Abstract

We evaluated and compared sixteen combinations of commonly used storage and extraction methods for faecal DNA from two Australian marsupial herbivores, two marsupial carnivores and an introduced carnivorous mammal. For all species the highest amplification and lowest genotyping error rates were achieved using dried faeces extracted via a surface wash followed by spin column purification. The highest error rates were seen in the two Dasyurus spp. and the lowest in Vulpes vulpes. The rates observed for each species were incorporated into computer simulations to identify the number of PCR replicates required to achieve accurate genotyping of DNA isolated via the optimised protocol. Three replicates per sample were sufficient for V. vulpes, Thylogale billardierii and Petrogale penicillata. However, further replicates may be required for marsupial carnivores, as their faeces yielded DNA that amplified substantially less often and less reliably, for all preservation and extraction methods tested, than did the other species. Although pilot studies remain vital for evaluating the feasibility of non-invasive sampling prior to undertaking any in-depth study the availability of a thoroughly tested storage and DNA extraction combination protocol known to be optimal for five different species should make that process much simpler.