Le rôle des ongulés dans les forêts tempérées actuelles. Une réponse à Fløjgaard et al. (DOI: 10.1111 / gcb.14029)

[Departement_IRSTEA]Territoires [ADD1_IRSTEA]Dynamique et fonctionnement des écosystèmesInternational audienceIn Boulanger et al. (2018), we investigated the effects of ungulates on forest plant diversity. By suggesting a revisit of our conclusions regarding ecosystem dynamics since the late Pleistocene, Fløjgaard et al. (2018) came to the conclusion that moderate grazing in forest should be a conservation target. Since major points of our paper were mis‐ or over‐ interpreted, we put the record straight on our study system and on the scope of our conclusions. Finally, we advocate for an assessment of the conservation issues of ungulates in forests not only regarding hypothetical and still debated states of past ecosystems but also considering timely challenges for forest ecosystems. In the study by Boulanger et al. (2018), we assessed the role of ungulates' presence on short‐term understory plant dynamics, using a nationwide network composed of paired exclosure‐control plots. We demonstrated that ungulates are key drivers of understory vegetation in managed forests since they alter the richness and abundance of the shrub layer, increase the richness of the herb layer, and favor open habitat plant species. Our results show that local scale mechanisms, here involving indirect facilitation, lead to global dynamics in forest‐ungulate systems. Our conclusions underline that an increase in species richness, often simplistically viewed as a positive change in ecosystems, can alter the community composition and thus require a more subtle interpretation regarding conservation issues. Fløjgaard et al. (2018) questioned them from a macroecological and evolutionary perspective with respect to general nature conservation issues. They argued that an "evolutionary informed baseline" would change our conclusions based on short‐term and experimental features. We need to specify three over‐ or misinterpretations. 1. Although ungulates affect many groups of organisms, depending or not on the vegetation layer, our results were strictly limited to plant species. 2. Our investigations were not restricted to "ungulate grazing" (Fløjgaard et al., 2018), since the ungulate species generally co‐occurring at our study sites are not only obligate grazers (Gordon, 2003): roe deer is a browser, red deer an intermediate grazer-browser, and wild boar a frugivore-omnivore. Moreover, our interpretations included combined effects of different ecological processes in which sympatric forest ungulates are involved, notably herbivory, zoochory, and physical engineering (Persson, Danell, & Bergström, 2000; Wilby, Shachak, & Boeken, 2001). 3. Fløjgaard et al. (2018) argued that our results were conditional on "mainly plantation" forests, in "highly managed landscapes." Our sampling design does not include intensively managed forest plantations isolated in an agricultural landscape matrix, but even‐aged stands mainly issued from natural regeneration, located in forest‐dominated landscapes, and mostly with a long history of forest continuity (Table 1). While Fløjgaard et al. (2018) consider that the landscape context may bias our results, recent independent (Jaroszewicz, Pirożnikow, & Sondej, 2013; Picard, Chevalier, Barrier, Boscardin, & Baltzinger, 2016), but also older studies (von Oheimb, Schmidt, Kriebitzsch, & Ellenberg, 2005; Schmidt, Sommer, Kriebitzsch, Ellenberg, & von Oheimb, 2004) on seed dispersal by wild ungulates suggest that there is no overall effect of the landscape matrix on the traits of the plants dispersed by our three model ungulates; these animals disperse preferentially plants typical of open habitats even in forest‐dominated landscape matrices. These exchanges of views revive the unsolved and current debate on the role of ungulates as drivers of Anthropocene forest dynamics, as potential contributors to forest biodiversity in primary or secondary forests and more globally as dynamic actors of ecosystem functioning through direct, indirect, trophic, and nontrophic cascading effects. More specifically, the openness of Pleistocene forest stands and the primary habitat of light‐demanding species in ancient landscapes remain a matter of discussion (Mitchell, 2005). Consequently, from a conservation point of view, it seems unwise to recommend a uniform "moderate grazing" baseline without stronger scientific evidence and without considering current and future socioeconomic issues (Webster et al., in press)