Phenotypic effects of the nurse<i> Thylacospermum caespitosum </i>on dependent plant species along regional climate stress gradients

25 July 2017

Jiang, Xingpei; Michalet, Richard; Shuyan, Chen; Zhao, Liang; Wang, Xiangtai; Wang, Chenyue; An, Lizhe; Xiao, Sa

Contrasting phenotypes of alpine cushion species have been recurrently described in several mountain ranges along small-scale topography gradients, with tight competitive phenotypes in stressful convex topography and loose facilitative phenotypes in sheltered concave topography. The consistency of phenotypic effects along large-scale climate stress gradients have been proposed as a test of the likely genetic bases of the differences observed at small-scale. Inversely, plastic phenotypic effects are more likely to vanish at some points along climate stress gradients. We tested this hypothesis for two phenotypes of the alpine cushion species Thylacospermum caespitosum at four points along regional gradients of cold and drought stress in northwest China. We measured the traits of the two cushion phenotypes and quantified their associated plant communities and environmental variables along the regional temperature and aridity gradients. Cushion height, convexity and stem density overall showed significant effect of phenotypes. Difference in tightness of cushions between phenotypes was consistent across climate conditions, whereas differences in cushion convexity and height between phenotypes increased with increasing cold stress. Phenotypic effects on species richness and abundance were consistent along both climate gradients but not effects on species composition, while there were no phenotypic effects on environmental variables. Additionally, RII (relative interaction index) curves were linear along the drought gradient but unimodal along the temperature gradient, likely due to the occurrence of contrasting species pools at the different sites. We conclude that the consistency of phenotypic effects of T. caespitosum was high for species richness and abundance and mainly explained by differences in interference mediated by likely heritable differences in cushion tightness. Additionally, our study shows that the shapes of the relationship between plant responses to neighbours and environmental stresses are not necessarily driven by niche-based deterministic factors.