Search

AIM: To assess the genetic structure, biogeography, and the potential for speciation, of a highly host‐specific insect pest with mainland and island populations. LOCATION: East coast of Australia, Lord Howe Island (LHI), and New Zealand. METHODS: We focussed on Mycopsylla fici, a plant sap‐feeding insect host‐specific to the fig tree Ficus macrophylla. We genotyped 152 insects from across the natural and extended host plant range at 14 microsatellite loci and analysed the data using standard population genetics statistics, Discriminant Analysis of Principal Components, genetic autocorrelation, and two Bayesian clustering approaches. RESULTS: Genetic analyses revealed that the northeastern Australian mainland population (Brisbane) is the centre of genetic diversity. Northeastern and southeastern (Sydney) mainland populations are genetically differentiated and interconnected in a stepping‐stone pattern. The LHI population is the most distinct genetically and Bayesian estimates indicated that the most recent colonization occurred c. 2,000‐17,500 years ago from a northeastern mainland origin. In contrast, the New Zealand population is little differentiated from the Sydney population and probably diverged by colonization within the past 200 years. MAIN CONCLUSIONS: The strong differentiation in nuclear microsatellites mirrors previous evidence for divergence of the LHI population from both mtDNA and endosymbiont DNA. The LHI population may be undergoing speciation from the mainland populations, with an oceanic barrier to gene flow. In contrast, the geographically isolated population in Auckland represents a far more recent colonization reflecting the contemporary naturalization of the plant host in New Zealand.

The ecology, diversity, and parasitoid complex of plant–sap feeding insects of the family Homotomidae (Hemiptera: Psylloidea) specialised on fig trees (Ficus) have so far received little research attention. They are ecologically important, however, as occasional outbreaks of the homotomid Mycopsylla fici may cause complete defoliation of its host plant, the Moreton Bay fig (Ficus macrophylla). Mycopsylla proxima, the only other species reported from Australia, feeds on F. rubiginosa without any recorded outbreaks. We searched for homotomids and their parasitoids on eight Ficus species on the east coast of Australia, Lord Howe Island (LHI), and in Auckland, New Zealand, and detected them on three Ficus species. Using mitochondrial and nuclear DNA sequences, we delimited three Mycopsylla species, including a putative new species on F. watkinsiana. We also characterised six (including one previously described) parasitoid species of the genus Psyllaephagus (Hymenoptera: Encyrtidae) based on congruent morphological characters and molecular data. Each of the homotomid species was highly host specific to a single fig species, whereas parasitoid species varied in host specificity: three host specific to M. fici and three host generalists. Geographic distribution varied among parasitoid species; e.g. one host-specific species was found on both the mainland and LHI, but a second species only on LHI. Our study revealed previously unrecognised diversity in fig homotomids and especially in their parasitoids. The herbivores and parasitoids showed contrasting patterns of host specificity. Interestingly, M. fici, the only outbreak species, had the highest diversity of associated parasitoid species and was the only species with host-specific parasitoids.

Most insects are associated with bacterial symbionts. The bacterial diversity and community composition within hosts may play an important role in shaping insect population biology, ecology and evolution. We focussed on the bacterial microbiome of the Australian fig homotomid Mycopsylla fici (Hemiptera: Psylloidea), which can cause defoliation of its only host tree, Ficus macrophylla. This sap-feeding insect is native to mainland Australia and Lord Howe Island (LHI) but also occurs where its host has been planted, notably in New Zealand. By using a high-throughput 16S rDNA amplicon sequencing approach, we compared the bacterial diversity and community composition in individual adult males of four host populations, Sydney, Brisbane, LHI and Auckland. We also compared males, females and nymphs of the Sydney population. The microbiome of M. fici was simple and consisted mostly of the following three maternally inherited endosymbiont species: the primary endosymbiont Carsonella, a secondary (S-) endosymbiont and Wolbachia. However, the relative abundance of their sequence reads varied between host populations, except for similarities between Sydney and Auckland. In addition, insects from Sydney and Auckland had identical bacterial strains supporting the hypothesis that Sydney is the source population for Auckland. In contrast, mainland and LHI populations harboured the same S-endosymbiont, co-diverged Carsonella but different Wolbachia strains. Besides detecting endosymbiont-specific patterns of either co-evolution or horizontal acquisition, our study highlights that relative abundance of maternally inherited endosymbionts should also be taken into account when studying bacterial communities across host populations, as variations in bacterial density may impact host biology and ecology.

While obligate primary (P-) endosymbionts usually cospeciate with their insect hosts, less is known about codiversification of secondary (S-) endosymbionts that are generally considered facultative. Typically, insects of the superfamily Psylloidea harbour one P- ( Carsonella ) and at least one S-endosymbiont, thought to compensate for Carsonella genome reduction. Most codiversification studies have used phylogenies of psyllids and their endosymbionts across and within host families or genera, but few have explored patterns within species. We focussed on P- and S-endosymbionts of three Mycopsylla (Homotomidae) species to explore whether they have congruent phylogenies and within-species geographic structures. The P-endosymbiont Carsonella , a S-endosymbiont and Wolbachia all had 100% prevalence, while Arsenophonus was only found in one species at low prevalence. Congruent phylogenies of Mycopsylla and P-endosymbionts across populations and species support strict cospeciation. S-endosymbiont phylogenies were also congruent across host species but low genetic variation in the S-endosymbiont was not correlated with host phylogeography, possibly due to a shorter evolutionary association. Between species, Wolbachia and Mycopsylla phylogenies were incongruent, probably due to horizontal transmission events. Our study is the first to explore endosymbionts of Mycopsylla and further supports the codivergence of Psylloidea hosts and P-endosymbionts, with obligate host interactions for both P- and S-endosymbionts.

Explanations for the evolution of pathogen-induced fecundity reduction usually rely on a common principle: the trade-off between host longevity and reproduction. Recent advances in nutritional research have, however, challenged this assumption and shown that longevity and reproduction are not inextricably linked. In this study, we showed that beetles infected by cysticercoids of the tapeworm Hymenolepis diminuta increased their total food intake and, more particularly, their carbohydrate consumption compared with uninfected insects. This increased intake was only pronounced during the first 12 days p.i., when the parasite grows and develops into a mature metacestode. Despite consuming more nutrients, infected individuals sustained lower levels of body lipid and were less efficient at converting ingested protein to body protein. However they demonstrated a capacity to compose a diet that sustained high levels of reproductive output unless confined to foods that were nutritionally dilute. We did not find any indication that macronutrient intakes had an effect on host pro-phenoloxidase activity; however, phenoloxidase activity was significantly affected by protein intake. Our results showed that when offered nutritionally complementary diets, infected hosts do not systematically suffer a reduction in fecundity. Thus, in our view, the assumption that a reduction in host reproduction represents an adaptive response by the host or the parasite to divert resources away from reproduction toward other traits should be reassessed.