The hydrologic balance (Urbanik 2007—unpublished data) shows that within the 3 years of study the hydraulic inflow to the reservoir was compensated by the outflow, so the calculated element retention is not the result of hydrologic factors, but biogeochemical factors. Mass balance of N, C, P and Si, conducted separately for both the reservoirs and the entire complex enabled identification of the classical biogeochemical cycle of conversion of the analysed elements. The results of the mass balance show that the SMCR retains a significant amount of biogenic elements (Table 1). The major part of elements was retained mostly in the Solina reservoir. The biogenic compounds were retained sporadically in the Myczkowce reservoir due to the hydrologic factors, i.e. feeding of a hypolimnion with N, C, P and Si-reach waters. The relationships presented in Table 2 show that retention of biogenic elements within the reservoir complex (mostly the Solina reservoir) results not only from the hydrologic or functional factors of the power station (storage of waters during the spring period and discharging during low water periods in summer and autumn), but also from inclusion of easily assimilative forms to the trophic chain and various chemical transformations. The fact that correlation between TN, TP and DSi, which are retained mostly in the easily assimilative forms (Table 1), is significant, it confirms that mechanisms of assimilation by water organisms are crucial for retention of elements in the studied reservoirs.
The distinctive feature of the mass balance was DSi depletion from the water body in the reservoirs, mostly in the Solina reservoir (Fig. 2). Approximately 20 % of the inflowing load of DSi was retained in the studied reservoirs. Humborg et al. (2006) reports that the DSi load flowing off from 1 km2 of the River Vistula basin to the Baltic Sea amounts to 0.8 t per year. Average annual load of DSi feeding the SMCR amounts to 1947 t, which is equivalent to 1.5 t of flow off from 1 km2 of the basin. Hence, anticipating that silicon originates only from sources connected with soil erosion (Garnier et al. 1999; Humborg et al. 2002) and that the DSi load level produced in other areas of the Vistula basin is similar, ca. 50 % of dissolved silicon is retained within the area of the Vistula basin, unfavourably reducing loads feeding the Baltic sea. This decrease leads to deterioration of seawater quality due to the deficiency of DSi when compared to other biogenic compounds from anthropogenic sources. The said phenomenon leads to imbalance between diatoms and other algae (Humborg et al. 2000). A similar decrease in DSi load in dammed reservoirs was described by Garnier et al. (1999), while Humborg et al. (2006) conclude that the cascade design of dammed reservoirs on rivers increases HRT and favours retention of dissolved silicon, depleting it from downstream waters. 20 % of DSi retention is a distinctive feature of oligotrophic waters (Garnier et al. 1999), but the data available in the literature (Garnier et al. 1999; Humborg et al. 2002, 2006) also show that similar levels of DSi retention were seen in both oligo- and eutrophic ecosystems. In the studied case, depletion of dissolved silicon in the surface lake body, related in turn to DSiret affects the water quality, stimulating growth of algae. The correlations shown on Fig. 3 may confirm that increase of chlorophyll level can be related with emergence of non-diatomic (green) algae. Diatoms are seen in lake and reservoir waters mostly in spring, but also even in late winter (Humborg et al. 2000; Lehmann et al. 2004). By analogy to the condition of Lake Lugano (Lehmann et al. 2004) it can be concluded that DSi level exceeding 0.7 g m−3 in the epilimnion of the Solina reservoir contributes to chl a level related with the presence of diatoms and green algae. Below this level, in summer, a rapid increase in chl a, reaching even as much as 12 mg m−3, is observed which, in turn, may lead to occurrence of thermophilic cyanobacteria with concomitant disappearance of diatoms. Despite low water temperature in the Myczkowce reservoir, an elevated level of chl a was seen, but the index of >2.5 mg m−3 was noted only in 2005, while in 2006 it was low. In general, phytoplankton production in this reservoir is minor. However, due to the poor silicon feed, silicon shortages can occur, which lead to minor tides of algae, mostly in the warmer water area of the dam, analogous to those present in the Solina reservoir (Koszelnik 2013). A decrease in water DSi below 1 g m−3 was seen in 2005, but not in 2006. In addition, when silicon was almost completely depleted from the water body, a decrease in Si:N and Si:P ratios (see Fig. 4) was seen, and DSi became the limiting element. With silicon shortage present, phosphorus and nitrogen are the main substrates utilised in production of organic matter in the reservoirs. In turn, the value of N:P molar ratio (Fig. 4) significantly exceeding 16:1 proves the stoichiometric excess of nitrogen versus phosphorus.
Approximately 28 % of phosphorus supplied to the studied reservoirs is retained. The process is occurring mostly in the Solina reservoir. This result depends on various factors allowing storage of phosphorus from the water body in the benthic deposits, related in general with an affinity to specific metals, presence of aerobic conditions or with pH (Golterman 1998). Phosphorus retention is a result of sedimentation of solid particles introduced to the reservoir with affluents and assimilative forms incorporated to the biomass of phytoplankton and transferred to sediments (Hejzlar et al. 2009; Dunalska et al. 2013). A part of such retained phosphorus can be released again to the water body due to resuspension or decomposition of bonds with iron or other metals in anaerobic conditions. The benthic deposits of the Solina reservoir are rich in metals with affinity to phosphorus, mostly in iron. Retention volume of these deposits is very high (Bartoszek et al. 2009), and favourable aerobic conditions make release of phosphorus from the deposit practically absent in both the reservoirs. On that basis it can be concluded that the principal phosphorus retention mechanism in the reservoir is a direct sedimentation of phosphorus contained in suspension and intermediate sedimentation of mineral forms, after their assimilation into the trophic chain. Storage properties of the benthic deposits are large enough so that the calculated Pret value could be greater, but the morphometry of the Solina reservoir (high depth, low area of an active bed) determines the identified level. Consequently, upon balancing of average phosphorus mass retained in the reservoir it was anticipated that the phosphate phosphorus retention will be equal to the amount of this element utilised by phytoplankton, and the difference between retention of P-PO4
3− and TP will be equal to the amount of element originating from external sources and accumulated in the deposit:
For the complex of reservoirs (t year−1)
|
Inflow
|
Retention
|
Partial P sedimentation
|
Sedimentation P accumulated in biomass
|
|
91
|
23
|
11
|
12
|
A complexity of biogeochemical nitrogen changes in the water environment affects the retention level of this element. In contrast to Si and P, nitrogen has a larger gaseous phase, and denitrification leading to change in state of aggregation affects the mass balance of this element in water ecosystems. In the studied period of 3 years, decrement of load inflowing to the reservoirs amounted only to 9 % per year. This value was affected by a significant element elimination in 2006. In 2005 Nret% was equal to ca. 22 % of the introduced load. Studies on nitrogen retention, carried out between 1999 and 2003, reveal that the retention of this element was varying significantly, amounting to from 12 to 36 % of the annual load (Tomaszek and Koszelnik 2003). Previous studies show that the rate of denitrification in the Solina reservoir is stable and amounts to ca. 5 g m−2 year−1 (Koszelnik et al. 2007), which corresponds to ca. 20 % of retention and 5 % of nitrogen load. Empirical models of denitrification in conditions occurring in the Solina reservoir, contingent on presence of nitrates and temperature (Gruca-Rokosz 2005, PhD thesis, unpublished data), were utilised to analyse various nitrogen retention mechanism. Estimated denitrification rate was 4 g N m−2 year−1, and on that basis it was calculated that, on average, 70 t of N is denitrified annually. Nitrogen sedimentation calculated from the balance mass equals to 110 t year−1, hence the balance of retained nitrogen is as follows:
For the complex of reservoirs (t year−1)
|
Inflow
|
Retention
|
Sedimentation
|
Denitrification
|
Indefinite
|
|
1957
|
174
|
110
|
70
|
−6
|
Share of denitrification process in nitrogen retention amounted to 40 % and was twice as high as that calculated for previous years (Koszelnik et al. 2007). However, load reduction is significantly lesser (similar to Nret%) and was only 3.6 %.
Influence of denitrification on the nitrogen mass balance depends on various factors. Seitzinger et al. (2002) describes that for North American estuaries 50 % of nitrogen is supplied from denitrification. Estuaries are bodies of water similar in many cases to dammed reservoirs; mostly due to the ratio between areas of basin and water table, retention time or biogenic compounds load. The said value can be real, but in many cases—including estuaries—significantly lower values are seen, i.e. 5–30 % (Dudel and Kohl 1992; Koszelnik et al. 2007; Povilaitis et al. 2012), but also values as high as 70 % are reported (Mengis et al. 1997). A comprehensive analysis of the available data presented in previous papers (Koszelnik et al. 2007) enables one to conclude that in water regions with high hydraulic dynamics contribution of nitrification to the mass balance is minor when compared to natural lakes, where this process can be significant.
Nitrogen sedimentation, in a way similar to phosphorus, is a result of its consumption. Thus, calculation of Nsed and Ncons contributions can be difficult. Jickells et al. (2000) states that in inland waters, retention mainly consists of storage of organic nitrogen produced within the ecosystems in the benthic deposits. In the studied case retention of assimilative forms, mostly nitrates, is equal to ca. 100 t year−1. Nevertheless, it should not be anticipated that the overall mass of retained NO3
− will be assimilated. Some part of it will be denitrified, as the main substrate for the said process, occurring in the anoxic layer of the benthic deposits, are nitrates(V) diffusing from water (Tomaszek and Gruca-Rokosz 2007). A surplus nitrogen seen in the above balance (−6 t) can result from utilisation of nitrogen stored in the deposits in the nitrification process, which, after various transformations, can be denitrified.
The mass balance of total organic carbon calculated for the reservoir complex shows that annually approximately 442 t of TOC is retained. The calculated TOC sedimentation is three times higher and amounts to 1300 t year−1. Hence, it should be recognised that a significant part of sedimentation matter is produced within the ecosystems:
For the complex of reservoirs (t year−1)
|
Inflow
|
Retention
|
Sedimentation
|
|
2206
|
442
|
1300
|
Lentic waters are characterised by a high carbon retention capability, as the major part of TOC supplied to lakes and reservoirs is respirated and included in the trophic chain (Garnier et al. 1999). The above is true for both deep and shallow reservoirs. Anderson and Sobek (2006) provide an example of a shallow lake in Sweden, in which the annual carbon load amounts to approximately 3 t. Calculated phytoplankton production for the lake is as high as 53 t C per year, and the macrophyte production—16 t C per year, while carbon sedimentation is three times greater than the carbon inflow to the ecosystem. Unlike the other elements, TOC retention in the Myczkowce reservoir was fairly high (7–8 % of the load), which can be explained by carbon uptake by macrophytes after its respiration. Rate of decomposition for these forms is low, and annually they release only 40 % of the retained organic carbon (Gessner 2001).