Thermal physiology and population dynamics of bollworm (Helicoverpa armigera) in South African fruit orchards

Predicting how climate affects pest insects such as the Bollworm (Helicoverpa armigera) requires an understanding of the environments to which they are exposed, and so we plan to develop a mechanistic model to translate broad-scale meteorological datasets into their microclimatic conditions. We will then run model simulations, at different spatial resolutions, across deciduous fruit growing regions within South Africa. These results will provide an indication of the spatial scale at which future studies focusing on Bollworm, and on insect pests in general, should be conducted. These models will also provide a first indication of likely climate change impacts of Bollworm in South African agricultural landscapes.

Over the past three years we have made considerable progress in understanding how climate and topography affect the performance and survival of the African Bollworm (Helicoverpa armigera). We have developed a biophysical model that translates spatially explicit data of climate and terrain into the microclimates encountered by the different life-history stages of the Bollworm on its host apple tree. Once conditions in the Bollworm’s immediate environment are determined, the model calculates the animal’s core-body temperature, and then draws on established datasets to ascertain its capacity to develop and survive.

Model simulations across the Western Cape province (centering on Villiersdorp) and found that, as well as climate, local topographic variation has significant impacts on the Bollworm’s population dynamics (voltinism), phenology and exposure to thermal stress. More specifically, individuals on northern facing slopes were predicted to have significantly warmer body temperatures, reduced thermal stress, and a greater number of completed generations. As such, the fitness of Bollworm may vary considerably between adjacent orchards that differ only in topography. These findings have been published in an international journal, Austral Entomology, and have received considerable interest from researchers in the field.

In a second study, we ran model simulations across local (Western Cape) and global extents at three spatial resolutions (0.5km, 1km and 1.5km grid cells) to ascertain how spatial scale affects our predictions of Bollworm’s core-body temperature and fitness. We found significant interactions between elevation of the site and model resolution such that the coarse scale under-, and then over-predicted core-body temperature (in comparison to the fine-scale model) as elevation increased.

These systematic discrepancies in our body-temperature predictions, and associated fitness traits, were detected across local and global extents. Our findings help explain why many previous studies detect either larger or smaller distributions when comparing model outputs that have been run at multiple resolutions.