How much does it take to be old? Modelling the time since the last harvesting to infer the distribution of overmature forests in France

Abstract

Abstract Aim The distribution of overmature forests in metropolitan France is poorly known, with only a few well-studied prominent sites, and has never been evaluated countrywide. Here, we modelled French forest reserves’ time since the last harvesting operation - a proxy for forest maturity - then inferred the current statistical distribution of overmature forests (i.e. forests over 50 years without harvesting) in France. Location Metropolitan France Methods We used inventories from forest reserves and managed forests to calibrate a generalised linear mixed model explaining the time since the last harvesting with selected structural attributes and environmental variables. We then projected this model on the independent National Forest Inventory dataset. We thus obtained an updated estimation of the proportion and a rough distribution of overmature forest stands in metropolitan France. Results We found that high basal area of very large trees, high volumes of standing and downed deadwood, high diversity of tree-related microhabitats and more marginally diversity of decay stages best characterized the time since the last harvesting. Volumes of stumps and high density of coppices translating legacy of past forest management also distinguished more overmature plots. Our projection yielded an estimated 3% of French forests over 50 years without harvesting mostly located in more inaccessible areas (i.e. mountainous areas) and a promising proportion of future overmature forests if left unharvested. Main conclusions Our study showed that the time since the last harvesting is a good proxy for a combination of stand structure attributes key in characterising overmature temperate forests. It gives the first robust statistical estimate of the proportion of overmature forests and may serve to report on their status in metropolitan France. Our method could be implemented at a larger spatial scale, notably in countries with accessible National Forest Inventory and calibration data, to produce indicators at international level.