Background: Maternally inheritedWolbachiabacteria infect many insect species. They can also be transferred horizontally into uninfected host lineages. AWolbachiaspillover from an infected source population must occur prior to the establishment of heritable infections, but this spillover may be transient. In a previous study of tephritid fruit fly species of tropical Australia we detected a high incidence of identicalWolbachiastrains in several species as well asWolbachiapseudogenes. Here, we have investigated this further by analysing field specimens of 24 species collected along a 3,000 km climate gradient of eastern Australia.Results: Wolbachiasequences were detected in individuals of nine of the 24 (37 %) species. Seven (29 %) species displayed four distinctWolbachiastrains based on characterisation of full multi locus sequencing (MLST) profiles; the strains occurred as single and double infections in a small number of individuals (2–17 %). For the two remaining species all individuals had incomplete MLST profiles andWolbachiapseudogenes that may be indicative of lateral gene transfer into host genomes. The detection ofWolbachiawas restricted to northern Australia, including in five species that only occur in the tropics. Within the more widely distributedBactrocera tryoniandBactrocera neohumeralis,Wolbachiaalso only occurred in the north, and was not linked to any particular mitochondrial haplotypes.Conclusions: The presence ofWolbachiapseudogenes at high prevalence in two species in absence of complete MLST profiles may represent footprints of historic infections that have been lost. The detection of identical low prevalence strains in a small number of individuals of seven species may question their role as reproductive manipulator and their vertical inheritance. Instead, the findings may be indicative of transient infections that result from spillover events from a yet unknown source. These spillover events appear to be restricted to northern Australia, without proliferation in host lineages further south. Our study highlights that tropical fruit fly communities containWolbachiapseudogenes and may be exposed to frequent horizontalWolbachiatransfer. It also emphasises that global estimates ofWolbachiafrequencies may need to consider lateral gene transfer andWolbachiaspillover that may be regionally restricted, transient and not inherited.

Bactrocera curvipennis (Froggatt) is a polyphagous pest fruit fly endemic to New Caledonia that is weakly attracted to the male lure cue-lure (CL). Effective male lures are important for the monitoring and management of numerous pest species of Dacinae fruit flies. However, if a species is weakly responsive to these lures its detection and control is difficult. Recently in Oceania and Asia, more attractive male lures (isoeugenol, methyl-isoeugenol, dihydroeugenol, and zingerone) were identified for several weakly CL- and methyl eugenol (ME)-responsive species. To determine if these lures may be more attractive to B. curvipennis, we field tested them in comparison to CL and ME in New Caledonia. Bactrocera curvipennis catch with isoeugenol-baited traps (mean 20.3 ± 3.0) was 15 times greater than with CL (1.3 ± 0.8) and catch with dihydroeugenol (5.7 ± 1.6) was four times greater than with CL. This is the first record of B. curvipennis responding to these lures. It was also the only species that responded to isoeugenol in this study. Bactrocera fulvifacies (Perkins) (Diptera: Tephritidae), a rarely encountered species 'nonresponsive' to male lures, was attracted to zingerone with its trap catch (2,574 flies) approaching that of Bactrocera tryoni (Froggatt) (Diptera: Tephritidae) at CL (2,724 flies). Another nonresponsive species, Dacus aneuvittatus (Drew) (Diptera: Tephritidae), was also trapped by zingerone-baited traps. This is the first record of these species responding to a male lure. The significantly greater response of B. curvipennis to isoeugenol would make it a considerably more effective attractant for use in surveillance and control programs.

The male fruit fly attractants, cue-lure (CL) and raspberry ketone (RK), are important in pest management. These volatile phenylbutanoids occur in daciniphilous Bulbophyllum Thouar (Orchidaceae: Asparagales) orchids, along with zingerone (ZN) and anisyl acetone (AA). While these four compounds attract a similar range of species, their relative attractiveness to multiple species is unknown. We field tested these compounds in two fruit fly speciose locations in north Queensland, Australia (Lockhart and Cairns) for 8 wk. Of 16 species trapped in significant numbers, 14 were trapped with CL and RK, all in significantly greater numbers with CL traps than RK traps (at least in higher population locations). This included the pest species Bactrocera tryoni (Froggatt) (Diptera: Tephritidae) (CL catches ca. 5× > RK), Bactrocera neohumeralis (Hardy) (Diptera: Tephritidae) and Bactrocera bryoniae (Tryon) (Diptera: Tephritidae) (CL catches ca. 3× > RK), and Bactrocera frauenfeldi (Schiner) (Diptera: Tephritidae) (in Cairns—CL catches ca. 1.6× > RK). Seven species were trapped with AA, and all were also caught in CL and RK traps in significantly greater numbers, with the exception of B. frauenfeldi. For this species, catches were not statistically different with CL, RK, and AA in Lockhart, and RK and AA in Cairns. Seven species were trapped with ZN, two at this lure only, and the remainder also with CL or RK but in significantly greater numbers. This is the first quantitative comparison of the relative attractiveness of CL, RK, AA, and ZN against multiple species, and supports the long-held but untested assumption that CL is broadly more attractive lure than RK.

Male fruit fly attractants, cue-lure (CL) and methyl eugenol (ME), are important in the monitoring and control of pest fruit fly species. Species respond to CL or ME but not both, and there are conflicting reports on whether combining CL (or its hydroxy analogue raspberry ketone) and ME decreases their attractiveness to different species. Fruit fly monitoring programs expend significant effort using separate CL and ME traps and avoiding lure cross-contamination, and combining the two lures in one trap would create substantial savings. To determine if combining lures has an inhibitory effect on trap catch, CL and ME wicks placed in the same Steiner trap were field tested in comparison to CL alone and ME alone in Australia and Papua New Guinea (PNG). In Australia, 24 out of 27 species trapped were significantly more attracted to CL or ME alone than the combination ME/CL lure, including the pests Bactrocera bryoniae (Tryon), B. frauenfeldi (Schiner), B. kraussi (Hardy), B. neohumeralis (Hardy), B. tryoni (Froggatt) (CL-responsive), and B. musae (Tryon) (ME-responsive). In PNG, 13 out of 16 species trapped were significantly more attracted to CL or ME alone than the ME/CL combination, including the pests B. bryoniae, B. frauenfeldi, B. neohumeralis, B. trivialis (Drew), Zeugodacus cucurbitae (Coquillett) (CL-responsive) and B. dorsalis (Hendel), B. musae, and B. umbrosa (Fabricius) (ME-responsive). This study shows that combining CL and ME in the one trap in equal parts significantly reduces catches of most species of Dacini fruit flies in Australia and PNG.