Mock Plant Communities and a Large Mammal Case Study Reveal ITS2 Primer Bias Against Graminoids

DNA fecal metabarcoding has revolutionized the field of herbivore diet analyses, offering deeper insight into plant-herbivore interactions and more reliable ecological inferences. However, due to PCR amplification bias, primer selection has a major impact on the validity of these inferences and insights. Using two pooling approaches on four mock communities and a case study examining diets of four large mammalian herbivores (LMH), we evaluated the efficacy of two primer pairs targeting the internal transcribed spacer 2 (ITS2) region: the widely used ITS-S2F/ITS4 pair and the UniPlant F/R pair, designed specifically for DNA metabarcoding. Both primer pairs consistently underrepresented graminoids, where > 40% of graminoid species did not amplify in vitro. However, the UniPlant F/R primer pair more accurately amplified mock plant communities, whereas the ITS-S2F/ITS4 pair underestimated graminoid relative abundance by at least twofold more than UniPlant F/R primers. Furthermore, in the LMH case study, UniPlant F/R primers identified graminoids as the dominant plant group for at least one LMH, indicating diet niche partitioning, while ITS-S2F/ITS4 primers largely failed to amplify graminoid DNA, potentially overestimating LMH diet overlap. Our findings underscore the importance of incorporating mock community analyses into DNA metabarcoding protocols to identify and mitigate primer bias, thereby enhancing the accuracy of ecological conclusions and supporting more effective conservation and management decisions.