Species richness of the genera Trachelomonas and Strombomonas ( pigmented Euglenophyceae ) in a subtropical urban lake in the Porto Alegre Botanical Garden , RS , Brazil

The objective of this study was to evaluate the taxonomic composition and richness of the genera Trachelomonas and Strombomonas (pigmented Euglenophyceae) in a subtropical urban lake (30°03'05"S; 51°10'34"W) within the Botanical Garden, in the city of Porto Alegre, located in the state of Rio Grande do Sul, Brazil, in order to contribute to the floristic knowledge of Euglenophyta in artificial urban lakes. From July 2007 to June 2008, we collected samples monthly from two sampling stations, using a 25-μm mesh plankton net. The samples were preserved in 4% formaldehyde. We evaluated the following environmental variables related to the water in the lake: ammonia content, silica content, organic matter content, dissolved oxygen concentration, temperature, pH, depth, transparency, electrical conductivity and rainfall. To study the species composition in relation to environmental factors, we used cluster analysis, principal components analysis, canonical correspondence analysis and indicator species analysis. At the specific and infraspecific level, 22 taxa were identified, Trachelomonas elliptica (Playf.) Defl. and T. gracillima Bal. & Dast. representing new records for the state of Rio Grande do Sul. The cluster analysis indicated seasonal variation, species richness being highest in the spring of 2007. Seven indicator species were characterized as acid-tolerant.


Introduction
Euglenophyceae is a class of microalgae composed of pigmented and colorless organisms, which, except for the genus Colacium, are all single-celled flagellates found in moist soil, brackish water, fresh water and marine environments.The pigmented Euglenophyta of the order Euglenales have two flagella, and only one emerges from the reservoir at the anterior pole of the cell, they show a pellicle composed of protein striae and have an energy reserve of paramylon.The pigmented genera Trachelomonas Ehr emend.Defl.and Strombomonas Defl.were studied in the Lago da Ponte, which is an artificial urban lake within the Botanical Garden, in the city of Porto Alegre, located in the state of Rio Grande do Sul, Brazil.Trachelomonas is the genus with the largest number of species within the division Euglenophyta, encompassing approximately 250 species that inhabit freshwater environments worldwide (Bourrelly, 1970), including marine species (Leedale, 1967), whereas Strombomonas encompasses approximately 50 species found only in freshwater environments (Bourrelly, 1970).
The Porto Alegre Botanical Garden, with an area of 39 ha, is located within an urban area and includes two lakes known locally as the Lago das Tartarugas ("Turtle Lake") and Lago da Ponte ("Bridge Lake").We studied the Lago da Ponte, which was previously a swamp.Through the removal of grasses and the addition of more water, it was transformed into a lake in 2003.It is a shallow lake with grass carp, two swans and a number of tortoises.In its reflecting pool, Salvinia auriculata Abl. can be found.The decomposition of aquatic macrophytes associated with the food fed to swans and animal feces propitiate the existence of Euglenophyta, a group of algae that has been widely cited as a major indicator of water with high concentrations of organic matter, nitrogen and phosphorus (Round, 1983).

Study area and sampling
The geographic coordinates of the Lago da Ponte are 30°03'05"S and 51°10'34"W.It is an artificial shallow water environment with an approximate area of 0.5 ha.We collected samples of Trachelomonas and Strombomonas species at two locations: Station 1, with an average depth of 54.1 cm (max., 100 cm); and Station 2, with an average depth of 22.8 cm (max., 30 cm).From July 2007 to June 2008 encompassing all four seasons, monthly samples were collected from the two sampling stations, totaling 24 samplings.Operationally, the nomenclature used hereafter is: E1 and E2 = sampling

Biological composition
The water samples for biological analyses were collected using a 25-μm mesh plankton net and were preserved in 4% formaldehyde.We analyzed slide-mounted live samples using a light microscope (Leica DMLS; Leica Microsystems.Inc., Depew, NY, USA) with a grid eyepiece, which was coupled to a camera lucida in order to make the drawings.
For the identification of taxa at the specific and infraspecific levels, we referred to the following basic works: Huber--Pestalozzi (1955), Németh (1980), Starmach (1983), Tell & Conforti (1986) and Shi et al. (1999).The samples were deposited in the collection of the Herbário Prof. Dr. Alarich R. H. Schultz (HAS, Prof. R. Alarich H. Schultz Herbarium) at the Museum of Natural Science of the Zoobotanical Foundation of the state of Rio Grande do Sul (Tab.1).

Environmental variables
Simultaneously to the collection of biological material, water was collected in order to conduct measurements of physical and chemical variables important as parameters for environmental evaluation.The concentrations of ammonia (μg L −1 ), silica (mg L −1 ), organic matter (mg L −1 O 2 ) and dis-solved oxygen (mg L −1 ) were measured in the laboratory in accordance with the standards established by the American Public Health Association (APHA, 2005).We measured temperature (°C mercury thermometer), pH (DMPH pHmeter; Digimed, Santo Amaro, Brazil), depth (in cm), transparency (Secchi disk) and electrical conductivity (in μS cm −1 , CD-28 conductometer; Digimed) on site.The rainfall data were obtained from the National Meteorology Institute -8th Meteorology District, which has a weather station near the study area.

Information Processing
Multivariate species composition data on Trachelomonas and Strombomonas in relation to environmental factors were analyzed in two different ways.First, data related to the presence and absence of taxa, at different times of the year, were analyzed using cluster analysis.Subsequently, we applied indicator species analysis to the array of biological data, defining indicator species as those that showed significance on the Monte Carlo test (p<0.05).
Principal component analysis was applied to the physicochemical data matrix, in order to determine the significance of the accumulated variance of the explanatory environmental variables in the ordination.Finally, to reveal the main gradients of change in the composition of the indicator species, in relation to environmental processes, we applied canonical correspondence analysis (Ter Braak, 1986).The physical and chemical data were logarithmically transformed (log x), and analyses were processed using the program PC-ORD (McCune and Mefford, 1999).

Geographic distribution
Of the 22 taxa identified in the lake, 54.5% have a cosmopolitan distribution, whereas the others have been recorded on two to three continents.Trachelomonas armata var.steinii, T. hispida var.hispida, T volvocina var.volvocina and T. volvocinopsis var.volvocinopsis are taxa with broader distribution in Brazil.
All species and infraspecies had already been cited in Brazil, although T. elliptica and T. gracillima represented new records for Rio Grande do Sul.
The most representative taxa for their occurrence in the urban lake studied were Trachelomonas curta var.minima, T. hispida var.coronata, T. volvocina var.volvocina and T. volvocinopsis, the last taxon being distinctive in that it was present in 96% of the analyzed samples.With the exception of T. curta var.minima, the other species are distributed worldwide.The least representative taxa in the lake were T. hemisphaerica, T. lemmermannii var.lemmermannii, T. sculpta var.sculpta, T. varians var.varians and T. volvocina var.derephora, all of which occurred in less than 11% of the samples.

Species richness
The highest number of taxa was recorded in the spring of 2007 (Fig. 40).The richness ranged from one to 17 taxa, peaking in November 2007, when 17 taxa were cited at Station 2, corresponding to 73% of the taxa recorded.

Cluster analysis
Cluster analysis, based on data regarding the presence and absence of Trachelomonas and Strombomonas, as well as the categorical variable "seasonality" (in function of the different times of year), revealed three main clusters: the first comprising Winter samplings; the second comprising Spring samplings; and the third comprising samplings from Summer and Fall (Fig. 41).The analysis revealed seasonal differences in the richness of the two genera.

Indicator species analysis
Figure 42 shows the species that showed significance (p<0.05) on a Monte Carlo test, having been selected for additional tests using canonical correspondence analysis in order to reveal the main gradients of change in species composition.

Principal component analysis
The principal component analysis revealed that the first eight physical and chemical variables reached a cumulative percentage of variance of the explanatory order correspon-ding to the analysis cut-off point, 95% according to the variable selection methodology described by Delevati et al. (2005).The other two variables whose cumulative percentage of variance exceeded the cut-off were excluded from a second analysis to avoid the problem of multicollinearity between environmental variables, particularly in the case of multivariate analysis (Ter Braak, 1986).

Ordination: canonical correspondence analysis
The canonical correspondence analysis revealed 58.7% of the total variability of the data on its first 3 axes, with eigenvalues for axes 1, 2 and 3 of 41.6, 10.0 and 7.2, respectively.This relatively low explicability, however, is expected in the ordination analysis of ecological data, confirming the complexity of factors acting in determining the composition of ecosystems (Ter Braak and Prentice, 1988).However, the species-environment correlations for axes 1, 2 and 3 (r = 0.939, r = 0.847 and r = 0.780, respectively) indicated a strong relationship between species distribution and the environmental variables used in the ordination.The Monte Carlo permutation test revealed that sequence axis 1 was statistically significant (p<0.05),indicating a high likelihood of not being a result of chance.
With respect to canonical correspondence analysis axis 1 (which explained 41.6% of the variance), the most important environmental variable in the sequence was, in conformity with the intra-set correlation coefficients, ammonia (r = 0.581), which correlated with the positive quadrant of axis 1 (Fig. 43), revealing an environmental gradient for this nutrient, which ranged from 33.2 ± 38.8 μg L −1 in Winter to 56.7 ± 12.1 μg L −1 in the Spring, highlighting that the pH remained relatively acid throughout the year, with an average value of 6.0 ± 0.05 (coefficient of variation=0.8%).The ammonia concentrations observed, however, are relatively low in comparison with those reported by Baumgarten et al. (2003), who, working in the Lagoa dos Patos, found that ammonia concentrations ranged from 240 to 300 μg L −1 in areas without anthropogenic disturbance.It is of note that values as low as those observed in the present study were reported by Nunes et al. (2006) for the Baia dos Coqueiros, in the Pantanal region along the coast of Brazil.Those authors reported the presence of macrophytes and found that the mean 24h ammonia concentration was 33.0 μg L −1 , with a pH of 5.4-7.0.
The ammonium ion is recognized as a good marker of recent urban pollution, mainly from domestic sewage (Pereira & Mercante, 2005).Therefore, the low values observed in the present study could be explained by the fact that the lake evaluated is an artificial aquatic environment located within a protected area, the nitrogen content coming from rain, organic/inorganic material of allochthonous origin or fixation of molecular nitrogen within the lake itself.Because the pH of the lake is low (annual average of 6.0 ± 0.05), the ammonia formed is unstable and is converted by hydration to the ammonium ion, which becomes predominant in this   condition (Esteves, 1998;Pereira & Mercante, 2005).We also found that, along the ammonia gradient, there was a clear differentiation of species that comprised the Spring subcluster, which were characterized as indicator species (Fig. 42).There is therefore a high probability of there being species tolerant of low pH (acidic water, pH < 7.0), given that this condition was essential for the prevalence of the ammonium ion in the gradient detected, even at low concentrations, characterizing a natural environment.Alves-da- Silva & Avila (1995) recorded 22 taxa of Trachelomonas in shallow lakes of the Zoological Garden and the Botanical Garden in the city of Porto Alegre, and 17 of those taxa were also identified in the present study.Those authors stated that species richness was greatest when the pH was acidic to neutral.
Alves-da-Silva & Torres (1994b), Alves-da-Silva & Bridi ( 2004) and Alves-da- Silva & Bicudo, (2006) noted that genus Strombomonas tolerates a wide range of pH values, occurring in slightly acidic to alkaline waters.In this context, the results of the present study corroborate those obtained by other researchers: Trachelomonas and Strombomonas achieve their greatest richness in water that is slightly acidic to slightly alkaline.These data increasing the ecological knowledge of this group of algae.

Figure 40 .
Figure 40.Distribution of the richness (no.de taxa) of Trachelomonas and Strombomonas in the Lago da Ponte, in the state of Rio Grande do Sul, Brazil, the two sampling stations (ST1= Station 1 and ST2= Station 2 ), between July 2007 and June 2008.

Figure 41 .
Figure 41.Cluster analysis based on monthly data related to the presence and absence of Trachelomonas and Strombomonas species and physicochemical variables for the Lago da Ponte, in the state of Rio Grande do Sul, Brazil, with seasonality as the categorical variable.G1 -Winter cluster; G2 -Spring cluster; G2-P -Spring subcluster; G3 -Summer/Autumn cluster.