Methods Tree-ring samples were collected at low (2898 m), intermediate (3309 m) and high (3639 m) altitudes in Yulong Snow Mountain. Residual chronologies were established by using the tree-ring width data. Relationships between the residual chronologies and climatic factors were determined by using response function analysis (RFA), redundancy analysis (RDA) and moving interval response analysis (MIRA).

Important findings The radial growth in P. likiangensis was influenced by both temperature and precipitation along an altitudinal gradient in Yulong Snow Mountain, but the seasonal response patterns varied with altitudes. It was significantly and positively correlated with precipitation from January through March of the current year at all the three altitudinal sites. The precipitation in post-growing season of the current year imposed a negative influence on radial growth at the sites of low and intermediate altitudes; whereas the effect was positive at the high altitude site. Spring drought of the current year was found to be another important factor affecting tree growth at the low and intermediate altitudes, while the current July temperature accelerated the radial growth at the high altitude. Results of RDA were generally consistent with that of RFA, indicating the effectiveness of RDA for quantifying the relationships between tree-ring width index and climatic factors. The results of MIRA indicated that variations in temperature and precipitation on a short-term scale also influenced tree growth. Based on the responses of tree growth to climate at the three altitudes and future climate predictions, the radial growth in P. likiangensis would likely be enhanced at the high altitude in the Yulong Snow Mountain, but the response patterns are uncertain at the low and intermediate altitudes.