Land-use change usually results in a landscape consisting of fragmented remnant habitats surrounded by a matrix of more-or-less intensively managed agricultural, monocultural forest or developed land. During the last couple of decades there has been an increasing awareness that the landscape surroundings, i.e. the matrix, influence the plant community and the diversity of focal habitats (Cousins & Aggemyr 2008; Jules & Shahani 2003; Murphy & Lovett-Doust 2004). Matrix quality, e.g. management intensity or the amounts of remnant habitats left in the landscape, as well as size and age of the focal habitats, may influence species composition and ecological assembly (Jamoneau et al. 2012; Aragon & Morales 2003; Morgan & Farmilo 2012). Even though habitat age and management continuity are such important factors for species composition in forest remnants (Peterken & Game 1984), there is a lack of landscape or community ecology studies that use historical spatial data (Cousins 2009). When species disperse between patches in a fragmented landscape, habitat edges play a particularly important role as they are often the first ‘filter’ that organisms encounter (Wiens 1992; Fagan et al. 1999). In the agricultural landscape, forest edges can be sharp, with a short transition zone, or soft, with a broad heterogeneous successional zone. In the latter case, the forest border can contain habitat specialists from varying types of habitat, e.g. from both semi-natural grassland and closed deciduous forest. Also, new forests created by afforesting agricultural fields, where forests have been cleared in the past, might provide important habitats for the colonization of typical forest specialists, provided that they are still left somewhere in the landscape. Remnant habitats in managed landscapes are important for many reasons, both as buffers and as potential facilitators of dispersal between remnant and restored habitats, but there is still a lack of studies on how the matrix quality affects plant species distribution and dispersal in fragmented landscapes.
In this issue of the Journal of Vegetation Science, Chabrerie et al. (2013) investigate plant species richness and turnover in forest edges in different agricultural landscapes. Vascular plants were recorded along 54 transects in different forest edges where local environmental, landscape and historical effects on species composition were analysed. Beside recording local environmental variables, such as light intensity, litter thickness and percentage of bare soil, at increasing distances from the forest edge, the past and present landscape were analysed in a 1-km radius around each transect. More than 300-yr-old historical maps and recent aerial photographs were used to determine edge age and the amount of different land-cover types, including linear habitat elements (hedgerows, roads) in the matrix.
The authors found that species richness decreased with increasing management intensity (grazing or ploughing) in the matrix and with distance from the edge. Proximal disturbance, both mechanical and chemical, and the lack of seed sources in more intensively managed landscapes clearly influence the forest edge community. Interestingly, the age of the forest edge had a negative effect on species richness, as the decline of species richness with increasing distance from the edge was even more pronounced in old edges than in younger ones (Fig. 1). Although the forest edge age negatively impacted total species richness, so-called “ancient forest species” increased (Verheyen et al. 2003; Fagan et al. 1999), as light- and nitrogen-demanding species were progressively replaced by more shade-tolerant, mesotrophic species with increasing distance from the edge. In young forests and in intensively-managed landscapes, forest edges became a sharper transition zone, hosting species of lower conservation value (Chabrerie et al. 2013). This highlights the importance of looking beyond species numbers into plant species identity, as areas with many species might still be of little conservation value for that particular habitat type.
As the authors point out, many landscape and local variables are intrinsically correlated. Nonetheless, the authors' effort to incorporate variables from different geographical and temporal scales is essential for understanding the complexity of ecological patterns. Few who are not familiar with the techniques will understand the labour intensity required to incorporate historical data, when such data just become a ‘simple’ variable or constraint in the final analysis and results. However, historical spatial data are crucial for our understanding of ecological processes in the landscape. Based on their results, Chabrerie et al. (2013) determine that two forces, matrix management intensity and age, regulate forest edge plant communities in fragmented agricultural landscapes. Depending on management in the matrix, older forest edges can contribute to conserving threatened forest specialists in an agricultural landscape because they contain more ancient forest specialists.
The study by Chabrerie et al. (2013) provides an excellent example where landscape ecology, plant community ecology and geography are incorporated to answer relevant questions regarding ecological processes and the management of remnant habitats in our everyday managed landscapes. The Journal of Vegetation Science welcomes more good theoretical and empirical studies that bridge the gaps between broad- and fine-grained ecological studies.