Correlations between frogs and pond attributes in central New South Wales, Australia: What makes a good pond?

摘要

We assessed if certain attributes of human created ponds in a forest environment make them more favourable for use by frogs. We recorded 24 habitat variables associated with 45 ponds in a forested area of central New South Wales and compared these with counts of frogs at the ponds. We used Canonical Correspondence Analysis (CCA) to explore which variables influenced community structure and Generalised Linear Modelling techniques to determine the variables influencing the total numbers of frogs and/or species using the ponds. We also used Zero Inflated Poisson modelling to assess which variables influenced the presence and abundance of individual species. We recorded 24 species, but only nine used a third or more of the ponds. CCA indicated that our measured variables had little influence on community structure and species did not cluster into regular associations. Species richness was significantly, but only weakly related to the measured variables, which displayed little explanatory value. Variables that significantly influenced the presence and/or abundance of individual species varied widely, but depth was most often important for the former and bank and emergent vegetation for the latter. Predicting a pond type that benefits a select group of species is not possible, but pond types beneficial to some species were identified. The results suggest that we can construct or modify ponds in forests to benefit selected species, but a range of ponds are required to benefit permanent pond breeding species.