Temperate and boreal forests in the Northern Hemisphere cover an area of about 2 x 10(7) square kilometres and act as a substantial carbon sink (0.6-0.7 petagrams of carbon per year). Although forest expansion following agricultural abandonment is certainly responsible for an important fraction of this carbon sink activity, the additional effects on the carbon balance of established forests of increased atmospheric carbon dioxide, increasing temperatures, changes in management practices and nitrogen deposition are difficult to disentangle, despite an extensive network of measurement stations. The relevance of this measurement effort has also been questioned, because spot measurements fail to take into account the role of disturbances, either natural (fire, pests, windstorms) or anthropogenic (forest harvesting). Here we show that the temporal dynamics following stand-replacing disturbances do indeed account for a very large fraction of the overall variability in forest carbon sequestration. After the confounding effects of disturbance have been factored out, however, forest net carbon sequestration is found to be overwhelmingly driven by nitrogen deposition, largely the result of anthropogenic activities. The effect is always positive over the range of nitrogen deposition covered by currently available data sets, casting doubts on the risk of widespread ecosystem nitrogen saturation under natural conditions. The results demonstrate that mankind is ultimately controlling the carbon balance of temperate and boreal forests, either directly (through forest management) or indirectly (through nitrogen deposition).

Gas exchange measurements were carried out on ash and oak trees in a forest plantation during three whole growing seasons characterized by different water availability (2001, 2002 and 2003). A quantitative limitation analysis was applied to estimate the effects of drought and leaf ontogeny on stomatal ( S L ) and non-stomatal limitations ( NS L ) to light-saturated net photosynthesis ( A max ), relative to the seasonal maximum rates obtained under conditions of optimal soil water content. Furthermore, based on combined gas exchange and chlorophyll fluorescence measurements, NS L was partitioned into a diffusive (due to a decrease in mesophyll conductance, MC L ) and a biochemical component (due to a decrease in carboxylation capacity, B L ). During the wettest year ( ) about two-thirds of the decline in A max was attributable to S L . However, with increasing drought intensity, NS L increased more than S L and nearly equalled it when the stress was very severe (i.e. with g sw < < < < 60 mmol H 2 O m). Within NS L , MC L represented the main component, except at the most severe water stress levels when it was equalled by B L . It is concluded that diffusional limitations (i.e. S L + + + + MC L ) largely affect net assimilation during most of the year, whereas biochemical limitations are quantitatively important only during leaf development and senescence or with severe droughts.Key-words : diffusional limitations; leaf age; stomatal and non-stomatal limitations; water stress.Abbreviations : A max , light-saturated net assimilation at ambient CO 2 ; g sw , stomatal conductance to water vapour; V cmax ( C i ), maximum rate of carboxylation on the basis of intercellular CO 2 concentration; V cmax ( C c ), maximum rate of carboxylation on the basis of chloroplastic CO 2

The leaf area to sapwood area ratio (A :A) of trees has been hypothesized to decrease as trees become older and taller. Theory suggests that A :A must decrease to maintain leaf-specific hydraulic sufficiency as path length, gravity, and tortuosity constrain whole-plant hydraulic conductance. We tested the hypothesis that A :A declines with tree height. Whole-tree A :A was measured on 15 individuals of Douglas-fir (Pseudotsuga menziesii var. menziesii) ranging in height from 13 to 62 m (aged 20-450 years). A :A declined substantially as height increased (P=0.02). Our test of the hypothesis that A :A declines with tree height was extended using a combination of original and published data on nine species across a range of maximum heights and climates. Meta-analysis of 13 whole-tree studies revealed a consistent and significant reduction in A :A with increasing height (P<0.05). However, two species (Picea abies and Abies balsamea) exhibited an increase in A :A with height, although the reason for this is not clear. The slope of the relationship between A :A and tree height (ΔA :A/Δh) was unrelated to mean annual precipitation. Maximum potential height was positively correlated with ΔA :A/Δh. The decrease in A :A with increasing tree size that we observed in the majority of species may be a homeostatic mechanism that partially compensates for decreased hydraulic conductance as trees grow in height.

We investigated the impact of drought on the physiology of 41-year-old Scots pine (Pinus sylvestris L.) in central Scotland. Measurements were made of the seasonal course of transpiration, canopy stomatal conductance, needle water potential, xylem water content, soil-to-needle hydraulic resistance, and growth. Comparison was made between drought-treated plots and those receiving average precipitation. In response to drought, transpiration rate declined once volumetric water content (VWC) over the top 20 cm of soil reached a threshold value of 12%. Thereafter, transpiration was a near linear function of soil water content. As the soil water deficit developed, the hydraulic resistance between soil and needles increased by a factor of three as predawn needle water potential declined from -0.54 to -0.71 MPa. A small but significant increase in xylem embolism was detected in 1-year-old shoots. Stomatal control of transpiration prevented needle water potential from declining below -1.5 MPa. Basal area, and shoot and needle growth were significantly reduced in the drought treatment. In the year following the drought, canopy stomatal conductance and soil-to-needle hydraulic resistance recovered. Current-year needle extension recovered, but a significant reduction in basal area increment was evident one year after the drought. The results suggest that, in response to soil water deficit, mature Scots pine closes its stomata sufficiently to prevent the development of substantial xylem embolism. Reduced growth in the year after a severe soil water deficit is most likely to be the result of reduced assimilation in the year of the drought, rather than to any residual embolism carried over from one year to the next.

We present a new synthesis, based on a suite of complementary approaches, of the primary production and carbon sink in forests of the 25 member states of the European Union (EU-25) during 1990–2005. Upscaled terrestrial observations and model-based approaches agree within 25% on the mean net primary production (NPP) of forests, i.e. 520±75 g C m−2 yr−1 over a forest area of 1.32 × 106 km2 to 1.55 × 106 km2 (EU-25). New estimates of the mean long-term carbon forest sink (net biome production, NBP) of EU-25 forests amounts 75±20 g C m−2 yr−1. The ratio of NBP to NPP is 0.15±0.05. Estimates of the fate of the carbon inputs via NPP in wood harvests, forest fires, losses to lakes and rivers and heterotrophic respiration remain uncertain, which explains the considerable uncertainty of NBP. Inventory-based assessments and assumptions suggest that 29±15% of the NBP (i.e., 22 g C m−2 yr−1) is sequestered in the forest soil, but large uncertainty remains concerning the drivers and future of the soil organic carbon. The remaining 71±15% of the NBP (i.e., 53 g C m−2 yr−1) is realized as woody biomass increments. In the EU-25, the relatively large forest NBP is thought to be the result of a sustained difference between NPP, which increased during the past decades, and carbon losses primarily by harvest and heterotrophic respiration, which increased less over the same period