Optimising the establishment, persistence and impact of Trichogramma on the Liverpool Plains in northern New South Wales

Abstract

<p>Egg parasitoids prevent larval hatch of insect eggs, making them ideal natural predators for economically destructive hosts. <i>Trichogramma</i> Westwood (Hymenoptera: Chalcidoidea) is utilised in crops around the world as a biological control agent for this purpose. In Australia, <i>Helicoverpa</i> Hardwick (Lepidoptera: Noctuidae) is a major lepidopteran pest of several spring and summer crops and has rapidly become resistant to broad-spectrum insecticides and <i>Bacillus thuringiensis</i> (Bt) proteins in genetically modified (GM) crops. However, little is known about native egg parasitoids of <i>Helicoverpa</i> on the Liverpool Plains, a highly productive and extensive cropping region on the north-western slopes of New South Wales. This thesis aimed to elucidate the potential of egg parasitoids such as <i>Trichogramma</i> to manage <i>Helicoverpa</i> species and how to optimise their effectiveness so they can be best utilised as a part of an integrated insect pest management program.</p> <p>To establish the presence, impact and habitats of <i>Helicoverpa</i> egg parasitoids on the Liverpool Plains, we conducted a 3-year landscape-scale survey to determine the abundance of egg parasitoids and their seasonal phenology in crop and non-crop vegetation. We assessed egg parasitoid species abundance using water-pan traps (WPT) (white and yellow) and determined their impact by measuring mean parasitism rate of <i>H. armigera</i> egg cards in various habitats. To determine variations in annual egg parasitoid populations, we surveyed spring crops (canola, chickpea, faba bean, linseed, lucerne, sunflower and wheat) and native vegetation (forest, agroforestry and grassland/native pasture). In summer, we used a landscape approach to identify major crop (maize, cotton and sorghum) and forest habitats in two different but connected landscapes. Using the same methods, we surveyed potential overwintering habitats in forest.</p> <p>We identified <i>Trichogrammatoidea bactrae</i>, <i>Trichogramma pretiosum</i>, <i>T. australicum</i>, <i>Telenomus</i> species and unidentified species of <i>Trichogramma</i>tidae.</p> <p><i>Trichogramma</i>tids and <i>Trichogrammatoidea bactrae</i> were the most abundant parasitoids in spring. Whilst <i>T. pretiosum</i> was the dominant egg parasitoid in late summer crops. <i>Trichogramma</i>tidae were more attracted to yellow traps than white whilst <i>Telenomus</i> spp. were only attracted to white traps. Egg parasitoid abundance and species differed significantly when compared between the Piallaway–Breeza and Caroona–Quirindi landscapes, indicating the composition of the landscape had an effect on parasitism impact of <i>Helicoverpa</i>. The abundance of different egg parasitoid taxa caught in water pan traps in the two landscapes was highly significant as was the interaction between individual taxon and landscape. In the forest habitats <i>T. australicum</i> and <i>Trichogrammatoidea bactrae</i> were equally abundant in the two landscapes, whereas <i>T. pretiosum</i>, <i>Telenomus</i> spp, and <i>Trichogramma</i>tids were more abundant in crops and non-crop vegetation in the Caroona–Quirindi landscape. The Caroona–Quirindi landscape had a greater proportion of native vegetation as well smaller and more diverse cropping systems, which conceivably leads less area affected by insecticides that greatly affect the potential of predators and parasitoids.</p> <p>Egg parasitoids were abundant and diverse in crop and non-crop habitats on the Liverpool Plains. However, low abundance in spring crops could limit <i>Helicoverpa</i> control in these crops, and the late abundance of <i>T. pretiosum</i> in summer crops could limit its effectiveness as a reliable natural biological control agent. To address this, we pursued two different experimental approaches: one at a plant level and the other at a broader landscape level.</p> <p>Highly variable rates of parasitised lepidoptera eggs have been reported between different cropping habitats, between different plants in the same crop and between different parts of the same plant. To investigate the effects of crop plant hosts, their combinations, and plant parts, we measured levels of parasitism by <i>T. pretiosum</i> of <i>H. armigera</i> eggs on a range of crop hosts that are regularly grown in northern New South Wales. Individual and mixed-species groups of three crop plants were presented in small-cage trials to determine how plant mixes and egg position on the plant affected parasitism rates. We separated winter/spring crops into two groups (1: linseed, canola and wheat, and 2: lucerne, adzuki and faba bean) and investigated a third group of summer crops (cotton, maize and sorghum). The mean percentage parasitism across the range of crop plants varied from 7% (adzuki) to 42% (canola and faba bean). The group 1 crop plants had similar mean percentage parasitism levels: linseed (40%), canola (42%) and wheat (37%). Egg parasitism on group 2 crop plants differed between adzuki (7%), lucerne (26%) and faba bean (42%). Parasitism of summer crops also varied, with cotton (34%) preferred to sorghum (19%) or maize (9%). Crop configuration (i.e. different combinations of the three crops in each group) had no effect on parasitism rate, but plant part had a significant effect, with more egg parasitism occurring on leaves and stems than flowers in each group. Our results indicated that preferred crops are canola, faba bean, wheat and linseed.</p> <p>To investigate <i>Trichogramma</i> utilising a landscape level approach, we identified the main cropping habitats on the Liverpool Plains in spring to act as a nursery crop to build early populations of these egg parasitoids. We then monitored <i>T. pretiosum</i> populations in summer crops adjacent to and 2 km from the release sites. In this approach, rather than employing inundative releases to control <i>Helicoverpa</i> (the main method used worldwide to combat pest outbreaks), we used inoculative releases to build up natural <i>T. pretiosum</i> populations to act as a preventative measure against seasonal variations of <i>Helicoverpa</i> attacks. Of the four winter crops preferred by T. pretiosum to parasitise <i>H. armigera</i> eggs in the cage experiment, wheat is the predominant crop grown on the Liverpool Plains followed by canola. We tested rates of early-season inoculative <i>T. pretiosum</i> releases at 30,000 wasps/ha and 60,000 wasps/ha in 5-ha plots every week for 3 weeks in wheat and canola across 12 farms in the Piallaway–Breeza and Caroona–Quirindi landscapes. We measured movement away from the release area in 50-m and 100-m intervals, and measured impact using <i>H. armigera</i> sentinel egg cards. We found high levels of egg card parasitism in these crops within the release area. Between releases, average parasitism of <i>H. armigera</i> egg cards in high and low-release sites combined was higher in canola (57%) than wheat (35%), averaging 51% overall. Parasitism of egg cards peaked in the low-rate treatments at 88% in canola and 78% in wheat. However, we found that <i>T. pretiosum</i> did not move much further than 50 m from the release area. Distance from release area significantly affected egg-card parasitism, with average parasitism of 16% of egg cards at 50 m declining to 2% at 100 m.</p> <p>We trapped egg parasitoids in white and yellow water traps and measured <i>Helicoverpa</i> sentinel egg card parasitism in cotton, sorghum, native forest and native grassland habitats in the subsequent summer in January and February in adjacent and more distant habitats. In January and February 2009, average egg card parasitism in four cotton crops within 50 m of the spring release crops was 27% and 33%, respectively, and in five sorghum crops it was 38.5% and 56.5%, respectively.</p> <p>The main findings of this study were that the Liverpool Plains has a resident population of diverse <i>Helicoverpa</i> egg parasitoids in different habitats and landscapes. We found that the introduced but naturalised <i>T. pretiosum</i> parasitised H. armigera in wheat and canola and, in the absence of insecticides, were suitable nursery crops to facilitate localised population enhancement. Low inoculative release rates of <i>Trichogramma</i> may be the most cost-effective method of utilising <i>Trichogramma</i> as part of an IPM strategy. The limited application of this method in Australia highlights a significant future management option to control <i>H. armigera</i> in these crops. After the early season releases of <i>T. pretiosum</i>, increased abundance of this egg parasitoid facilitated natural pest control in nearby summer crops, cotton and sorghum, which are readily attacked by <i>H. armigera</i>. Whilst continued Bt protein (GM) advancements have limited the impact of <i>Helicoverpa</i> spp. in cotton, egg parasitoids should be an integral component of local insecticide and GM resistance management strategies, particularly for Australian grain crops. Furthermore, we have demonstrated that the impact of a very small parasitoid is comparable to other predatory (beneficial) insects, which require larger scales than a single field. </p> <p>Consequently, individual farms and regions should be evaluated independently for effective targeted management of economic pests.</p>