Landownership is an unexplored determinant of forest understory plant composition in Northern France

Few studies have investigated the influence of landownership on biodiversity. Therefore we analysed how the presence of forest understory plant species varied according to landownership in a broad geographical context and assessed which plant traits discriminated between species associated with state, other public or private forests. We also quantified the degree to which differences in soil type, forest management and landscape structure among ownership categories could explain the differences in species composition. Landownership, climate, soil, forest stand and landscape variables were collected on 38,751 plots located in temperate forests (5.1 Mha) in northern half of France using the French National Forest Inventory and GIS analyses. First, logistic regressions were used to determine species response to landownership after controlling for month of the plant survey and spatial autocorrelation. Relationships between plant traits and species association with landownership were then tested. Second, climate, soil, forest stand and landscape differences among ownerships were investigated. Third, species and trait responses to landownership were re-examined after controlling for month of the plant survey, spatial autocorrelation and climate, soil, forest stand and landscape variations. Of the 276 species, 69 were associated with state forests, 38 with other public forests and 85 with private forests. Species associated with state and other public forests were more often urbanophobic, ancient-forest species, barochores and myrmecochores whereas private forest species were more often nutrient-, light-demanding, urbanophilic and endozoochorous species. Differences among landownerships were detected for all the soil, climate, forest management and landscape variables investigated. However, the ownership effect remained or became significant for 131 plants after controlling for environmental variability. Landownership is not simply an administrative classification of land without ecological signification but was evidenced as a driver shaping understory plant community composition and plant traits in Northern France. These differences were partly explained by soil, management, edge and patch size effects, but land use history may help to explain the residual landownership effect. Our results have major implications on biodiversity monitoring and large-scaled conservation strategies. There is a need for deeper investigation into the impacts of land policy on biodiversity.