Disparate effects of global-change drivers on mountain conifer forests: warming-induced growth enhancement in young trees vs. CO2 fertilization in old trees from wet sites.

Theory predicts that the postindustrial rise in the concentration of CO2 in the atmosphere (c(a)) should enhance tree growth either through a direct fertilization effect or indirectly by improving water use efficiency in dry areas. However, this hypothesis has received little support in cold-limited and subalpine forests where positive growth responses to either rising ca or warmer temperatures are still under debate. In this study, we address this issue by analyzing an extensive dendrochronological network of high-elevation Pinus uncinata forests in Spain (28 sites, 544 trees) encompassing the whole biogeographical extent of the species. We determine if the basal area increment (BAI) trends are linked to climate warming and increased c(a) by focusing on region- and age-dependent responses. The largest improvement in BAI over the past six centuries occurred during the last 150 years affecting young trees and being driven by recent warming. Indeed, most studied regions and age classes presented BAI patterns mainly controlled by temperature trends, while growing-season precipitation was only relevant in the driest sites. Growth enhancement was linked to rising ca in mature (151-300 year-old trees) and old-mature trees (301-450 year-old trees) from the wettest sites only. This finding implies that any potential fertilization effect of elevated c(a) on forest growth is contingent on tree features that vary with ontogeny and it depends on site conditions (for instance water availability). Furthermore, we found widespread growth decline in drought-prone sites probably indicating that the rise in ca did not compensate for the reduction in water availability. Thus, warming-triggered drought stress may become a more important direct driver of growth than rising ca in similar subalpine forests. We argue that broad approaches in biogeographical and temporal terms are required to adequately evaluate any effect of rising c(a) on forest growth.