Distribution of soil properties along forest-grassland interfaces: Influence of permanent environmental factors or land-use after-effects?

Soil properties vary spatially according to land use; both because land users have selected specific soil properties for specific land uses, and land uses modify the soil properties. However, permanent environment factors and land-use effects are unlikely to display the exact same spatial patterns. Study of the spatial and historical patterns of distribution of soil properties could help to separate between these two causes. In this aim, we studied 22 forest-grassland interfaces with controlled historical configurations in northeast France. In each land use (forest and grassland), three distances to the edge (edge, periphery and core) and two land-use histories (ancient and recent) were studied. Along forest-grassland interfaces, forests were usually located slightly upslope of grasslands, and mainly because this non-random topographic position the topsoil texture was significantly more silty in forests, and clayey in grasslands. After statistically controlling for the effects of topography and soil texture, we observed two main gradients of variation in soil properties according to the distance-to-edge (acidity in forest and nutrient content in grassland). In forest, pH and Ca dropped from the edges to the peripheries (15 m distance), while in grassland, C, N, P and Na sharply increased from the edges to the cores (25 m distance). These results demonstrate, through the edge effect, the strong influence of the land use on a part of soil properties. Furthermore, less than two centuries after grassland afforestation or deforestation, we observed that soil properties in recent forests and recent grasslands were respectively closer to their current land use than to their former land use. These results demonstrate a rapid change in soil properties after land-use change. However, recent forests and recent grasslands kept a legacy of soil texture from their former land use, respectively. Recent grasslands also kept a lower soil density, N and Na content compared to ancient grasslands. Hence, this study of forest-grassland interfaces show strong and short-scale relationships between land use and soil properties and suggest that they express both original choices of land users for specific soil properties and land-use after-effects. The non-random topographic position of the forest-grassland interfaces indicates a conscious choice of this positioning by the land users, for agronomic reasons. Beyond that, land use, through vegetation composition and management practices, also has a strong impact on soil properties. The fact that land-use changes affect most soil properties after only a few decades confirms the existence of land-use effects over time.