Habitat fragmentation may disrupt original patterns of gene flow and lead to drift-induced differentiation among local population units. Top predators such as the jaguar may be particularly susceptible to this effect, given their low population densities, leading to small effective sizes in local fragments. On the other hand, the jaguar's high dispersal capabilities and relatively long generation time might counteract this process, slowing the effect of drift on local populations over the time frame of decades or centuries. In this study, we have addressed this issue by investigating the genetic structure of jaguars in a recently fragmented Atlantic Forest region, aiming to test whether loss of diversity and differentiation among local populations are detectable, and whether they can be attributed to the recent effect of drift. We used 13 microsatellite loci to characterize the genetic diversity present in four remnant populations, and observed marked differentiation among them, with evidence of recent allelic loss in local areas. Although some migrant and admixed individuals were identified, our results indicate that recent large-scale habitat removal and fragmentation among these areas has been sufficiently strong to promote differentiation induced by drift and loss of alleles at each site. Low estimated effective sizes supported the inference that genetic drift could have caused this effect within a short time frame. These results indicate that jaguars' ability to effectively disperse across the human-dominated landscapes that separate the fragments is currently very limited, and that each fragment contains a small, isolated population that is already suffering from the effects of genetic drift.

The elusive nature and endangered status of most carnivore species imply that efficient approaches for their non-invasive sampling are required to allow for genetic and ecological studies. Faecal samples are a major potential source of information, and reliable approaches are needed to foster their application in this field, particularly in areas where few studies have been conducted. A major obstacle to the reliable use of faecal samples is their uncertain species-level identification in the field, an issue that can be addressed with DNA-based assays. In this study we describe a sequence-based approach that efficiently distinguishes jaguar versus puma scats, and that presents several desirable properties: (1) considerably high amplification and sequencing rates; (2) multiple diagnostic sites reliably differentiating the two focal species; (3) high information content that allows for future application in other carnivores; (4) no evidence of amplification of prey DNA; and (5) no evidence of amplification of a nuclear mitochondrial DNA insertion known to occur in the jaguar. We demonstrate the reliability and usefulness of this approach by evaluating 55 field-collected samples from four locations in the highly fragmented Atlantic Forest biome of Brazil and Argentina, and document the presence of one or both of these endangered felids in each of these areas.

No abstract