A spatio-temporal modelling approach to understand the effect of urban fruit fly outbreaks on peri-urban orchards

Abstract

Urban areas are well-known sources of fruit fly introduction and unmanaged populations pose a high risk of dispersal to surrounding agricultural areas. Although management of urban fruit fly populations is recognised as being important for area-wide control programmes, its relative importance is rarely quantified. Identifying possible sources of fruit fly populations is important to the success of area-wide programs, as well as to on-farm management. We developed a modelling approach to analyse the spatio-temporal connectivity of fruit fly across urban-agriculture boundaries. For a case study, we examine Mediterranean fruit fly, Ceratitis capitata, outbreaks in two fruit growing regions in south-west Western Australia. Connectivity between population sources was defined using a dispersal kernel and routinely collected surveillance data from traps located across urban and agricultural areas. Spatio-temporal Bayesian models were then developed to quantify the extent to which unmanaged urban populations contribute to commercial orchards across each region. Our model results from two fruit growing regions indicate that population connectivity is a strong predictor of urban and commercial orchard Mediterranean fruit fly density. Models demonstrate that unmanaged urban outbreaks can contribute to increased pest pressure in surrounding orchards up to 2 km, but the urban contribution quickly declines over this distance. We also demonstrated that our models provide a means of predicting changes in Mediterranean fruit fly density across a region. Regional scale inference from our models may be particularly useful for sites where surveillance trapping data is unavailable. With an increasing risk of fruit fly incursions from expanding urban areas into agricultural regions, it is important to understand the processes driving population connectivity to target pest management strategies. By modelling surveillance trapping data and population connectivity, we can quantify how unmanaged urban pest infestations can contribute to pest pressure in nearby commercial orchards. Such information can help pest managers quantify the risk posed by unmanaged urban infestations when designing and implementing area-wide management.