RESUMEN
The study of diet composition is required to understand the interactions between animal and plant ecosystems. Different non-invasive techniques applied on faecal samples have commonly been used for such purposes, with cuticle microhistological analysis (CMA) and emerging DNA-based methods being the most relevant. In this work, we refined and optimized a qualitative DNA-based approach combining PCR amplification of long trnL(UAA) and ITS2 fragments and capillary electrophoresis (PCR-CE), instead of short trnL(UAA) fragments and massive sequencing technologies commonly reported. To do so, we developed a controlled diet assay using a stabled Pyrenean chamois specimen (Rupicapra pyrenaica pyrenaica), which included representative herbaceous and shrubby plant species. We also assessed the impact of sample freshness on the diet determination of this mountain caprinae by exposing faecal samples to the outdoor environment for three weeks. Faecal samples from both experiments were analysed by qualitative PCR-CE and semi-quantitative CMA in order to compare the pros and cons of both approaches. Our results show that all of the offered plant species were detected by both methodologies although CMA over-detected shrubs compared to herbaceous species. At the same time, sample degradation due to sustained climate exposure is a limiting factor for molecular analysis, but not for CMA. Taken all together, our results suggest that the qualitative information obtained by CMA and PCR-CE can be interchangeable when faecal samples are fresh (less than one week after deposition) but, afterwards, molecular analysis underestimates diet composition probably due to DNA degradation. CMA, however, can accurately be used at least three weeks after defecation. Moreover, by combining the results of simultaneous PCR amplification of two complementary genes, this optimized PCR-CE methodology provides a reliable, feasible and more affordable alternative for multiple and routine analyses of complex samples. Neither CMA nor PCR-CE seems to solve comprehensively the quatification of herbivore diets and thus further research needs to be done.