4.2 Human impact on SY
Our results indicate that the cropland fraction in a catchment is significantly negatively correlated with the SSL trend (Spearman r = −0.64, p < 0.001). This is likely indicating that theSSL decline in highly cultivated basins tends to be sharper than the corresponding sediment load under pristine conditions. Therefore, this study opts for a cropland fraction as a proxy of landuse impact on suspended sediment load. One may argue that not only arable land but also other anthropogenic land use types (e.g., pasture) can affect the suspended sediment yield. However, previous studies (Vanmaercke et al. , 2015) suggest that land use impacts on suspended sediment load are mainly caused by agriculture (i.e., arable land and permanent crops).
Nevertheless, the influence of anthropogenic activity on the sediment flux of the Caucasus rivers can be multidirectional in the coming decades. On the one hand, the construction of new ski pistes contributes to an increase in sediment export from the catchment area. This is also facilitated by the construction of hotel complexes and other infrastructure in resort areas. On the other hand, a decline of pressure on mountain pastures due to a livestock amount decrease is typical for this part of the Greater Caucasus. For example, land abandonment is also common for the mid-mountainous and low-mountainous zones in many regions of the Mediterranean (Rodrigo-Comino et al. , 2018). As a rule, these were arable lands, and therefore, in the first years after the abandonment of lands, an increase in sediment runoff occurred. However, as the slopes were overgrown, the sediment runoff sharply decreased (Lizaga et al. , 2018).
There is some evidence that cropland and glacier change have minimal impact on the rivers regulated with HPP. On the other hand, they possibly have influenced suspended sediment fluxes of other rivers. We can suppose that most of the SSD reduction in the Mal-Ka, Kam-Ol, Cheg-Nc was related to cropland changes between 1987 and 2015, while the precipitation regime changes have a weaker effect (see Table 5 ). Cropland area was much more stable in the Ard-Ta, Bel-Ko, Che-Ba, and Uru-Kh basins, suggesting the leading role of deglaciation inSSD reduction (precipitation change explains up to 22% ofSSD reduction, cf. Table 5 ) in combination with other climatic factors. In the Sunzha basin (Sun-Br and Sun-Gr), both cropland and glacier were much more variable, and the amount of cultivated cropland dropped significantly from 2010 to 2015. While precipitation changes explain only 7% of the SSD reduction, the other 93% are mostly connected to the land reforms (cropland decrease) and glacier shrinkage.
At the same time, in the foothill zone, after a slight decline in agriculture in the 1990s, there is a gradual expansion of the area of cultivated land. In particular, the number of orchards is increasing and the area under cultivation of raw crops. In addition, the creation of a network of small ponds and reservoirs in the foothill zone contributes to the redeposition of a part of the sediment coming from the slopes into permanent streams. The influence of dams on the sediment runoff of mountain rivers is typical for many mountain areas. Most of the rivers in the Alps are regulated by a system of dams, which significantly affects the trends in the change in sediment runoff and does not allow the influence of climatic changes on them to be revealed with the necessary detail (Costa et al. , 2018). Also, it can be hypothesized that suspended sediment yield flux will continue to decline as the number of hydropower plants increases worldwide and in North Caucasus in particular (Fel’dman, 1985; Zarfl et al. , 2015).