Abstract
The genus Bactrocera (Diptera: Tephritidae) is endemic to the monsoonal rainforests of South-east Asia and the western Pacific where the larvae breed in ripe, fleshy fruits. While most Bactrocera remain rainforest restricted, species such as Bactrocera dorsalis, Bactrocera zonata and Bactrocera tryoni are internationally significant pests of horticulture, being both highly invasive and highly polyphagous. Almost universally in the literature it is assumed that Bactrocera breed continuously if temperature and hosts are not limiting. However, despite that, these flies show distinct seasonality. If discussed, seasonality is generally attributed to the fruiting of a particular breeding host (almost invariably mango or guava), but the question appears not to have been asked why flies do not breed at other times of the year despite other hosts being available. Focusing initially on B. tryoni, for which more literature is available, we demonstrate that the seasonality exhibited by that species is closely correlated with the seasons of its endemic rainforest environment as recognised by traditional Aboriginal owners. Evidence suggests the presence of a seasonal reproductive arrest which helps the fly survive the first two-thirds of the dry season, when ripe fruits are scarce, followed by a rapid increase in breeding at the end of the dry season as humidity and the availability of ripe fruit increases. This seasonal phenology continues to be expressed in human-modified landscapes and, while suppressed, it also partially expresses in long-term cultures. We subsequently demonstrate that B. dorsalis, across both its endemic and invasive ranges, shows a very similar seasonality although reversed in the northern hemisphere. While high variability in the timing of B. dorsalis population peaks is exhibited across sites, a four-month period when flies are rare in traps (Dec-Mar) is highly consistent, as is the fact that nearly all sites only have one, generally very sharp, population peak per year. While literature to support or deny a reproductive arrest in B. dorsalis is not available, available data is clear that continuous breeding does not occur in this species and that there are seasonal differences in reproductive investment. Throughout the paper we reinforce the point that our argument for a complex reproductive physiology in Bactrocera is based on inductive reasoning and requires specific, hypothesis-testing experiments to confirm or deny, but we do believe there is ample evidence to prioritise such research. If it is found that species in the genus undergo a true reproductive diapause then there are very significant implications for within-field management, market access, and biosecurity risk planning which are discussed. Arguably the most important of these is that insects in diapause have greater stress resistance and cold tolerance, which could explain how tropical Bactrocera species have managed to successfully invade cool temperate regions.