Aged but withstanding: Maintenance of growth rates in old pines is not related to enhanced water-use efficiency

Growth of old trees in cold-limited forests may benefit from recent climate warming and rising atmospheric CO2 concentrations (ca) if age-related constraints do not impair wood formation. To test this hypothesis, we studied old Mountain pine trees at three Pyrenean high-elevation forests subjected to cold-wet (ORD, AIG) or warmer-drier (PED) conditions. We analyzed long-term trends (1450–2008) in growth (BAI, basal area increment), maximum (MXD) and minimum (MID) wood density, and tree-ring carbon (δ13C) and oxygen (δ18O) isotope composition, which were used as proxies for intrinsic water-use efficiency (iWUE) and stomatal conductance (gs), respectively. Old pines showed positive (AIG and ORD) or stable (PED) growth trends during the industrial period (since 1850) despite being older than 400 years. Growth and wood density covaried from 1850 onwards. In the cold-wet sites (AIG and ORD) enhanced photosynthesis through rising ca was likely responsible for the post-1850 iWUE improvement. However, uncoupling between BAI and iWUE indicated that increases in iWUE were not responsible for the higher growth but climate warming. A reduction in gs was inferred from increased δ18O for PED trees from 1960 onwards, the warmest site where the highest iWUE increase occurred (34%). This suggests that an emergent drought stress at warm-dry sites could trigger stomatal closure to avoid excessive transpiration. Overall, carbon acquisition as lasting woody pools is expected to be maintained in aged trees from cold and high-elevation sites where old forests constitute unique long-term carbon reservoirs.

We are very grateful to several projects financed by “Organismo Autónomo de Parques Nacionales” (projects 12/2008 387/2011). E.G. was funded by a Juan de la Cierva post-doctoral research contract (FJCI-2014-19615, MEC, Spain). Spanish (AMB95-0160, CGL2011-26654) and EU projects ISONET (contract EV K2-2001-00237) and MILLENNIUM (017008–2) also supported this study by contributing additional datasets.