Carnivore ecology and diet assessment using DNA-based approaches: the elusive black-footed cat (Felis nigripes) as a case study.

Abstract

Information obtained from the assessment of diet has been used to reconstruct food webs of elusive and shy predators that are difficult to observe in their natural habitats. The information from these studies has been crucial in developing conservation strategies. Predator-prey dynamics in particular, can also be used as a proxy of ecosystem health. Traditional approaches for diet analysis often involve direct observation or morphological identification of prey remains from scat. These approaches can be challenging for smaller predators that are difficult to track, and are generalist, opportunistic or scavenging carnivores. This is because traditional methods are not well adapted to analyze and identify such a wide variety of prey species, making them liable to bias and inaccuracy. DNA-based methods for the identification of diet from scats have been shown to provide better resolution, accuracy and consistency when compared to all other methods. Species identification of prey consumed can be achieved using DNA barcoding; identification of species through the amplification of specific genetic markers linked to taxonomically identified species. DNA metabarcoding is a recent technological advancement which allows for identification of multiple species from a single sample using high-throughput sequencing (HTS). Using the shy and elusive black-footed cat (Felis nigripes, Burchell 1824) as a case study, in this dissertation I aimed to determine if DNA-based methods could be used to provide important population-level information (number of individuals in population, relatedness of individuals, dietary components, and parasite load). The first data chapter (Chapter 2) aimed to provide an in-depth review of the state of DNA reference libraries for small South African mammals. Small mammals constitute a large portion of small and medium carnivore diet. As such, the abundance of DNA reference libraries for small mammals gives a good indication if DNA metabarcoding studies are feasible. Analysis of DNA records revealed that the majority of small mammals are represented by at least one of the five mitochondrial genes. This study supports the use of multiple gene regions when performing scat metabarcoding particularly when wanting to determine the small mammal component of the diet. Chapter 3 tested in silico if published metabarcoding primers for 12S rRNA, 16S rRNA, COI and cyt b could reliably delimit common prey items eaten by black-footed cats. Successful species delimitation using metabarcoding rests on the presence of a “DNA barcode gap”. This gap is the difference between intraspecific and interspecific genetic distances within a group of organisms. Eight metabarcoding primer pairs were chosen from the literature. Using an alignment of DNA sequences from South African small mammals, each amplicon produced by the primer pairs was tested to determine if species level identification could be made. An optimal set of primers were developed for use in black-footed cat (and other small and medium carnivores) diet analyses. Chapter 4 assessed if microsatellites previously designed for use in domestic cat (Felis silvestris catus) could be successfully amplified and used to identify individual black-footed cats from scats. All nine microsatellite loci used in this study were amplified successfully and were polymorphic. The loci were found to have sufficient discriminatory power to distinguish individuals and identify clones and could therefore be used in parentage assignments. Chapter 5 aimed at identification of prey items of black-footed cats in the Benfontein Nature Reserve, Kimberley, South Africa using DNA metabarcoding. Of the four DNA regions selected for high-throughput sequencing, the cyt b region did not sequence well and was therefore excluded from the study. The result from this study were compared to a previous study conducted which was based on direct observations of black-footed cat feeding. Unsurprisingly, our DNA data supported previous studies which showed that the diet of black-footed cat consists mainly of rodents. Importantly, the DNA metabarcoding results identify some prey items that have not been recorded in black-footed cat diet previously highlighting the sensitivity of the DNA-based method. For example larger prey items such as antelope were also identified suggesting that the cats may also opportunistically scavenge. Chapter 6 assessed the presence of parasites found in black-footed cat wild populations. Five pathogens known to be of veterinary importance in domestic cats were found in the black-footed cat scats. These pathogens had previously not been documented in wild African felids, but are well documented in other taxa. These results add to the growing knowledge of diseases that could possibly contribute to the declining populations of black-footed cats in South Africa. In conclusion, DNA based methods have been shown to improved species resolution when compared to traditional methods of diet assessments. More so when coupled with highthroughput sequencing technologies. The results of this study show that DNA metabarcoding can be applied successfully to study the diet of South Africa carnivores. The results also indicate that DNA metabarcoding can be used in identification of species that are endemic to southern Africa.