Comparison of the zooplankton-based RCC to Carlson’s trophic state indices and water quality parameters

Zooplankton is not included as a biological quality element, even though it is a key component of pelagic food webs. The data on zooplankton community structure can be used to evaluate the trophic state of lakes through an RCC Index, based on quantitative proportions between the main zooplankton groups. As previous investigations have related the index only to the Secchi disc transparency, the aim of this study was to compare the Index to a set of water quality parameters and trophic state indices. Two sandpit lakes with contrasting characteristics were selected as model lakes for the study. A total of 57 samples were collected simultaneously with the physicochemical water quality parameters between 2016 and 2020. Carlson’s trophic state indices were used to assess the trophic state of the lakes. Pearson correlation coefficient and Principal Component Analysis were used to assess the observed correlations between the measured parameters and to distinguish what factors regulate zooplankton community structure. Overall, 86 zooplankton taxa were identified. RCC Index ranged from 2 in the small sandpit lake to 87 in the big sandpit lake. A highly significant negative correlation was found between the RCC Index and the values of TP and COD. The results complement the arguments for the inclusion of zooplankton as a BQE for the assessment of the status of lakes.


Introduction
Since 2000, the Water Framework Directive has been the main law for water protection in Europe applied to control the quality of surface water ecosystems.According to this directive, the ecological state of the water bodies should be estimated by hydro morphological and physical-chemical elements simultaneously with a set of biological quality elements (BQE): phytoplankton, macrophytes, phytobenthic species, invertebrate benthic animals, and ichthyofauna.The zooplankton community is not included in the list as an obligatory biological quality element, even though it is considered a key component of pelagic food webs and no scientific reasoning exists to exclude it as a BQE [1,2].Moreover, zooplankton abundance correlates well with the total phosphorus concentration, and the top-down control of zooplankton on phytoplankton may mask high production situations in lakes if only phytoplankton are considered [3] To be identified as a good ecological indicator the BQE has to respond quickly to changes in the environmental conditions.The position of the zooplankton community in the food webs determines its role as a regulator of the relationship between upper and 1305 (2024) 012007 IOP Publishing doi:10.1088/1755-1315/1305/1/012007 2 lower trophic levels.Due to the differences in life cycles of the three main zooplankton groupsrotifers, cladocerans, and copepods, they respond within weeks to changes in nutrients and phytoplankton abundance [4].Under eutrophication, the zooplankton community reorganizes its structure towards a predominance of rotifers [5,6].The rotifers have great importance in freshwater ecosystems as their reproductive rates are the fastest among the metazoan zooplankton.Because of their adaptability, they are widely distributed at densities that can exceed 1×10 6 ind.m -3 .Rotifers are also known to be an important link between the microbial loop and higher trophic levels in the ecosystems [7].Therefore, many authors suggest that zooplankton should be used as an indicator in assessing the trophic state of lakes [2,3,6,8].
Kozuharov et al. [9] proposed an index (RCC) based on the structure of the zooplankton community for the evaluation of the trophic state of lentic ecosystems.The index is easy to apply as it is based on the main groups in the zooplankton community, it's cost-effective and time-saving as it does not rely on tedious species determination.
Several investigations applied the index to assess the trophic state of a diverse set of water bodies, such as the study of Ilieva et al. [10] who used it in glacial lakes and confirmed the biological classification of mountain lakes concerning their evolution, published by Naidenow [11].Stanachkova et al. [12] and Stefanova et al. [13] used the RCC Index to describe the process of purification in the ecotone zone between Iskar River and Iskar Reservoir.All investigations applied the index to a diverse set of water bodies but correlated the results only with Secchi disc transparency.According to Carlson [14], chlorophyll-a and total phosphorus are preferred indicators when predicting trophic state, and transparency should only be used if there are no better methods available.The aim was to study the relationship between the RCC Index and the water quality parameters in relation to the trophic state of water bodies, defined by Carlson's trophic state indices (TSI-Chl and TSI-TP).

Materials and Methods
Two sandpit lakes (the Big and the Small Negovan Lakes) were selected as model lakes for the study, based on their different hydrological environment, exploitation history, and different degree of degradation.The lakes are situated near Sofia city, western Bulgaria (Figure 1) on the territory of the former Negovan marsh, which was drained and converted into arable land by the end of the 1930s.This region is the lowest point in the watershed of the Lesnovska River, which is characterized by a winter-spring to spring high water period (February or March and March-June) from snowmelt and rains and a long summer autumn low water period (July-November).The groundwater level is at 0.5-1.5 m below ground and fluctuates insignificantly during the year -30-60 cm.After heavy rains, groundwater rises to the surface and causes overflow of the lakes and seasonal swamping of the adjacent areas.The groundwater level is influenced by water extraction and the presence of sand and gravel pit lakes.Under natural conditions, the groundwater level rises from the autumn-winter to the spring months (May/June).In the past, the most significant changes in river discharge and groundwater level occurred as a result of seasonal changes in rain and snowfall, often resulting in floods, thus maintaining the Negovan marsh.After the canalization and embankment of the river, the marsh was drained and parts of it converted into sand and gravel pit lakes.The Small Negovan Lake (SNL) is an abundant sand pit quarry with a maximum depth of less than 4 meters, while the Big Negovan Lake (BNL) is a deep (up to 18 m), and still in operation, sandpit lake.In 2022 the SNL was declared a protected area which is home to rare plant and animal species.
The study was conducted between 2016 and 2020 on three sampling points in the SNL and one in the BNL.A total of fifty-seven samples were collected (Table 1).On each occasion, samples were collected for physicochemical water quality elements, according to Regulation H-4 [15], and the Water Framework Directive.Field measurements of temperature, dissolved oxygen, pH, and specific conductance were made at each sampling station.Carlson's trophic state indices (TSI) were used to assess the trophic state of the lakes [14,16].Water samples were taken for laboratory determination of total alkalinity (ISO 9963-1) and nutrient concentrations: total (TP) and phosphate phosphorus (PO4-P), ammonium (NH4-N) nitrogen, nitrate (NO3-N) and total nitrogen (TN).All determinations were done with MERCK test kits on the WTW photoLab7100VIS spectrophotometer.Biological variables included chlorophyll-a (ISO 10260) and zooplankton samples collected, determined to species level, counted, and divided into the three main groups: rotifers, copepods, and cladocerans.The RCC Index [9], calculated based on the zooplankton community data was compared to the Carlson strophic state indices.Differences in biotic and abiotic parameters between the lakes were estimated by a t-test.Pearson correlation coefficient was used to measure the degree and direction of the correlation between the measured parameters.Also, Principal Component Analysis (PCA) was applied to find what factors regulate zooplankton community structure, expressed by the RCC Index, and its relation to the trophic state of the lakes.
Permission for the field study conducted on public land was provided by Sofia Municipality contract CO.RD55-276/07.05.2015.

Species composition
Overall, 86 taxa were identified belonging to 3 phyla and 1 subphylum divided into three main groups: the highest richness was found among the rotifers -64 species, cladocerans were 13 species from four families, and copepods were presented by one calanoid and three cyclopoids (Table 2).In addition, veliger larvae from zebra mussel Dreissena polymorpha (Pallas, 1771) [17] were found in both lakes.The dominant species complex includes cosmopolitan rotifers such as Keratella cochlearis and Keratella quadrata.The species from the family Synchaetidae: Synchaeta pectinata, Polyarthra vulgaris, Polyarthra dolichoptera, and Polyarthra minor, which are planktonic and eurytherm species [18], also have a significant share in the composition of the zooplankton community in both lakes.Anureopsis fissa was not very common during the studied period, but in the samples from June 2020 was dominant in the small lake.An important feature in the taxonomic structure of the zooplankton in the SNL is the significant prevalence in the rotifer complex of fam.Lecanidae, represented by six species and of gen.Brachionus is represented by four species, all typical for eutrophic lakes.From the cladocerans, Bosmina longirostris was found in both lakes, but often dominated in the small lake.The family Chydoridae and Ceriodaphnia quadrangula were observed only in the small lake, as the extensive macrophyte beds [19] and shallow waters favour the development of those species [4].An interesting feature of the SNL is the complete lack of species of the genus Daphnia, while Daphnia galeata, D. longispina, and D. cucullata were found regularly in the BNL.A possible reason could be the inflow of agricultural chemicals, nutrients, and fertilizers from the cultivated areas surrounding the lake, as the species of the genus Daphnia are highly sensitive to pollutants [20].The appearance of the predatory Leptodora kindtii in the big lake further affects the zooplankton community structure.The relationship between copepods and the trophic state has been less studied [21].In continental waterbodies, copepods are usually less diverse than the other groups; in this study, only 3 species were found -Eucyclops serulatus, Acanthocyclops robustus, and Cyclops vicinus vicinus.In general terms, cyclopoid copepods are more abundant in eutrophic systems, while, in general terms, calanoids prefer oligotrophic clear water [22,23].Eudiaptomus gracilis from Calanoida was dominant in the BNL, while Cyclopoida occurs predominantly in the SNL indicating a difference in the trophic state between the lakes [24].

Zooplankton community, RCC Index, and relation to the environmental characteristics
RCC Index ranged significantlyfrom 2 in the SNL to 87 (April 2018) in the BNL.The values of the index were always higher in the BNL compared to those for the same period in the SNL.The average values of RCC in the BNL are 76, suggesting a good ecological state of the water body, as high values of the index are typical for oligo-mesotrophic conditions [9,10,12,13].The share of Copepoda and Cladocera was higher than rotifers in BNL throughout the entire investigation period due to the presence of the calanoid Eudiaptomus gracilis and Cyclops vicinus vicinus from order Cyclopoida.Cladocerans D. galeata, and D. longispina appeared only in BNL, as they are associated with deeper water bodies.By contrast, Bosmina longirostris, Alona spp., and Chydorus sphaericus preferred eutrophic environments like the SNL.The lowest RCC values (28) in the BNL were registered in June 2020 (Figure 2), due to the abnormally wet spring months, with an average rainfall of 150% of the monthly norm.7 lake and have similar taxonomic and physicochemical characteristics.The stable conditions in the old parts of the lake support a stable zooplankton community dominated by rotifers, leading to analogy, also in the RCC and trophic state indices.The low RCC values in SNL suggest an increased trophic state in the lake, throughout most of the year.A good zooplankton indicator species, supporting the higher trophic state are Keratella cochlearis, Anureopsis fissa, Conochilus sp., and the genus Testudinella [7].Table 3. Correlation matrix between the studied variables: NS: number of zooplankton species; zooplankton abundance: ind.m -3 ; RCC Index; Physical and chemical parameters measured in water samples for each site along the sampling period: PO4P; NH4N; NO2N; TP; NO3N; TN; Alkalinity; Water Temperature; O2%; O2 mg/l; Electroconductivity; pH; Chl a; COD; TSI-TP; TSI-Chl; Phyto volume; Cyanobacteria; A correlation matrix, showing the relationship between the biotic communities and the environmental variables is given in Table 3.The results show a strong negative correlation between TP and RCC, COD and RCC, as well as TSI and RCC indices.The significant values are marked in bold.Environmental parameters outline the frame of ecological state but the assessment would be more complete if a biological quality element is included in the evaluation of the water quality.Phytoplankton has been documented as being highly sensitive to alterations in the nutrient concentrations in the water and, it is so far the only planktonic biological element set by WFD for lentic ecosystems.The TSI-Chl-a shows a good correlation with RCC, as well as with most of the observed shifts in the zooplankton communities, which coincide with changes in the trophic status.The relation between the two indices was significant with r = -0.68 and p-value<0.001.In general, Carlson's trophic state indices (SD, total phosphorus (TP), and Chl-a) should produce the same results in well-balanced, deep lakes.However, in shallow lake ecosystems, Carlson's trophic state indices often show high discrepancies between each other, due to the stronger influences of factors that limit algal biomass or affect the measured variables [16,21].One such variable is the zooplankton community, as it is a grazer on algae and bacteria, influencing their abundance, but at the same time, they also provide phytoplankton with nitrogen and phosphorous through nutrient recycling [25].
The correlation matrix (Table 3) shows that the RCC index and TP were strongly and negatively correlated (r = -0,72; p <0.001).Similar results, where TP was negatively correlated to the zooplankton diversity and rotifer species richness, have been previously reported [26].The chemical oxygen demand, commonly used to indirectly measure the amount of organic compounds, negatively correlated (r = -0.67;p <0.001) with RCC in the studied lakes.Dependence between COD and zooplankton community is weakly studied and gives reason for future research.Temperature was also significantly correlated with RCC, with r = 0.68; p <0.001).
Zooplankton abundance was also strongly correlated with the nitrogen (TN, NH4-N, NO2-N, and NO3-N) and PO4-P, but not with the water temperature and TP.This is because the lakes are surrounded by agricultural fields [27] supplying nutrient reach inflows which influence the zooplankton abundance [28,29], namely the observed spring peaks in the abundance of the rotifers in the SNL [30].The zooplankton abundance showed a positive significant relationship to the phytoplankton abundance, expressed as Chl-A (Table 3).The lack of daphniids and the prevalence of small-sized zooplankton are probably responsible for the decreased relationships between the parameters in comparison to the observed correlations in big lakes and reservoirs [31,32].The size structure and dynamics of the zooplankton community are affected not only by the trophic resources but also by the relation between zooplankton and fish.The cascading effect of selective fish predation changes zooplankton community structure and thus may influence the phytoplankton abundance [25].
There is a lack of information on fish populations in both lakes.Uzunova et al [33] have studied the ichthyofauna with a focus on the population dynamics of the invasive species Lepomis gibbosus L., concluding that the fish populations in the BNL are unstable in the operational sand pit lake due to the unstable environmental conditions, the lack of macrophyte vegetation in the littoral and suitable breeding habitats.On the other hand, the SNL supports fish populations with relatively stable numbers and growth rates, but with skewed toward small body size classes due to the intense angling.According to the PCA ordination, the samples from BNL are separated from the SNL data set.The sampling points in the SNL are further separated based on restored vs. old habitats.Thus, both stations in the original sandpit lake -NM1 and NM2, form a cluster with similar RCC and trophic state indices.This group is due to the similar conditions in the old parts of the lake, supporting a stable zooplankton community dominated by rotifers (low values of RCC).The station located in the newly created habitats (NM3) is significantly different and according to the physic-chemical parameters and the trophic state is positioned between the old eutrophic habitat stations and the deeper, more mesotrophic station in the BNL.The results reflect the gradual shifts in the trophic state of the newly constructed habitats with the corresponding succession of the biotic community.

Conclusion
Due to anthropogenic influences, water quality in many lakes is deteriorating, subsequently affecting zooplankton community structure, namely their composition and abundance.The obtained results showed the ability of the RCC index to reflect the changes in the trophic state in the studied IOP Publishing doi:10.1088/1755-1315/1305/1/0120079 lakes.Although additional work is required to verify the relationship in water bodies from different lake types, the preliminary results complement the arguments for the inclusion of zooplankton as a biological quality element for the assessment of the ecological status of lakes.

Figure 1 .
Figure 1.Sampling sites in the Big Negovan Lake and Small Negovan Lake.Google Earth history August 2020

Figure 2 .
Figure 2. RCC Index of the studied lakes for the entire period Trophic state indices varied widely in SNL, especially at the sampling point located in the newly constructed habitats (MN3), due to the more dynamic environmental conditions in the first years after the restoration of the lake.The other two points: MN1 and MN2 are situated in the old parts of the

Table 1 .
Sampling calendar of the survey