No abstract

Drought affects more people than any other natural disaster but there is little understanding of how ecosystems react to droughts. This study jointly analyzed spatio-temporal changes of drought patterns with vegetation phenology and productivity changes between 1999 and 2010 in major European bioclimatic zones. The Standardized Precipitation and Evapotranspiration Index (SPEI) was used as drought indicator whereas changes in growing season length and vegetation productivity were assessed using remote sensing time-series of Normalized Difference Vegetation Index (NDVI). Drought spatio-temporal variability was analyzed using a Principal Component Analysis, leading to the identification of four major drought events between 1999 and 2010 in Europe. Correspondence Analysis showed that at the continental scale the productivity and phenology reacted differently to the identified drought events depending on ecosystem and land cover. Northern and Mediterranean ecosystems proved to be more resilient to droughts in terms of vegetation phenology and productivity developments. Western Atlantic regions and Eastern Europe showed strong agglomerations of decreased productivity and shorter vegetation growing season length, indicating that these ecosystems did not buffer the effects of drought well. In a climate change perspective, increase in drought frequency or intensity may result in larger impacts over these ecosystems, thus management and adaptation strategies should be strengthened in these areas of concerns.

Aim The present study models the response of vegetation to drought between 1982 and 2011 focusing on the growing season, considering: (1) vulnerable versus resistant ecosystems, (2) ecosystem resilience, (3) the delayed response of vegetation to accumulated precipitation deficits and (4) reduction in productivity due to drought. Location Continental Europe. Methods Vegetation condition was analysed by anomalies of the fraction of photosynthetically active radiation absorbed by vegetation (FAPAR) within the growing season. FAPAR was modelled as a function of: (1) monthly deviation from long‐term average drought conditions, (2) past monthly deviation from long‐term average drought conditions, and (3) past monthly deviation from long‐term average FAPAR conditions. Significant regression coefficients were mapped as ecosystem resistance as well as ecosystem resilience to drought. Annual and long‐term growing season productivity were assessed over ecosystems vulnerable to drought within diverse biomes and plant functional types and were studied together with the extent, timing, frequency, duration and intensity of drought events. Results Mediterranean ecosystems showed the least resistance to drought but were resilient. Steppic ecosystems showed weak resistance and also weak resilience. Results also indicated weak resistance against drought of the Atlantic and Continental regions with various recovery rates. Productivity decline was observed when the timing, intensity, frequency and extent of drought events targeted the growing season of the vegetation cover. Main conclusions This study presents an improved method for modelling the impact of drought on the productivity of vegetation by considering the timing, intensity, duration and location of drought events within the growing season, and maps and analyses the resistance and resilience of plant functional types and biomes.

Habitat heterogeneity contributes to the maintenance of diversity, but the extent that landscape-scale rather than local-scale heterogeneity influences the diversity of soil invertebrates-species with small range sizes-is less clear. Using a Scottish habitat heterogeneity gradient we correlated Collembola and lumbricid worm species richness and abundance with different elements (forest cover, habitat richness and patchiness) and qualities (plant species richness, soil variables) of habitat heterogeneity, at landscape (1 km(2)) and local (up to 200 m(2)) scales. Soil fauna assemblages showed considerable turnover in species composition along this habitat heterogeneity gradient. Soil fauna species richness and turnover was greatest in landscapes that were a mosaic of habitats. Soil fauna diversity was hump-shaped along a gradient of forest cover, peaking where there was a mixture of forest and open habitats in the landscape. Landscape-scale habitat richness was positively correlated with lumbricid diversity, while Collembola and lumbricid abundances were negatively and positively related to landscape spatial patchiness. Furthermore, soil fauna diversity was positively correlated with plant diversity, which in turn peaked in the sites that were a mosaic of forest and open habitat patches. There was less evidence that local-scale habitat variables (habitat richness, tree cover, plant species richness, litter cover, soil pH, depth of organic horizon) affected soil fauna diversity: Collembola diversity was independent of all these measures, while lumbricid diversity positively and negatively correlated with vascular plant species richness and tree canopy density. Landscape-scale habitat heterogeneity affects soil diversity regardless of taxon, while the influence of habitat heterogeneity at local scales is dependent on taxon identity, and hence ecological traits, e.g. body size. Landscape-scale habitat heterogeneity by providing different niches and refuges, together with passive dispersal and population patch dynamics, positively contributes to soil faunal diversity.

No abstract