Rushes as a phytoindicator of a lake’s surrounding land use

Rushes belong to the group of plants commonly found in water bodies, fulfilling many important functions. However, their presence depends both on features of the immediate surroundings and morphology of the lake. At the same time, rush presence can be an indicator of important features of the surrounding habitats. The aim of this study was to determine the impact of shoreline and buffer zone development on the attributes of rushes. Indications of lake features affected the rushes most strongly. We analysed 65 lakes located in the eastern part of Poland, within the boundaries of the Łęczna-Włodawa Lake District. The multidimensional analyses identified factors that most strongly determined the development of lake rushes in Polesie Lubelskie. The presence of peat bogs, scrubland communities, and wet meadows favoured the development of rushes. Among the morphometric characteristics of lakes, the depth and frequency of water mixing were significant, as well as the trophic level of the lake. Moreover, models describing scenarios for further development of rushes were developed. They predicted that the surface area of the rushes in the lake will increase (average of 0.288 ha) when the peatlands become overgrown 1 ha, and the share of dry (drained) meadows will increase (average of 0.996 ha) in the buffer zone. The increase in the density of rushes will take place if the share of buildings and scrub communities increases both in the buffer zone and shoreline area.


Introduction
The lake and its surroundings create an ecological system with a dynamic character. The allochthonous matter is supplied from the outside by surface run-off with rainfall or underground sources, which enrich the reservoir with organic substances (Gutry-Korycka 1999;Kallf 2001;Wetzel 2001;Søndergaard et al. 2003). The lake system itself creates a mutually regulating mechanism based on the close relationship between biocenosis with its internal interactions and a biotope representing elements of the inanimate environment (Esch and Fernandez 2013;Wetzel 2001).
An important element in the assessment of the limnological condition of lakes is its trophic level, which determines the abundance of nutrients (Zacharias et al. 2002). The magnitude of the pollutant load can affect the lake in multiple ways and vary with lake morphometry, catchment area, and land use (Grochowska and Gawrońska 2005;Soszka 2010). Anthropogenically transformed areas, especially for agricultural purposes areas, are the main source of nutrients for water reservoirs (Misztal et al. 1992). Agriculture changes the landscape and chemistry of the environment, which is why it is listed as the first in global assessments of the causes of the impoverishment of diversity (Kędziora et al. 2012;Boulanger et al. 2018;Adhikari et al. 2019). Tourism and recreation are also a source of nutrients that affect the productivity of lake ecosystems. Many authors claim that this impact is small (Bajkiewicz-Grabowska 1987;Kalff 2001;Ławniczak et al. 2010); however, the supply of even small amounts of nutrients affects the water quality and may destabilise the basic components of the lake's biocenosis, especially the coastal and littoral vegetation (Hadwen et al. 2005;Soszka 2010).
Nutrients are also indirectly determined by hydro-technical modifications, regulating water relations in the catchment area. Irrigation projects in the form of regulating riverbeds, digging drainage ditches, and including or excluding lakes from the outflow network cause fundamental changes in shaping the balance of nutrients supplying water reservoirs (Wilgat et al. 1991;Sender 2018).
The interaction between a lake and a shoreline is very important. The last and only 'dam' that can counteract a negative impact of a lake surroundings is often shaped in the coastal zone. A buffer zone is defined as an area with vegetation or a strip of terrestrial vegetation adjacent to the reservoir. Its vegetation naturally filters pollutants that arise in its surroundings and then, flows into the lake, protecting against erosion, and is an important place for aquatic insects, microorganisms, fish, and other animals that help maintain the biological balance in the aquatic ecosystem. Unfortunately, many lakes are deprived of a natural coastal zone as a consequence of the construction of seasonal cottages, permanent residences, farms, and other land uses by humans (Fischer and Fischenich 2000).
Shrubs, coastal vegetation, and rushes constitute the last barrier capable of capturing biogenic substances of various origins generated due to the nature of the lake's surrounding development, which can be called a phytogeochemical barrier (Caracciolo et al. 2014;Vodyanitskii and Shoba 2016;Boulanger et al. 2018). Limiting the presence of a buffer zone in the form of shrubs or its absence in the coastal zone means that the rushes are the first to be subjected to the influence of the environment.
Their area, range, density, and composition in relation to the type of buffer zone development may also give preliminary information about the ecological status of lake water. The prevalence of rushes is another criterion determining their usefulness in biodiagnostics of lake health.
Due to the progressing processes of overgrowing and degradation of lakes, research was carried out in the Łęczyńsko-Włodawskie Lake District, which is a homogenous region with a large trophic and morphometric diversity of lakes. The object of the research was rushes, which can be phytoindicators of land development. So far, no research in this aspect has been carried out for this region, and the developed models can be the basis for forecasting the development of rushes for lakes with similar conditions. The aim of this study was to determine the influence of the nearest coastal zone development and a slightly more distant buffer zone on features describing the rushes (density, depth, area, number of species, share of rushes in the lake surface, width of the rushes belt). Therefore, in all morphologically and trophically diversified lakes of Łęczna-Włodawa Lake District, analyses of the buffer and coastal zone management were carried out, as well as studies of the rushes inhabiting the littoral zone of these lakes. Attributes that may have an impact on rushes development were also considered, such as the decline of a buffer zone, the source feeding the lake, the type of mix, development of the shoreline, surface and depth of the lake. Multidimensional analyses (multivariate principal component analysis) identified factors that most strongly determine the development of lake rushes in Łęczna-Włodawa Lake District.
The authors proposed models (multiple-step regression with backward elimination) describing the relationships between the studied indicators and describing scenarios for the further development of rushes, depending on the use of lake surrounding habitats. Significant information was sought about the interrelationship of the analysed variables with the features of rushes and indication of the best management method as an effect of colonisation by rushes.

Study area
Polesie Lubelskie is the western part of the great region of Western Polesie; it occupies a part of the Vistula and Bug inter-rivers. In terms of administration, it is located in the north-eastern part of the Lublin Province. The peculiarity of the area is its flatness, with very shallow groundwaters and a large variety of lakes, marshes, peat bogs, and wetlands. Lakes constitute a distinctive element of the hydrographic network of the Łęczna-Włodawa Plain (acc. Kondracki 2002) subregion. Lakes are characterised by slow water exchange and represent various hydrochemical types (Appendix Table 7). There are both glacial lakes and lakes with karst origin. Within the Lake District area, there are 67 lakes, 61 of which have retained their natural character, whereas 6 have been converted into storage reservoirs (Michalczyk et al. 2011) (Fig. 1). Our cartographic and field studies let us identify sixty-five lakes, located within the borders of the Łęczna-Włodawa Lake District.
In general, the width of the buffer zone should be adequate to the manner and intensity of use in the areas surrounding the reservoir, the slope of the land, and soil type. The study presents management analyses in relation to the shoreline and the 200 m buffer zone. Such width is considered optimal for the performance of its functions (bank stability, maintenance of the benthic community, nutrient reduction, sediment removal, and wildlife habitat) (Muskoka Watershed Council 2013).
In 2017-2019 (from April to October), research was carried out to determine the depth, area, method of supply, and fertility of lakes in Łęczna-Włodawa Lake District. Most of the parameters were analysed during field studies. In some lakes, data on their depth and mix type were obtained from the literature (Harasimiuk et al. 1998).

Methods of floristic study
Measurement of helophyte density, colonisation depth, and species composition and number were collected from vegetated belts within the lake area. Maximum depth of rushes occurrence and GPS location were determined using sonar Elite no. 5 of LOWRENCE.
Aquatic transects radiated inward towards the centre of the lake for the length of helophyte occurrence and were uniformly distributed along the entire coastline every 100 m for small lakes and 250 m for larger ones (Appendix Table 7). Macrophyte density was measured using a floristic frame with a sampling area of 0.25 m 2 placed randomly along transects. Helophyte belts were measured using a measuring tape and the rangefinder BOSCH DLE 50 Professional.
We monitored physical parameter of water [Secchi disc visibility (SD)] twice a year in the spring and summer. At a similar time, the phosphorous content [Total phosphorous (TP)] was obtained according to the methods of Hermanowicz et al. (1976), using the microchip Slandii photometer LF 205. The Carlson trophic state index (Carlson TSI) was obtained by considering the TP and SD (Carlson 1977).
The following floristic indicator was used in the biocenotic analysis of macrophytes density: (1) Species density index (G) (Kasprzak and Niedbała 1981) where G-species density (pcs/m 2 ), n g -species number (n), S-sampling area (m 2 ).

Morphological methods
(2) The index of shoreline development (K) was expressed by the formula: where K-shoreline development, L-length of the lake shoreline (m), A-lake surface (m 2 ). An index in the range of 1.0-1.5 indicated low diversification of the shoreline, and a value above 3.01 meant a very varied shoreline.
(3) Another factor that significantly affects the surface run-off rate is terrain decline (S), which was calculated based on the formula: where S-terrain decline, Δh-difference in height between the end points of the drop line expressed in metres, L-terrain length of the section of the slope line calculated based on the measurement of this section on the map according to the map scale.

Cartographic methods
Data regarding management types with the buffer zone in lake surroundings were collected from areas reaching approximately 200 m from the lakeshore outward, as well as all shorelines. The percentage cover of specific land use types (urban, i.e. buildings; arable, i.e. agricultural fields and natural forests and meadows) was calculated within the surveyed area. Littoral zone areas of the lakes and their buffer zone were determined by the interpretation of aerial photos, orthophotomaps, and topographic maps. Orthophotomaps from 2018, in the scale 1:10,000 were purchased by the Department of Landscape Ecology and Nature Conservation at the University of Life Sciences in Lublin. On the orthophotomap, land-use forms were mapped. Analyses were performed using the MicroStation Power Draft V8i program.

Statistical methods
The statistical evaluation of the data was carried out with Statistica 13.1 software. To obtain synthetic information about the indicators describing rushes, coastline, and buffer zone management, the statistical measures of position and dispersion were determined and described. Data visualisation on a two-dimensional scatter plot, defining relations between features and indicating the most significant features in the data set, was possible by multivariate principal component analysis (PCA). Within it, only variables significantly correlated with the features describing the rushes were included (Pearson's correlation coefficient > 0.3). At the same time, the Bartlett test was performed to justify the use of PCA. The data were standardised prior to PCA analysis. Apart from the influence of the lake surroundings on the occurrence of rushes, the influence of morphometric properties of the studied lakes was also analysed (flow and mixing type, TSI Carlson, decline, depth, size). For this purpose, the nonparametric Kruskal-Wallis test was used. To group the lakes into homogeneous clusters in terms of the values of the indicators describing the shoreline and buffer zones, the cluster analysis method was used. The parameters having significance for created clusters were indicated, and their central tendencies in groups were compared.
To predict the future impact of parameters surrounding the lake on the development of the rush zone, three mathematical models were developed using multiple-step regression with backward elimination. For this, it was necessary to use the logarithm of the compaction of rushes. The statistical significance of each model was confirmed by the F test (p < 0.05) and the significant influence of each predictor by the Student's t-test. Moreover, the coefficient of determination (R 2 ) was calculated to assess the degree of model fitting to experimental data.

Characteristics of studied lakes
The study included 65 lakes of the Łęczna-Włodawa Lake District, which varied in terms of trophic level, sizes, depths, power supplies, and diversified use of the lake's surroundings.
Among the four distinguished trophic groups of the studied lakes, the vast majority were eutrophic lakes, comprising as many as 40 lakes (62%). Due to the predominance of shallow and very shallow lakes, non-stratified lakes dominated among them. Most of them were flow-through lakes, with an area not exceeding 100 ha. Only 8 of them occupied an area larger than 100 ha. The differentiation of the shoreline development index was small, but among the examined lakes, 3 groups could be distinguished. Among the morphological parameters, the terrain slope in the buffer zone was analysed. It was characterised by great diversification, with values ranging from 67‰ to just 0.15‰ (Table 1).

Lake rushes in Łęczna-Włodawa Lake District
Rushes were present in all lakes of the studied area. The average area inhabited by rushes in the lakes of Łęczna-Włodawa Lake District was 11.05 ha, which was on average 25% of the lake's area. With an overwhelming number of small and medium-sized lakes, up to 30 ha, it is a great value. Among all the parameters describing the rushes, their surface was characterised by the greatest variability (V > 171%). The average number of species representing rushes was less than 7, in some lakes it was only 3, and in some even 14 species. Despite the small species diversity of the rushes of the Polesie lakes, their density, ranging from 20 to even 100 pcs/m 2 , was relatively high. The maximum depth of rush occurrence ranged from 0.5 m to a maximum of 1.6 m. It was the least diversified feature among rushes (V = 12.18%, Table 2).

Use of the shoreline and buffer zones
Both the shore zone of the lakes and the buffer zone showed significant differences in the type of use. Scrub communities and meadows constituted the most frequent use of the coastal zone in the lakes of the Polesie Lubelskie, while the share of fallows and water was the lowest. All examined shoreline indicators were characterised by a very high variability (V > 100%). The highest variability was found in the case of fallow and the lowest for shrub vegetations (Table 3).
Meadows, forests, and arable lands dominated the use of the buffer zone, covering a 200 m zone around the lakes. Buildings, especially summer resorts, of many lakes were also relatively numerous. The lowest share was recorded for ungrown bogs and waters. As in the case of the shoreline and buffer zones, the examined indices were characterised by a very high variability (V > 100%). The highest variability was found for water share and the lowest for arable lands (Table 4).

Relationships between the development of the shoreline zone and features of rushes
To determine the interrelationships between the indicators describing the use of the shoreline zone and features of rushes, multivariate PCA was performed. The analysis included only those marginal variables whose correlation coefficient with the parameters describing the rushes exceeded 0.3 (dry meadows, wet meadows, scrub communities, peat bogs with succession, water, and buildings). The results of PCA analysis for the shoreline showed that the four Eigen values exceeded a value of 1 (Kaiser criterion). Instead of 12 parameters, four new factors (PC1, PC2, PC3, PC4) were obtained, explaining over 68% of the variability of the edge-rush system (Fig. 2).
If the absolute value of factor loadings between the parameters describing the studied system (edgerush) and the new factors is between 0.75 and 1.00, there is a strong correlation if in the range of 0.5-0.75 (mean). The first two PCA axes explained 45.6% of the total variation in data. The first factor (PC1) was strongly correlated with the width of rushes and moderately correlated with surface area, density, and share of rushes in lake cover, wet meadows, scrub, and waters. The second factor (PC2) was strongly correlated with depth of rushes and moderately correlated with species number and buildings. The next two factors were correlated with waters (PC3) and dry meadows and peat bogs with succession (PC4). On this basis, seven parameters: width, depth, and area of rushes, as well as their presence in the shoreline zone of wet meadows, water, scrub communities, and buildings, were selected as the most important for explaining the variation in the data set (edge-rush). This means that there was a strong, positive relationship between the parameters: share of rushes in lake cover and their density, width of the rush belt, and the presence of wet meadows and scrub communities. Conversely, there was a negative relationship between the number of species and the area occupied by buildings in the lakes' shoreline zone (Fig. 2, Appendix Table 8).

Relationships between development of the buffer zone and the features of rushes
Similarly, the variables in the buffer zone included those whose correlation coefficient with the parameters describing the rushes exceeded 0.3 (coniferous  forest, dry meadows, scrub, wet meadows, ungrown bogs, peat bogs with succession, buildings, and fallow lands). PCA analysis showed five eigenvalues that exceeded a value of 1 (Kaiser criterion). Thus, instead of 14 parameters, five new components were obtained (PC1, PC2, PC3, PC4, PC5), which explained as much as 71% of the variability of the buffer-rush system (Fig. 3). The first two PCA axes explained 42.6% of the total variation in the data (buffer-rush). The first factor (PC1) was strongly correlated with surface area and width of rushes and moderately correlated with density and share of rushes in lake cover, scrub, and wet meadows. The second factor (PC2) was strongly correlated with buildings and moderately correlated with depth of rushes, species number, and dry meadows. The next three factors were correlated with fallows (PC3), peat bogs with succession (PC4), and coniferous forest (PC5).
On this basis, five parameters, surface area and width of rushes, buildings, scrub communities, and wet meadows, were selected as the most important in explaining the variation in the data set (buffer-rush). This means that there was a strong, positive relationship between the indicators: the presence of buildings and the depth of rushes, surface area of the rushes and presence of scrub communities, and width of the rushes and wet meadows. Moreover, there was a strong correlation between the presence of ungrown fens in this zone and the number of species, range, and cover by rushes (Fig. 3, Appendix Table 9).

Factors influencing the occurrence of rushes
Apart from the use of the lake surroundings, the analysis also included the dependence of selected hydromorphological features of lakes on the occurrence of rushes. The results of the Kruskal-Wallis test indicated that the type of lake supply and development of the shoreline were insignificant. Conversely, the frequency of water mixing in the lake, trophic level, and depth of the lake significantly influenced most of the analysed features of the rushes, except for the inhabited area and width of the rushes. The decline in surface area had a significant effect on the ratio of the rushes area to the lake area and on the range of the rushes. The size of the lake had a significant effect on the range, area, and width of the rushes belt (Table 5).

An attempt to classify lakes
Based on the analysis of the profile values of the average features of the buffer-rushes system in relation to  The lakes characterised by small values of the features: density, width of rushes, and the share of rushes in the lake, as well as a significant share in the use of the buffer zone by coniferous forests and peat bogs, were mesotrophic lakes (M). Unfortunately, in the case of these lakes, the largest building area in the buffer zone was also found.
The next group were dystrophic lakes (D), which appeared as lakes with a small share of buildings, dry meadows, scrubland communities, and fallows in the buffer zone. There was, however, the highest share of coniferous forests and ungrown peat bogs. Rushes, however, were characterised by a small area but many species.
The lakes with the highest values of width and share of rushes on the lake's surface were mostly eutrophic lakes (E).
Lakes with a high degree of water surface coverage by rushes and high density of rushes, but a small number of species, were predominantly hypertrophic lakes (Hy). Dry meadows were the most common use of the buffer zone of lakes with this trophic level.
The last group was represented by lakes in which the buffer zone was dominated by scrub communities, wet meadows, peat bogs with succession, and fallow lands. The lakes belonging to this group were generally large in terms of area and with a significant slope of land. These were slightly eutrophic lakes (LE) (Fig. 4).

Rushes development scenarios
The development of rushes significantly depends on the use of the buffer and shoreline zones. The indicators characterising lakes (e.g. surface area) are also important. To determine the predicted impact of lake management on the development of rushes, three regression models were developed to identify the scale of the impact of one indicator on the other. The statistical significance of each model was demonstrated (p < 0.05), and the standard error of Se estimation was sufficiently low.

Model 1
Rushes area = − 3.453 + 0.185 * lake area + 0.288 * peat bogs with succession in the buffer zone + 0.996 * dry meadows in the buffer zone + 0.004 * deciduous forest in the shoreline zone.
The improved coefficient of determination (R 2 ) indicated that Model 1 explains 84% of the variability (R 2 = 0.84, Se = 7.48, p < 0.0001) of the rushes area, depending on the lake surface, peat bogs with succession, wet meadows in the buffer zone, and deciduous forests in the shoreline zone. The estimated coefficients in Model 1 indicated the following scenario of the development of the rushes surface: • If the area of the lake increases by 1 ha, the area of the rushes will increase on average by 0.185 ha (if the other variables remain unchanged = ceteris paribus). • If the area of peat bogs in the buffer zone is subject to succession increases by 1 ha, the area of the rushes will increase by an average of 0.288 ha (ceteris paribus). • If the area of dry meadows in the buffer zone increases by 1 ha, the area of rushes will increase by an average of 0.996 ha (ceteris paribus). • If the share of deciduous forests in the shoreline zone increases by 1 m, the area of the rushes will increase by an average of 0.004 ha (ceteris paribus).

Model 2
Log (density) = 3.344 + 0.0003 * wet meadows in the shoreline zone + 0.0001 * shrub communities in the shoreline zone-0.0002 * buildings in the shoreline zone.
The improved coefficient of determination (R 2 ) indicated that Model 2 explains 24% of the variability in the density of rushes, depending on the presence of wetlands, scrubland communities, and buildings in the shoreline zone.
The estimated Model 2 coefficients, converted according to the logarithm definition, indicate the following scenario for the development of rushes density: • The density will increase on average by 10 pcs per m 2 , if the share of wet meadows in the bank increases by 1 m, (ceteris paribus). • The density will increase by an average of 10 pcs per m 2 , if the share of shrub communities in the bank increases by 1 m (ceteris paribus).
• The density will increase on average by 10 pcs per m 2 , if the share of buildings in the bank increases by 1 m (ceteris paribus).

Model 3
Width of rushes = 0.653 * water in the shoreline zone.
The coefficient of determination (R 2 ) indicated that Model 3 explains 43% of the variability in the width of the rushes, depending on the water content in the shore.
The regression coefficient predicted that with an increase in waters in the share of shoreline by 1 m, the width of the rushes will increase by 0.653 m.

Discussion
Polesie Lubelskie is a unique region in terms of nature in the scale of Poland, and even Europe, but has been struggling for years with increasing anthropopressure. Among them, the development of agriculture, recreation causing the development of shores and penetration of coastal areas, and mining, which through its activity, affects the hydrological regime of the entire region and water drainage, has been changing the landscape for decades (Michalczyk et al. 2011). These activities contribute an increase in erosion, sedimentation rate, changes in the composition and re-suspension of bottom sediments, trampling and destruction of coastal and littoral vegetation, influence on the rate of water exchange and lowering its levels, and changing biological elements (phytoplankton, macroinvertebrates, littoral communities, fish) (Chmielewski 2009).
Macrophytes inhabiting the littoral zone, especially rushes, constitute one, sometimes the only, biogeochemical barrier. Emergent macrophytes (rushes) are rooted in bottom substrates, but the plant extends above the water surface. They occur in shallow water and often include species of sedges, grasses, and bulrush. Due to the place of their occurrence in lakes, they are assigned many functions, including stabilisation and oxygenation of the bottom through an extensive root system (Brix and Schierup 1990), reduction in undulations, habitat of many aquatic organisms (Cazzanelli et al. 2008;Choi et al. 2014), and participation in a circulation of elements (Ławniczak et al. 2010;Dybiec et al. 2021). However, their occurrence and differentiation depend on the use of the reservoir catchment area (Radomsky and Goeman 2001;Alahunta et al. 2012;Sender and Garbowski 2016;Sender et al. 2020), water level fluctuations (van Geest et al. 2005;Zohary and Ostrovsky 2011), nutrients, and trophic status (Malthus et al. 1990;Locoul and Freedman 2006). The most important benefits of their presence in the lake are competition for phytoplankton in the uptake of nutrients, which are accumulated in plants for a much longer time (calculated in years), a refuge for zooplankton and other invertebrates feeding on filtered seston, and a feeding, shelter, and reproduction site for vertebrates. Rushes also take part in the process of matter circulation, oxygenation, and consolidation of bottom sediments. In addition, rushes in larger clusters play the role of a buffer protecting water against the run-off of harmful compounds. They produce metabolites (by some species), some of which can inhibit phytoplankton development (Blindow et al. 2014;Gołdyn et al. 2003;Szmeja 2006;Pełechaty and Pronin 2015).
Lakes of the Polesie Lubelskie region constitute an excellent scientific base due to their trophic and morphological diversity. Among 65 lakes in the Polesie, the average area covered by rushes was 25%, whereas the average for Polish lakes in the 1970s was 9.3-12.3% (Sender and Garbowski 2016). This means that nearly a quarter of the Polesie lakes are covered with rushes. This is mainly due to the small average area of these lakes, usually low depths, and atrophic type (fertility) of the lake. In small eutrophic lakes from the Wielkopolska area, rushes covered 32% of the area (Ławniczak-Malińska and Achtenberg 2018). For the group of eutrophic lakes in Polesie, 76.4% of the area, on average, was occupied by rushes. Unfortunately, around Polesie Lubelskie, fertile and eutrophic lakes dominated. In these lakes, the rushes reached a large area and width, while the surroundings were dominated by shrub communities and meadows. In the next distinguished trophic types of Polesie lakes, both rushes and the surroundings showed specific tendencies. Therefore, macrophyte development scenarios dictated by appropriate management in their surroundings may be a tool for lake valorisation and planning their proper management.
Almost 70% of the coastline of the examined lakes was inhabited by rushes, despite its slight diversification. These values were higher in small hypertrophic and eutrophic lakes, while the lowest were in mesotrophic and slightly eutrophic lakes, which are intensively used for tourism.
Scrub communities and wet meadows constituted the most frequent use of the shoreline zone, while drained meadows, forests, and arable fields dominated the buffer zone. Buildings, especially the summer resorts of many lakes, were also relatively numerous. Scrub communities often inhabit transition zones between lakes and various types of ecosystems (ecotones), constituting an excellent biogeochemical barrier for various types of pollutants (Izydorczyk et al. 2010;Carluer et al. 2011;Caracciolo et al. 2014;Łaskawiec 2015;Boulanger et al. 2018). They are especially valuable where the vicinity of water reservoirs is arable fields. Plants as bioretention organisms can retain 10-50% of nutrients by creating their own biomass; the rest is physically and chemically bound by other components of the ecosystem, e.g. by accumulation in the surface layer of the soil (Krauze 2004). Plants additionally interact with soil and microorganic edaphone, stimulating many biochemical processes (Srivastava et al. 2017;Carluer et al. 2011;Izydorczyk et al. 2013). In the case of the shores of water reservoirs, plant buffer zones additionally stabilise the shoreline by preventing water erosion of the banks, reducing run-off, and improving the soil structure (Jørgensen et al. 2002;Blanco-Canqui and Lal 2008;Sahu and Gu 2009;Serafin et al. 2019). The width of the zone is an important issue, according to the assumption that the wider the zone, the more effective it is in bioresorption of nutrients (Petersen et al. 1992;Jørgensen et al. 2002). In the studied lakes of Polesie Lubelskie, if there were scrub communities and peat bogs in the vicinity of the lakes, rushes developed most abundantly.
In addition to the presence of scrub communities, trees (forests), grasses, and perennials (meadows and peat bogs) in the buffer zone are very important. These areas are characterised by a high bioretention capacity of phosphorus (Ryszkowski et al. 1999) and nitrogen (Gu et al. 1999;Blanco-Canqui and Lal 2008). In the studied lakes, the presence of forests in the buffer and shoreline zones resulted in the analysed features of rushes reaching the lowest values. The grasses growing on meadow areas have an intense, bundled root system, significantly increasing the effectiveness of biogeochemical barriers; however, it is important to move them at the end of the growing season (Lemkowska et al. 2010). Among the grasses in all analysed Polesie lakes, the nitrophilic Phragmites australis (Izydorczyk et al. 2018) dominated. It has low ecological requirements, and although it indicates an intensification of eutrophication of ecosystems, it intensively captures biogens, heavy metals, and pesticides (Hootsmans and Vermaat 1991;Bonanno and Giudice 2010).
Water level, wave action, water current, physicochemical characteristics of the sediment, and intensity of grazing are key factors determining the distribution and composition of emergent vegetation along shorelines (Coops and van der Velde 1996;Coops et al. 2004;Horppila et al. 2013). The research on rushes in Łęczna-Włodawa Lake District lakes indicated a group of further factors influencing the development of rushes, including trophic levels, depth, frequency of water mixing, and lake size, as well as land slope.
The proposed models made it possible to evaluate and forecast further development processes of the rushes. The evolutionary processes of the lakes are irreversible and continuous; like any living organism, they mature. Their maturity is based on enriching the diversity that normally lasts for centuries. However, the current global changes, mainly related to human activity, significantly accelerate this process. That is why prevention seems to be so important, especially in the management of the lake's surroundings. Increasing the area inhabited by rushes, as a natural process, seems to guarantee the maintenance of the proper water level in lakes and a large share of forested areas in its surroundings. The evident increase in the density of rushes will increase the development of the lake's shoreline, and the share of drained meadows in the buffer zone will increase the area occupied by rushes.

Conclusions
Rushes are an excellent research object due to their common occurrence. Their presence is shaped by the diverse features of the shoreline and buffer zone and the morphometric properties of the lake.
This study, involving a morphologically and trophically diversified group of 65 lakes in Łęczna-Włodawa Lake District, indicated preferential features both in the use of the shoreline and the buffer zones. The presence of peat bogs, scrubland communities, and wet meadows favoured the development of rushes. Among the morphometric features of the lakes, the depth and frequency of water mixing were significant, as well as trophic level.
A comprehensive analysis of the morphometric features of the lakes and type of development distinguished 4 classes of Łęczna-Włodawa Lake District, significantly differentiated in terms of rush features. In addition, the trophic level of lakes and the dominance of a specific management strategy for the surroundings can be an indicator of the appropriate development of certain features of rushes.
The regression models allowed to determine that the area of rushes is significantly dependent on the surface of the lake, the presence of overgrowing peat bogs and meadows in the buffer zone and the presence of deciduous forest in the shoreline zone. On the other hand, a greater density of reeds is favoured by the presence of wet meadows and bushes as well as buildings in the coastal zone.
The applied models can be the basis for forecasting the conditions for the development of rushes for lakes with similar morphometric and geo-climatic conditions. Author Contribution All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Weronika Maślanko, Monika Różańska-Boczula and Joanna Sender. The first draft of the manuscript was written by Joanna Sender, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding
The authors did not receive support from any organization for the submitted work.

Data availability
The datasets generated during the current study are available from the corresponding author on reasonable request.

Conflict of interest
The authors have no financial or proprietary interests in any material discussed in this article.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.