**1. Introduction**

Understanding the impact of land occupation (land use/cover) on basin processes, such as rainfall-runoff and soil erosion, is an integral part of land and water management-oriented decisions [1,2]. The processes of soil erosion and sediment transport take part in van Rijn's (1993) [3] sedimentary cycle, and often are the main causes of soil loss in basins [4]. Although these processes are of natural origin [5], the interaction between climate, soil, topography, land use, and land cover significantly influences erosion rates and sediment loads [6,7]. Soil loss due to these processes is a frequent problem that hydrologists, land planners, and basin managers will need to contend with [8]. Accordingly, soil erosion quantification in erosion-prone areas, with the highest accuracy possible, provides a complete knowledge of soil loss hotspots and allows prioritized treatment measures for supporting land processes

in the concerned basin [9]. Subsequently, the reliable assessment and representation of sediment yields—which depend on the cascading effect of soil erosion—allows an in-depth understanding of the soil erosion-sediment link at the basin scale [9].

For representing both processes, several models have been developed and used extensively to replace the conventional assessment methods, i.e., the Water Erosion Prediction Project (WEPP) [10,11], the Universal Soil Loss Equation (USLE) [12], the Modified Universal Soil Loss Equation (MUSLE) [13], the Revised Universal Soil Loss Equation (RUSLE) [14,15] and the Soil and Water Assessment Tool (SWAT) [16]. SWAT is one of the most widely used basin models. It has been applied extensively in modeling the impact of land occupation changes, under different scenarios and different contexts [17,18]. The widespread use of SWAT can be justified by its sensitivity and flexibility towards the land occupation input [19], its adaptability to different contexts—even to those with data scarcity [20]—its simple data requirements and ease of computation [21], as well as the straightforward calibration through its stand-alone SWATCUP interface [22].

Despite abundant research regarding the impact of land occupation on soil loss, few studies focus on the particular case of small water bodies and their effect as a land occupation class [23]. Small water bodies, like ponds and wetlands, are considered as the most amplified form of human-induced modifications to the hydro-sedimentological system of basins [24]. Ponds represent a total of over 90% of global standing water bodies, 30% of global standing waters by surface area [25], and form the most widespread aquatic habitat dominating the continental standing waters in Europe [26]. Despite their well-documented significance [27], abundant numbers, and increasing proliferation [28], ponds have not received considerable scientific attention compared to rivers and lakes [26]. It is worth mentioning that research regarding ponds in Europe has tripled in the last decade [29], where results showed that ponds contribute significantly to several basin related processes [30]. Examples of these processes are sediment interception [31], removal of pollutants for river protection [32], nutrient recycling [33], greenhouse gas emission [34], regulation of hydrological flows [35], biogeochemistry [30], and climate [36]. In addition to their environmental role, ponds have a well-known value for housing and sustaining biodiversity, supporting livelihoods, local economies, and taking part in the socio-cultural heritage of the settings in which they are located [37].

Under the hydro-sedimentological scope, particularly, ponds have shown to retain as much as 90% of sediments transported in basins [38]. Consequently, ponds have been heavily blamed for rupturing the ecological and sedimentary continuum of the basins to which they belong [39]. The disruptive effect of ponds is due to the increase of residence time of waters, resulting in a decline in the temporal variation of the main discharge [40]. Accordingly, the deceleration of overland flow allows suspended particles to settle under the effect of their weight, causing a reduction in the amount of sediments entrained by water, making ponds sediment sinks [40]. However, this effect strongly depends on their position in the basin, their depth, volume, slope [41], as well as the surrounding land occupation [42]. Winfield Fairchild and Velinsky (2006) [43] showed that ponds located upstream of rivers—considering their sediment retention capacity—are capable of creating a state of imbalance in the geochemical and hydro-sedimentary status of the underlying rivers. Consequently, the Directive Cadre sur l'Eau (DCE) [44] stresses the need to assess the impact of hydromorphological elements that are capable of influencing hydrologic pathways, river morphology, width, and continuity.

Beyond the contribution of isolated ponds, connected networks of ponds were found to contribute to basin processes at higher rates than lakes or even rivers [45,46]. Particularly in France, ponds are mainly concentrated in three regions: the Sologne region, Brenne (Central France), and Dombes (Eastern France). In response to DCE recommendations, this study aims to assess the impact of man–made ponds on soil erosion and sediment transport, at the scale of the Indre portion of the Claise basin. This part corresponds to the Brenne Natural Regional Park that houses 4500 waterbodies (ponds, marshes, and small water surfaces), 2179 of which are located in Claise, being part of an interconnected network. To evaluate erosion risks in the Claise, the Coordination of Information on the Environment CORINE (1992) Erosion Risk model will be used, since it presents a simplification of the reliable USLE

model [1] and given that no-erosion field data, for the Claise, were available for the study. SWAT is employed to assess the impact of ponds on the Claise's hydro-sedimentary regime. This choice is owed to SWAT's ability to simulate the physical processes that occur in ponds, which in turn allows an accurate representation of pond containing basins [38]. This is achieved through the SWAT Pond (.pnd) input file that makes SWAT one of the few hydrological models having an input for ponds [47]. Related studies have assessed the behavior of ponds using SWAT [38,48] and highlighted the efficiency of SWAT on this part.

The impact of the Claise's ponds on erosion and sediment transport will be assessed by testing alternative scenarios, where the land occupation input for both models will be simulated with and without ponds. By this approach, a quantification of the pond impact can be obtained. The presented work serves as a decision-oriented tool for basins similar to the Claise, where pond proliferation has been halted until a proper understanding of their effect is established. In addition, analysis of soil erosion risks and sediment transport is useful for conservation measures that aim towards prolonging the useful life of these small water bodies or for ceasing their proliferation.
