Question: A set of easily-measured ('soft') plant traits has been identified as potentially useful predictors of ecosystem functioning in previous studies. Here we aimed to discover whether the screening techniques remain operational in widely contrasted circumstances, to test for the existence of axes of variation in the particular sets of traits, and to test for their links with 'harder' traits of proven importance to ecosystem functioning. Location: central-western Argentina, central England, northern upland Iran, and northeastern Spain. Recurrent patterns of ecological specialization: Through ordination of a matrix of 640 vascular plant taxa by 12 standardized traits, we detected similar patterns of specialization in the four floras. The first PCA axis was identified as an axis of resource capture, usage and release. PCA axis 2 appeared to be a size-related axis. Individual PCA for each country showed that the same traits remained valuable as predictors of resource capture and utilization in all of them, despite their major differences in climate, biogeography and land-use. The results were not significantly driven by particular taxa: the main traits determining PCA axis 1 were very similar in eudicotyledons and monocotyledons and Asteraceae, Fabaceae and Poaceae. Links between recurrent suites of 'soft' traits and 'hard' traits: The validity of PCA axis 1 as a key predictor of resource capture and utilization was tested by comparisons between this axis and values of more rigorously established predictors ('hard' traits) for the floras of Argentina and England. PCA axis 1 was correlated with variation in relative growth rate, leaf nitrogen content, and litter decomposition rate. It also coincided with palatability to model generalist herbivores. Therefore, location on PCA axis 1 can be linked to major ecosystem processes in those habitats where the plants are dominant. Conclusion: We confirm the existence at the global scale of a major axis of evolutionary specialization, previously recog-nised in several local floras. This axis reflects a fundamental trade-off between rapid acquisition of resources and conservation of resources within well-protected tissues. These major trends of specialization were maintained across different environmental situations (including differences in the proximate causes of low productivity, i.e. drought or mineral nutrient deficiency). The trends were also consistent across floras and major phylogenetic groups, and were linked with traits directly relevant to ecosystem processes.

Abstract.Research into interspecific variation in functional traits is important for our understanding of trade-offs in plant design and function, for plant functional type classifications and for understanding ecosystem responses to shifts in species composition. Interspecific rankings of functional traits are a function of, among other factors, ontogenetic or allometric development and environmental effects on phenotypes. For woody plants, which attain large size and long lives, these factors might have strong effects on interspecific trait rankings. This paper is the first to test and compare the correspondence of interspecific rankings between laboratory grown seedlings and field grown adult plants for a wide range of functional leaf and stem traits. It employs data for 90 diverse woody and semiwoody species in a temperate British and a (sub)Mediterranean Spanish flora, all collected according to a strict protocol. For 12 out of 14 leaf and stem traits we found significant correlations between the species ranking in laboratory seedlings and field adults. For leaf size and maximum stem vessel diameter > 50 % of variation in field adults was explained by that in laboratory seedlings. Two important determinants of plant and ecosystem functioning, specific leaf area and leaf N content, had only 27 to 36 and 17 to 31 % of variation, respectively, in field adults explained by laboratory seedlings, owing to subsets of species with particular ecologies deviating from the general trend. In contrast, interspecific rankings for the same traits were strongly correlated between populations of field adults on different geological substrata. Extrapolation of interspecific trait rankings from laboratory seedlings to adult plants in the field, or vice versa, should be done with great caution. Keywords: Allometry; Interspecific variation; Intraspecific variation; Ontogeny; Shrub; Trait; Tree.Nomenclature: Castroviejo et al. (1986Castroviejo et al. ( -2000; Stace (1991).Abbreviations: DM/SM = Leaf dry mass/saturated mass ratio; LD = leaf density; PAR = Photosynthetically active radiation; SLA = Specific leaf area; SSLM = Specific saturated leaf mass; SVD max = Maximum stem vessel diameter.

Drought stress is the main cause of mortality of holm oak (Quercus ilex L.) seedlings in forest plantations. We therefore assessed if drought hardening, applied in the nursery at the end of the growing season, enhanced the drought tolerance and transplanting performance of holm oak seedlings. Seedlings were subjected to three drought hardening intensities (low, moderate and severe) for 2.5 and 3.5 months, and compared with control seedlings. At the end of the hardening period, water relations, gas exchange and morphological attributes were determined, and survival and growth under mesic and xeric transplanting conditions were assessed. Drought hardening increased drought tolerance primarily by affecting physiological traits, with no effect on shoot/root ratio or specific leaf mass. Drought hardening reduced osmotic potential at saturation and at the turgor loss point, stomatal conductance, residual transpiration (RT) and new root growth capacity (RGC), but enhanced cell membrane stability. Among treated seedlings, the largest response occurred in seedlings subjected to moderate hardening. Severe hardening reduced shoot soluble sugar concentration and increased shoot starch concentration. Increasing the duration of hardening had no effect on water relations but reduced shoot mineral and starch concentrations. Variation in cell membrane stability, RT and RGC were negatively related to osmotic adjustment. Despite differences in drought tolerance, no differences in mortality and relative growth rate were observed between hardening treatments when the seedlings were transplanted under either mesic or xeric conditions.

Climatic dryness imposes limitations on vascular plant growth by reducing stomatal conductance, thereby decreasing CO uptake and transpiration. Given that transpiration-driven water flow is required for nutrient uptake, climatic stress-induced nutrient deficit could be a key mechanism for decreased plant performance under prolonged drought. We propose the existence of an "isohydric trap," a dryness-induced detrimental feedback leading to nutrient deficit and stoichiometry imbalance in strict isohydric species. We tested this framework in a common garden experiment with 840 individuals of four ecologically contrasting European pines (Pinus halepensis, P. nigra, P. sylvestris, and P. uncinata) at a site with high temperature and low soil water availability. We measured growth, survival, photochemical efficiency, stem water potentials, leaf isotopic composition (δ C, δ O), and nutrient concentrations (C, N, P, K, Zn, Cu). After 2 years, the Mediterranean species Pinus halepensis showed lower δ O and higher δ C values than the other species, indicating higher time-integrated transpiration and water-use efficiency (WUE), along with lower predawn and midday water potentials, higher photochemical efficiency, higher leaf P, and K concentrations, more balanced N:P and N:K ratios, and much greater dry-biomass (up to 63-fold) and survival (100%). Conversely, the more mesic mountain pine species showed higher leaf δ O and lower δ C, indicating lower transpiration and WUE, higher water potentials, severe P and K deficiencies and N:P and N:K imbalances, and poorer photochemical efficiency, growth, and survival. These results support our hypothesis that vascular plant species with tight stomatal regulation of transpiration can become trapped in a feedback cycle of nutrient deficit and imbalance that exacerbates the detrimental impacts of climatic dryness on performance. This overlooked feedback mechanism may hamper the ability of isohydric species to respond to ongoing global change, by aggravating the interactive impacts of stoichiometric imbalance and water stress caused by anthropogenic N deposition and hotter droughts, respectively.

1. Forest area is increasing in temperate biomes through active and passive restoration of old fields. Despite the large extension of restored forests, the success of contrasting restoration strategies (active -planted forests -vs. passive -secondary forests -) over time has never been evaluated in Mediterranean forests.2. We studied how restoration strategy determined forest restoration success. We evaluated which restoration strategy resulted in forests more like references (i.e. forests with continuous canopy cover since at least the 1940s) in terms of structure, diversity, functional composition, and dynamics. We then assessed whether active restoration accelerated forest recovery compared to passive restoration.3. We studied a chronosequence of recovery in four forest types (mountain and Mediterranean pine forests and mesic and Mediterranean oak forests) using the data of the Spanish Forest Inventory in central Spain. Each plot was classified as planted, secondary or reference forest. We modelled the response ratios of 11 forest attributes and a multifunctionality index as a function of restoration strategy, forest age, and abiotic and biotic constraints. 4. Secondary forests showed a greater likeness to references than planted forests in oak forests while minor differences between secondary and planted forests were found in pine forests. The recovery speed of most forest attributes in secondary and planted forests was similar. Multifunctionality was higher, and increased more rapidly, in planted than in secondary forests in Mediterranean oak forests. However, multifunctionality was similar for both restoration strategies in the other forest types. Synthesis and applications. The long-term assessment of forest recovery inMediterranean abandoned fields indicated that both planted forests and natural forest succession are successful restoration strategies, depending on the aim and the forest type. In our research, restoration strategy did not influence the magnitude and speed of forest recovery in pine forests. However, in oak forests, natural forest succession led to forests more alike to references, but planted forests can maximize and accelerate recovery of forest multifunctionality. K E Y W O R D Sabandoned fields, forest age, forest inventory, forest recovery, forest succession, Mediterranean, multifunctionality, restoration strategy | 747Journal of Applied Ecology CRUZ-ALONSO et AL.