Factor Assessment of Pond Silting in Forest-Steppe Agrolandscapes of the Central Russian Upland, European Russia

This paper presents the results of assessing the influence of siltation factors in 23 ponds in one of the most agriculturally developed macro-regions of European Russia. Key natural and anthropogenic factors determining the intensity of pond siltation have been identified, and a typification of ponds has been developed to predict the rate of accumulation of bottom sediments in them. For the typification, statistical methods such as correlation analysis (Spearman's coefficient), cluster and factor analysis, as well as the Random Forest machine learning algorithm were used. Correlation analysis revealed that the percentage of catchment cultivation has a significant effect (0.55) on the volume of bottom sediments, while soil loss (0.47) and vertical terrain dissection (0.43) have a moderate effect. Forest cover (0.23) has a weak effect. The most important factors in the siltation process are the average slope of the catchment (24.5%), the percentage of soil cultivation (18.8%), and the average annual soil loss (14.1%). All factors were grouped into three clusters, which explained 77.8% of the variance. As a result, four pond types were identified, differing in their dominant limiting factors: pond hydrological characteristics, catchment morphometry, and the degree of anthropogenic transformation of the catchment. Verification of the typification was carried out based on the calculation of annual soil losses considering the sediment delivery coefficient; the discrepancies between the calculated and actual pond sediment volumes were 1.2–10.0%. The proposed approach allows, based on remote sensing and mapping data, to effectively identify the most degraded small water bodies and plan restoration measures without the need for costly field surveys at the initial stage.