摘要：Locust population outbreaks have been a longstanding problem for Australian agriculture. Since its inception in the mid-1970s, The Australian Plague Locust Commission (APLC) is responsible for monitoring, forecasting and controlling populations of several locust pest species across inland eastern Australia (ca. two million km2). Ground surveys are typically targeted according to prevailing environmental conditions. However, due to the sheer size of the region and limited resources, such surveys remain sparse. Here we develop daily time-step statistical models of populations of Chortoicetes terminifera (Australian plague locust) that can used to predict abundances when observations are lacking, plus uncertainties. We firstly identified key environmental covariates of locust abundance, then examined their relationship with C. terminifera populations by interpreting the responses of Generalized Additive Models (GAM). We also illustrate how estimates of C. terminifera abundance plus uncertainties can be visualized across the region. Our results support earlier studies, specifically, populations peak in grasslands with high productivity, and decline rapidly under very hot and dry conditions. We also identified new relationships, specifically, a strong positive effect of vapour pressure and sunlight, and a negative effect of soil sand content on C. terminifera abundance. Our modelling tool may assist future APLC management and surveillance effort.
其他摘要：Abstract Locust population outbreaks have been a longstanding problem for Australian agriculture. Since its inception in the mid-1970s, The Australian Plague Locust Commission (APLC) is responsible for monitoring, forecasting and controlling populations of several locust pest species across inland eastern Australia (ca. two million km 2 ). Ground surveys are typically targeted according to prevailing environmental conditions. However, due to the sheer size of the region and limited resources, such surveys remain sparse. Here we develop daily time-step statistical models of populations of Chortoicetes terminifera (Australian plague locust) that can used to predict abundances when observations are lacking, plus uncertainties. We firstly identified key environmental covariates of locust abundance, then examined their relationship with C. terminifera populations by interpreting the responses of Generalized Additive Models (GAM). We also illustrate how estimates of C. terminifera abundance plus uncertainties can be visualized across the region. Our results support earlier studies, specifically, populations peak in grasslands with high productivity, and decline rapidly under very hot and dry conditions. We also identified new relationships, specifically, a strong positive effect of vapour pressure and sunlight, and a negative effect of soil sand content on C. terminifera abundance. Our modelling tool may assist future APLC management and surveillance effort.
