Meadow degradation on the Qinghai-Tibet Plateau has become increasingly severe. As a keystone species, plateau pika ( Ochotona curzoniae ) exerts density-dependent impacts on vegetation and soil through burrowing and related activities. This study was conducted to elucidate the effects of plateau pika activity on the landscape pattern of alpine meadows on the Qinghai-Tibet Plateau in the Haergai River Basin, Qinghai Province. Healthy meadow, bare land, and pika burrows were identified from high-resolution unmanned aerial vehicle (UAV) imagery using support vector machine (SVM) and neural network (NN) classification approaches. The results indicated that SVM achieved a superior classification performance, with an overall accuracy of 99.05% and a Kappa coefficient of 0.9766. Landscape pattern analysis revealed that pika burrows were characterized by a small patch size, high abundance, and a regular shape with a dispersed distribution, substantially increasing landscape heterogeneity and reducing meadow connectivity. The number of pika burrow patches was positively correlated with meadow patch density, the number of patches, and the shape index, but negatively correlated with patch area and the aggregation index, indicating that pika activity accelerates meadow fragmentation. Correlation analysis and structural equation modeling further demonstrated strong coupling relationships among meadow, bare land, and pika burrows, forming a dynamic process of “pika burrow expansion–meadow fragmentation–bare land increase.” Based on these findings, a landscape-pattern-based meadow degradation classification was established, in which pika burrow patch densities of <580 patches·hm −2 indicated slight degradation, 580–1268 patches·hm −2 indicated moderate degradation, and ≥ 1268 patches·hm −2 indicated severe degradation. The critical degradation thresholds were identified as approximately 10,000 pika burrow patches and 2.7 × 10 4 meadow patches per hectare, beyond which the rate of meadow degradation tended to slow, and fragmentation shifted from declines in patch area and aggregation to internal reorganization and proliferation of small patches. The degradation classification framework proposed in this study effectively captures both the status of meadow degradation and underlying regime shifts, providing a scientific basis for monitoring and zoned management of alpine meadow degradation.
Zhang et al. (Tue,) studied this question.