Novel strategies to exploit existing natural infections: synergisms between baculoviruses and other toxins. Research Project Final Report

EXECUTIVE SUMMARY Baculoviruses are a well studied group of insect viruses which have known advantages as biological control agents of insect pests in agriculture and forestry; strains used are very specific to the insect species they can infect, they can be formulated and sprayed conventionally and they are easily broken down by UV thereby leaving minimal non-toxic residues. They are the major group of viruses which infect insects but they do not replicate in vertebrates, plants or other micro-organisms. However, the very fact that they have a restricted host range and may take several days to kill an insect has meant baculoviruses have not been rapidly taken up or developed as alternatives to synthetic pesticides. The aim of this project was to investigate whether any other pathogens/toxins could synergise with baculoviruses and improve baculovirus efficacy against common Brassica pests such as the cabbage moth (Mamestra brassicae). This is a viable assumption as quiescent, symptomless baculovirus infections are common in caterpillars in the field and these have been shown to become fully lethal infections when triggered by another baculovirus. In addition, many pathogens and toxins (e.g. antifeedants) act by restricting growth or retarding development. This could act to extend the “window of vulnerability” of many pest caterpillar species to baculoviruses by delaying the onset of developmental resistance. This work contributes directly to Defra’s aim of promoting sustainable methods of pest control by reducing inputs of chemical pesticides and the exploitation of biological alternatives. Prospective pathogens/toxins were identified as potential triggers of persistent baculovirus infections. Representatives from each group of pathogen/toxin were tested in bioassays against caterpillar pests from Brassica crops (namely the cabbage moth (Mamestra brassicae), the silver Y moth (Autographa gamma), the large white (Pieris brassicae) and the cabbage looper (Trichoplusia ni)). The ability of pathogens/toxins to trigger persistent baculovirus infections in caterpillars and the nature of the interaction of potential triggers with an applied baculovirus PaflNPV (Panolis flammea nucleopolyhedrosis virus) were tested in laboratory bioassays in small dishes and on whole plants in microcosms (laboratory insect rearing cages). Although it is known that the addition of a second baculovirus is a consistent trigger of persistent infections in M. brassicae our results suggest that other chemicals/pathogens do not exhibit the same effect with any degree of consistency. However, the chemical Spinosad and insect bacterium Bacillus thuringiensis (Bt) exhibited synergism with applied PaflNPV as they consistently retarded host growth which in turn increased the likelihood that an insect would die of a PaflNPV infection. It was hypothesised that this was due to the well recorded antifeedant effect of both Spinosad and Bt delaying the onset of developmental resistance and allowing more insects to become infected with baculovirus. When insects were exposed to Spinosad and Bt alongside PaflNPV in microcosm experiments, the relationship between antifeedant and baculovirus became less clear. Difference between the results on whole plants and in contained dishes is suspected to be due to the ability of the caterpillars to avoid contaminated food or perform compensatory feeding. Results investigating host immune responses indicated that PO activity was not significantly increased in the presence of DETC or Bt, despite the fact that DETC has been shown to suppress the host immune system. The presence of persistent baculovirus infections in several moth species was confirmed using molecular methods developed during the project. Differences in occurrence of persistent infections were shown, both geographically and within species of Lepidoptera; a fact that has previously only been shown for M. brassicae and its persistent baculovirus infection. The screening of large numbers of Lepidoptera is essential to answer questions regarding the evolution and maintenance of these infections in nature and the possibility that triggering of persistent infections may occur in non-target caterpillars that are exposed to applied baculovirus could be an important factor in testing of the side effects of baculovirus biocontrol agents. It was not within the scope of this project to answer these questions but the development of protocols within the current project means we are now in a position to advance our knowledge in this area. The transmission dynamics in M. brassicae of an applied baculovirus PaflNPV, the synergists Spinosad & Bt and the persistent infection MbNPV were investigated in a series of small plot caged field trials on spinach plants. The transmission of a baculovirus can be modelled using the mass action model which is based on the assumption that infection is acquired at a constant rate, so that eventually all individuals that are exposed become infected with virus i.e there is a linear relationship between infection and viral mortality. However, baculoviruses are known to have a non-linear transmission pattern in the field because some individuals are able to escape infection by being in a pathogen-free refuge. This refuge may be due factors such as a patchy distribution of virus in the field whereby individuals never encounter virus or a proportion of individuals are less susceptible to viral infection. We tested the hypothesis that the addition of Bt or Spinosad would reduce the pathogen-free refuge and result in a greater probability of viral infection within field caterpillar populations. There was no difference between applying the virus concurrently with Bt or with a 2 or even 4 day delay between applications on the likelihood that insects would die of a baculovirus infection but overall, there was increased host mortality due to virus in the presence of Bt suggesting that the toxin enhanced viral mortality. Combined mortality from the two pathogens was double that when only virus was applied and this increase in mortality was almost entirely due to increased viral deaths. The mechanism sustaining this synergism is not clear but further field studies showed there were significant reductions in host mass gain in the presence of Bt which in turn may delay the onset of virus developmental resistance, thereby keeping individual larvae in a “susceptible” pool. The slight synergism shown between Spinosad and PaflNPV in laboratory tests did not extend to field conditions. The likelihood that an insect would succumb to a lethal virus infection was not significantly different in the presence or absence of Spinosad. The ability of a baculovirus to self-perpetuate and cause lethal infections in subsequent generations of larvae is one of the long-term benefits of using baculoviruses for biocontrol. We hypothesised that reduction in host mass gain in the presence of Bt may negatively impact on the amount of virus produced per cadaver (i.e. viral yield) but there were likely to be more patches of virus over the plants (i.e. proportionally more insects die of virus) and each patch would be smaller than when Bt was absent (i.e. viral yield per cadaver was reduced in the presence of Bt). We thought this may be beneficial to viral transmission as an increase in patches of virus may increase the likelihood that an insect would encounter a viral patch. Within the UK, M. brassicae generally have two generations within an average cropping season. To test the hypothesis that the combined application of Bt and baculovirus would achieve season long control of M. brassicae, we produced predictive models and simulated two generations of larvae in field trials on cabbage plants over a period of 5 weeks. The field experiment evaluated the transmission of virus from applied sprays to insects in the first generation and the subsequent transmission of virus from viral first generation insect cadavers to second generation larvae. First generation larvae were exposed to Bt, PaflNPV or a mixture of the two pathogens. When these larvae died or pupated, a second generation was simulated by the introduction of eggs into the crop. The transmission dynamics of the virus in the presence of Bt in the first generation of larvae were the same as the previous trial; the proportion of insects dieing due to a viral infection was increased in the presence of Bt. The majority of insects died due to an MbNPV infection suggesting this was due to triggering of the persistent infection by the applied baculovirus PaflNPV. As the two trials were done using the same methods but in different years and on different host plants (spinach and cabbage) it suggests that the synergism between Bt and the baculovirus may not be affected by abiotic factors or host plant. However, more evidence would be needed to corroborate this implication. However, reductions in host mass gain in the presence of Bt in the first generation of larvae incurred a penalty to the virus in terms of reduced viral yield per cadaver and transmission to the second generation. Mortality due to virus in the second generation was lower in plots where first generation larvae had been exposed to Bt. This suggested that the effect of Bt on yield in the first generation was detrimental to the ability of the virus to cycle to the second generation. It is probable that insects die in areas where they have been feeding (as was observed in several plots during experiments) and this increases the likelihood that other feeding larvae will come into contact with virus. Many, small patches would seem to be more beneficial in this case. However, the virus is subject to abiotic factors such as UV and precipitation which can inactivate and remove infectious material from the crop. This is likely to have a more significant effect on the potential infectivity of a patch of virus when it is a small patch compared to larger patches which can afford to lose some infective material. Interestingly, the fact that only two larvae succumbed to Bt infection in the second generation suggests that the efficiency of viral transmission was due to baculovirus alone and that the synergist was only acting in the first generation on host mortality. Similarly, the fact that the presence of Bt in the first generation had no impact on the weight of larvae in the second generation suggests that even sublethal effects were not occurring. As far as we are aware, this is the first time that the secondary transmission of a baculovirus in the presence of Bt has been demonstrated. Conclusions This study has shown that there is potential to achieve improved control of caterpillar pests in Brassica crops using combinations of insect control agents approved for used in organic crop production and baculoviruses. Our research considered a series of intricate interactions in both the laboratory and field between an applied baculovirus, a persistent baculovirus and potential synergist pathogens/toxins. Persistent baculovirus infections were not reliably triggered by the range of pathogens/toxins tested which has important implications as non-target Lepidoptera tested were shown to carry persistent baculovirus infections and may be exposed to low levels of toxins/pathogens applied for biocontrol. Field trials against the pest M. brassicae showed that suppression of first generation larvae could be improved by co-application of a baculovirus (PaflNPV) and the bacterium Bt. Experimental evidence suggests that this is due to a delay in the onset of larval developmental resistance, induced by the presence of Bt. However, the presence of Bt reduced viral yield from first generation larvae and this was detrimental to secondary transmission of the virus, implying that longer term control of pests by the self-perpetuating virus would be reduced. These experiments have been undertaken with a specific life stage of M. brassicae and work would be needed to investigate the interactions at different instars and in more than one generation of the pest in order to make any recommendations regarding control of M. brassicae with a combination of virus and Bt. Overall, this research provides evidence that there is potential to reduce synthetic pesticide use or improved control achieved by organically approved products through the development of combinations of biopesticides as alternatives to, or in conjunction with, currently registered agronomic products for control of agricultural and horticultural pests