Decametric-scale buffering of climate extremes in forest understory within a riparian microrefugia: the key role of microtopography.

Riparian corridors often act as low-land climate refugia for temperate tree species in their southern distribution range. A plausible mechanism is the buffering of regional climate extremes by local physiographic and biotic factors. We tested this idea using a 3-year-long microclimate dataset collected along the Ciron river, a refugia for European beech (Fagus sylvatica) in southwestern France. Across the whole network, canopy gap fraction was the main predictor for spatial microclimatic variations, together with two other landscape features (elevation above the river and woodland fraction within a 300m radius). However, within the riparian forest only (canopy gap fraction < 25%, distance to the river < 150m), variations of up to -4°C and + 15% in summertime daily maximum air temperature and minimum relative humidity, respectively, were still found from the plateau to the cooler, moister river banks, only ~ 5-10m below. Elevation above the river was then identified as the main predictor, and explained the marked variations from the plateau to the banks much better than canopy gap fraction. The microclimate measured near the river is as cool but moister than the macroclimate encountered at 700-1000m asl further east in F. sylvatica's main distribution range. Indeed, at all locations, we found that air relative humidity was higher than expected from a temperature-only effect, suggesting that extra moisture is brought by the river. Our results explain well why beech trees in this climate refugium are restricted to the river gorges where microtopographic variations are the strongest and canopy gaps are rare.

[Beech (Fagus sylvatica) germination and seedling growth under climatic and allelopathic constraints]. / Germination et croissance des plantules de hêtre (Fagus sylvatica) sous contraintes climatiques et allélopathiques.

Recruitment is a key process for forest sustainability, especially in warm margin of distribution area. The influence of climate (temperate or warm), of soil water availability, and of allelopathic interactions from different forest species have been tested on the germination of Fagus sylvatica in controlled climatic conditions. Germination rates of non-dormant Fagus seeds were improved by relatively warm temperatures (20°C), but reversibly stopped under heat constraint (27°C). The relative growth rate of Fagus seedlings was better under temperate climatic conditions. Foliar extracts of Hedera helix showed the highest allelopathic effect on Fagus recruitment, especially in temperate conditions. Our results suggest a limitation of Fagus recruitment in warm margin of its distribution area, and a modulation of recruitment success according to the identity of plant neighbourhood.