The colonisation of canola crops by the diamondback moth, Plutella xylostella L., in southern Australia

Abstract

Background: Mobile insect pests cause substantial losses to the yields and economic value of annual crops throughout the world. A sound knowledge of the movement patterns and origins of seasonal pest populations is essential for tactical management, yet these are difficult to identify and often poorly characterised. Further complexities for management arise when pest species carry insecticide-resistant genotypes or when cryptic diversity creates uncertainty around species identity. The diamondback moth, Plutella xylostella L., the most destructive pest of Brassica crops worldwide, seasonally colonises winter canola crops in southern Australia leading to intermittent outbreaks. Little is known about regional movement and crop colonisation patterns of P. xylostella in the Australian canola agro-ecosystem, or the contribution of movement to seasonal outbreaks and the spread of insecticide resistance. The recent discovery of an endemic cryptic species, Plutella australiana Landry & Hebert, raised questions about the relative pest status of the two Plutella species in Australian Brassica crops. Sympatric populations of Australian P. xylostella and P. australiana were compared using multiple complementary approaches that included genomic datasets, Wolbachia infections, inter-species crosses and insecticide bioassays. The seasonal colonisation of Australian canola by P. xylostella was investigated using molecular markers and field studies over three years to infer the most likely source populations and the management implications. Results: Although some laboratory hybridisation occurred, deep divergence between the genomes and biology of the two Plutella species was clear, implying contrasting colonisation histories and pest status. Population genetic structure in Australian P. xylostella was assessed using RAD-seq, first using nine field and laboratory populations to refine methods, then in 59 field populations collected throughout southern Australia in two separate years from wild and cultivated brassicas. A statistically powerful genome-wide SNP marker set revealed no spatial, temporal or host-related genetic structure among Australian P. xylostella populations despite a geographic sampling scale > 3000 kilometres, reflecting a recent range expansion and/or high gene flow and confirming a previous microsatellite assessment. In a three year study, wild brassicaceous potential host plants were sampled for P. xylostella in autumn, then subsequent crop colonisation was measured across a regional network of sentinel canola crops and using a derived temperature-dependent development model to back-predict initial oviposition in each crop. Each year, wild brassicas harboured P. xylostella in canola-growing areas prior to sowing, then most canola crops were colonised soon after germination. The autumn abundance of the insect and its hosts, and the subsequent colonisation timing in canola, varied widely between years driven by pre-season rainfall. A CLIMEX model explored spatio-temporal fluctuations in the potential distribution and abundance of P. xylostella based on climate suitability. CLIMEX predictions and light trapping at four locations confirmed that P. xylostella most likely persists in canola production areas throughout the year. Conclusions: Plutella xylostella is the predominant pest in Australian Brassica crops while P. australiana is of minor importance. Low genetic diversity in Australian P. xylostella precludes the use of neutral markers to identify dispersal patterns. In South Australia, insecticide-resistant P. xylostella populations persisting in the local area, rather than arriving from distant source regions, appear to be the major source of seasonal colonisation of winter canola, suggesting local factors may determine outbreak risk.