Ecology of microhylid frogs in the Australian Wet Tropics and implications for their vulnerability to chytridiomycosis

Chytridiomycosis is an emerging infectious disease of amphibians caused by the chytrid fungus Batrachochytrium dendrobatidis (Bd), and is responsible for causing mass mortality, population declines and extinctions of amphibian species in much of the world. The extent of pathological effects of chytridiomycosis varies amongst hosts, and terrestrial frogs and salamanders that have parental care of direct-developing eggs may tend to be less susceptible. The frog family Microhylidae is characterised by terrestrially-breeding species that reach their highest diversity in Australia in the Wet Tropics (WT) bioregion of northern Queensland. In this region, frogs of other families suffered severe declines and extinctions in association with outbreaks of chytridiomycosis. There is a lack of information on the responses of Australian microhylids to this emerging disease, and this project was carried out to gain explicit information on the interactions of microhylids with Bd. The temporal calling patterns of two microhylid species (Austrochaperina robusta and Cophixalus ornatus) were examined at one site within the WT, using automated recordings every night for five entire wet seasons, over an 11-year period. Environmental variables were also recorded, to examine correlations between calling and weather conditions. The calling activity of C. ornatus and A. robusta fluctuated only slightly from year to year; the observed fluctuations were less than two-fold. There was no downward trend in average and maximum calls for either species over time, suggesting that population densities of these species have not suffered from major losses, and that variation in calling is unlikely to be due to climate change or an external factor such as Bd. To determine if microhylids could be infected with Bb in the wild, I examined 595 samples from nine species, and found that none showed evidence of infection by Bd. When this data were regarded as a single sample representative of Australian microhylids, the upper 95% binomial confidence limit for presence of infection was 0.0062 (less than 1%). This suggests that microhylids have a very low prevalence of Bd in nature, and are either not susceptible, or are only slightly susceptible, to chytridiomycosis under natural conditions. The susceptibility of the most common species of microhylid (Cophixalus ornatus) to Bd was tested in a series of laboratory experiments. Seven C. ornatus and five Litoria wilcoxii (susceptible controls) were exposed to increasing numbers of Bd zoospores and tested for infection using quantitative PCR assays. All C. ornatus, and four of the five L. wilcoxii, became infected by Bd at some point during the experiment. The mean intensity of infection, as measured by number of zoospore equivalents in infected individuals, was significantly higher in L. wilcoxii than in C. ornatus. All C. ornatus individuals eliminated their infections by the end of the experimental trials, whereas L. wilcoxii individuals still retained relatively intense infections. Innate immune defenses in the form of antimicrobial peptides (AMPs) may be particularly important in providing resistance to chytridiomycosis, and highly effective AMPs could be responsible for the extremely low prevalences of Bd infection I found in nature. To test this hypothesis, the AMPs of 81 microhylids from six species were examined. Secretions containing skin peptides were collected by norepinephrine induction, and used in growth inhibition assays to measure their effectiveness against Bd. Sixty two percent of samples contained AMPs with at least some activity against Bd, and 17% showed 100% inhibition of Bd growth at the levels tested. Microhylid species produced peptides in quantities similar to two sympatric hylid frogs (Litoria genimaculata and L. rheocola). Mean protein secretion of microhylids did not differ significantly from these species, however, the overall protection of microhylids provided by AMPs was significantly lower than that of hylids. This suggests that AMPs are not likely to be responsible for the low prevalence of infection by Bd in Australian microhylids. Thermal and hydric environmental variables are known to affect the prevalence of Bd infections and the occurrence of epidemic outbreaks of chytridiomycosis, as Bd requires water to reproduce, and has a thermal optimum of 17-25°C. Measurements of the body temperatures of frogs were combined with longer-term temperature and moisture data collected from permeable and impermeable agar models. These models were placed in frog retreat sites to document the thermal and hydric environments that microhylids experience in the field. Models never lost more than 20% of weight after periods exceeding 96hrs, suggesting that retreat sites provide insulation from dehydration. Models produced an accurate outline of the thermal envelope of microhylids as 90% of frogs had surface temperatures that were inside the range of the models. Surface temperature readings of 197 frogs were within the growth range for Bd, and model thermal data suggested that at two of the four transects (Carbine uplands and Paluma), temperature levels during the period of data collection were entirely within the optimal range for Bd growth. However, at Bellenden Ker, models were below the optimal range for Bd growth 70 percent of the time, and at the Atherton uplands, model temperatures reached levels that were above the range for optimal growth of Bd 24 percent of the time. Model temperatures also reached levels above 28°C, which may cause the fungus to stop growing. Given that data were collected over a narrow time window, and that Bd has a narrow range of tolerance for temperature and moisture conditions, the data indicate that microenvironments used by microhylids are likely to affect the overall dynamics of the host-pathogen system. Some individuals and species may be less susceptible to Bd infection due to the environments they use. In summary, the results of this study provide new levels of understanding of the interactions between Australian frogs of the family Microhylidae and the amphibian chytrid fungus. Despite inhabiting environments where Bd has caused declines in other frog species, they have not suffered population losses associated with Bd, and Bd infections are either at very low prevalence or are entirely absent in nature. This apparent resistance to infection does not appear to be solely due to AMPs, as samples collected from microhylids in the field were no more effective against Bd in microhylids than are those of hylid species. Microhylids are not constitutively immune to infection by Bd from some other, unknown mechanism intrinsic to themselves, since C. ornatus readily became infected in laboratory experiments. Field data showed that the thermal environments experienced by microhylids may contribute to their near immunity to Bd infection in nature, but are unlikely to explain it entirely. One potentially important source of resistance to Bd infection that was not evaluated in the present study is the contribution of symbiotic skin microbes, and future work should evaluate this, as it is possible that the Australian microhylids have a skin microbiota that is particularly effective at combating Bd infection in the field