Author:

Keywords:

human impact, erosion, soil, modelling, AquaCrop, Watem Sedem, carrying capacity, crop yield

Abstract:

Humans have impacted their environment throughout history, especially since the introduction of widespread agriculture and the associated forest logging activities. For the Mediterranean region the idea existed that the soil erosion following the cultivation of land degraded the landscape to such an extent it caused crisis in ancient societies. Our research has focused in trying to find sound scientific evidence or counterarguments for this hypothesis. In order to quantify the impact of ancient societies on the landscape a simple water balance driven crop yield model is coupled to a soil erosion model. The soil erosion model was validated based on a detailed 4000 year long chronology of sediment deposition in a small mountainous catchment of Gravgaz (1250 m a.s.l.) in the territory of the ancient city of Sagalassos (SW Turkey). The historic sediment dynamics are modeled reasonably well, with a model efficiency of 0.75 and a relative root mean squared error of 0.23, based on 250 year averages. The model is capable of simulating the soil erosion phase after major deforestation from the Iron Age onwards, as well as the depletion of soil reservoirs on limestone lithologies on the hillslopes and the resulting decrease in sediment delivery towards the central valley around the Roman period. The crop yield model that is integrated with the soil erosion model allows 4 main input parameters: 1) Climate, 2) Crop specifications, 3) Field management (Irrigation, stone bunds, mulching) and 4) Soil (Soil profile & groundwater). Combining modelled soil depths and the crop yield model, 3 main conclusions pop out of the analysis: 1) Importance of ancient soil erosion and soil thickness on crop yield Although the catchment-wide average crop yield decrease over time, the central valley bottoms still allow high yields (~3 ton/ha/yr) for winter barley due to an accumulation of sediments and an increase in soil thickness. Spatial variation of crop yield indeed strongly mimics soil thickness, which results in an almost full loss of crop yield on the hillslopes from 2400BP onwards after the major forest clearance from the Early Bronze Age onwards. 2) Climate generally less important compared to soil thickness Analysis show how soil thickness plays a threshold role for crop yield around the ~50-100cm soil thickness range. During phases of soil erosion and deposition in the past, especially in a mountainous Mediterranean environment, soils are prone to pass the ~50-100cm soil thickness range, in both directions. The effect of climate on crop yield is most visible on the more shallow soils for summer barley, due to the water scarcity during summer. But generally, the climatical effect only imposes a small variation on top of this strong soil thickness effect. 3) The model as a tool to reveal need for field management in the past Applying a simple water balance model of Budyko (Woodward et al., 2015), we calculated that deforestation in the Gravgaz catchment would have resulted in a 10% increase in runoff. This extra runoff would have found its way to the central marsh area, increasing the area of the marsh, and hence decrease the area available for agriculture and ultimately also decrease crop yield. A substantial part of the catchment with the high yields (around 3 ton/ha/yr) this way disappears. This example, amongst others, shows how the model allows us to quantify the effect of human environment interactions on agricultural carrying capacity. Furthermore, depending on the size of the population that had to be sustained in the past, decreases in crop yield could have resulted in a need for water management to sustain crop yield over areas with a depleting soil reservoir. The model can give an indication of when, where and how high this need for adaptive field management techniques (irrigation, manuring, crop rotation,…) was. Although the modeled processes are kept rather simple, and some model fine-tuning is still necessary, these novel methods of quantifying agricultural carrying capacity should allow to nuance traditional views of direct links between landscape degradation and societal crisis, and open the debate on the higher resilience of these societies in the past.