Below the canopy: global trends in forest vertebrate populations and their drivers.

Global forest assessments use forest area as an indicator of biodiversity status, which may mask below-canopy pressures driving forest biodiversity loss and 'empty forest' syndrome. The status of forest biodiversity is important not only for species conservation but also because species loss can have consequences for forest health and carbon storage. We aimed to develop a global indicator of forest specialist vertebrate populations to improve assessments of forest biodiversity status. Using the Living Planet Index methodology, we developed a weighted composite Forest Specialist Index for the period 1970-2014. We then investigated potential correlates of forest vertebrate population change. We analysed the relationship between the average rate of change of forest vertebrate populations and satellite-derived tree cover trends, as well as other pressures. On average, forest vertebrate populations declined by 53% between 1970 and 2014. We found little evidence of a consistent global effect of tree cover change on forest vertebrate populations, but a significant negative effect of exploitation threat on forest specialists. In conclusion, we found that the forest area is a poor indicator of forest biodiversity status. For forest biodiversity to recover, conservation management needs to be informed by monitoring all threats to vertebrates, including those below the canopy.

Plant movements and climate warming: intraspecific variation in growth responses to nonlocal soils.

Most range shift predictions focus on the dispersal phase of the colonization process. Because moving populations experience increasingly dissimilar nonclimatic environmental conditions as they track climate warming, it is also critical to test how individuals originating from contrasting thermal environments can establish in nonlocal sites. We assess the intraspecific variation in growth responses to nonlocal soils by planting a widespread grass of deciduous forests (Milium effusum) into an experimental common garden using combinations of seeds and soil sampled in 22 sites across its distributional range, and reflecting movement scenarios of up to 1600 km. Furthermore, to determine temperature and forest-structural effects, the plants and soils were experimentally warmed and shaded. We found significantly positive effects of the difference between the temperature of the sites of seed and soil collection on growth and seedling emergence rates. Migrant plants might thus encounter increasingly favourable soil conditions while tracking the isotherms towards currently 'colder' soils. These effects persisted under experimental warming. Rising temperatures and light availability generally enhanced plant performance. Our results suggest that abiotic and biotic soil characteristics can shape climate change-driven plant movements by affecting growth of nonlocal migrants, a mechanism which should be integrated into predictions of future range shifts.

Use of an airborne lidar system to model plant species composition and diversity of Mediterranean oak forests.

Airborne lidar is a remote-sensing tool of increasing importance in ecological and conservation research due to its ability to characterize three-dimensional vegetation structure. If different aspects of plant species diversity and composition can be related to vegetation structure, landscape-level assessments of plant communities may be possible. We examined this possibility for Mediterranean oak forests in southern Portugal, which are rich in biological diversity but also threatened. We compared data from a discrete, first-and-last return lidar data set collected for 31 plots of cork oak (Quercus suber) and Algerian oak (Quercus canariensis) forest with field data to test whether lidar can be used to predict the vertical structure of vegetation, diversity of plant species, and community type. Lidar- and field-measured structural data were significantly correlated (up to r= 0.85). Diversity of forest species was significantly associated with lidar-measured vegetation height (R(2) = 0.50, p < 0.001). Clustering and ordination of the species data pointed to the presence of 2 main forest classes that could be discriminated with an accuracy of 89% on the basis of lidar data. Lidar can be applied widely for mapping of habitat and assessments of habitat condition (e.g., in support of the European Species and Habitats Directive [92/43/EEC]). However, particular attention needs to be paid to issues of survey design: density of lidar points and geospatial accuracy of ground-truthing and its timing relative to acquisition of lidar data.