The representation and analysis of agricultural landscape mosaics has been an important step for landscape ecology in recent years (1). However, the use of broad (often aggregated) landcover groupings limits our understanding about how different species respond to variations in landscape heterogeneity.
Semi-natural habitats and linear features are known to provide important resources and functional connectivity for biodiversity (2). However, these are rarely analysed as separate habitat types. Nor are they represented in a manner that captures their different ecological value or ‘quality’ for particular species; e.g. a mature hedge can provide very different ecological functions compared with a flailed hedge, an orchard or patch of heathland. This limits our understanding about where, and how best to implement practical conservation measures.
We analysed the effect of landscape heterogeneity (composition and configuration) on woodland carabid communities across 36 2 x 2 km agricultural landscape mosaics in southern Britain. We categorised hedgerows based on their vegetation structure from in-situ surveys and aerial imagery (H1-H3; Fig. 1). We also classified different types of semi-natural habitat such as scrub encroachment, orchard, managed heath and rough grassland (Fig. 2). Other habitats in the mosaic were represented using high-resolution landcover data.
Our results showed three main findings. 1) By considering different types of semi-natural habitat, we could explicitly identify those which provided the greatest ecological value for different species. 2) The best information about how species respond to the landscape was derived from those where we incorporated a measure of habitat ‘quality’. 3) The ability for species to actually utilise the benefits provided by a habitat was highly dependent on the spatial configuration of the landscape. For example, ‘mature hedgerows with trees’ (H3) were significantly important for flightless, slow dispersing carabids in landscapes with a high degree of woodland fragmentation, but ‘managed hedgerows’ (H1) were not of perceptible value to these species, even where they acted as connections between woodland patches.
Considering habitats as separate components rather than broad landcover groups and incorporating qualitative habitat indicators alongside quantitative landscape metrics conveys a deeper understanding about the ecological processes acting at the local and landscape scale. There is great scope to delve deeper into within-patch heterogeneity in European wooded-agricultural landscapes. This will help us to ensure that conservation actions take account of habitat types, how they are managed and where they are located in a landscape in order to provide the greatest benefits to biodiversity.
