Grassland-associated bacterial and eukaryotic DNA signals in plateau pika feces across sampled natural and degraded grasslands

Grassland-associated habitat differences may be reflected in fecal microbial and eukaryotic DNA signatures of wild small mammals, but such host-associated responses remain poorly characterized in alpine grasslands. The plateau pika (Ochotona curzoniae), a keystone small mammal on the Qinghai-Tibetan Plateau, links vegetation, soil, burrow systems, predators and parasite-associated interfaces. We evaluated whether fecal bacterial and broad eukaryotic DNA signatures differed between sampled natural and degraded grasslands, with particular attention to nematode-like 18 S signals. We analyzed fecal 16 S rRNA V3-V4 and 18 S rRNA V4 profiles from 30 matched plateau pika samples collected at one sampled natural-grassland site and one sampled degraded-grassland site in Seda County, China. After rarefaction to 26,106 reads per sample for 16 S and 21,298 reads per sample for 18 S, no alpha-diversity metric differed significantly between grassland states. In contrast, bacterial and overall eukaryotic composition differed significantly, although grassland state explained only a small fraction of multivariate variation. Broad 18 S profiles were fungal-dominated. Nematode-like, broad parasite-associated and strict parasite-associated 18 S signals were higher in the sampled degraded grassland. PR2/SILVA searches supported broad Nematoda assignments for key amplicon sequence variants but did not support Lamotheoxyuris or Eimeria as main-text lineage-level conclusions. Cross-domain analyses did not support strong bacterial-eukaryotic coupling. Fecal bacterial and eukaryotic DNA signatures captured grassland-associated compositional reorganization in plateau pika, while within-sample diversity remained similar between the two sampled grassland states. Elevated Nematoda/nematode-like 18 S DNA signals in the sampled degraded grassland provide a non-invasive molecular signal that warrants targeted parasitological validation. Because each grassland state was represented by one site and 18 S rarefaction curves remained incompletely saturated, the findings are restricted to the sampled sites and should not be interpreted as landscape-wide effects, infection prevalence or parasite burden.