Advancing Dietary Analysis in Herbivores

: testing, validating, and deploying faecal DNA metabarcoding for accurate diet composition assessment

Abstract

Grazing animals are selective feeders, making continuous choices as to where and what to eat. Differences in nutritional value of plant species and plant parts consumed can directly affect performance, reproduction, survival, and overall population dynamics. However, quantifying the diets of grazing animals, especially those in extensive management systems, has been challenging and has limited our understanding of plant-herbivore interactions. Traditional methods such as grazing observation and microhistology of faecal samples offer limited taxonomic precision. The advancement of next-generation sequencing and DNA metabarcoding, particularly faecal DNA (fDNA) metabarcoding, offers a non-invasive, high-resolution approach to determining diet composition. Despite its advantages, questions about the quantitative accuracy of metabarcoding in identifying diet components to the species level persist. This project aims to test, validate, and apply fDNA metabarcoding techniques to determine ruminant diet composition in both controlled environments and field trials. Controlled diet experiments with domesticated Soay sheep involved non-invasive faecal sampling and vegetation collection for DNA metabarcoding analysis, focusing on the accuracy of metabarcoding in detecting differences in digestibility, target species, and the effectiveness of a study-specific reference database. The research question, "Can fDNA metabarcoding accurately determine the quantitative composition of herbivore diets to the species level?" encompasses identifying major and minor diet components, detecting plant species mixtures with varying digestibility, and assessing the impact of a custom DNA reference library on species-level taxonomic assignment. This includes in silico testing of primers for optimal PCR amplification of targeted vegetation species and evaluating DNA metabarcoding as a supplement to traditional botanical surveys for plant community characterisation. The resulting barcode reference library covers 203 species and highlights species-level discrimination challenges, particularly within the Poaceae family. Despite primer bias and differential digestibility issues, the study revealed consistent dietary trends and demonstrates the utility and challenges of metabarcoding in assessing diet composition, the impact of grazing on biodiversity within grasslands, and the need for optimised experimental designs. The integration of DNA metabarcoding with traditional botanical surveys in long-term field studies highlighted both its value and the necessity for further research in order to enhance our understanding of plant-herbivore interactions, ecosystem dynamics and improve grassland management for biodiversity conservation. Validating the use of fDNA metabarcoding approaches for herbivore populations will hold significance for both wild and domesticated herbivores. The insights gained will revolutionise our understanding of the implications of the dietary choices, and knowledge gained will aid in developing optimal grazing regimes to improving animal nutrition via grazing management, while also helping to prevent over-exploitation and habitat resource depletion.

Date of Award	2024
Original language	English
Awarding Institution	Aberystwyth University
Supervisor	Mariecia Fraser (Supervisor) & Helen Hipperson (Supervisor)