Since the end of World War II, marginal European landscapes with unfavourable environmental conditions for cultivation have experienced severe land-use changes. In many cases, large portions of arable land have been successively abandoned in favour of grassland or fallow land. This general trend of marginalisation in turn affected ecological landscape functions and processes with far-reaching consequences for biodiversity and natural resources. Furthermore, the land-use pattern in marginal European landscapes is expected to undergo further major changes in the future, particularly in the course of EU agricultural policy. Given this background, this multiple-paper thesis (A) analysed agricultural landscape change in a marginal agricultural landscape and (B) developed methods that may support landscape change research at multiple spatio-temporal scales. Both aims were addressed in three studies. Our study area was the Lahn-Dill Highlands (1270 km²), a marginal agricultural landscape in Hesse (Germany) with a pronounced land-use change in the past decades.
In the first study, we developed an approach to identify types of land-cover patterns and dynamics (TLPDs) at the rural district scale. By the combination of recent satellite data with historic agricultural statistics, and the application of k-means cluster analysis, we identified six TLPDs and characterised their physical settings. We found a general trend of abandonment of cultivation at the landscape scale, which is governed by significant differences in current land-cover patterns and the directions of land-cover change at the district scale: In the eastern part of the area, where elevation is low, and the proportions of steep slopes and dry soils are small, land cover remained relatively stable. Slight to dramatic changes occurred, in contrast, in the remaining districts with comparatively unfavourable conditions for cultivation.
In the second study, we developed a 3-step methodological approach to systematically assess the spatial distribution of grassland age in a marginal agricultural landscape. The approach is based on the combination of an a-priori two-stage landscape stratification with conventional aerial photograph interpretation of selected patches, and the subsequent spatial extrapolation of the determined grassland age. Results proved that our approach provides a realistic estimation of grassland age at the scale of districts and over a time period of five decades. We found that the derived probabilities of grassland age classes are specific for grassland types in areas with a homogenous pattern of land-cover change. Furthermore, the results indicated a predominance of old grassland patches (>47 years). Occurrences of mid-aged grassland (18-47 years) were concentrated in districts with a pronounced land-cover change in this time period, whereas young grassland (<18 years) is apparently evenly distributed across the study area.
In the third study, we analysed the potential effects of three alternative transfer payment schemes on the farmland habitat diversity in a marginal agricultural landscape. We defined (1) a scenario with direct transfer payments coupled to production, (2) a scenario with direct transfer payments decoupled from production, and (3) a scenario phasing out all direct transfer payments. We characterised habitat diversity with three indices: habitat richness, evenness, and rarity. The habitat pattern in 1995 served as reference for comparison. All scenarios predicted a general trend of homogenisation of the farmland habitat pattern, yet to a differing extent. Transfer payments coupled to production (Scenario 1) supported spatially segregated land use with fallow land primarily in low-productive areas and arable land use in the more productive sites. The scenario predicted intermediate values for habitat richness and habitat evenness. Decoupling transfer payments from production (Scenario 2) favoured grassland as the most profitable farming system. This led to a grassland-dominated landscape with low values of all habitat diversity indices. Phasing out transfer payments (Scenario 3) resulted in complete abandonment or afforestation of agricultural land and extremely low values in all habitat diversity indices. Scenario results revealed that the payment of subsidies may prevent cessation of agricultural production, but may not fully counteract the homogenisation in marginal landscapes.