Possibility of Metal Accumulation in Reed Canary Grass (Phalaris arundinacea L.) in the Aquatic Environment of South-Western Polish Rivers

A four-year research study was conducted on aquatic plants (reed canary grass) growing in the beds of three rivers and their tributaries in Lower Silesia, Poland. Metal contents (Cu, Cd, Ni, Pb, Zn, Fe, Mn) were determined in plant samples, metal accumulation in water (BCFw) and sediment (BCFB), Metal Pollution Index (MPI) and Enrichment Factor (EF) were calculated. The highest contents of copper, lead, nickel and cadmium were found in reed canary grass sampled from the Nysa Szalona River. The highest values were recorded for zinc in the Bystrzyca River, and for iron and manganese in the Strzegomka River. The series of metals were as follows: Nysa Szalona and Strzegomka: Cd < Ni < Pb < Cu < Zn < Mn < Fe, Bystrzyca: Cd < Ni < Cu < Pb < Zn < Mn < Fe. Throughout the study period, the lowest values of metals in plants were recorded in 2015 and 2018, and the highest in 2017. The general picture of MPI in aquatic plants is arranged in the series Bystrzyca < Strzegomka < Nysa Szalona. These values classify the studied material at a high level of pollution in all rivers. In the comparison of the two extreme sites, i.e., source–mouth, higher values were found at the mouth of the reservoir, which suggests that metals move with the water current and accumulate more with the direction of the river flow, which is most likely a consequence of the influence of the catchment area as the source of metals. The series of EF enrichment factor values were as follows: Bystrzyca—Ni < Cd < Fe < Cu < Zn < Mn < Pb, Nysa Szalona—Ni < Fe < Zn < Cd < Mn < Cu < Pb, Strzegomka—Ni < Cd < Fe < Zn < Cu < Pb < Mn. For all the samples studied, the values found in spring were much higher than in autumn, which indicates the great importance for research in that area. The levels of copper and iron were within the range of moderate values, lead and manganese reached very high and exceptionally high values, and the remaining metals were within the values described as significant. Bioaccumulation of metals determined relative to bottom sediments was highest in 2017 and lowest in 2018, while bioaccumulation relative to water was highest in 2018 and lowest in 2016. The four-year study found that the metal content in reed canary grass was mostly within the range of mean values presented in the literature from moderately polluted areas. Also, no significant deviation was found from levels that have been recorded for the same rivers for more than two decades.


Introduction
Natural levels of metals in the environment are disrupted by industrial, agricultural and municipal activities. Each type of activity often results in a different metal load to the air, soil, groundwater and surface water. Elevated levels of some metals are not the cause of significant changes in the environment and do not have a strong toxic effect  (Table 2), 2-Dobromierz reservoir-research sites on the Strzegomka River and its tributaries (Table 3), 3-Lubachów reservoir-research sites on the Bystrzyca River and its tributaries (Table 4). Table 2. Research sites-the Nysa Szalona River and its tributaries above the Słup dam reservoir [19].

No
Site-Geographical Coordinates Surface Water Types, Water Categories-Type Code SWB * Status Upland silicate stream with coarse-grained sub-  (Table 2), 2-Dobromierz reservoir-research sites on the Strzegomka River and its tributaries (Table 3), 3-Lubachów reservoir-research sites on the Bystrzyca River and its tributaries (Table 4). Table 3. Research sites-the Strzegomka River and its tributaries above the Dobromierz dam reservoir [19].

No
Site-Geographical Coordinates Surface Water Types, Water Categories-Type Code SWB * Status  Table 4. Research sites-the Bystrzyca River and its tributaries above the Lubachów dam reservoir [19].

No
Site-Geographical Coordinates Surface Water Types, Water Categories-Type Code SWB * Status Reed canary grass was harvested whole (root, stem, leaves, and inflorescence). Immediately at the test site after collection, plants were washed with river water. In the laboratory, they were dried at room temperature to an air-dry state and then cut, crushed, and homogenized.
For the determination of metals, 0.5 g of air-dry and homogenized sample was weighed in an HP-500 Teflon dish. Then, 10 cm 3 of concentrated HNO 3 (Sigma-Aldrich, Poznań, Poland) was added and the samples were left at room temperature for 24 h. After this time, the samples were placed in a Mars 5 microwave oven (CEM) and mineralized using a 3-stage mineralization. After cooling to room temperature, the mineralizates were transferred to test tubes and diluted with distilled water to 25 cm 3 [24][25][26].
232 plant samples were collected and results are given in mg·kg −1 dry weight. The levels of lead, copper, nickel, zinc, cadmium, iron and manganese in aquatic plants were determined by flame atomic absorption spectrometry (FAAS) using a Spectra AA-110/220 from Varian [27].
Test results were verified using certified reference materials for aquatic plants-IAEA-336 International Atomic Energy Agency-Analytical Quality Control Services Austria and CRM 482-Commission of the European Communities, Community Bureau of Reference-BCR.
Accumulation of metals in aquatic plants was determined by: metals bioaccumulation factor (BCF B ) as a ratio of its content in aquatic plant (C P ) to its concentration in bottom sediment (C B ) [28] BCF B = C P C B metals bioaccumulation factor BCF W as a ratio of its content in aquatic plant C P to its concentration in water C W [28] BCF W = C P C W The assessment of the state of plants contamination with metals was carried out using the metal pollution index (MPI) [29]. MPI = (Cf 1 × Cf 2 . . . Cf n ) 1/n where Cf 1 , Cf 2 . . . Cf n -concentration of first metal, second metal, n-th metal.
EF values less than 2 indicate no or minimal enrichment of the plant with particular metals, range 2-5 as moderate, 5-20 as significant, 20-40 as very high, and above 40 as extremely high [30].

Statistical Analysis
The analysis of results was performed using Microsoft Office Excel 2019 and Statistica 13.0 (StatSoft Poland, Krakow, Poland). Calculations were performed using R version 3.6.0. The Shapiro-Wilk test was applied to check the normality of the distribution. Spearman correlations were used due to the distribution of samples. Spearman correlations were calculated in Statistica, and box and whisker plots were also created in this program. All statistically significant differences were calculated at p < 0.05. Because the data had a non-normal distribution, the Kruskal-Wallis test with post-hoc analysis was used. A PCA test using the r-groups statistic (RStudio Version 1.1.442-© 2022-2018, Rstudio, Inc., Boston, MA, USA) was used to visualize differences between groups. It was based on all data and presented the differences in the parameters of the studied plant depending on the year, season of study and river.

Metals in Reed Canary Grass
The highest contents of Cu, Ni, Cd, Pb were recorded in reed canary grass sampled from the Nysa Szalona River (Table 5). Samples from the Bystrzyca River had the highest zinc levels, while samples from the Strzegomka River had the highest iron and manganese levels. The lowest values of Fe, Mn and Zn were recorded in plants obtained from the Nysa Szalona River, Pb and Cd from the Strzegomka River, and Cu and Ni from the Bystrzyca River. The series of metals were as follows: Nysa Szalona and Strzegomka: Cd < Ni < Pb < Cu < Zn < Mn < Fe, and Bystrzyca: Cd < Ni < Cu < Pb < Zn < Mn < Fe.  NS-Nysa Szalona River; B-Bystrzyca River; S-Strzegomka River; P-metal content in the plant; BCF B -metals bioaccumulation factor-bottom sediment; BCF W -metals bioaccumulation factor-water.

Copper
Throughout the study cycle, the range of copper content in reed canary grass was from 2.33 mgCu·kg −1 to 396.55 mgCu·kg −1 , with a range for mean values of 7.92 mgCu·kg −1 -31.02 mgCu·kg −1 ( Table 5). The bioaccumulation factor with respect to bottom sediments ranged from BCF B = 0.0090 to BCF B = 13.13, and with respect to water it was higher, at a level of BCF W = 30.47-82,142.29 (Table 5). Statistically significant differences in copper content of reed canary grass were recorded between all rivers in each year with minor exceptions (Figure 2). These exceptions indicate that each year was different, and therefore inferences about environmental quality should be made during long-term studies. The long-term data allowed the identification of the Nysa Szalona River as having statistically higher levels of this metal. sediments ranged from BCFB = 0.0090 to BCFB = 13.13, and with respect to water it was higher, at a level of BCFW = 30.47-82,142.29 (Table 5). Statistically significant differences in copper content of reed canary grass were recorded between all rivers in each year with minor exceptions (Figure 2). These exceptions indicate that each year was different, and therefore inferences about environmental quality should be made during long-term studies. The long-term data allowed the identification of the Nysa Szalona River as having statistically higher levels of this metal.  The content of copper in hydromacrophytes examined in the study compared to the results from previous years indicates a relatively unchanged level maintained for years in the case of the Strzegomka and Nysa Szalona Rivers (300-500 mgCu•kg −1 ) [16]. Plants collected from the Słup dam reservoir, into which the Nysa Szalona River flows, and from the Bystrzyca River supplying the reservoir in Lubachowo fall into a similar range [2,4]. Also in other reservoirs of flowing and standing waters of that region and not strongly influenced by economic activity, the level of copper in plants was similar to that presented in this study [31][32][33][34][35][36][37][38][39]. However, in the reservoirs of the same region, whose catchment areas are strongly exposed to greater pollution or even contamination (copper mine), an increase in the level of this metal in aquatic plants can be observed [40].

Cadmium
Cadmium content in reed canary grass ranged from 0.0400 mgCd•kg −1 to 10.89 mgCd•kg −1 ( Table 5). Mean values ranged from 0.8969 mgCd•kg −1 to 1.65 mgCd•kg −1 . Cadmium bioaccumulation factor with respect to sediment (BCFB = 0.0209-41.07) was lower than with respect to water (BCFW = 5.35-18,155.50) ( Table 5). In studies on the Strzegomka, Nysa Szalona and Bystrzyca Rivers conducted in previous years, the cadmium content in plants was similar to the results obtained in this study, with levels The content of copper in hydromacrophytes examined in the study compared to the results from previous years indicates a relatively unchanged level maintained for years in the case of the Strzegomka and Nysa Szalona Rivers (300-500 mgCu·kg −1 ) [16]. Plants collected from the Słup dam reservoir, into which the Nysa Szalona River flows, and from the Bystrzyca River supplying the reservoir in Lubachowo fall into a similar range [2,4]. Also in other reservoirs of flowing and standing waters of that region and not strongly influenced by economic activity, the level of copper in plants was similar to that presented in this study [31][32][33][34][35][36][37][38][39]. However, in the reservoirs of the same region, whose catchment areas are strongly exposed to greater pollution or even contamination (copper mine), an increase in the level of this metal in aquatic plants can be observed [40].

Cadmium
Cadmium content in reed canary grass ranged from 0.0400 mgCd·kg −1 to 10.89 mgCd·kg −1 ( Table 5). Mean values ranged from 0.8969 mgCd·kg −1 to 1.65 mgCd·kg −1 . Cadmium bioaccumulation factor with respect to sediment (BCF B = 0.0209-41.07) was lower than with respect to water (BCF W = 5.35-18,155.50) ( Table 5). In studies on the Strzegomka, Nysa Szalona and Bystrzyca Rivers conducted in previous years, the cadmium content in plants was similar to the results obtained in this study, with levels reaching a maximum of 8.80 mgCd·kg −1 [2,4,5,16,31,32,35,[37][38][39][40][41][42][43][45][46][47][48][49]51,52]. This indicates a persistently stable level of cadmium in the environment and relatively unchanged water chemistry conditions. Higher concentrations were recorded in submerged plants or those growing in intensively polluted reservoirs such as post-mining pits, for example [33,34,36,53]. Statistically significant differences in cadmium levels in aquatic plants were found between all rivers only in the last year of the study (Figure 3). In contrast, no such differences were recorded between any rivers in 2016. The multiannual data obtained indicated that both the Bystrzyca and Strzegomka have statistically the lowest cadmium content compared to the Nysa Szalona. or those growing in intensively polluted reservoirs such as post-mining pits, for example [33,34,36,53]. Statistically significant differences in cadmium levels in aquatic plants were found between all rivers only in the last year of the study (Figure 3). In contrast, no such differences were recorded between any rivers in 2016.
The Bystrzyca River and the Słup Dam reservoir, into which the Nysa Szalona flows, were characterized in previous years by similar nickel levels to those presented in this study [2,4]. This indicates a fairly stable level of nickel in the studied environment, although in studies on the Nysa Szalona and Strzegomka carried out in earlier years the level of nickel was higher [34]. Higher levels of nickel were also found in submerged plants and those growing in artificial reservoirs, but also in mountainous and lowland areas subjected to anthropopression [36,40,50].
Similar to cadmium, statistically significant differences in nickel levels between plants from different rivers were found in 2018 ( Figure 4). In the remaining years, such differences were not observed, although in sum, without division into years, statistical differences were observed between the Bystrzyca, and the Nysa Szalona.
The Bystrzyca River and the Słup Dam reservoir, into which the Nysa Szalona flows, were characterized in previous years by similar nickel levels to those presented in this study [2,4]. This indicates a fairly stable level of nickel in the studied environment, although in studies on the Nysa Szalona and Strzegomka carried out in earlier years the level of nickel was higher [34]. Higher levels of nickel were also found in submerged plants and those growing in artificial reservoirs, but also in mountainous and lowland areas subjected to anthropopression [36,40,50].
Similar to cadmium, statistically significant differences in nickel levels between plants from different rivers were found in 2018 ( Figure 4). In the remaining years, such differences were not observed, although in sum, without division into years, statistical differences were observed between the Bystrzyca, and the Nysa Szalona.
Similar values to the present results were noted for the same rivers in previous years, which as for nickel, copper and cadmium confirms relatively stable conditions present in the rivers for many years [2,4,16]. However, higher values were found in plants from mountainous regions, but also from highly industrialized and urbanized areas [34,38,40,48,53].   (Table 5). Bioaccumulation with respect to water was BCFW = 6.03-219,934.00, and much less with respect to sediment BCFB = 0.0013-3.79 (Table 5).
Similar values to the present results were noted for the same rivers in previous years, which as for nickel, copper and cadmium confirms relatively stable conditions present in the rivers for many years [2,4,16]. However, higher values were found in plants from mountainous regions, but also from highly industrialized and urbanized areas [34,38,40,48,53].
Only in 2016, statistically significant differences in lead content in reed canary grass were found ( Figure 5) between all studied rivers. In 2015 and 2018 such differences were found only between the Bystrzyca and the Nysa Szalona, and between the Nysa Szalona and the Strzegomka. The multiannual data collected allowed for the observation of the lowest amount of lead in the Strzegomka and the highest in the Nysa Szalona.   Only in 2016, statistically significant differences in lead content in reed canary grass were found ( Figure 5) between all studied rivers. In 2015 and 2018 such differences were found only between the Bystrzyca and the Nysa Szalona, and between the Nysa Szalona and the Strzegomka. The multiannual data collected allowed for the observation of the lowest amount of lead in the Strzegomka and the highest in the Nysa Szalona.   (Table 5). Bioaccumulation with respect to water was BCFW = 6.03-219,934.00, and much less with respect to sediment BCFB = 0.0013-3.79 (Table 5).
Similar values to the present results were noted for the same rivers in previous years, which as for nickel, copper and cadmium confirms relatively stable conditions present in the rivers for many years [2,4,16]. However, higher values were found in plants from mountainous regions, but also from highly industrialized and urbanized areas [34,38,40,48,53].
Only in 2016, statistically significant differences in lead content in reed canary grass were found ( Figure 5) between all studied rivers. In 2015 and 2018 such differences were found only between the Bystrzyca and the Nysa Szalona, and between the Nysa Szalona and the Strzegomka. The multiannual data collected allowed for the observation of the lowest amount of lead in the Strzegomka and the highest in the Nysa Szalona.   . Statistically significant differences are marked in the graph with letters a, b and in the table these differences are shown by year between rivers-T-differences were present, F-no statistical difference (p-value ≤ 0.05).
The level of zinc, similarly to the above-mentioned metals, in previous studies carried out in the area of the Słup reservoir into which the Nysa Szalona flows, was within a similar range as at present (8.00-80.10 mgZn·kg −1 ) [2]. However, for the Nysa Szalona, Strzegomka and Bystrzyca Rivers studied separately, the values found were higher (up to 700 mgZn·kg −1 ) [4,16]. Higher values were also found in the submontane regions of Lower Silesia and other rivers in southern Poland, as well as in the western border areas of Germany [32][33][34]38]. Higher values were also found in more industrialized regions [40,49]. Generally, in the majority of studies covering the European area and other parts of the world, the level of copper is similar to that recorded in this study [2,5,31,37,39,[41][42][43][44][45][46][47][48][50][51][52]54,55]. Statistically significant differences in zinc content were recorded with one exception (year 2018, Nysa Szalona and Strzegomka) between all rivers ( Figure 6). All the obtained data allowed this study to single out the Nysa Szalona River as the watercourse with the statistically lowest level of zinc.
The level of zinc, similarly to the above-mentioned metals, in previous studies carried out in the area of the Słup reservoir into which the Nysa Szalona flows, was within a similar range as at present (8.00-80.10 mgZn•kg −1 ) [2]. However, for the Nysa Szalona, Strzegomka and Bystrzyca Rivers studied separately, the values found were higher (up to 700 mgZn•kg −1 ) [4,16]. Higher values were also found in the submontane regions of Lower Silesia and other rivers in southern Poland, as well as in the western border areas of Germany [32][33][34]38]. Higher values were also found in more industrialized regions [40,49]. Generally, in the majority of studies covering the European area and other parts of the world, the level of copper is similar to that recorded in this study [2,5,31,37,39,[41][42][43][44][45][46][47][48][50][51][52]54,55]. Statistically significant differences in zinc content were recorded with one exception (year 2018, Nysa Szalona and Strzegomka) between all rivers ( Figure 6). All the obtained data allowed this study to single out the Nysa Szalona River as the watercourse with the statistically lowest level of zinc.   Table 5). Bioaccumulation of iron calculated relative to its level in water was BCFW = 1.23-16,874.94, and relative to bottom sediments it was lower, BCFB = 0.0001-6.92 (Table 5).
The level of iron and manganese in plants studied in the Bystrzyca River remained within a similar range as in previous years [4]. Also in other rivers in this region of Poland the iron content in hydromacrophytes was similar [31][32][33]35,39,54]. Apart from that  (Table 5). Bioaccumulation of iron calculated relative to its level in water was BCF W = 1.23-16,874.94, and relative to bottom sediments it was lower, BCF B = 0.0001-6.92 (Table 5).
The level of iron and manganese in plants studied in the Bystrzyca River remained within a similar range as in previous years [4]. Also in other rivers in this region of Poland the iron content in hydromacrophytes was similar [31][32][33]35,39,54]. Apart from that area, the recorded metal contents in aquatic plants were similar in other reservoirs [49,52,56].
However, significantly higher amounts were found in plants from the Nysa Szalona and Strzegomka Rivers studied several years earlier and in industrial areas, which indicates the anthropogenic origin of these metals [16,32,34,37,40,50]. Statistically significant differences in manganese content were noted between the Nysa Szalona and Strzegomka in each year of study (Figure 7). Statistically significant differences were recorded between the Bystrzyca and the Nysa Szalona. The exception was the year 2018, where these differences were not found, exactly the opposite statistical relationship was observed between the Bystrzyca and Strzegomka Rivers. and Strzegomka Rivers studied several years earlier and in industrial areas, which indicates the anthropogenic origin of these metals [16,32,34,37,40,50]. Statistically significant differences in manganese content were noted between the Nysa Szalona and Strzegomka in each year of study (Figure 7). Statistically significant differences were recorded between the Bystrzyca and the Nysa Szalona. The exception was the year 2018, where these differences were not found, exactly the opposite statistical relationship was observed between the Bystrzyca and Strzegomka Rivers. In the case of iron, statistically significant differences in its content in the studied samples recorded between the three rivers do not allow to determine the regularity (Figure 8). The four-year study made it possible to show the Nysa Szalona as the river with the statistically lowest level of iron.  In the case of iron, statistically significant differences in its content in the studied samples recorded between the three rivers do not allow to determine the regularity (Figure 8). The four-year study made it possible to show the Nysa Szalona as the river with the statistically lowest level of iron.

Mn
However, significantly higher amounts were found in plants from the Nysa Szalona and Strzegomka Rivers studied several years earlier and in industrial areas, which indicates the anthropogenic origin of these metals [16,32,34,37,40,50]. Statistically significant differences in manganese content were noted between the Nysa Szalona and Strzegomka in each year of study (Figure 7). Statistically significant differences were recorded between the Bystrzyca and the Nysa Szalona. The exception was the year 2018, where these differences were not found, exactly the opposite statistical relationship was observed between the Bystrzyca and Strzegomka Rivers. In the case of iron, statistically significant differences in its content in the studied samples recorded between the three rivers do not allow to determine the regularity (Figure 8). The four-year study made it possible to show the Nysa Szalona as the river with the statistically lowest level of iron.  The differences between the rivers may be noted using PCA plot. Nysa Szalona and Strzegomka were the most different in respect of the examined parameters, and Bystrzyca had parameters intermediate between the other rivers ( Figure 9).

Mn
The differences between the rivers may be noted using PCA plot. Nysa Szalona and Strzegomka were the most different in respect of the examined parameters, and Bystrzyca had parameters intermediate between the other rivers ( Figure 9).   The differences between the rivers may be noted using PCA plot. Nysa Szalona and Strzegomka were the most different in respect of the examined parameters, and Bystrzyca had parameters intermediate between the other rivers ( Figure 9).

Comparison of Metals Content-Source and Mouth of the River
Comparison of metal levels within the extreme sites on the main rivers shows that, for the most part, the estuaries of the rivers to the reservoirs in all study years had higher contents of the metals studied than the site below the springs (Table 6). This indicates enrichment with metal compounds of the main rivers by successively escaping tributaries. A similar relationship was found for the accumulation coefficient in relation to sediment. On the other hand, for the coefficient calculated in relation to metal concentration in water, sometimes higher values occurred at the site downstream of the springs.    NS-Nysa Szalona River; B-Bystrzyca River; S-Strzegomka River; 1-river below its source; 2-river mouth to the reservoir; P-metal content in the plant; BCF B -metals bioaccumulation factor-bottom sediment; BCF Wmetals bioaccumulation factor-water.
Metal contents also varied with the seasons. In the Strzegomka River at both extreme sites, higher amounts of Cu, Ni, Cd, Pb and Mn were recorded in autumn, and Zn and Fe in spring ( Table 7). The Bystrzyca River had higher values for all metals at the reservoir outlet in spring (Table 8). Below the springs, there was a differentiation: in spring, higher levels of cadmium, lead and zinc were recorded, whereas the remaining metals predominated in autumn. On the other hand, in the Nysa Szalona River at both sites, copper and cadmium were more abundant in the spring, and higher amounts for the other metals were recorded in the autumn (Table 9). Table 7. Metals (mg·kg −1 ) and metal bioaccumulation factors relative to water (BCF W ) and bottom sediment (BCF B ) in reed canary grass in successive tributaries of the Strzegomka River in spring and autumn (mean values).  s-spring, a-autumn, P-metal content in the plant; BCF B -metals bioaccumulation factor-bottom sediment; BCF W -metals bioaccumulation factor-water Table 8. Metals (mg·kg −1 ) and metal bioaccumulation factors relative to water (BCF W ) and bottom sediment (BCF B ) in reed canary grass in successive tributaries of the Bystrzyca River in spring and autumn (mean values).  s-spring, a-autumn, P-metal content in the plant; BCF B -metals bioaccumulation factor-bottom sediment; BCF W -metals bioaccumulation factor-water Furthermore, differences between the spring and autumn seasons are noticeable. Spring was the most differentiated season with respect to the studied parameters ( Figure 11).

Metals in Tributaries of Major Rivers
The analysis of metal contents in plants taken from successive tributaries of the three main rivers shows that in the Strzegomka River, the level of Cu, Ni, Cd, Fe, Mn increased with the river course and in the middle part the content of these metals was the highest. For lead, no major changes were observed in the whole river, and the range of values was 7.12 to 9.58 mgPb•kg −1 (Table 7). For zinc, values decreased with the river course from 89.90 mgZn•kg −1 to 63.46 mgZn•kg −1 in spring, and from 56.49 mgZn•kg −1 to 47.89 mgZn•kg −1 in autumn.
In the Bystrzyca River, copper, nickel, and cadmium levels increased downstream (Table 8). In the case of zinc, the upper and lowland tributaries contributed higher amounts than the midstream rivers. For iron and manganese, decreasing values were observed with the direction of water flow in the river.
In the case of the Nysa Szalona River, the amount of copper and lead in plant samples increased steadily with successive tributaries (Table 9). It was also observed that the midstream rivers contributed higher amounts of cadmium, zinc and manganese, and lower amounts of iron. Nickel remained at an even level in all tributaries.
Two groups of elements clearly emerge from the study. The first one includes Pb, Cu and Cd, and they turned out to be statistically highly correlated (Figure 12). The second group of metals includes Zn, Ni, Fe, Mn which are a group of elements correlated with each other. More detailed values of correlations are given in Table 10.

Metals in Tributaries of Major Rivers
The analysis of metal contents in plants taken from successive tributaries of the three main rivers shows that in the Strzegomka River, the level of Cu, Ni, Cd, Fe, Mn increased with the river course and in the middle part the content of these metals was the highest. For lead, no major changes were observed in the whole river, and the range of values was 7.12 to 9.58 mgPb·kg −1 (Table 7). For zinc, values decreased with the river course from 89.90 mgZn·kg −1 to 63.46 mgZn·kg −1 in spring, and from 56.49 mgZn·kg −1 to 47.89 mgZn·kg −1 in autumn.
In the Bystrzyca River, copper, nickel, and cadmium levels increased downstream (Table 8). In the case of zinc, the upper and lowland tributaries contributed higher amounts than the midstream rivers. For iron and manganese, decreasing values were observed with the direction of water flow in the river.
In the case of the Nysa Szalona River, the amount of copper and lead in plant samples increased steadily with successive tributaries (Table 9). It was also observed that the midstream rivers contributed higher amounts of cadmium, zinc and manganese, and lower amounts of iron. Nickel remained at an even level in all tributaries.
Two groups of elements clearly emerge from the study. The first one includes Pb, Cu and Cd, and they turned out to be statistically highly correlated (Figure 12). The second group of metals includes Zn, Ni, Fe, Mn which are a group of elements correlated with each other. More detailed values of correlations are given in Table 10.  The investigated river ecosystem covers a considerable area of the south-western part of Poland. In this area, besides groundwater, rivers as surface waters are the main source of drinking water supply for the population. The content of metals in plants growing in the riverbeds which supply the reservoirs from which drinking water is obtained suggests that they do not pose a threat to human life. The level of metals in hydromacrophytes is within the average limits for these types of waters in this climatic zone. This is confirmed by the results of the studies by the authors cited above [2,4,17,18,35,39,54]. Also, the concentrations of metals in river water recorded throughout the four-year study cycle are low enough to meet the criteria for water extracted by water  The investigated river ecosystem covers a considerable area of the south-western part of Poland. In this area, besides groundwater, rivers as surface waters are the main source of drinking water supply for the population. The content of metals in plants growing in the riverbeds which supply the reservoirs from which drinking water is obtained suggests that they do not pose a threat to human life. The level of metals in hydromacrophytes is within the average limits for these types of waters in this climatic zone. This is confirmed by the results of the studies by the authors cited above [2,4,17,18,35,39,54]. Also, the concentrations of metals in river water recorded throughout the four-year study cycle are low enough to meet the criteria for water extracted by water production plants for water production [57,58]. If the natural environment continues to be so stable, and there are no changes in the reaction of water and bottom sediments, it can be assumed that the current level of metals in water and plants will not pose a threat to human health in this area in the future

Metal Pollution Index (MPI) and Enrichment Factor (EF)
The metal pollution index (MPI) of aquatic plants was used to compare the metal content of samples from different sites [29]. The general pattern of MPI in aquatic plants was arranged in a series B < S < NS, and the same pattern was found for water, however a slightly different order occurred for sediments: S < B < NS (Table 11) [17,18]. During the four-year cycle of the study, higher MPI values were recorded in 2016 and 2018 than in 2015 and 2017 (Table 12). pollution degree high highest no effect high highest no effect high high/very high no effect These values classify the studied material at high pollution degree in all rivers. Against this background, previously published data for sediment reached the highest and very high grades [18]. In the case of the MPI for water, no effect of pollution was found [17]. With this comparison, it can be seen that it is only when metal levels in sediments and plants are studied that the accumulation of metals in the environment becomes apparent. Hence, conducting a comprehensive study is justified. In the comparison of the two extreme sites: source-mouth, the predominance of higher values at the reservoir mouth in the first three years of the study is evident, indicating the origin of these metals from the catchment. On the other hand, in the last year (2018) higher MPI values were found at the sources in the Nysa Szalona and Bystrzyca Rivers (Table 12).
The enrichment factor (EF) calculated for all plant samples reached the lowest values for nickel and the highest values for lead and manganese (Tables 13-15). The series of increasing values were as follows: Bystrzyca-Ni < Cd < Fe < Cu < Zn < Mn < Pb, Nysa Szalona-Ni < Fe < Zn < Cd < Mn < Cu < Pb, Strzegomka-Ni < Cd < Fe < Zn < Cu < Pb < Mn. For all studied plants in each of the three rivers, the values were definitely higher in spring than in autumn. In the Bystrzyca River, higher values of the enrichment factor EF were found at the site immediately upstream of the reservoir and in the Nysa Szalona River at the site downstream of the springs. In the Strzegomka River, a differentiation was observed: higher values for Cu, Ni, Fe and Mn were recorded downstream of the source, while the remaining metals were observed at the upstream site.
For most samples, Cu and Fe are within the range of moderate values. Lead and manganese reach very high and extremely high values. The remaining metals fall in the values described as significant. This assessment suggests that the amount of metals absorbed by reed canary grass varies. While for the time being there is no concern about the levels of copper and iron, for the rest of the metals a more frequent, regular study should be carried out.

Conclusions
A four-year study of reed canary grass in rivers of southwestern Poland ranked metal levels as follows: Nysa Szalona and Strzegomka: Cd < Ni < Pb < Cu < Zn < Mn < Fe, and Bystrzyca: Cd < Ni < Cu < Pb < Zn < Mn < Fe. Throughout the study period, the lowest values of metals in plants were recorded in 2015 and 2018, and the highest in 2017.
The general picture of MPI in aquatic plants is arranged in the series Bystrzyca < Strzegomka < Nysa Szalona. These values classify the studied material at a high level of pollution in all rivers. In the comparison of two extreme sites, i.e., source-mouth, higher values are found at the mouth of the reservoir, which suggests that metals move with the water current and accumulate more with the direction of the river flow, which is most likely a consequence of the influence of the catchment area as the source of metals.
The series of enrichment factor (EF) values were as follows: Bystrzyca-Ni < Cd < Fe < Cu < Zn < Mn < Pb, Nysa Szalona-Ni < Fe < Zn < Cd < Mn < Cu < Pb, Strzegomka-Ni < Cd < Fe < Zn < Cu < Pb < Mn. For all the samples studied, the values found in spring were much higher than in autumn, which would indicate the great importance of conducting research in that area.
Bioaccumulation of metals determined relative to bottom sediments was highest in 2017 and lowest in 2018, while bioaccumulation relative to water was highest in 2018 and lowest in 2016.
The four-year study found that the metal content in reed canary grass was mostly within the range of the mean values presented in literature data from moderately polluted areas. Also, no significant deviation was found from levels that have been recorded for the same rivers for more than two decades.