Field measurements showed that the\u00a0150 surveyed dams retain\u00a0a\u00a0total of\u00a031,321\u00a0m\u00b3 of\u00a0surface water, with 89 dams in\u00a0the\u00a0K\u0159emeln\u00e1 catchment retaining 12,389\u00a0m\u00b3 across an\u00a0area of\u00a0169 km\u00b2. The\u00a0distribution of\u00a0retained volumes exhibits pronounced right-skewness (median\u00a041.0\u00a0m\u00b3, mean\u00a0208.8\u00a0m\u00b3), confirming that the\u00a0vast majority of\u00a0dams retain\u00a0relatively small volumes, while a\u00a0small number of\u00a0large beaver ponds account for a\u00a0substantial proportion of\u00a0the\u00a0total retained volume \u2013 the\u00a0largest 3\u00a0% of\u00a0ponds account for 54\u00a0% of\u00a0the\u00a0total volume. A\u00a0similar distribution pattern was observed for the\u00a0inundated areas of\u00a0beaver ponds, with a\u00a0total area of\u00a098,586\u00a0m\u00b2. The\u00a0median\u00a0inundated area of\u00a0an\u00a0individual beaver pond was only 168.9\u00a0m\u00b2, compared with a\u00a0mean\u00a0of\u00a0657.2\u00a0m\u00b2, and the\u00a0four largest ponds accounted for 53\u00a0% of\u00a0the\u00a0total inundated area. Modelling in\u00a0the\u00a013.8\u00a0km\u00b2 Chl\u00edstovick\u00fd Stream catchment identified nearly 6 km of\u00a0watercourses suitable for beaver colonisation. Under full colonisation, the\u00a0model predicts the\u00a0formation of\u00a013\u201343 dams. Based on the\u00a0volumes measured at the\u00a0experimental sites, surface water accumulation in\u00a0the\u00a0Chl\u00edstovick\u00fd Stream catchment can\u00a0be expected to amount to hundreds or several thousand cubic metres. A\u00a0significant proportion of\u00a0retention, however, also occurs below ground, further increasing the\u00a0overall hydrological benefit. The\u00a0formation of\u00a0beaver ponds covering a\u00a0total area of\u00a0several thousand to tens of\u00a0thousands of\u00a0square metres could also provide an\u00a0important ecological benefit by creating opportunities for the\u00a0development of\u00a0wetland habitats. The\u00a0study demonstrates that beaver dams represent an\u00a0important component of\u00a0nature-based water retention in\u00a0the\u00a0landscape; however, their importance extends far beyond surface water accumulation. Beaver dams also contribute to sediment deposition, stabilisation of\u00a0runoff patterns, and increased ecological stability of\u00a0the\u00a0landscape, and should therefore be given adequate space in\u00a0the\u00a0landscape.<\/span><\/p>\n The\u00a0Eurasian\u00a0beaver is native to the\u00a0European\u00a0landscape and historically occupied extensive areas from the\u00a0British Isles to eastern Siberia. Intensive hunting, however, brought it close to extinction, and by the\u00a0end of\u00a0the\u00a019th century the\u00a0European\u00a0population was estimated at only about 1,200 individuals in\u00a0eight isolated populations\u00a0[1]. Species protection and reintroduction programmes have enabled its gradual return since the\u00a020th century, and by around 2020 the\u00a0European\u00a0population was estimated at approximately 1.5\u00a0million individuals\u00a0[2]. A\u00a0similar development occurred in\u00a0the\u00a0Czech Republic, where, following its complete extirpation, the\u00a0beaver was recorded again\u00a0in\u00a01977 at the\u00a0confluence of\u00a0the\u00a0Dyje and Kyjovka rivers, and its population has grown rapidly since the\u00a0beginning of\u00a0the\u00a021st century\u00a0[3, 4]. In\u00a02020, the\u00a0population in\u00a0the\u00a0Czech Republic was estimated at approximately 15,000 individuals\u00a0[5, 6].<\/span><\/p>\n Beavers are often described as \u2018ecosystem engineers\u2019 because of\u00a0their ability to modify their environment by building dams\u00a0[7]. As populations grow rapidly, this can\u00a0lead to conflicts with people, but also produces a\u00a0range of\u00a0positive effects. A\u00a0beaver dam is a\u00a0structure made of\u00a0natural materials that impounds a\u00a0watercourse. A\u00a0beaver pond forms upstream of\u00a0the\u00a0dam, comprising accumulated water that changes the\u00a0character of\u00a0the\u00a0watercourse from flowing to standing water. Dams and ponds often form cascades that cause water to spread across the\u00a0surrounding floodplain. These complexes subsequently form what is known as a\u00a0beaver wetland (Figs. 1 and 2<\/span>)\u00a0[6, 8]. Although dams and the\u00a0ponds they create primarily provide protection from predators and access to food\u00a0[9], they also have a\u00a0major influence on the\u00a0hydrological regime of\u00a0the\u00a0landscape. They slow water flow, promote sedimentation, enhance interactions between surface water and groundwater, and increase surface water and groundwater storage\u00a0[10\u201312].<\/span><\/p>\n The\u00a0benefits of\u00a0beaver activity, however, extend beyond water retention alone. Dam building contributes to the\u00a0restoration of\u00a0near-natural river corridors and the\u00a0natural structural diversity of\u00a0floodplains, promotes the\u00a0spontaneous renaturation of\u00a0technically modified watercourses, and reduces the\u00a0adverse effects of\u00a0agricultural and forestry drainage systems\u00a0[6, 13\u201315]. Newly created wetland habitats increase habitat diversity and biodiversity, aided by the\u00a0fact that beaver activity increases the\u00a0amount of\u00a0ecologically valuable dead wood in\u00a0watercourses\u00a0[6, 13, 16, 17].<\/p>\n The\u00a0principal focus of\u00a0this study is to determine the\u00a0volume of\u00a0surface water retained by beaver dams. The\u00a0vast majority of\u00a0studies addressing this topic come from North America and show that beaver dams substantially increase the\u00a0volume of\u00a0water stored in\u00a0watercourses and wetlands\u00a0[18\u201320]. Beedle\u00a0[18], for example, found that beaver ponds in\u00a0Alaska retained up to 37 times more water than\u00a0natural stream reaches, with individual pond volumes ranging from tens of\u00a0cubic metres to more than\u00a011,000\u00a0m\u00b3. Karran\u00a0et al.\u00a0[19] determined pond volumes of\u00a01\u20139,001\u00a0m\u00b3 at four sites in\u00a0North and South America. The\u00a0few European\u00a0studies available also confirm a\u00a0substantial retention effect of\u00a0beaver dams; for example, the\u00a0total volume of\u00a0water retained by 22 beaver dams in\u00a0the\u00a0Krzemianka catchment in\u00a0Poland was estimated at 15,000\u00a0m\u00b3\u00a0[21].<\/p>\n In\u00a0addition to retained volumes, this study also quantifies the\u00a0inundated area of\u00a0beaver ponds. This is important not only hydrologically but also ecologically, as larger beaver wetlands are associated with greater bird species richness and abundance and also increase the\u00a0availability of\u00a0lentic and littoral habitats within\u00a0the\u00a0floodplain\u00a0[22, 23].<\/p>\n Although beaver behaviour and the\u00a0development of\u00a0their territories cannot be predicted reliably, certain\u00a0parameters can\u00a0be used to assess the\u00a0suitability of\u00a0an\u00a0area for potential beaver colonisation\u00a0[24]. Research has demonstrated relationships between beaver occurrence and watercourse characteristics such as slope, channel width, and the\u00a0availability of\u00a0riparian\u00a0vegetation\u00a0[25]. These findings have been used to develop spatial models that predict the\u00a0capacity of\u00a0the\u00a0environment to support dam building\u00a0[26]. Such models are also used to estimate potential water retention in\u00a0the\u00a0landscape. For example, a\u00a0study in\u00a0the\u00a0Sierra Nevada, California, estimated that beaver dams could store up to 120\u00a0million m\u00b3 of\u00a0surface water while also helping to increase landscape resilience to fire\u00a0[27].<\/p>\n In\u00a0this study, precise geodetic methods were used to determine the\u00a0volume of\u00a0surface water retained by 150 beaver dams in\u00a0the\u00a0K\u0159emeln\u00e1 catchment and at ten other selected wetlands in\u00a0the\u00a0Plze\u0148 Region, predominantly in\u00a0the\u00a0\u0160umava Mountains, with the\u00a0aim of\u00a0contributing quantitative evidence to assessments of\u00a0the\u00a0importance of\u00a0beavers in\u00a0the\u00a0Czech landscape. The\u00a0inundated area of\u00a0the\u00a0beaver ponds was also evaluated. At the\u00a0same time, we modelled the\u00a0potential number of\u00a0dams that could develop in\u00a0a\u00a0selected fourth-order catchment following beaver colonisation, assuming minimal human\u00a0restriction of\u00a0beaver activity. These results were then used to quantify the\u00a0amount of\u00a0water that beavers could retain\u00a0in\u00a0the\u00a0catchment through their activities.<\/p>\n The\u00a0aim of\u00a0the\u00a0study was therefore to improve understanding of\u00a0the\u00a0positive effects of\u00a0beaver colonisation by quantifying surface water retention in\u00a0the\u00a0landscape, which is one of\u00a0the\u00a0prerequisites for effective management of\u00a0the\u00a0Eurasian\u00a0beaver and objective assessment of\u00a0conflicts associated with its occurrence. Surface water retention is only one benefit of\u00a0beaver dams and is linked to a\u00a0range of\u00a0other hydrological and ecological processes. In\u00a0addition to retaining water and sediment, beaver structures can, for example, help sustain\u00a0flows during dry periods, improve the\u00a0microclimate, create valuable wetland habitats, and increase the\u00a0amount of\u00a0dead wood in\u00a0watercourses\u00a0[11, 18, 28].<\/p>\n In\u00a0the\u00a0K\u0159emeln\u00e1 catchment in\u00a0the\u00a0\u0160umava Mountains, a\u00a0geographically and hydrologically coherent area containing 89 beaver dams at the\u00a0time of\u00a0the\u00a0survey in\u00a0spring 2023, and at ten selected beaver wetlands in\u00a0the\u00a0Plze\u0148 Region, precise field surveying was used to determine the\u00a0volumes of\u00a0surface water retained by the\u00a0dams and the\u00a0inundated areas of\u00a0the\u00a0beaver ponds. In\u00a0the\u00a0Chl\u00edstovick\u00fd Stream catchment, the\u00a0potential occurrence of\u00a0beaver dams was also modelled according to the\u00a0suitability of\u00a0individual watercourse reaches for dam construction.<\/p>\n The\u00a0study covered three areas, shown in\u00a0Fig.\u00a03<\/span><\/em>. Field measurements and determination of\u00a0the\u00a0actual volumes of\u00a0water retained by beaver dams and the\u00a0inundated areas of\u00a0their ponds were carried out at 89 beaver dams in\u00a0the\u00a0K\u0159emeln\u00e1 catchment and at ten beaver wetlands in\u00a0the\u00a0\u017delezn\u00e1 Ruda area, on the\u00a0\u00daslava river, and on the\u00a0Voldu\u0161sk\u00fd stream.<\/span><\/p>\n The\u00a0K\u0159emeln\u00e1 catchment was selected in\u00a0cooperation with \u0160umava National Park for the\u00a0study Field measurement and mapping of\u00a0beaver dams in\u00a0the\u00a0K\u0159emeln\u00e1 catchment<\/span>\u00a0[29] as an\u00a0example of\u00a0a\u00a0landscape in\u00a0which the\u00a0presence of\u00a0beavers is subject to only minimal human\u00a0restriction. All known beaver dams in\u00a0the\u00a0catchment were examined, from small ponds within\u00a0watercourse channels to extensive ponds forming wetlands on stream floodplains. The\u00a0research in\u00a0the\u00a0K\u0159emeln\u00e1 catchment therefore provided a\u00a0comprehensive picture of\u00a0the\u00a0characteristics of\u00a0beaver dams and ponds at the\u00a0regional scale.<\/span><\/p>\n The\u00a0ten selected wetlands in\u00a0the\u00a0Plze\u0148 Region were chosen in\u00a0cooperation with the\u00a0Nature Conservation Agency of\u00a0the\u00a0Czech Republic (NCA CR) for the\u00a0study Effects of\u00a0beavers on river-landscape ecosystems<\/span>\u00a0[30] as examples of\u00a0areas representing the\u00a0target state of\u00a0development of\u00a0a\u00a0watercourse and its immediate surroundings following beaver colonisation. The\u00a0wetlands contain\u00a0varying numbers, types, and sizes of\u00a0beaver dams and associated ponds, although on average the\u00a0ponds are generally larger than\u00a0those in\u00a0the\u00a0K\u0159emeln\u00e1 catchment.<\/span><\/p>\n The\u00a0third area, the\u00a0Chl\u00edstovick\u00fd Stream catchment, was used to model beaver dam capacity. This work was carried out as part of\u00a0the\u00a0project Water systems and water management in\u00a0the\u00a0Czech Republic under climate change<\/span>, in\u00a0which the\u00a0Chl\u00edstovick\u00fd Stream catchment was already being used for other modelling of\u00a0landscape water retention and rainfall\u2013runoff relationships and was therefore also selected for modelling potential beaver dam capacity.<\/span><\/p>\n Surveying and determination of\u00a0the\u00a0actual volume of\u00a0water retained by beaver dams and the\u00a0inundated areas of\u00a0beaver ponds were carried out in\u00a0the\u00a0K\u0159emeln\u00e1 River catchment in\u00a0\u0160umava National Park. The\u00a0study area, whose principal watercourse is the\u00a0K\u0159emeln\u00e1 river, more than\u00a030 km long, lies in\u00a0the\u00a0north-western part of\u00a0the\u00a0park, predominantly within\u00a0the\u00a0Czech Republic but extending marginally into Germany. The\u00a0catchment has a\u00a0total area of\u00a0approximately 169\u00a0km\u00b2, corresponding to about one quarter of\u00a0the\u00a0area of\u00a0\u0160umava National Park.<\/span><\/p>\n Orographic conditions result in\u00a0long-term above-average precipitation totals in\u00a0the\u00a0catchment (the\u00a0Pr\u00e1\u0161ily rain-gauge station recorded mean\u00a0annual precipitation of\u00a01,413.7\u00a0mm in\u00a01991\u20132020 period), resulting in\u00a0abundant water and a\u00a0dense hydrographic network. This includes both torrential reaches in\u00a0incised mountain\u00a0valleys and small meandering watercourses on upland plateaus.<\/span><\/p>\n Thanks to the high level of protection and historically limited human activity, the watercourses and their floodplains have retained a relatively natural character. These conditions provide suitable habitat for the Eurasian beaver,\u00a0<\/span>whose population is thriving and whose activities further increase the\u00a0water-retention capacity of\u00a0the\u00a0landscape.<\/span><\/p>\n The\u00a0\u0160umava National Park Administration has conducted long-term monitoring of\u00a0Eurasian\u00a0beaver activity in\u00a0the\u00a0area. Data from this monitoring were used to identify sites containing beaver dams, which were subsequently surveyed in\u00a0the\u00a0field.<\/span><\/p>\n One of\u00a0the\u00a0beaver dams in\u00a0the\u00a0K\u0159emeln\u00e1 catchment near the\u00a0settlement of\u00a0Mechov is shown in\u00a0Fig.\u00a04<\/span><\/em>.<\/span><\/p>\n Actual volumes of\u00a0water retained by beaver ponds were also determined at ten selected beaver wetlands in\u00a0the\u00a0Plze\u0148 Region. Eight were selected around \u017delezn\u00e1 Ruda, in\u00a0the\u00a0\u0158ezn\u00e1 and \u00dahlava catchments, the\u00a0latter adjoining the\u00a0K\u0159emeln\u00e1 catchment. These wetlands lie outside the\u00a0national park and are more affected by human\u00a0activity than\u00a0sites in\u00a0the\u00a0K\u0159emeln\u00e1 catchment, although nature conservation is strengthened by their location within\u00a0the\u00a0\u0160umava Protected Landscape Area.<\/span><\/p>\n The\u00a0remaining two wetlands studied were on the\u00a0\u00daslava River near the\u00a0village of\u00a0Pr\u00e1dlo and on the\u00a0Voldu\u0161sk\u00fd stream near Litohlavy. Both sites lie outside protected areas in\u00a0a\u00a0typical Czech agricultural landscape. While the\u00a0wetland on the\u00a0\u00daslava developed in\u00a0open countryside, the\u00a0wetland on the\u00a0Voldu\u0161sk\u00fd stream is partly within\u00a0woodland. The\u00a0ten wetlands studied contained a\u00a0total of\u00a061 individual beaver ponds, often directly adjoining one another.<\/span><\/p>\n One of\u00a0the\u00a0four dams forming the\u00a0beaver wetland on the\u00a0\u00daslava river is shown in\u00a0Fig.\u00a05<\/span><\/em>.<\/span><\/p>\n Potential beaver dam occurrence was modelled in\u00a0the\u00a0Chl\u00edstovick\u00fd Stream catchment. This is a\u00a0fourth-order catchment covering 13.8 km\u00b2 and is predominantly agricultural in\u00a0character. It is located in\u00a0the\u00a0Central Bohemian\u00a0Region near the\u00a0village of\u00a0Male\u0161ov. The\u00a0principal watercourse is the\u00a0Chl\u00edstovick\u00fd stream, a\u00a0left-bank tributary of\u00a0the\u00a0Vrchlice river, which it joins near Male\u0161ov approximately 3 km upstream of\u00a0the\u00a0upstream boundary of\u00a0the\u00a0Vrchlice Reservoir.<\/span><\/p>\n The\u00a0volume of\u00a0water retained by existing dams at the\u00a0experimental sites and the\u00a0inundated area of\u00a0the\u00a0beaver ponds were determined by analysing the\u00a0geodetic survey outputs using GIS tools.<\/span><\/p>\n Geodetic surveying of\u00a0beaver dams and ponds in\u00a0the\u00a0K\u0159emeln\u00e1 catchment was carried out in\u00a0spring 2023, while the\u00a0beaver wetlands in\u00a0the\u00a0\u017delezn\u00e1 Ruda area, on the\u00a0\u00daslava river, and on the\u00a0Voldu\u0161sk\u00fd stream were surveyed in\u00a0spring 2024 and 2025. Three high-precision GPS instruments were used. Emlid Reach RS+ and Emlid Reach RS2 GPS receivers were used at all sites, with a\u00a0CHC X91 GPS receiver additionally used in\u00a0the\u00a0K\u0159emeln\u00e1 catchment.<\/span><\/p>\n A\u00a0total of\u00a089 beaver dams and their inundated areas were surveyed in\u00a0the\u00a0K\u0159emeln\u00e1 catchment. For the\u00a0great majority of\u00a0dams, three points on the\u00a0dam were surveyed (left abutment, right abutment, and centre of\u00a0the\u00a0crest), together with the\u00a0channel bed downstream of\u00a0the\u00a0dam, the\u00a0pond depth upstream of\u00a0the\u00a0dam, and the\u00a0upstream boundary of\u00a0inundation. Several dams with complex pond shapes or dam alignments were surveyed in\u00a0greater detail. <\/span><\/p>\n For five selected dams, the\u00a0ponds, including water depths, were also surveyed in\u00a0detail to validate the\u00a0GIS method used to determine retained volumes (Fig.\u00a06<\/span><\/em>). A\u00a0total of\u00a01,445 points were surveyed in\u00a0the\u00a0K\u0159emeln\u00e1 catchment.<\/span><\/p>\n At the\u00a0ten selected wetlands in\u00a0the\u00a0\u017delezn\u00e1 Ruda area, on the\u00a0Voldu\u0161sk\u00fd stream, and on the\u00a0\u00daslava river, a\u00a0total of\u00a061 beaver dams and their inundated areas were surveyed. Dam polylines, inundation boundaries, and water depths in\u00a0the\u00a0ponds were measured. Owing to the\u00a0morphological conditions, these were generally larger ponds than\u00a0those in\u00a0the\u00a0K\u0159emeln\u00e1 catchment. A\u00a0total of\u00a03,716 points were surveyed at these sites.<\/span><\/p>\n Based on the\u00a0field GPS measurements described in\u00a0the\u00a0previous section, GIS analyses were applied at the\u00a0two experimental study areas to determine the\u00a0volume of\u00a0water retained by beaver dams. The\u00a0approaches used share the\u00a0principle of\u00a0combining precise geodetic data with a\u00a0digital terrain\u00a0model (DTM), while their specific implementation was adapted to the\u00a0characteristics of\u00a0the\u00a0individual sites and the\u00a0data available.<\/p>\n For the\u00a089 dams surveyed in\u00a0the\u00a0K\u0159emeln\u00e1 catchment in\u00a0spring 2023, a\u00a0GIS-based volume-determination method was used that combined beaver dam lines surveyed by high-precision GPS with a\u00a0detailed digital terrain\u00a0model representing conditions before the\u00a0ponds formed. Dam lines were derived from the\u00a0geodetic survey, and a\u00a0high-accuracy, fine-resolution DTM capturing pond-bed morphology was used (a\u00a050 \u00d7 50 cm LiDAR DTM from 2017, produced for \u0160umava National Park by German\u00a0partners as part of\u00a0a\u00a0joint cross-border project).<\/p>\n In\u00a0the\u00a0GIS environment, the\u00a0dam lines were converted to raster format and used to determine the\u00a0water level in\u00a0each inundated area as the\u00a0highest terrain\u00a0elevation along the\u00a0beaver dam line (usually at the\u00a0left or right dam abutment). The\u00a0contributing catchments were then delineated so that individual ponds could be assessed separately even where they immediately adjoined one another in\u00a0cascades. The\u00a0extent of\u00a0each inundated area was subsequently defined as the\u00a0intersection between the\u00a0contributing catchment and areas with terrain\u00a0elevations below the\u00a0specified water level.<\/p>\n Volume was calculated by comparing the water-surface elevation with the digital terrain model within the delineated inundated area, from which\u00a0the\u00a0spatial distribution of\u00a0water depths was derived. Total water volume was then calculated as the\u00a0sum of\u00a0the\u00a0partial volumes corresponding to individual raster cells. The\u00a0procedure was implemented as a\u00a0Python script in\u00a0ArcGIS\u00a0Pro, enabling efficient processing of\u00a0a\u00a0large number of\u00a0dams and generation of\u00a0accompanying raster outputs.<\/p>\n The\u00a0method was validated using detailed geodetic surveys of\u00a0five dams. For these dams, water depths within\u00a0the\u00a0inundated area and the\u00a0perimeter and crest of\u00a0the\u00a0dam were surveyed in\u00a0detail. Volume was calculated using the\u00a0Thiessen polygon method and compared with the\u00a0value obtained from GIS analysis of\u00a0the\u00a0dam lines and terrain\u00a0model. The\u00a0comparison confirmed that the\u00a0method was suitable for determining the\u00a0volume of\u00a0surface water retained by beaver dams.<\/p>\n The\u00a0method was modified slightly to determine the\u00a0volume retained by the\u00a061 dams within\u00a0beaver wetlands in\u00a0the\u00a0\u017delezn\u00e1 Ruda area, on the\u00a0\u00daslava river, and on the\u00a0Voldu\u0161sk\u00fd stream, for two reasons. First, at some of\u00a0these sites the\u00a0dams had created extensive wetlands with several water-surface levels across the\u00a0wetland floodplain, which the\u00a0original method could not represent. Second, the\u00a0inundation boundaries and water depths had been surveyed in\u00a0greater detail, and the\u00a0aim was to incorporate these values into a\u00a0more accurate calculation. To make optimal use of\u00a0the\u00a0survey outputs and account for variable water levels, a\u00a0modified method was applied combining point measurements of\u00a0water depth with the\u00a01 \u00d7 1\u00a0m DMR 5G digital terrain\u00a0model.<\/p>\n Water-surface elevation at individual points was derived from the\u00a0measured depths and corresponding elevations in\u00a0the\u00a0digital terrain\u00a0model and was then interpolated into a\u00a0continuous water-surface raster. Comparing this surface with the\u00a0terrain\u00a0model yielded the\u00a0spatial distribution of\u00a0water depths and subsequently the\u00a0water volume and inundated area of\u00a0individual ponds. The\u00a0method thus maintains high accuracy where direct measurements are available while minimising uncertainties in\u00a0sparsely surveyed areas where terrain\u00a0is difficult to access or water is too deep for the\u00a0survey team to work safely.<\/p>\n Based on models developed abroad, a\u00a0beaver dam capacity model adapted to Czech data sources was created in\u00a0ArcGIS Pro. The\u00a0model classifies individual watercourse reaches according to the\u00a0potential number of\u00a0beaver dams that could develop following colonisation. Each reach is assigned a\u00a0category together with an\u00a0estimate of\u00a0the\u00a0minimum and maximum expected number of\u00a0dams.<\/p>\n The\u00a0beaver dam capacity model used for the\u00a0Chl\u00edstovick\u00fd Stream catchment is based on a\u00a0British model derived from the\u00a0American\u00a0BRAT (Beaver Restoration Assessment Tool) model. The\u00a0American\u00a0model was developed by a\u00a0research team led by William W. Macfarlane\u00a0[26], subsequently validated repeatedly at several sites in\u00a0North America\u00a0[14, 31], and used in\u00a0numerous studies\u00a0[27, 32, 33]. For European\u00a0conditions, the\u00a0model was adapted by a\u00a0team led by Hugh A. Graham of\u00a0the\u00a0University of\u00a0Exeter\u00a0[24], and the\u00a0model applied to the\u00a0Chl\u00edstovick\u00fd Stream catchment is based on this adaptation.<\/p>\n Graham et al.\u00a0[24] present the\u00a0BFI (Beaver Foraging Index) and BDC (Beaver Dam Capacity) models, which were developed, validated, and applied in\u00a0the\u00a0British catchments of\u00a0the\u00a0Tay, Otter, and Coombeshead. The\u00a0BFI model classifies the\u00a0availability of\u00a0vegetation for food and dam construction in\u00a0two key zones: immediately adjacent to the\u00a0bank (10\u00a0m) and within\u00a0a\u00a0wider riparian\u00a0zone (40\u00a0m). BFI outputs are subsequently used as one of\u00a0the\u00a0inputs to the\u00a0BDC model, which also analyses longitudinal channel slope and stream power at both low (Q80% \u2013 80% quantile) and high (Q2% \u2013 2% quantile) flows, thereby assessing the\u00a0ability of\u00a0beavers to build dams and the\u00a0dams\u2019 potential resistance to floods. The\u00a0model integrates these variables using a\u00a0fuzzy inference system (FIS). The\u00a0principal BDC output is a\u00a0classification of\u00a0reaches into theoretical maximum-capacity categories expressed as ranges in\u00a0the\u00a0number of\u00a0dams per kilometre of\u00a0watercourse.<\/p>\n Graham et al.\u00a0[24] validated the\u00a0model by comparing its predictions with field-survey data covering 258 km of\u00a0watercourses and 89 identified beaver dams.<\/p>\n The model used for the Chl\u00edstovick\u00fd Stream catchment retains the same parameters and fuzzy-logic principles as the British model but differs in its input data, some of which in the original version were specific to Great Britain. A further modification was the introduction of two threshold values: maximum watercourse width and minimum Q80% flow. According to Hartman et al. [34], beavers build dams almost exclusively on watercourses less than 6 m wide. Although watercourses of this width do not occur in the Chl\u00edstovick\u00fd Stream catchment, the threshold serves to exclude small ponds and reservoirs. The second threshold was minimum flow, reflecting the fact that beavers build\u00a0dams only on perennial watercourses, whereas intermittent watercourses are unsuitable. Watercourse reaches that failed to meet either threshold were automatically classified by the model\u00a0as \u2018none\u2019, i.e. with zero capacity for beaver dam construction. The\u00a0individual parameters and input data used are summarised in\u00a0Tab.\u00a01<\/span><\/em>.<\/span><\/p>\n Application of\u00a0the\u00a0beaver dam capacity model divided individual watercourse reaches in\u00a0the\u00a0Chl\u00edstovick\u00fd Stream catchment into four categories according to the\u00a0potential density of\u00a0beaver dams following colonisation of\u00a0the\u00a0catchment. Each category represents a\u00a0range between the\u00a0minimum and maximum number of\u00a0beaver dams per kilometre that can\u00a0be expected following beaver colonisation of\u00a0a\u00a0watercourse. The\u00a0resulting categories and ranges of\u00a0expected beaver dam density were adopted from Graham et al.\u00a0[24] and MacFarlane et al.\u00a0[26]:<\/span><\/p>\n The\u00a0150 beaver dams at the\u00a0experimental sites retain\u00a0a\u00a0total of\u00a031,321\u00a0m\u00b3 of\u00a0water. Measurements of\u00a0the\u00a0volumes of\u00a0surface water retained by the\u00a0dams show a\u00a0strongly right-skewed distribution with high variability. Mean\u00a0volume is 208.8\u00a0m\u00b3, median\u00a0volume 41.0\u00a0m\u00b3, and the\u00a0standard deviation 681.5\u00a0m\u00b3, demonstrating considerable variation in\u00a0beaver pond size. Recorded volumes range from 0.02 to 6,583\u00a0m\u00b3. High skewness (6.83) and kurtosis (53.78) confirm the\u00a0pronounced asymmetry of\u00a0the\u00a0distribution and the\u00a0presence of\u00a0extreme values, while a\u00a0normality test showed a\u00a0statistically significant departure from a\u00a0normal distribution (p < 0.001).<\/span><\/p>\n The\u00a0largest pond forms part of\u00a0the\u00a0beaver wetland on the\u00a0\u00daslava river and retains 6,583\u00a0m\u00b3 of\u00a0water. The\u00a0entire wetland, formed by four dams, stores a\u00a0total of\u00a09,991\u00a0m\u00b3 of\u00a0water. The\u00a0distribution of\u00a0water depths within\u00a0this wetland is shown in\u00a0Fig.\u00a07<\/span><\/em>.<\/span><\/p>\n The data show that most beaver dams retain relatively small volumes of\u00a0water, while only a small number of sites accumulate substantially larger volumes and have a major influence on the overall distribution of the data.<\/span><\/p>\n Figs. 8<\/span><\/span><\/em> and 9<\/span><\/em> show that the\u00a0five largest ponds, representing 3\u00a0% of\u00a0the\u00a0total number, stored 54\u00a0% of\u00a0the\u00a0total water volume. Conversely, the\u00a0smallest 80\u00a0% of\u00a0ponds, each with a\u00a0capacity of\u00a0up to 200\u00a0m\u00b3, retained only 15\u00a0% of\u00a0the\u00a0total volume.<\/span><\/p>\n This pattern is accentuated by the\u00a0inclusion of\u00a0several extensive beaver wetlands, particularly the\u00a0wetland on the\u00a0\u00daslava river. It nevertheless remains pronounced when the\u00a0dataset is restricted to the\u00a0K\u0159emeln\u00e1 catchment, a\u00a0geographically and hydrologically coherent area in\u00a0which retained volumes were determined for all known beaver dams. Here too, the\u00a0dataset of\u00a089 dams shows a\u00a0strongly right-skewed, non-normal distribution. Mean\u00a0volume is 139.2\u00a0m\u00b3, median\u00a0volume 33.9\u00a0m\u00b3, and the\u00a0standard deviation 404.9\u00a0m\u00b3. Skewness (6.14) and kurtosis (41.27) again\u00a0confirm the\u00a0strong asymmetry of\u00a0the\u00a0data and the\u00a0influence of\u00a0several extreme values.<\/span><\/p>\n In\u00a0the\u00a0K\u0159emeln\u00e1 catchment, the\u00a0largest pond accounts for 26\u00a0% of\u00a0the\u00a0total retained volume, specifically 3,274\u00a0m\u00b3 out of\u00a012,389\u00a0m\u00b3. The\u00a0largest 10\u00a0% of\u00a0dams retain\u00a0approximately two thirds of\u00a0the\u00a0total water volume. On average, beaver dams in\u00a0the\u00a0K\u0159emeln\u00e1 catchment retain\u00a073.3\u00a0m\u00b3 of\u00a0water per square kilometre of\u00a0catchment area.<\/span><\/p>\n The\u00a0results show that the\u00a0great majority of\u00a0dams at the\u00a0experimental sites retain\u00a0only small volumes of\u00a0water: 73\u00a0% of\u00a0dams had pond volumes of\u00a00\u2013100\u00a0m\u00b3. These volumes are not substantial in\u00a0terms of\u00a0the\u00a0overall total. Instead, a\u00a0small proportion of\u00a0extensive ponds that store thousands of\u00a0cubic metres of\u00a0water plays the\u00a0decisive role. Although small ponds make only a\u00a0limited contribution to total retained volume, they may be highly important for biodiversity and the\u00a0other related aspects discussed above.<\/span><\/p>\n The\u00a0150 beaver dams at the\u00a0experimental sites create a\u00a0total inundated area of\u00a098,586\u00a0m\u00b2. As with retained water volumes, the\u00a0inundated areas of\u00a0the\u00a0beaver ponds show a\u00a0strongly right-skewed distribution, dominated by smaller ponds with a\u00a0small number of\u00a0extremely extensive ponds. Mean\u00a0inundated area is 657.2\u00a0m\u00b2, whereas the\u00a0median\u00a0is only 168.9\u00a0m\u00b2, with recorded values ranging from 0.5 to 27,941\u00a0m\u00b2. High skewness (8.84) and kurtosis (85.73) confirm the\u00a0strong influence of\u00a0extreme values on the\u00a0distribution; compared with retained water volumes, the\u00a0higher values of\u00a0these statistics indicate an\u00a0even greater dominance by a\u00a0few very extensive beaver ponds. The\u00a0distribution of\u00a0inundated beaver pond areas is clearly shown in\u00a0Fig.\u00a010<\/span><\/em>.<\/span><\/p>\n The\u00a0largest inundated area was recorded for the\u00a0same pond on the\u00a0\u00daslava river that also had the\u00a0greatest retained water volume, reaching 27,941\u00a0m\u00b2. The\u00a0entire beaver wetland, formed by four beaver dams, has a\u00a0total inundated area of\u00a041,740\u00a0m\u00b2, representing more than\u00a040\u00a0% of\u00a0the\u00a0total inundated area in\u00a0the\u00a0dataset (the\u00a0\u00daslava river wetland shown in\u00a0Fig.\u00a011<\/span><\/em>).<\/span><\/p>\n The\u00a0results show that, as with retained water volumes, the\u00a0total inundated area is strongly influenced by a\u00a0few of\u00a0the\u00a0largest ponds. The\u00a0four largest ponds, each with an\u00a0inundated area exceeding 4,000\u00a0m\u00b2 and representing only 2.7\u00a0% of\u00a0the\u00a0sites studied, account for 53\u00a0% of\u00a0the\u00a0total inundated area.<\/span><\/p>\nINTRODUCTION<\/h2>\n
![\"\"](\"https:\/\/www.vtei.cz\/wp-content\/uploads\/2026\/10\/Babuljak-obr-1.jpg\")
<\/a>\nFig. 1. Drone image of a beaver wetland on the Litohlavsk\u00fd stream, showing two beaver dams and associated ponds forming a cascade, and a wetland that developed on the right bank of the stream (photo: T. Laburda)<\/h6>\n
<\/a>\nFig. 2. Beaver wetland on the Litohlavsk\u00fd stream, viewed upstream; the main beaver dam (right, among the trees) has caused water to spread beyond the stream channel, creating\u00a0a wetland (left and centre) (photo:\u00a0A.\u00a0Babuljak)<\/h6>\n
METHODOLOGY<\/h2>\n
Study areas<\/h3>\n
<\/a>\nFig. 3. Study areas \u2013 left: K\u0159emeln\u00e1 catchment; centre: ten selected beaver wetlands in the Plze\u0148 Region; right: Chl\u00edstovick\u00fd catchment (basemaps: Basic\u00a0Topographic Map of\u00a0the\u00a0Czech Republic, Czech Office for Surveying, Mapping and Cadastre; \u010c\u00daZK)<\/h6>\n
K\u0159emeln\u00e1 catchment<\/h3>\n
<\/a>\nFig. 4. One of the largest beaver dams and associated ponds in the K\u0159emeln\u00e1 catchment, forming part of the wetland near the settlement of Mechov (photo: A. Babuljak)<\/h6>\n
Ten beaver wetlands in\u00a0the\u00a0Plze\u0148 Region<\/h3>\n
<\/a>\nFig. 5. One of\u00a0the\u00a0four dams that make up the\u00a0beaver wetland on the\u00a0\u00daslava river near the\u00a0village of\u00a0Pr\u00e1dlo (photo: A. Babuljak)<\/h6>\n
Chl\u00edstovick\u00fd Stream catchment<\/h3>\n
Evaluation of\u00a0retained volumes and inundated areas<\/h3>\n
Surveying<\/h3>\n
<\/a>\nFig. 6. Surveying of\u00a0beaver ponds and dams \u2013 left: GPS-based surveying and water-depth measurement; centre: detailed survey of a beaver pond; right: simplified survey of beaver dams and inundation boundaries (photo: J.\u00a0Uhl\u00edkov\u00e1; basemap: Orthophoto of\u00a0the\u00a0Czech Republic, Czech Office for Surveying, Mapping and Cadastre; \u010c\u00daZK)<\/h6>\n
GIS tools<\/h3>\n
Modelling potential beaver dam occurrence<\/h3>\n
Tab.\u00a01. Input parameters for\u00a0the\u00a0beaver dam capacity model<\/h5>\n
<\/a>\n\n
RESULTS<\/h2>\n
Actual volume retained by beaver dams at\u00a0the\u00a0experimental sites<\/h3>\n
<\/a>\nFig. 7. Water depths in\u00a0a\u00a0beaver wetland on the\u00a0\u00daslava river formed by four dams and\u00a0their associated beaver ponds (basemap: Orthophoto of\u00a0the\u00a0Czech Republic, Czech Office for Surveying, Mapping and Cadastre; \u010c\u00daZK)<\/h6>\n
<\/a>\nFig. 8. Water volumes retained by beaver dams at the experimental sites<\/h6>\n
<\/a>\nFig. 9. Contribution of different pond size classes to the total retained volume<\/h6>\n
Inundated area of\u00a0beaver ponds at the\u00a0experimental sites<\/h3>\n
<\/a>\nFig. 10. Inundated areas of beaver ponds at the experimental sites<\/h6>\n
<\/a>\nFig. 11. Beaver wetland on the \u00daslava river, which accounts for more than 40 % of the total area of all 150 beaver ponds (photo: T. Laburda)<\/h6>\n