the management of M.sjostedti on a cowpea crop in two seasons. Plants in the treatment plots where a spot spray application of M. anisopliae was done five days after the placement of Lurem-TR recorded the lowest densities of M. sjostedti. Fungal viability and thrips conidial acquisition did not differ between the two application methods. Compared to the control treatment plots, both application strategies resulted in yield increases of 34.1 and 46.2% with spot and cover spray treatments. The cost benefit analysis suggests that the spot spray application was more profitable due to the reduction in labour and the uantity of inoculum used.