The development of biopesticides for integrated management of sucking pests

Abstract

Sucking pests have emerged as major pests in agricultural crops in Australia due to the reduction in chemical usage with the adoption of Integrated Pest Management (IPM), including the wide-scale use of microbial insecticides against Helicoverpa spp. (Lepidoptera: Noctuidae), and Bollgard II® cotton. Few selective insecticides are available for the control of sucking pests in both grains and cotton crops. Application of broad-spectrum insecticides to control sucking pests such as mirids ((Creontiades spp. (Stal) (Hemiptera: Miridae)) and green vegetable bug (Nezara viridula (Linnaeus) (Hemiptera: Pentatomidae)) is not compatible with the IPM strategies developed to control Helicoverpa spp. The development of selective insecticides is critical for the development of IPM strategies and successful management of these pests. We assessed whether applying ‘insurance’ applications of the organophosphate Dimethoate® impacts on natural enemies and on other major pests in mungbeans. The outcome was that a single application of Dimethoate® tripled Helicoverpa spp. in three days while only halving the population of natural enemies. The trial demonstrated that even a moderate reduction in natural enemies is sufficient to increase Helicoverpa spp. populations in mungbeans. No biopesticides for sucking pests are currently registered in Australia; however entomopathogenic fungi have potential to be developed into biopesticides. Isolates from the species Metarhizium anisopliae (Metschnikoff) and Beauveria bassiana (Balsamo) Vuillemin (Hypocreales: Clavicipitaceae) were selected to assess for pathogenicity and virulence against green vegetable bug in the laboratory. Four Australian isolates showed potential, achieving 90% or greater mortality in five days. We investigated the efficacy of commercial biopesticides based on entomopathogenic fungi and compared them with the native isolates of B. bassiana and M. anisopliae we identified in the laboratory, against mirids in the field. The most effective field rate was 1 x 1013/Ha; higher rates (5 x 1013/Ha) of the M. anisopliae isolate had an impact on natural enemies. At the lower rate there was minimal impact on natural enemies. Formulation and application were also assessed. The native isolate Beauveria 36 was most effective applied in oil at very low volume (VLV) using spinning disc equipment. However, no benefit could be found from applying the commercial biopesticide isolate as a VLV in oil or as an ES either by VLV or knapsack. The native isolates reduced mirids in three days and continued to do so in most cases until the final sampling date; however mirid numbers increased in Dimethoate® treated plots between three and six days after treatment. Mirid nymphs become as damaging as adults at 4th instar, implying that earlier instars should be targeted. Results from these trials show that 3rd and even the earlier instars are susceptible to fungal pathogens. The use of ‘insurance’ sprays in mungbeans is based on the assumption that mirids immigrate into mungbeans at flowering, but the high numbers of nymphs present at flowering suggest they may arrive earlier. One advantage of biopesticides is that they do not disrupt natural enemy populations, and as such, they could potentially be used early in the crop cycle without triggering pest outbreaks. However, fungal pathogens take between three and six days to kill compared to three for chemical insecticides. We thus quantified the mirids in crops prior to flowering, and used intensive sampling to develop a model of population structure and development to identify potential windows for control using biopesticides. Intensive sampling was conducted in soybeans, cotton and mungbeans at regular intervals through the crop development from emergence to early reproductive stages. The results showed that mirids immigrate into crops and start breeding from vegetative stages 2 or 3 in mungbeans and can reach threshold prior to flowering. Researchers have established development times for mirids as well as the threshold temperature at which development ceases. These regression equations were used to develop two models, the first to depict the developmental time of mirids at constant temperatures, and the second uses the population data at a sampling time to estimate the population structure for sampling time n + 1. The modelling suggests that fungal entomopathogens can be used to target 2nd and 3rd instar mirids prior to flowering. In summary, the work in this thesis shows that fungal entomopathogens have potential to be used for microbial control of mirids with little impact on natural enemies. We showed that B. bassiana and M. anisopliae are effective in both mungbeans and cotton. We found that mirids immigrate into, and breed in mungbeans from early vegetative stages. The model we developed from intensive sampling suggests that fungal pathogens could be applied to 2nd or 3rd instar mirids prior to flowering in mungbeans, however more work is needed to confirm if 1 or 2 applications of a fungal entomopathogen can keep mirids under threshold, without triggering an outbreak of Helicoverpa spp. or with reduced yield compared to conventional control methods.