부산대학교 도서관

Aim: Plant functional traits are broadly used to quantify and predict impacts of climate change on vegetation. However, high intraspecific trait variation can bias mean values when few measurements are available. Here, we determine the extent of individual leaf trait variation and covariation across a highly heterogeneous environmental gradient for a widely distributed subtropical pine. We demonstrate the implications of trait variation for characterising species by assessing data availability and variability across the Pinus genus. Location: Central Mountain Range, Taiwan. Taxon Pinus taiwanensis Hayata (Pinaceae). Methods: We measured eight functional traits suggested to reflect plant strategies: needle length, area, thickness, dry and fresh mass, stomatal row density (SD), leaf dry matter content (LDMC) and specific leaf area (SLA). We examined trait variation in response to climatic and physiographic factors across an elevational gradient of 495–3106 m a.s.l. using linear mixed effects models (LMMs). Intraspecific trait covariation was explored using principal component analyses (PCAs) and LMMs. Descriptive statistics were calculated for Pinus records in the global TRY plant trait database. Results: Intraspecific variability among traits was high (CV 20%–44%) and predictable with elevation (generally p < 0.05, with declining needle size and LDMC with elevation and increasing SD). However, 41%–92% of variance was un‐explained by topography. Sixty‐five percent of variation was explained by two trait covariation axes, with predictable changes with elevation (p < 0.001). Pinus data availability in TRY was low. Across traits, only 12.5%–53% of species had sufficient sample sizes for intraspecific analyses. Main conclusions: We show substantial trait variation for a single species, here likely driven by temperature differences and additional biotic and abiotic drivers across the elevational range. Improved understanding of the extent and implications of intraspecific variability is necessary for reliable quantifications and predictions of the impacts of environmental change, especially in understudied, hyper‐diverse ecosystems such as tropical forests. [ABSTRACT FROM AUTHOR]