Unexpected asymmetric Qinghai-Tibetan wildlife dietary response to anthropause: Rapid initial shifts and a slow return.

Abstract

Understanding how wildlife adapts to changes in human activity intensity is crucial for effective conservation in human-modified ecosystems. The 3-year COVID-19 related anthropause in China created an unprecedented experimental setting to examine details of the ecological impact of human absence from otherwise disturbed habitats. We investigated the ecological impact of the anthropause on roadside small mammals by testing if and how their diets changed over a 7-year period spanning before (2018-2019), during (2020-2022) and after (2023-2024) the anthropause, and whether these changes reflected temporary or lasting shifts in human activity. We assessed human impact on a keystone Qinghai-Tibetan Plateau wildlife taxon, pikas (Ochotona spp.), by recording and comparing their diets (using faecal DNA metabarcoding) from roadside and road-distant (non-disturbed) populations. We coupled the dietary differences results with vegetation structure and traffic density records to assess the effects of human activity intensity and infrastructure on dietary composition. Before the anthropause, roadside populations consumed dicot-rich diets distinct from nearby undisturbed meadow populations. During the reduced human activity of the anthropause, roadside pika diets rapidly converged with those of undisturbed populations within a single year, decoupling from local roadside vegetation compositions. After human activity resumed, dietary divergence between roadside and natural populations reappeared slowly across generations. These data reveal an asymmetric, multigenerational response: rapid convergence under reduced disturbance, followed by a slow divergence with the resumption of disturbance. Regression analyses show that traffic density was associated with dietary divergence between habitats and with diet-vegetation dissimilarity in roadside populations. These multigenerational data demonstrate that wildlife in human-modified landscapes remain sensitive to changes in human activity, even after prolonged exposure. Importantly, these asymmetric responses suggest that temporary reductions in human disturbance, such as traffic management, can yield lasting biodiversity benefits in increasingly human-dominated landscapes, and these responses should be explored as part of broader conservation efforts in human-modified ecosystems.