Effects of nutrients enrichment on algal communities : an experimental in mesocosms approach

Aim: The effect of different concentrations of nutrients on algal communities living at the bottom of mesocosms was examined; Methods: The experiment was carried out in the tributary of Rosana reservoir (Corvo river) from November 22nd to December 16th of 2003. Mesocosms were set up in triplicate, with four treatments: nitrogen (N); phosphorus (P); nitrogen plus phosphorus (NP); plus a control treatment (C), consisting in addition of water from the reservoir; Results: Results of the Principal Components Analysis applied on the abiotic variables have shown that two axes explained 93.6% of the total variability (69.6 and 23.9%, respectively). It was possible to identify a group of P (positive) and N (negative) treatments of axis 1. The addition of nutrients in the experiment resulted in changes in algal composition. Mean values of density ranged from 8.383 to 34.510 ind.mL–1. Higher densities occurred in the phosphorus and nitrogen plus phosphorus treatments. The most representative classes were Zygnemaphyceae, Chlorophyceae, Bacillariophyceae and Cyanobacteria, both in density and species richness; Conclusions: Phosphorus seems to be beneficial to green algae in general, since a marked increase in Zygnemaphyceae and Chlorophyceae was found in the treatments containing phosphorus and nitrogen plus phosphorus. In the nitrogen treatment, however, green algae decreased dramatically, while Cyanobacteria remained constant, indicating their capacity to live in changed environments and confirms the limiting role played by phosphorus on the algal groups.


Material and Methods
The Rosana reservoir is located at the final stretch of the Paranapanema river, locally known as "Pontal do Paranapanema".The selected site to conduct this study is a tributary of this reservoir (Corvo river 22° 39' S; 052° 46' W, Figure 1), which is located near the reservoir dam, in the State of Paraná.
The experiment was carried out from November 22 nd to December 16 th 2003.Twelve mesocosms were designed for four treatments (triplicate, stage 1, Figure 2) built of polyethylene bags and fixed randomly on the lake 2 m distant one from another to prevent contamination among them.They were suspended in the water column (1 m depth) by a float attached to a permanent buoy.Mesocosms with a capacity of 1000 L and a depth of 1 m were installed on November 21 st , without contact with the sediment, filled with water from the reservoir itself (stage 2-3, Figure 2).The mesocosms were set up in triplicate, with four treatments: a control treatment (C) consisting on the addition of water from the reservoir = 2 µg N-NO 3 -L -1 and 4 µg P-PO 4 3-L -1 ; Nitrogen treatment (N) added = 1000 µg N-NO 3 -L -1 ; phosphorus treatment (P) added = 40 µg P-PO 4 3-L -1 and nitrogen plus phosphorus treatments (N+P) added = 1000 µg N-NO 3 -L -1 + 40 µg P-PO 4 3-L -1 .Nutrients were introduced in the treatments only in November 22 th (stage 4, Figure 2); in the initial phase of experiment and control treatment were these existing concentrations in water of the environment (Figure 2).
For determination of biological variables (algal density and species composition) samples were obtained by taking 5-10 mL from the bottom of each mesocosm after emptying them carefully, during the last day, on December 16 th (stage 5, Figure 2).The samples were fixed and preserved with Lugol's solution.Quantitative study was carried out according to Utermöhl (1958), using an Olympus brand CK-2 inverted microscope.The algae were quantified (cells per liter) in the sedimentation chamber using the random fields method, until the rare species curve was reached according to Bicudo (1990).Only living algae were counted for the densities calculation and algae filaments were counted as filament numbers.The taxa considered abundant were those with densities higher than the average of each sample (Lobo and Leighton, 1986).
Abiotic data were collected simultaneously with the biotic data, and were supplied by Laboratório de Limnologia Básica, from Núcleo de Pesquisa em

Introduction
In the decade of 1960 and 1970, two perspectives of the relationship between nutrient levels and algal growth (ecosystem and an autecological approach), propitiated new insights about nutrient cycling, anthropogenic eutrophization, algal community structure, and interactions among algae and other organisms (Borchadt, 1996).A perspective of the problems related with the eutrophization can be gotten from the analysis of raised resources required to development of algae, since phosphorus and nitrogen are probably the most likely to be limitant of algal growth (Borchadt, 1996).
Experimentally, biomass has a positive response to increased water nutrients availability, considering the status of nutrient from periphyton a good predictor (França et al., 2009).The structure of the periphytic algae communities can be altered by artificial addition of nutrients (Murakami and Rodrigues, 2009).In addition the species structure (chlorophytes and diatoms) were sensible to phosphorus availability and P-addition (Ferragut and Bicudo, 2009), showing that the phosphorus, especially, may be considered a driver of periphytic algae (Ferragut and Bicudo, 2009;Murakami and Rodrigues, 2009).Preliminary synthesis of the evaluation of nutrient limitation to algal growth reveal phosphorus as the most limiting element in reservoirs (Huszar et al., 2005).
In Brazil, the effect of enrichment with N and P on attached algae (periphyton), either in biomass response (Cerrao et al., 1991;Engle and Melack, 1993;Mendes and Barbosa, 2002;França et al., 2009;Oliveira et al., 2010), or on the community structure approaches (Murakami and Rodrigues, 2009;Ferragut andBicudo, 2009, 2010) are still scarce.Thus, this study broaden the understanding of the influence of concentration of nutrients in the structure of algae at the bottom of mesocosms artificially enriched with individual or combined actions of N and P.
Considering that nitrogen and phosphorus constitute the most important limiting nutrients to algal growth, we aimed to assess the effects of different concentrations of nutrients on algal communities living at the bottom of mesocosms.If algal growth is dependent upon nutrients, then we hypothesize that the algal community -density and species composition, should be positively correlated with nutrient concentrations.out to verify if the variables satisfied the assumptions of normality and homoscedasticity, respectively, and were performed through the software Statistica 7.1 (StatSoft, 2005).When such differences were observed, Tukey's Test was carried out.

Results
The water physical and chemical parameter values, water temperature, conductivity, dissolved oxygen, and turbidity showed steadier values, while other variables, such as nutrients, showed greater variation.Total nitrogen and ammonium ion were higher at mesocosms with nitrogen and nitrogen plus phosphorus treatments while total phosphorus was higher at mesocosms with phosphorus and nitrogen plus phosphorus treatments (Table 1).
Results of the Principal Components Analysis applied on the abiotic variables are shown in Figure 3 and Table 2. Two axes were retained for interpretation and together they showed 93.6% of the total variability (69.6 and 23.9%, respectively).It is possible to identify a group of P (positively) and N (negatively) treatments of axis 1.The highest values of orthophosphate and total dissolved phosphorus, corresponding to the P treatment while Limnologia, Ictiologia e Aquicultura -Universidade Estadual de Maringá (NUPELIA/UEM).The following physical and chemical water variables were analyzed: water temperature, electric conductivity, dissolved oxygen (Golterman et al., 1978), turbidity, alkalinity (Golterman and Clymo, 1971).Total suspended material (Wetzel and Likens, 1991), total nitrogen (Valderrama, 1981), ammonium (Solorzano, 1969), total dissolved phosphorus and orthophosphate (Golterman et al., 1978), and total phosphorus (Valderrama, 1981).
Principal Components Analysis (PCA) was used to reduce dimensionality of abiotic (12 variables) data, which were log-transformed, according to equation log (n+1).The software used was PC-ORD version 5.0 for Windows (McCune and Mefford, 2006).Principal components axes retained for interpretation were those that presented eigenvalues higher than the eigenvalues produced by the Broken-Stick model (Jackson, 1993).
In order to compare the mean values for density in the different treatments, we applied an analysis of variance (one-way ANOVA) with significant level (p) = 0.05, using the software StatSoft 7.1.Kolmorov-Smirnov and Bartlett Tests were carried the highest values of molar ratio and ammonium ion corresponding to the N treatment (Figure 3).
The second PCA axis separated the treatments mainly with respect to C (positively) and NP (negatively).The highest values of total phosphorus and total suspended material corresponding to NP treatment, while orthophosphate correlated with highest values in C treatment (Figure 3).
A total of 200 taxa were identified, distributed among 11 taxonomic classes (Bacillariophyceae, density values were observed in the N+P treatment and also in P treatment (Figure 4).
In the treatment involving the combined addition of N+P, the most abundant species corresponded when comparing treatment with P and N, in which both classes had opposed density.Chlorophyceae presented more density in P treatment, decreasing drastically in N treatment (Figure 5).

D r e p a n o c h l o r i s u h e r k o v i c h i i M a r v a n ,
Cosmarium regnesii Reinsch, Staurastrum tetracerum (Kütz.)Ralfs, Gloeocapsa sp. and Staurodesmus dejectus (Bréb.)Teil.were the most abundant species in the community, contributing with 52% (Figure 6).1999; Stelzer and Lamberti, 2001).In the study, addition of nutrients N and P, either individually or in combination, affected the density of algae community.
Treatment with the addition of P was associated with greater availability of P, since that ration N:P was equal 7:1 and Zygnemaphyceae was more predominant.Despite the ration N:P add much (19:1) in the treatment combined with N+P, the green algae continued more predominant, represented primarily by Chlorophyceae and then Zygnemaphyceae.In literature there is a discussion regarding to what is optimum for algae growth in relation at the N:P ration.Green algae attain their maximum biomass at the N:P ratios = 20:1 to 81.2% total density.Cosmarium regnesii, Monoraphidium cf.contortum, Fragilaria capucina, Drepanochloris uherkovichii, Staurastrum tetracerum, Pseudanabaena cf.moniliformes, Trachelomonas sp., Monoraphidium arcuatum (Korsh.)Hind., Gloeocapsa sp. and Nitzschia cf.palea (Kütz.)Smith, what contributed to 63% of the algal community (Figure 6).

Discussion
When algae compete for phosphorus and nitrogen separately, absolute quantities of these elements should be responsible for the dominance of some taxa and elimination of others (Grover, 1989;Borchardt, 1996; Bulgakov and Levich, Trachelomonas sp.

Cosmarium regenesii
Ind.mL -1 allows the Cyanobacteria to effectively absorb light energy and live in environments with low intensity of light.Therefore, this class has a competitive advantage in more turbid lentic environments, and is also capable of growing "in the shade" of other organisms (Chorus and Bartram, 1999).In Rosana reservoir tributary, without the addition of nutrients, Cyanobacteria were prominent and represented one of the most important components in the phytoplankton (Borges et al., 2008).Many Euglenophyceae species are inhabitants of interface (water-air, sediment-water), much probably with a competitive advantage due to their periphytic habit (Rosowski, 2003).This possibly favored the abundance of Trachelomonas sp. at the bottom of treatments with the addition of N and N+P.
Diatoms were the dominant algal group on periphyton community structure, with several taxa responding primarily to nutrient ratio and total nutrient concentration (Stelzer and Lamberti, 2001).During the experiment, in the competition of the species for nutrients, Grover (1989) suggested that Synedra radians and Fragilaria capucina were good competitors for P. In the present study, Fragilaria capucina was highly abundant, mainly in P and N+P treatment, when N:P = 7:1 and 19:1, respectively.In environments with greater availability of assimilable forms of nitrogen Stelzer and Lamberti (2001) registered higher densities of Gomphonema parvulum.In the present study, these species were more abundant at high total nutrient concentration.
Nitzschia cf.palea was one of the taxa abundant in the treatment with N+P addition.In experiment with the addition of nutrients in urban reservoir, Ferragut and Bicudo (2009) found that Nitzschia palea and Monoraphidium arcuatum were the most abundant species in the P treatment, while M. contortum was abundant in the N+P treatment.According to the author, Nitzschia palea was associated with greater phosphorus concentration.Van Dam et al. (1994) also commented on the preference of Nitzschia palea for more eutrophic environments.
Based on the results of this study we can conclude that P played a limiting role on the Bacillariophyceae and Cyanobacteria, while on green algae in general, P and N+P treatments showed a marked increase, especially of classes Zygnemaphyceae and Chlorophyceae.In the N treatment, green algae decreased dramatically, while Cyanobacteria remained steady, demonstrating that and 50:1, while Cyanobacteria and diatoms have an optimum at N:P ratio = 2-5:1 (Bulgakov and Levich, 1999).Schindler and Smith apud Levich and Bulgakov (1992) indicated that, at the atomic N:P ratio less than 25:1, the Cyanobacteria prevails in the phytoplankton, meanwhile the increase of this ratio up to 25:1 and more leads to predominance of green and diatom algae.We found a different relationship, once that green algae occurred in treatments with ration N:P that varied between 7-19:1.According to Huszar et al. (2000) and Biggs (1996), Chlorophyceae and Zygnemaphyceae are represented especially by ubiquitous species common in eutrophic environments.Consequently, the relative quantity of essential nutrients required for growth and reproduction is different among the species.
The inflow of phosphorus promotes an increase of Chlorophyceae, because it can assimilate and accumulate P from the sediment faster when resuspended, revealing that sediment is an important source of P (Havens et al., 1999;Dodds, 2003).This group contributed the most toward species richness in all treatments.The limiting role played by phosphorus in composition of the attached algae has already been demonstrated (Ferragut andBicudo, 2009, 2010;Murakami and Rodrigues, 2009), as well as on their biomass (Oliveira et al., 2010;Mendes and Barbosa, 2002).
A great abundance of Monoraphidium cf.contortum and D. uherkovichii can be related to the higher disponibility of phosphorus.The genus Monoraphidium is commonly associated with shallow and enriched environments (Reynolds et al., 2002), high phosphorus concentrations, and low N:P ratios (Kononen, 1988;Kuosa, 1988).
Cyanobacteria can store essential metabolic nutrients within their cytoplasm, especially under excessive supply conditions.Another advantage of the Cyanobacteria is nitrogen fixation, which provides them a simple nutritional requirement, even under conditions predominantly limited by nitrogen.Most Cyanobacteria accumulate polyphosphate granules when the cells are P-rich and they are often large enough to be distinguished easily with the light microscope.These features can be useful when assigning nutrient dynamics at a site (Kelly and Whitton, 1998).In order to optimize their position and thus find a suitable niche for survival and growth, Cyanobacteria use different environmental stimulis (e.g.light, chemical compounds, temperature).In addition, the protein phycobilin, together with chlorophyll-a, the Cyanobacteria has the capacity of maintaining itself in changed environments.In addition, the data confirmed that the relative amount of essential nutrients required for development are different and should benefit certain species, since the algal community (density and composition) were positively correlated with nutrient concentrations.

Figure 1 .
Figure 1.Location of the Corvo river, tributary of the Rosana reservoir (Paranapanema river basin), where were installed the mesocosms.

Figure 2 .
Figure 2. Schedule of the design experiments with different treatments of mesocosms installed in the Corvo river (Rosana reservoir).Stage 1 -mesocosms were randomly designed (in triplicate) with four treatments: Control (C); Nitrogen (N); phosphorus (P) and nitrogen plus phosphorus (N+P).Stage 2-3 -mesocosm were filled with water from the reservoir itself.Stage 4 -nutrients were introduced in the mesocosms.Stage 5 -In each mesocosms, after emptying carefully, samples were obtained by taking 5-10 mL from the bottom.Ilustração: C.Y. Joko, E.A. Murakami & S.A. Felisberto.

Figure 4 .
Figure 4. Algal density in different treatments of mesocosms installed in the Corvo river (Rosana reservoir).Control treatment (C); nitrogen treatment (N); phosphorus treatment (P) and nitrogen plus phosphorus treatments (NP).

Figure 5 .
Figure 5. Algal density of the most representative classes in different treatments of mesocosms installed in the Corvo river (Rosana reservoir).Control treatment (C); nitrogen treatment (N); phosphorus treatment (P) and nitrogen plus phosphorus treatments (NP).

Figure 6 .
Figure 6.Abundant species contribution in each treatment of mesocosms.Control treatment (C); Nitrogen treatment (N); phosphorus treatment (P) and nitrogen plus phosphorus treatments (N+P).

Table 1 .
Values (average) and standard deviation obtained for physical and chemical parameters in each treatment.Control treatment (C); nitrogen treatment (N); phosphorus treatment (P) and nitrogen plus phosphorus treatments (NP).

Table 2 .
Correlation and performance values obtained through Principal Component Analysis (PCA) to mesocosms installed in the Corvo River.