Influence of ecological filters on phytoplankton communities in semi-arid solar saltern environments

: Objective: This study analyzed the influence of ecological filters (abiotic variables) on the phytoplankton community in hypersaline ecosystems. Methods: The abiotic variables measured herein were: pH, water temperature, salinity, ammonium ion, total nitrogen, nitrite, nitrate, total phosphorus, and soluble reactive phosphorus. The phytoplankton community was studied for density, richness and Shannon-Wiener diversity index. Data were analyzed using analysis of variance, linear regression and Canonical correspondence analysis (CCA). Results: In total, 110 taxa were identified in 3 solar salterns, distributed into 9 taxonomic classes, with the predominance of Cyanobacteria (41.8%) and Bacillariophyceae (22.7%). The species Pseudanabaena galeata was the only species sampled in all saline environments. Salinity significantly influenced the richness in the three salterns, Augusto Severo (F (1,22) = 23.2, p <0.001), Francisco Menescal (F (1,22) = 50.02, p <0.001) and Santa Clara (F (1,22) = 66.33, p <0.001). The first two CCA axes explained 41.6% of total data variability, with a negative relationship with soluble reactive phosphorus for axis 2. Conclusion: The study showed the influence of the dynamics of salterns ecosystems on the phytoplankton community structure. This is associated with filters developed by the environment, where the increasing salinity, temperature and precipitation of salts influence the composition of these organisms in the ecosystem.


Introduction
Wetlands are among the most important ecosystems in the world (Cardoso et al., 2012).These systems are considered unique owing to their hydrology and role as ecotone between terrestrial and aquatic ecosystems (Mitsch & Gosselink, 2007).Scientifically, however, little is known about these systems, mainly concerning hypersaline wetlands, which need to be studied at the level of the dynamics of ecological and economic processes and hydrogeochemical characteristics (De Medeiros Rocha et al., 2012;Costa et al., 2014).
Solar salterns are artificial ecosystems located in tropical and subtropical regions of the world, consisting of a series of interconnected evaporators (evaporators and crystallizers), in which sea or estuarine water is abstracted and transferred from one pond to another by gravity or electrical pumping (Davis, 2000;Abid et al., 2008).These environments have cultural, scientific, historical and commercial values for the local society and worldwide societies (De Medeiros Rocha et al., 2012).Sola salterns are considered wetlands due to the ecological services provided, such as artisanal fisheries, refuge zones for migratory birds, great biological diversity and the presence of endemic species of hypersaline environments (Pedrós-Alió et al., 2000;Costa et al., 2014).
These hypersaline environments show different compositions of the planktonic flora because of geographic, climatic, physical, chemical and biological factors (Williams, 1998), which work as ecological filters by selecting species that cannot develop in the environment (Weiher & Keddy, 1995;Poff, 1997).Species selection occurs when biotic and abiotic factors limit the development of species (Poff, 1997).Biological filters are related to ecological interactions (competition and predation) that ultimately influence the dispersion in the local diversity.Moreover, abiotic filters are associated with species constraints, which can only develop or adapt under certain environmental conditions (Myers & Harms, 2009).Forbes (1887) was one of the first authors to address this relationship between biological influência da dinâmica do ecossistema salino na estrutura da comunidade fitoplanctônica.Esse fato está associado aos filtros desenvolvidos pelo ambiente, onde o aumento crescente da salinidade, da temperatura, da precipitação dos sais, acaba influenciando na composição desses organismos no ecossistema.
halotolerant or halophilic species that can thrive in the extreme environment of salterns.

Study area
The study was conducted in three salterns of the Brazilian semiarid region (Figure 1), located on the margins of the Apodi-Mossoró River estuary, northern coast of Rio Grande do Norte State, Brazil.Which is characterized by a longitudinal gradient of salinity from the mouth and along the estuary, with a progressive increase in salinity; this averages 30 ± 50 g.L -1 in the upper section of the estuary, and this dynamics of salinity characterizes it as a negative or hypersaline estuary (Medeiros et al., 2010).
Samplings of the phytoplankton community and limnological variables were carried out in the Initial, Intermediate I and II sectors for each salt.In situ, 2 evaporators were chosen from each sector (the first and the last with different salinity) for sampling the plankton community.Thus, 6 evaporators at each saltern were sampled, from which 4 samples were taken, averaging 24 samples per saltern, and totaling 72 samples in the 3 studied salterns.

Environmental variables
Water temperature was measured in situ with a thermometer (TermistoTenmars TM 744R RS-232), and salinity with a portable digital refractometer (Fischer).The pH was measured by a Hanna digital pH meter (HI9224).Analyses of ammonium ions, total nitrogen, nitrite, nitrate, total phosphorus and soluble reactive phosphorus were performed according to APHA et al. (1998).

Biological data
Phytoplankton samples were taken at the subsurface with a plankton net (20 μm) by filtering 200 liters for each sample, which were placed in polyethylene vials (50 mL) and preserved in 4% formaldehyde solution for qualitative analysis.The quantitative analysis of phytoplankton was obtained by counting 2 mL of the sample in a Sedgewick-Rafter chamber (Chellappa et al., 2008(Chellappa et al., , 2009)).The result was multiplied by the volume of the analyzed sample (50 mL) then divided by the volume filtered (200,000 mL -1 ), thereby obtaining the density (ind.mL -1 ) of individuals.According to the method of   Chellappa et al. (2008Chellappa et al. ( , 2009)), the count of the individuals was used, because it uses the qualitative sample to determine phytoplankton density.
Along the saltern system, the existence of ecological filters was predicted, promoted by the salt gradient and environmental conditions.For the analyses to show the influence of the filters, we considered sampling in each saltern compartment, diversity among sectors (Initial Sector, Intermediate I and II Sectors) and the geographical location of the salterns on the hypersaline estuary.
In order to check for differences between species richness and limnological variables, a oneway ANOVA was run for each variable (Gotelli & Ellison, 2011).A simple linear regression was employed to determine the influence of the salinity gradient in each studied salterns on the species richness.Canonical Correspondence Analysis (CCA) was applied to detect relationships between density of phytoplankton species and environmental variables analyzed in the three hypersaline zones.All of these analyses were run using the vegan package (Oksanen et al., 2012) in R software (R Core Team, 2012).
In the CCA, the first two axes explained 41.6% of total data variability.Axis 1 explained 21% and axis 2 explained 20.6% (Figure 4a and b); by Monte (1%), Conjugatophyceae (1%).The predominance of Cyanobacteria and Bacillariophyceae indicate their adaptation to the saline environment.Seventytwo taxa were found at the Augusto Severo Saltern, with the predominance of Cyanobacteria and Bacillariophyta (36 taxa and 22 taxa, respectively), 64 taxa at the Francisco Menescal Saltern, with predominance of Cyanobacteria and Bacillariophyta (41 taxa and 27 taxa, respectively) and 45 taxa at the Santa Clara Saltern, with a greater abundance of Cyanobacteria and Bacillariophyta (33 taxa and 15 taxa, respectively).
Pseudanabaena galeata (Böcher) was the only species sampled in all of the saline environments.Around 96 taxa were rare species distributed in the three salterns (Table 1).The phytoplanktonic density was 0.203 ind.mL -1 , 0.481 ind.mL -1 and 0.333 ind.mL -1 , respectively, for the salterns    Carlo analysis of the CCA test, this is significant (p=0.03).In axis 1, abiotic variables (salinity, temperature, nitrogen and nitrite) showed positive correlations with the sampling sites located in the sumulum (Psi) and Scripsiella tochoidea (Sto) showed a positive correlation with the 2 axes of the CCA.

Discussion
In the present study, the phytoplankton communities were influenced by ecological filters, which were characterized as being abiotic conditions existing in the environment.The gradient of salinity and temperature in particular are the environmental variables that most influence the richness and abundance of species, as evidenced in the study by Thiéry & Puente (2002).However, along the hypersaline ecosystem, abiotic variables and the interaction between species become ecological filters, inhibiting the development of species.
The abiotic conditions of each sector selected organisms that do not have biological characteristics that allow them to survive, develop and reproduce in increasingly restrictive environments (Townsend & Hildrew, 1994;Statzner et al., 2001).This effect was evidenced by Abid et al. (2008), who showed that physical and chemical parameters (salinity, pH, temperature, total phosphorus, soluble reactive phosphorus, total nitrogen, nitrate, nitrite, and total dissolved salts), as well as the geographical location of the saltern, influence the decrease in species richness.
The passage of the biological community through the sectors will result in the selection of species that are better adapted to local environmental conditions, as observed in this study by the prevalence of two groups of classes, Cyanobacteria and Bacillariophyceae, and the development of a single species (Pseudanabaena galeata) along the salt gradient; i.e. they had managed to overcome the ecological filters by adapting to environmental conditions.To overcome a filter, a species requires physiological characteristics (Poff, 1997) such as osmotic balance, nitrification, sulfate reduction with the formation of methane acetate, and processes of respiration, photosynthesis and fermentation (Oren, 2002).
Several authors (e.g.Pedrós-Alió et al., 2000;Oren, 2002;Ayadi et al., 2004), in their studies on phytoplankton communities and the microbial web in solar saltworks developed in Israel, Tunisia and Spain, respectively, have evidenced the influence of hypersaline ecosystem dynamics, such as abiotic factors, especially salinity and temperature, as a selective filter of species richness and density, influencing the structure of the biological community; these groups observed species that are able to develop along the salt gradient.Nevertheless, there is a loss of the pool of local species in each sector of the saltern, which was also observed in the present study, where the Initial Sector had greater richness and density of species, with a decrease in species richness and density along with the movement of brine to the other sectors (Sectors Intermediate I and II).This was observed in the regression analysis, and in the scatter plot for the three salterns.
The ecological filters developed in this environment led to the selection of many organisms and the emergence of new species adapted to such conditions (Oren, 2001).Thus, biological diversity will eventually reduce depending on the filters developed by the environment, such as high salinity, physical and chemical conditions, and resource availability, thus affecting the diversity and structure of the phytoplankton community (Ayadi et al., 2004;Winder & Sommer, 2012).As observed in the canonical correspondence analysis, the environmental variables (pH, total phosphorus, nitrogen, salinity and temperature) have a positive influence.
There was a clear positive influence (pH, ammonium, total nitrogen, nitrite, nitrate and total phosphorus) on the density and spatial variation of the plankton community along the hypersaline ecosystem, except for soluble reactive phosphate, which has a negative influence for some phytoplankton communities.Similarly, in hypersaline and marine environments, some authors observed that the phytoplankton community was influenced by environmental variables (ammonia, nitrite, nitrate, phosphate, temperature, oxygen) and salinity gradient (Giordano et al., 1994;Pedrós-Alió et al., 2000;Thiéry & Puente, 2002;Telesh et al., 2013).These abiotic factors inhibit the development of the pool of local species (Chalmandrier et al., 2013); thus, the species cannot develop to other sectors.
In the Shannon-Wiener index, the Santa Clara Saltern showed the highest value.Already in relation to richness, the Augusto Severo Saltern had the highest number of species identified, which is associated with its location in the estuary, and abstraction of water directly from the sea; hence, it has lower salinity levels in relation to the other salterns studied, thus contributing to the fact that it does not report great species loss.Regarding the phytoplankton density, the Francisco Menescal Saltern presented higher density, despite higher levels of salinity compared with the Augusto Severo Saltern.However, as evidenced by Pedrós-Alió et al. (2000) in two Spanish salterns, the increased salinity resulted in reduced abundance and a number of different groups of eukaryotic microorganisms, but an increase in biomass of prokaryotes.The significant reduction in species richness in the system was also verified by Thiéry & Puente (2002) in the Camargue Saltern, France, which showed the influence of temperature and salt concentrations on the reduction of plankton abundance (Williams, 1998).Chalmandrier et al. (2013) observed the effects of biotic interactions (predation, competition, resources) on the community structure, but these biotic interactions are difficult to investigate due to the selective effect of abiotic variables.The decrease in species richness and diversity in the salterns is related to ecological filters imposed along the saline environment.Thus, this allows the appearance of species adapted to this environment, in this case, halotolerant and halophilic species (Williams, 1998;Oren, 2001).
The decrease in richness and diversity of species in salt, too, is due to the inverse dynamics of the estuary, i.e. increased salinity of the mouth towards the mainland, which provides a rich biological diversity directly related to the estuary dynamics (Telesh et al., 2013).
Thus, the importance of studies that address the effect of ecological filters on biological communities is noticed.This is in order to have a better understanding of ecosystem dynamics, making it possible to recommend measures for their conservation and management by taking into account the study of biological communities.
The present study evidenced the dynamics of the saline environment and its influence on the structure of phytoplankton communities.This fact is associated with filters developed by the hypersaline environment, where increasing salinity, temperature and precipitation of salts influence the reduction in the richness and density of these organisms in the ecosystem.Therefore, the hypothesis was confirmed, since the dynamics of the hypersaline ecosystem (abiotic factors) becomes a selective filter of species due to the unsuitable abiotic conditions for the development of species other than those which are halotolerant or halophilic.logistic support for field samplings.To the Limnology Laboratory and to Benthos Ecology Laboratory of the UEPB for help in the analysis and identification.LUH received financial support from FAPERGS (Process #12/1354-0) and CNPq (Process #471572/2012-8).

Figure 3 .
Figure 3. Effects of salinity on algal richness in the salterns (Augusto Severo, Francisco Menescal and Santa Clara).

Table 2 .
Density (%) of the most representative species of the hypersaline wetlands.I = initial sector, II = intermediate sector, III = intermediate sector.( -) = absence of species.