Genetic structure and diversity of the endangered growling grass frog in a rapidly urbanizing region

Abstract

Two pervasive and fundamental impacts of urbanization are the loss and fragmentation of natural habitats. From a genetic perspective, these impacts manifest as reduced genetic diversity and ultimately reduced genetic viability. The growling grass frog (Litoria raniformis) is listed as vulnerable to extinction in Australia, and endangered in the state of Victoria. Remaining populations of this species in and around the city of Melbourne are threatened by habitat loss, degradation and fragmentation due to urban expansion. We used mitochondrial DNA (mtDNA) and microsatellites to study the genetic structure and diversity of L. raniformis across Melbourne's urban fringe, and also screened four nuclear gene regions (POMC, RAG-1, Rhod and CRYBA1). The mtDNA and nuclear DNA sequences revealed low levels of genetic diversity throughout remnant populations of L. raniformis. However, one of the four regions studied, Cardinia, exhibited relatively high genetic diversity and several unique haplotypes, suggesting this region should be recognized as a separate Management Unit. We discuss the implications of these results for the conservation of L. raniformis in urbanizing landscapes, particularly the potential risks and benefits of translocation, which remains a contentious management approach for this species.

Keywords: Litoria raniformis; genetic diversity; haplotype; urbanization.

Figure 1.

Geographical distribution and sample sites of L. raniformis around Melbourne. Dark grey, current MMA; light grey, proposed urban growth area. Closed circles indicate detection records between 2000 and the present. Open circles indicate the location of sampling sites included in this study. The four study regions around Melbourne are indicated and clusters of sites within these regions circled (dashed lines).

Figure 2.

Haplotype network for 495 bp of the COI gene. Each pie represents a unique genetic sequence (haplotype) and the area of the pie is proportional to haplotype frequency within the entire dataset. Each line represents one mutational step. Small black circles correspond to inferred alleles, missing from the dataset.

Figure 3.

Geographical association of COI haplotypes. Each pie represents the haplotypes found in that cluster and the area of the pie is proportional to sample size. Clusters have been assigned letters and haplotypes numbered. See included table for the number of individuals displaying each haplotype.

Figure 4.

Structure bar plot (K=2) with individuals organized by geographical region. Each vertical bar represents a single individual and estimates Q: the probability an individual belongs to a population of the given colour.