Nighttime warming alleviates the incidence of juniper forest growth decline on the Tibetan Plateau.

Recent warming over the Tibetan Plateau (TP) is approximately twice the global-mean surface temperature increase and poses a threat to the healthy growth of forests. Although many studies have focused on whether recent climate warming has caused forest growth decline on the TP, it remains unclear how asymmetric warming, that is faster increasing nighttime temperature than daytime, impacts forest growth decline. We explored this question by using a ring-width index series from 1489 juniper trees (Juniperus prezwalskii and J. tibetica) at 50 sites on the TP. We calculated the percentage of trees with growth decline (PTD) to reconstruct historical forest growth decline and employed a piecewise structural equation meta-model (pSEM) and linear mixed model (LMM) to explore influencing factors. We found that the PTD has decreased since the late 19th century, with an abrupt decreasing trend since the 1980s. Results of the pSEM show that winter minimum temperature has a stronger indirect negative effect on the variation in PTD (β = -0.24, p < 0.05) compared to that of the weak indirect positive effect of summer maximum temperature (β = 0.16, p < 0.05). The results of LMM show that the variation in PTD is directly negatively (p < 0.001) affected by both winter minimum temperature and summer total precipitation, but the former has a greater independent contribution than the latter (with 17.7% vs 2.5% of variances independently explained, respectively). These results suggest that increased winter minimum temperature substantially mitigates the growth decline in juniper forests on the TP. As the minimum temperature generally occurs at night, we conclude that the asymmetric increase in nighttime temperature has decreased the incidence of juniper forest growth decline on the TP under climate warming. This alleviating effect of nighttime warming is likely due to reduced low-temperature constraints and reduced damage to tree growth.