Summary• The variability of branch-level hydraulic properties was assessed across 12 Scots pine populations covering a wide range of environmental conditions, including some of the southernmost populations of the species. The aims were to relate this variability to differences in climate, and to study the potential tradeoffs between traits.• Traits measured included wood density, radial growth, xylem anatomy, sapwoodand leaf-specific hydraulic conductivity (K S and K L ), vulnerability to embolism, leaf-to-sapwood area ratio (A L : A S ), needle carbon isotope discrimination (Δ 13 C) and nitrogen content, and specific leaf area.• Between-population variability was high for most of the hydraulic traits studied, but it was directly associated with climate dryness (defined as a combination of atmospheric moisture demand and availability) only for A L : A S , K L and Δ 13 C. Shoot radial growth and A L : A S declined with stand development, which is consistent with a strategy to avoid exceedingly low water potentials as tree size increases. In addition, we did not find evidence at the intraspecific level of some associations between hydraulic traits that have been commonly reported across species.• The adjustment of Scots pine's hydraulic system to local climatic conditions occurred primarily through modifications of A L : A S and direct stomatal control, whereas intraspecific variation in vulnerability to embolism and leaf physiology appears to be limited.

There is a missing link between tree physiological and wood-anatomical knowledge which makes it impossible mechanistically to explain and predict the radial growth of individual trees from climate data. Empirical data of microclimatic factors, intra-annual growth rates, and tree-specific ratios between actual and potential transpiration (T PET(-1)) of trees of three species (Quercus pubescens, Pinus sylvestris, and Picea abies) at two dry sites in the central Wallis, Switzerland, were recorded from 2002 to 2004 at a 10 min resolution. This included the exceptionally hot and dry summer of 2003. These data were analysed in terms of direct (current conditions) and indirect impacts (predispositions of the past year) on growth. Rain was found to be the only factor which, to a large extent, consistently explained the radial increment for all three tree species at both sites and in the short term as well. Other factors had some explanatory power on the seasonal time-scale only. Quercus pubescens built up much of its tree ring before bud break. Pinus sylvestris and Picea abies started radial growth 1-2 weeks after Quercus pubescens and this was despite the fact that they had a high T PET(-1) before budburst and radial growth started. A high T PET(-1) was assumed to be related to open stomata, a very high net CO2 assimilation rate, and thus a potential carbon (C)-income for the tree. The main period of radial growth covered about 30-70% of the productive days of a year. In terms of C-allocation, these results mean that Quercus pubescens depended entirely on internal C-stores in the early phase of radial growth and that for all three species there was a long time period of C-assimilation which was not used for radial growth in above-ground wood. The results further suggest a strong dependence of radial growth on the current tree water relations and only secondarily on the C-balance. A concept is discussed which links radial growth over a feedback loop to actual tree water-relations and long-term affected C-storage to microclimate.

An increasing number of studies have reported on forest declines and vegetation shifts triggered by drought. In the Swiss Rhone valley (Valais), one of the driest inner-Alpine regions, the species composition in low elevation forests is changing: The sub-boreal Scots pine (Pinus sylvestris L.) dominating the dry forests is showing high mortality rates. Concurrently the sub-Mediterranean pubescent oak (Quercus pubescens Willd.) has locally increased in abundance. However, it remains unclear whether this local change in species composition is part of a larger-scale vegetation shift. To study variability in mortality and regeneration in these dry forests we analysed data from the Swiss national forest inventory (NFI) on a regular grid between 1983 and 2003, and combined it with annual mortality data from a monitoring site. Pine mortality was found to be highest at low elevation (below 1000 m a.s.l.). Annual variation in pine mortality was correlated with a drought index computed for the summer months prior to observed tree death. A generalized linear mixed-effects model indicated for the NFI data increased pine mortality on dryer sites with high stand competition, particularly for small-diameter trees. Pine regeneration was low in comparison to its occurrence in the overstorey, whereas oak regeneration was comparably abundant. Although both species regenerated well at dry sites, pine regeneration was favoured at cooler sites at higher altitude and oak regeneration was more frequent at warmer sites, indicating a higher adaptation potential of oaks under future warming. Our results thus suggest that an extended shift in species composition is actually occurring in the pine forests in the Valais. The main driving factors are found to be climatic variability, particularly drought, and variability in stand structure and topography. Thus, pine forests at low elevations are developing into oak forests with unknown consequences for these ecosystems and their goods and services.

Tree water deficit estimated by measuring water-related changes in stem radius (DeltaW) was compared with tree water deficit estimated from the output of a simple, physiologically reasonable model (DeltaWE), with soil water potential (Psisoil) and atmospheric vapor pressure deficit (VPD) as inputs. Values of DeltaW were determined by monitoring stem radius changes with dendrometers and detrending the results for growth. We followed changes in DeltaW and DeltaWE in Pinus sylvestris L. and Quercus pubescens Willd. over 2 years at a dry site (2001-2002; Salgesch, Wallis) and in Picea abies (L.) Karst. for 1 year at a wet site (1998; Davos, Graubuenden) in the Swiss Alps. The seasonal courses of DeltaW in deciduous species and in conifers at the same site were similar and could be largely explained by variation in DeltaWE. This finding strongly suggests that DeltaW, despite the known species-specific differences in stomatal response to microclimate, is mainly explained by a combination of atmospheric and soil conditions. Consequently, we concluded that trees are unable to maintain any particular DeltaW. Either Psisoil or VPD alone provided poorer estimates of DeltaW than a model incorporating both factors. As a first approximation of DeltaWE, Psisoil can be weighted so that the negative mean Psisoil reaches 65 to 75% of the positive mean daytime VPD over a season (Q. pubescens: approximately 65%, P. abies: approximately 70%, P. sylvestris: approximately 75%). The differences in DeltaW among species can be partially explained by a different weighting of Psisoil against VPD. The DeltaW of P. sylvestris was more dependent on Psisoil than that of Q. pubescens, but less than that of P. abies, and was less dependent on VPD than that of P. abies and Q. pubescens. The model worked well for P. abies at the wet site and for Q. pubescens and P. sylvestris at the dry site, and may be useful for estimating water deficit in other tree species.

SummarySeparating continuously measured stem radius (SR) fluctuations into growth-induced irreversible stem expansion (GRO) and tree water deficit-induced reversible stem shrinkage (TWD) requires a conceptualization of potential growth processes that may occur during periods of shrinking and expanding SR below a precedent maximum. Here, we investigated two physiological concepts: the linear growth (LG) concept, assuming linear growth, versus the zero growth (ZG) concept, assuming no growth during periods of stem shrinkage.We evaluated the physiological mechanisms underlying these two concepts and assessed their respective plausibilities using SR data obtained from 15 deciduous and evergreen trees.The application of the LG concept produced steady growth rates, whereas growth rates varied strongly under the ZG concept, more in accordance with mechanistic expectations. Further, growth increased for a maximum of 120 min after periods of stem shrinkage, indicating limited growth activity during those periods. However, this extra growth was found to be a small fraction of total growth only. Furthermore, TWD under the ZG concept was better explained by a hydraulic plant model than TWD under the LG concept.We conclude that periods of stem shrinkage allow for very little growth in the four tree species investigated. However, further studies should focus on obtaining independent growth data to ultimately validate these findings.