Trajectories of land-system change

Spatial overlay between LSAs and ACTs

Cross-tabulating the spatial patterns of Land-System Archetypes and Archetypical Change Trajectories (Table SI V-5) revealed that the largest spatial overlap existed between stability (ACT17) and low-intensity forest (LSA12) with approximately 600,000 km². Stability had the largest spatial overlap with the majority of LSAs (11 out of 15), i.e. that these LSAs did not experience changes in comparison to the situation in 1990 (Figure V-4, upper panel). Only few LSAs were clearly influenced by only one dominant ACT, such as

expansion (ACT16) on urban built-up (LSA15), and to a lesser degree intensification of wood production (ACT11) on high-intensity forest (LSA11). Interestingly, cropland-grassland conversion (ACT14) co-occurred with all three intensity levels of cropland management (LSA03-05).

We found seven ACTs that led to one dominant LSA in 2006 (Figure V-4, lower panel), e.g. the intensification to intensity cropland (ACT03) that was associated with high-intensity arable cropland (LSA03) in 2006, and four ACTs that were mostly related to one LSA. The remaining ACTs did not exhibit any clear trends into specific LSAs, and most LSAs were thus influenced by multiple ACTs or remained stable compared to 1990.

Figure V-4: Spatial coverage [%] of each ACT per LSA (upper panel) and of each LSA per ACT (lower panel). Rows (upper) and columns (lower) sum up to 100% spatial extent. Circle sizes and colour gradient depict the magnitude of co-occurrence.

Land-System Archetypes and explanatory factors

We compared spatial patterns in Land-System Archetypes with those of explanatory factors (Table SI V-4 and Figure SI V-2). Managed land systems pertaining to croplands (LSA01-05), grasslands (LSA07-09), and forests (LSA11) were generally characterised by good accessibility and low terrain ruggedness. For these areas, arid climatic conditions were observed on less-intensively managed croplands and especially for permanent crops.

Aridity indices were above-average for managed grasslands and forests. Growing degree days (GDDs) for permanent crop areas and low-intensity cropland exceeded the study area average. On managed land systems, soil organic carbon was above-average only on managed grasslands. Socio-economic factors were generally high for all managed land systems, especially for croplands and forests. However, we observed decreasing trends of socio-economic values with less intensively managed areas (e.g., economic size, labour or capital input, and subsidies on croplands or economic size on grasslands). Contrasting with this, labour input exhibited an increasing trend with less-intensive grassland management.

High-intensity mosaics (LSA13), which were characterised by multiple and mostly intensively managed land systems, were similar to intensively managed croplands (LSA03-05) in terms of socio-economic factors, accessibility, and terrain ruggedness.

Land systems that were not managed showed distinct differences to managed land systems.

Fallow farmlands (LSA06) and semi-natural areas (LSA 10 and 12) revealed above average travel time, terrain ruggedness, aridity index, and soil organic carbon while growing degree days and most of the socio-economic factors were below the study area average. Two LSAs with distinct characteristics regarding their co-occurrence with explanatory factors were low-intensity mosaics (LSA14) that generally did not reveal marked deviations from the study area averages and urban built-up (LSA15) that exhibited the best accessibility, the lowest terrain ruggedness, and the highest economic activity among all LSAs. Most LSAs were mainly located in the Continental and Central Atlantic zone, except permanent crop cultivation areas (Mediterranean zones) and low-intensity forests (Boreal zone) (Figure SI V-4, top). Protected areas had the highest shares in low-intensity LSAs, especially on fallow farmland and low-low-intensity grassland areas. All agricultural areas revealed markedly lower shares of protected areas (Figure SI V-5, top).

We found substantial variability across LSAs in terms of explanatory factors. Aridity index and growing degree days revealed the lowest variability across LSAs whereas economic activity and population density showed the highest variability.

Archetypical Change Trajectories and explanatory factors

Comparing Archetypical Change Trajectories with explanatory factors (Table SI V-4 and Figure SI V-3) showed that stability (ACT17) was generally observed where conditions were close to the study area average for all investigated explanatory factors. Intensifying land systems (ACT01-04 for croplands and ACT11 for forests) were generally characterised by good accessibility, low terrain ruggedness, and above-average socio-economic conditions, except for areas with low-level fertiliser intensification. Along with these general patterns, we also observed land-use specific conditions. ACTs related to fertiliser intensification (ACT03-04) were found in regions with below-average soil organic carbon contents, travel time to major cities, and aridity indices. For areas of intensifying wood production (ACT11), especially subsidies and labour input were above the study area average.

De-intensifying land systems exhibited a more nuanced picture regarding the co-occurrence of ACTs and explanatory factors. De-intensifying croplands (ACT05-07) and grasslands (ACT08-10) were found to be associated with good accessibility and below-average terrain ruggedness, likely due to their former function as high- or medium-intensive croplands, while aridity indices and growing degree days were close to the study area average. Soil organic carbon was above average for both livestock-related ACTs (ACT09-10). Socio-economic conditions were generally unfavourable (e.g., low economic activity, high labour input) on areas of declining grassland yields (ACT08). Areas of low-level de-intensification of livestock farming (ACT10) were associated with generally above-average socio-economic factors, differing from average conditions on de-intensifying croplands and high-intensity livestock areas.

Land systems that underwent land-use conversions were characterised by specific environmental and socio-economic conditions. Forests expanding over grasslands (ACT12) were associated with below-average socio-economic conditions as well as unfavourable accessibility and terrain ruggedness. Forest loss for agricultural expansion (ACT13) was found in regions characterised by below-average aridity indices, subsidies, and capital input as well as above average growing degree days. Conversions from cropland to grassland (ACT14) were associated with below-average terrain ruggedness and travel time while socio-economic factors were generally above or close to the study area average. Areas where permanent crops declined (ACT15) were characterised by all socio-economic factors, soil organic carbon contents, aridity indices, and travel time being below average. Most ACTs were mainly located in the Continental zone, except permanent

cropland loss and deforestation for agricultural expansion (Mediterranean zones) and de-intensification of cropland (Pannonian zone) (Figure SI V-4, bottom). Protected areas had the highest shares in forest-related ACTs and were markedly low on croplands and in expanding urban areas. Areas of yield increases reveal relatively high shares of protected areas (Figure SI V-5, bottom).

4 Discussion

Disentangling the complexity of pattern, trajectories, and driving factors of land-use change is a major challenge in land-system science. In this regard, identifying high-level, archetypical patterns and trajectories in land systems is an important step for assessing the outcome of land change across larger regions and for a more context-specific, regionalised policy making. We mapped and characterised Land-System Archetypes (LSAs) and Archetypical Change Trajectories (ACTs) for Europe for the period between 1990 and 2006, using land-change indicators pertaining to both land-use extent and intensity, and overlaid these archetypes with a range of explanatory factors of land change.