Beware of Plant DNA in Animal Dietary Metabarcoding: Lessons From a Strictly Insectivorous Bat

• DOI: 10.1111/1755-0998.14100
• Publication Date: 2025-03-12T10:29:04Z
• Authors: Luis P. da Silva; Miguel Porto; Francisco Amorim; Pedro Beja; Vanessa A. Mata

ABSTRACTDNA metabarcoding is increasingly used in dietary studies, but it has limitations, such as detecting nonfood taxa. This issue is frequently mentioned in the literature but poorly understood, limiting interpretation of results and mitigation strategies. We evaluate the extent and sources of nonfood plant DNA in dietary metabarcoding, based on 281 faecal samples of a strictly insectivorous bat. We modelled plant taxa detections in relation to pollination syndromes, flowering and fruiting phenology and habitat associations, and we estimated co‐occurrences between plants and arthropods. The bat arthropod diet was consistent with previous studies. Plants were detected in 82.9% of samples, representing 148 taxa, and all pollination syndromes evaluated. Plant detections were more frequent during their flowering periods, particularly for those with mixed pollination syndromes, suggesting a relationship between flowering and detectability. Fruiting had a positive, albeit weaker, effect. There was a tendency for more frequent detection of forest plants and less frequent detection of plants associated with riparian and agricultural habitats. Co‐occurrences between arthropods and plants were weak and inconsistent. Our results highlight the potential for widespread detection of nonfood plant DNA in metabarcoding studies, calling for great care when analysing the plant component of diets. Specifically, we recommend: (i) implementing strategies for reducing plant contamination during field sampling; (ii) using multiple field and lab negative controls; and (iii) using ancillary information (e.g., sample visual inspection and literature review) to aid interpretation of metabarcoding results. Moreover, we recommend that studies reporting plant consumption results greatly diverging from dietary patterns obtained through other methods should include detailed explanations of methodological steps taken to exclude the confounding effects of nonfood plant DNA.