Conceptual diagram of the hypothesized facilitative effects of pikas (Ochotona curzoniae) on yak (Bos grunniens) growth performance on the Qinghai-Tibetan Plateau.

(A) In the absence of pikas, the poisonous forb Stellera chamaejasme exerts strong negative effects on the growth of yaks by outcompeting palatable grasses and sedges, thereby reducing forage availability and foraging efficiency. (B) In the presence of pikas, clipping of Stellera suppresses its abundance, mitigating its negative impact and promoting yak growth. Grey dashed line and “+” indicate livestock grazing promotes the proliferation of the poisonous forbs by suppressing palatable grasses and sedges. Black dashed lines indicate negative effects of poisonous plants on yaks, black solid lines indicate competition between plant groups, and red solid lines indicate pika suppression of Stellera. Line thickness represents the relative strength of putative species interactions. (C, D) show pika clipping activity and its effect on Stellera abundance at the study site. Credits: Xiaona Zheng (photographs).

Diet selection of pikas and yaks and their potential interactions mediated by poisonous plants based on field surveys in July 2021.

(A) Feeding and clipping frequencies of pikas, and (B) grazing frequencies of yaks on grasses, sedges, forbs, and Stellera across 10 2 × 2 m plots and 10 250 m transects, respectively. (C) Relationship between active pika burrow density and Stellera cover, and (D) between Stellera cover and an index of yak grazing activity (dung density) in 30 10 × 10 m plots. (E) Photographs showing (left to right) a pika at its burrow entrance, flowering Stellera, and a yak grazing among Stellera plants. Different letters above bars indicate significant differences at P < 0.05. Credits: Xiaona Zheng (photographs).

Combined effects of two-year (2022–2023) pika and Stellera removal treatments on yak growth, forage quantity, and quality in the field manipulative experiment.

(A) Pika density (indicated by active burrow density), (B) yak weight gain, (C) relationship between pika density and yak weight gain, (D) Stellera cover, (E) grass cover, (F) sedge cover, (G) crude protein (CP) content, (H) acid detergent fibre (ADF) content, and (I) neutral detergent fibre (NDF) of total forage (dry mass basis). Average values from both years were used in analyses, yielding one data point per 150 × 150 m plot. Significant interactions between pikas and Stellera were evaluated using post hoc tests; means not sharing letters differ significantly. For panels (A) and (F), only main effects were significant (table S3). Error bars indicate ± SE.

Combined effects of two-year (2022–2023) pika and Stellera removal treatments on yak foraging efficiency in the field manipulative experiment.

(A, B) Bite rates and (C, D) bites per step for sedges and grasses, respectively. Mean values from both years were used for analysis (one data point per 150 × 150 m plot). Significant interactions between pikas and Stellera were evaluated using post hoc tests; means not sharing letters differ significantly. Error bars indicate ± SE.

Combined effects of two-year (2022–2023) pika and Stellera removal on forb cover in the field manipulative experiment.

The average values of each variable in the two years were used for statistical analysis, providing a single data point for each variable in each 150 × 150 m plot. Error bars represent ± SE.

Combined effects of two-year (2022–2023) pika and Stellera removal on foraging efficiency of yaks in the field manipulative experiment.

(A) bite rates and (B) bites per step of yaks for forbs, respectively. The average values of each variable in the two years were used for statistical analysis, providing a single data point for each variable in each 150 × 150 m plot. Error bars represent ± SE.

Model summary for diet and clipping selections of pikas and yaks in the field surveys in July 2021.

Model contrasts for diet and clipping selections of pikas and yaks in the field surveys in July 2021.

Model summary for interactive effects of pikas and S. chamaejasme on pika burrow density, yak weight gain, S. chamaejasme cover, grass cover, sedge cover, and forb cover in the field manipulative experiment in 2022-2023.

Model contrasts for pika burrow density, yak weight gain, S. chamaejasme cover, and grass cover in the field manipulative experiment in 2022-2023.

Model summary for interactive effects of pikas and S. chamaejasme on crude protein (CP) %, acid detergent fibre (ADF) %, and neutral detergent fibre (NDF) % of total forage in the field manipulative experiment in 2022-2023.

Model contrasts for crude protein (CP) % and acid detergent fibre (ADF) % of total forage in the field manipulative experiment in 2022-2023.

Model summary for interactive effects of pikas and S. chamaejasme on yak bite rate (bites/h) and the bite to step ratio (bites/step) for grasses, sedges, and forbs in the field manipulative experiment in 2022-2023.

Model contrasts for yak bite rate (bites/h) and the bite to step ratio (bites/step) on grasses and sedges in the field manipulative experiment in 2022-2023.

Summary statistics from generalized linear mixed effects models testing for pre-treatment differences in response variables representing plant community composition (cover by group) and diversity in designated mesocosm locations, as measured in 2021 prior to initiation of pika (±) and S. chamaejasme (±) treatments.