Aim Leaf area index (LAI) is one of the key variables related to carbon, water and nutrient cycles in terrestrial ecosystems, but its global distribution patterns remain poorly understood. We evaluated the dependence of LAI on mean annual temperature (MAT) and wetness index (WI; a ratio of annual precipitation to potential evapotranspiration) for three plant functional types (PFTs: deciduous broadleaf, DB; evergreen conifer, EC; evergreen broadleaf, EB) at the global scale.Location Global. MethodsWe developed a new global database of unprecedented size (2606 published values) of field-observed LAI (site-specific maximum) values for vegetation of woody species. To maximize the generic applicability of our analysis, we standardized the definition of LAI, and corrected or excluded potentially erroneous data obtained from indirect optical methods. ResultsThe global dependence of LAI on MAT showed a reverse S-shaped pattern, in which LAI peaked at around 8.9 and 25.0°C and was lowest at around −10.0 and 18.8°C. The dependence on WI followed a saturation curve levelling off at around log WI = 0.30. LAI for evergreen forests increased linearly with increasing WI, but that for DB showed a curvilinear pattern saturating at log WI = 0.03. EC forests had higher LAI values than those of DB forests under cool conditions (MAT ≤ 8.9°C), but similar values under temperate conditions (MAT = 8.9-18.8°C).Main conclusions This analysis of global LAI−climate relationships supports the general belief that temperature limits LAI under cool conditions whereas water availability plays a predominant role under other conditions. We also found that these patterns differed significantly between PFTs, suggesting that the LAI of different PFTs may respond differently to climate change. Our study provides a broad empirical basis for predicting the global distribution of LAI and for analysing the effects of global climate change on vegetation structure and function.

Our results reveal that nitrogen distribution is mainly driven by the vertical light gradient but other factors such as LAI also have significant effects. Our equations contribute to an improvement in the projection of plant productivity and cycling of carbon and nitrogen in terrestrial ecosystems.

In Fagus, full-mast seeding years are invariably followed by at least one non-mast year. Both flower and leaf primordia develop during the summer within the same winter buds. Flower bud initiation occurs when the N content of developing seeds is increasing rapidly. We hypothesized that competition for nitrogen (N) between developing seeds and buds limits flower primordium formation in mast years and, hence, limits seed production in years following mast years. We tested this hypothesis in three Fagus crenata Blume forests at elevations of 550, 900 and 1500 m. Bud N concentration (N con), amount of N per bud (N bud) and dry mass per bud (DM) were compared between a mast year (2005) and the following non-mast year (2006), and between winter buds containing both leaf and flower primoridia (BF), which were formed during the non-mast year, and winter buds containing leaf primordia only (BL), which were formed in both mast and non-mast years. In addition, leaf numbers per shoot corresponding to the analyzed buds were counted, and the effect of masting on litter production was analyzed by quantifying the amounts of litter that fell in the years 2004 to 2007. The dry mass and N content of BF formed in 2006 by trees at both 550 and 1500 m were 2.1-3.4-fold higher than the corresponding amounts in BL, although the numbers of leaves per current-year shoot in 2007 that developed from the two bud types in the same individuals did not differ significantly. These results indicate that more N and carbohydrate are expended in producing BF than in producing BL. The amount of litter from reproductive organs produced in the mast year was similar to the amount of leaf litter at 900 and 1500 m, but three times as much at 550 m. Leaf numbers per shoot were significantly lower at all elevations in the mast year than in the non-mast years (and the amount of leaf litter at 550 and 1500 m tended to be lower in the mast year than in the non-mast years. In conclusion, preferential allocation of resources to seeds in the mast year reduced the availability of resources for flower primordium formation, and this may have accounted for the poor seed production in the following non-mast year.

It is generally assumed that the production of a large crop of seeds depletes stores of resources and that these take more than 1 year to replenish; this is accepted, theoretically, as the proximate mechanism of mast seeding (resource budget model). However, direct evidence of resource depletion in masting trees is very rare. Here, we trace seasonal and inter-annual variations in nitrogen (N) concentration and estimate the N storage pool of individuals after full masting of Fagus crenata in two stands. In 2005, a full masting year, the amount of N in fruit litter represented half of the N present in mature leaves in an old stand (age 190-260 years), and was about equivalent to the amount of N in mature leaves in a younger stand (age 83-84 years). Due to this additional burden, both tissue N concentration and individual N storage decreased in 2006; this was followed by significant replenishment in 2007, although a substantial N store remained even after full masting. These results indicate that internal storage may be important and that N may be the limiting factor for fruiting. In the 4 years following full masting, the old stand experienced two moderate masting events separated by 2 years, whilst trees in the younger stand did not fruit. This different fruiting behavior may be related to different "costs of reproduction" in the full masting year 2005, thus providing more evidence that N may limit fruiting. Compared to the non-fruiting stand, individuals in the fruiting stand exhibited an additional increase in N concentrations in roots early in the 2007 growing season, suggesting additional N uptake from the soil to supply resource demand. The enhanced uptake may alleviate the N storage depletion observed in the full masting year. This study suggests that masting affects N cycle dynamics in mature Fagus crenata and N may be one factor limiting fruiting.

Canopy photosynthetic capacity, measured as leaf maximum carboxylation rate (V (cmax)), is a key factor in ecosystem gas exchange models applied at different scales. We report seasonal and interannual variations in V(cmax) of natural beech stands (Fagus crenata Blume) along an altitudinal gradient in the temperate climate zone of Japan. Estimates are based on 6 years of gas exchange measurements. Pronounced seasonal and interannual variations in V(cmax) normalized to 25 degrees C (V(c,25)) were found for sun leaves. The seasonal pattern of V(c,25) generally followed an inverse parabolic curve, with an increase in spring, peak values in the middle of the growth period and a decline in autumn. Leaf nitrogen concentration (N(l)) and leaf mass per area were significantly related to V(c,25) during spring and summer, but were unrelated in autumn when V(c,25) declined. Annual peak V(c,25) ranged from 40.1 to 97.0 micromol m(-2) s(-1) and varied over as much as a twofold range at a particular site. Annual peak V(c,25) occurred about 28 days before annual peak N(l), with which it was poorly related. Our results show that it can be inappropriate to include constant values of photosynthetic parameters in ecosystem gas exchange models.