Life history and impact of non-native brown tree frog, Litoria ewingii, on New Zealand pond ecosystems

Abstract

Freshwater ecosystems worldwide are under increasing pressure from multiple threats, including invasive species and climate change. Some ponds naturally dry, but these threatened wetland ecosystems are vulnerable to increases in the magnitude of drying events associated with climate warming. To invade these dynamic habitats, especially those drying unpredictably, invaders like the Australian brown tree frog (Litoria ewingii) would be advantaged by both flexibility and tolerance. L. ewingii was introduced to New Zealand in 1922 and has since spread across the country, breeding in both urban and undisturbed ponds. Native pond ecosystems may be especially vulnerable to L. ewingii tadpoles because native frogs do not produce aquatic tadpoles, so pond organisms lack co-evolution with tadpoles. L. ewingii has several traits that may enhance invasive success, but this species has received little attention in New Zealand. I repeatedly surveyed ponds spanning a pond- permanence gradient in the Canterbury high country to determine whether drying limited their spread, and to evaluate factors that may regulate their densities or development. High tadpole densities (50/m2) occurred throughout the year, with a spring peak, in fishless ponds varying widely in conditions, suggesting wide tolerance (except to fish). Tadpoles also increased their rate of metamorphosis to more quickly escape pond drying, with a cost to size at metamorphosis. All of these traits mean L. ewingii could have strong ecological impacts. To determine the functional role of L. ewingii tadpoles and to uncover potential influences on pond community structure and function, I conducted two mesocosm experiments. The first crossed tadpole presence with predatory invertebrate presence. Tadpoles grazed down algae and macrophytes, and reduced the number of two invertebrate taxa (Culicidae and Chironomidae). However, predatory invertebrates did not affect tadpoles, so L. ewingii probably lack any top-down control in fishless ponds. I evaluated effects of tadpole density on ecosystem processes, in a second mesocosm experiment that also manipulated abiotic conditions. Half the mesocosms were shaded to simulate permanent ponds, since shade stabilised temperature fluctuations typical of temporary ponds. Phytoplankton increased with higher tadpole densities in unshaded mesocosms, possibly because nutrient release from tadpole excretion enhanced algal growth when sunlight was plentiful. Higher densities of tadpoles also decreased dissolved oxygen – potentially an indirect effect of grazing reducing photosynthesis. Overall, tadpoles are likely to have large influences on ecosystem processes like nutrient cycling in small lentic habitats, especially at the high densities they appear in. In light of their many invasive traits and potential effects on pond ecosystems, the spread of these frogs in New Zealand should be more carefully considered.