A feedback loop between management, intraspecific trait variation and harvesting practices.

Intraspecific variation in plants is a major ecological mechanism whose local determinants are still poorly understood. In particular, the relationship between this variation and human practices may be key to understanding human-nature relationships. We argue that it is necessary to consider how human practices both influence and depend on the phenotypic variability of species of interest. Arnica montana (arnica) is a good model to study the complex interactions between human actions and plant phenotype, as (i) its ecological niche is shaped by human management actions and (ii) its variability has consequences for harvesters. Using a functional trait approach, we examined feedback loops linking management actions, plant phenotype and harvesting practices. In 27 sites in southeastern France, we measured vegetative and reproductive functional traits of arnica of interest for harvesters, and recorded management actions (grazing; mowing) and ecological variables (including height of surrounding vegetation and tree cover). We examined their effects on plant traits with linear mixed models and used path analysis to test if the effects of human management on traits are mediated by the height of surrounding vegetation. Management actions affected functional traits of arnica. Biomass removal practices (grazing, mowing) were associated with smaller plants producing smaller leaves with reduced specific leaf area. We uncovered the core role of the height of surrounding vegetation in determining this phenotype. Tree cover was associated with reduced flowering. The observed intraspecific variation in response to management actions differentially impacts the two main harvesting practices. Flower-head harvesting depends on reproductive traits that are not impacted by mowing (which is done in winter) but adversely affected by tree cover. In contrast, traits associated with large biomass under tree cover or with high surrounding vegetation are favourable for whole-plant harvesters. Our trait-based approach unveiled clear links between management actions and plant phenotype, with impacts on both vegetative and reproductive traits. These changes induced by management also affect the practices of harvesters. We thus demonstrated a feedback loop between human actions and plant phenotype and provided a novel perspective on human-related causes and consequences of plant intraspecific variability.