Algal Assemblages Across a Wetland, from a Shallow Lake to Relictual Oxbow Lakes (Lower Paraná River, South America)

Hydrobiologia 511: 25–36, 2004. © 2004 Kluwer Academic Publishers. Printed in the Netherlands.

25

Algal assemblages across a wetland, from a shallow lake to relictual oxbow lakes (Lower Paran´a River, South America) Irina Izaguirre1 , In´es O’Farrell1 , Fernando Unrein1 , Rodrigo Sinistro1 , Mar´ıa dos Santos Afonso2 & Guillermo Tell1 1 Departamento

de Ecolog´ıa, Gen´etica y Evoluci´on, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, C1428EHA, Buenos Aires, Argentina 2 INQUIMAE, Departamento de Qu´ımica Inorg´ anica, Anal´ıtica y Qu´ımica F´ısica – FCEN – UBA, C1428EHA, Buenos Aires, Argentina Received 21 May 2002; in revised form 29 July 2003; accepted 24 August 2003

Key words: algal assemblages, Paran´a River floodplain, CCA

Abstract This study deals with the variation in the algal assemblages across the transversal dimension of a wetland of the Lower Paraná River from October 1998 to September 1999. The relationship between the algal composition and the physico-chemical variables is analyzed by means of Canonical Correspondence Analysis. The abundant floating macrophytes generated a severe reduction in light penetration in the relictual oxbow lakes (ROLs), almost anoxic conditions and high values of P, N and K. Such characteristics accounted for their very particular algal flora dominated by Cyanobacteria and several diatoms, species probably mixotrophic regarding their capacity to grow in darkness and tolerate very low oxygen contents. In the shallow lake, the phytoplankton comprised many small autotrophic green algae, accompanied by many flagellates of the classes Cryptophyceae, Euglenophyceae and Dinophyceae. Our results indicate that the macrophyte cover was probably the stirring factor in the selection of algal species along the transitional zone comprising a floodplain shallow lake and several ROLs.

Introduction Wetlands are areas characterized by periodical flooding and/or inundation and characteristically exhibit a great degree of genetic and environmental diversity (Wilson, 1988). Their spatial heterogeneity and richness in ecotones were adequately defined in Holland (1988). Neiff et al. (1994) proposed the following definition for South American wetlands: ‘systems of sub-regional extent in which the spatial and temporal presence of a variable cover of water causes characteristic biogeochemical fluxes, soils of accentuated hydromorphism, and a biota whose structure and dynamics are well adapted to a wide range of water availability. They can be considered macrosystems whose complexity grows with hydrosedimentological variability and geographic extent’. According to these authors, floodplains are classified as ‘large wetlands’.

Most South American large rivers are bordered by extensive alluvial plains, which can exceed 300 000 km2 and include a great number of shallow lakes with different limnological properties (Welcomme, 1985). As Junk et al. (1989) pointed out for floodplain wetlands, in these alluvial systems the regular flood pulse and the habitat heterogeneity favor a high biodiversity of terrestrial and aquatic plants and animals. Studies on the algal communities from floodplain shallow lakes are relatively scarce as compared to other world regions. Most of the present studies describe phytoplankton composition and dynamics that are subject to complex bi-directional interactions between the river and their marginal shallow lakes (García de Emiliani, 1997; Huszar & Reynolds, 1997; Train & Rodrigues, 1998; de Melo & Huszar, 2000). The first studies on the phytoplankton community of the Paraná River floodplain shallow lakes were

26 conducted in its middle stretch by Zalocar de Domitrovic (1990, 1992) and García de Emiliani (1993, 1997). Unrein & Tell (1994) and Izaguirre & O’Farrell (1999) focused on phytoplankton aspects such as nutrient limitation and primary production dynamics in water bodies of the Lower Paraná. As regards spatial analysis of algal communities along environmental gradients or ecotone areas, Izaguirre et al. (2001a) and Unrein (2001) described species replacement in some systems of the Lower Paraná thus contributing with the scarce information available in these terms for wetlands. In particular, an interesting algal flora adapted to almost anoxic conditions and extremely poor light intensities has been observed in the profusely vegetated water bodies comprised in the wetland selected for this study (Izaguirre et al., 2001b). The main aim of this research is to analyze the algal assemblages across the transversal dimension of a wetland located in the Otamendi Natural Reserve (Buenos Aires, Argentina), from a floodplain shallow lake to relictual oxbow lakes. The relationship between the algal composition and the physicochemical variables is analyzed by means of Canonical Correspondence Analysis techniques.

Study site The study area is located in a wetland from the Natural Reserve Otamendi, which is delimited by the Paraná de las Palmas and Luján Rivers, Buenos Aires Province, Argentina (34◦ 10 to 34◦ 17 S; 58◦ 48 to 58◦ 53 W) (Fig. 1). The aquatic environments are represented by shallow lakes and several semi-permanent water bodies comprising relictual oxbow lakes and very small ponds. The lakes are surrounded by marshy vegetation and temporarily and partially covered by floating macrophytes, whereas the relictual oxbow lakes are permanently and completely covered. On the other hand, the small pond sometimes presented a situation similar to the relictual oxbow lakes (complete vegetation cover) and in other cases the macrophyte cover was scarce or even absent as in the lake. The largest lake is the ‘laguna Grande’ with an area of about 28 ha. Regarding the definition given by Naiman & Décamps (1990), the study site can be described as an area of consecutive water-land ecotones. The dominant macrophytes in the wetland are the rooted Schoenoplectus californicus and Scirpus giganteus and several floating species such as Azolla filiculoides, Lemna minima, Wolffiella oblonga, Hy-

drocotyle sp., Salvinia rotundifolia and Pistia stratiotes. The area is almost permanently flooded by rainfall, as well as by the river floods, which account for the very poor drainage and consequent reducing conditions of the soils (Chichizola, 1993). The region has a temperate sub-humid climate, with a moderate effect due to the Atlantic masses, the Río de la Plata and Paraná de las Palmas Rivers. Precipitations occur during the whole year; with mean annual values of 950 mm. Mean summer and winter temperatures in this region are 22 ◦ C and 9.5 ◦ C, respectively.

Materials and methods Six sampling points were established across a transect at three relictual oxbow lakes (ROL1-ROL3, ROLs hereinafter), at a small pond (SP) and at the littoral (LSL) and pelagial (PSL) zones of the ‘Laguna Grande’ (SL) (Fig. 1). Samples were collected fortnightly from October 1998 to January, and monthly from February to September 1999. Temperature, pH and conductivity were measured in situ using Hanna HI 8314 and HI 8033 portable electronic meters. Dissolved oxygen was determined by the Winkler method. Samples for nutrient analyses were filtered through Whatman GF/F. The concentration of the main inorganic compounds (phosphates, nitrates, sulfates, chlorides, potassium, sodium, calcium and magnesium) were determined by ionic chromatography using DIONEX DX-100 instrument with a conductivity detector and DIONEX AS-4 or CS-10 cromatographic columns. Ammonia was analyzed following the indophenol method (APHA, 1975). Humic substances were estimated by spectrophotometry, reading the absorbance at 250 nm (Kronberg, 1999). Total phosphorus and nitrogen were determined by ionic chromatography after alkaline digestion according Koroleff (1983). Suspended solids were evaluated drying the non filtrable residue at 103–105 ◦ C until constant weight (APHA, 1975). Incident and underwater irradiance were measured with a Li-Cor PAR sphaerical quantum sensor (Li-250). Qualitative algal samples were obtained with a 15 µm pore net and fixed in 4% formalin. Water samples for quantitative analyses of the algal communities were preserved in PVC flasks with 1% Lugol’s iodine solution. Counts were performed according to Utermöhl (1958). Replicate chambers were

27

Figure 1. Map of the Paran´a River Basin showing the location of the study area and the six sampling sites: ROL, relictual oxbow lake; LSL, littoral shallow lake; PSL, pelagial shallow lake; SP, small pond.

left to sediment for at least 24 h. Counting errors were estimated according to Venrick (1978), accepting a maximum of 20% for the most frequent species. The spatial variation in limnological properties and algal assemblages across the wetland was analyzed by the ordination method CCA (Canonical Correspondence Analysis). Calculations were performed by the program CANOCO (ter Braak, 1988). The data set analyzed was based on quantitative samples from six sites during 11 months using abundance of phytoplankton species and the corresponding environmental variables. Those species that occurred less than 5 times along the study period were eliminated, only if they never exceeded 10% of the total density in each sample. Forward selection was used for adding environmental variables to the model. The significance of the ordination axes was assessed by Monte Carlo permutations.

Results Physical and chemical variables Table 1 summarizes physical and chemical variables of the sampling sites. Some of these variables presented marked differences between the ROLs and SL. The lowest mean temperature was measured in the ROLs and differed in at least 6 ◦ C from SL (Fig. 2a). Moreover, the dense macrophyte cover of the ROLs determined an extremely low photosynthetic active radiation, as nearly 4% of the incident radiation reached at 5 cm depth. On the contrary, nearly 50% of the incident radiation reached this depth at SL since most of the water surface was free of floating vegetation. Mean suspended solid concentration was at least twofold higher in the ROLs than in SL (Fig. 2b). The most striking difference between both systems is related to the oxygen content of their waters: ROLs remained anoxic during almost the entire study period, whereas

28

Figure 2. Mean values and variation coefficients for the main physical and chemical variables at the six sampling sites.

29

Figure 3. Algal density and species richness variations during the study period for the six sampling sites.

mean dissolved oxygen concentration in SL exceeded 4.9 mg l−1 (Fig. 2c). Water was slightly acidic in the ROLs and SP; pH ranged from 3.8 to 7.8 in these water bodies. On the other hand, SL waters showed a tendency to neutral and even alkaline waters as pH varied from 5.7 to 9.1 (Fig. 2d). Humic acid concentrations were higher in the ROLs than in SL where values from the vegetated littoral zone exceeded that of the pelagial area (Fig. 2e). The above mentioned spatial patterns were not registered for conductivity given that ROL3 was similar to LSL. Among ions, K+ , Ca2+ (Figs 2f and 2g) and Mg2+ revealed a trend similar to conductivity, the latter ion showed the strongest correlation with this variable (r = 0.84 and r = 0.79, respectively, p < 0.05). On the other hand, SO42− and Na+ concentrations increased from ROL1 to ROL3 and decreased at SL, specially in the pelagial zone. Cl− presented no definite spatial pattern. Nitrogen and phosphorus concentrations were lower in the ROLs than in SL (Table 1). In particular, ammonia exhibited a decreasing gradient from ROL1 to PSL (Fig. 2h), whereas nitrate presented a marked drop in SL with concentrations one order of magnitude lower. Total nitrogen concentrations were very high across the wetland with mean values from 1.68 to 2.6 mg l−1 ; figures were slightly higher in the ROLs (Fig. 2i). Although phosphate concentrations did not differ strongly between the ROLs and SL, total phosphorous clearly reflected the same decreasing spatial trend observed for nitrogen (Fig. 2j).

On the other hand, the separate analysis of SP revealed the widest fluctuations of some environmental variables. The more pronounced variations were registered for nutrients, conductivity and suspended solids (Fig. 2). SP was alternatively anoxic or with oxygen concentrations not exceeding 5.6 mg l−1 ; these fluctuations were associated with the persistence or absence of the macrophyte cover of the pond (Fig. 5). Algal assemblages A total of 305 algal taxa, among species and varieties, were identified for the study area (Table 2). Most of the species can be associated with benthic and periphytic habitats, and only a few taxa may be considered regular plankton organisms. Pennate diatoms with raphe were the best representatives of the benthic and periphytic habitats, whereas small Chlorococcales and centric diatoms were regular components of the plankton. The Bacillariophyceae Class contributed with the highest number of species in this wetland. Chlorophyceae, Cyanobacteria and Euglenophyceae were also well represented, while less taxa stood for Chrysophyceae, Tribophyceae, Cryptophyceae and Dinophyceace. The highest algal densities were registered at SP with a peak of 428 432 ind ml−1 , whereas the lowest values were observed at ROL3 with less than 10 000 ind ml−1 (Fig. 3). In general terms, phytoplankton abundance decreased from ROL1 to ROL3 and densities were usually higher at LSL as compared to PSL. The fluctuations recorded at each sampling

30 Table 1. Mean, minimum and maximum values of the physical and chemical variables for the aquatic systems comprised in the study area (ROL: relictual oxbox lake, LSL: littoral shallow lake, PSL: pelagial shallow lake, SP: small pond)

Temperature ( ◦ C) Suspended solids (mg l−1 ) Dissolved oxygen (mg l−1 ) pH Humic acids (Abs. 250 nm) Conductivity (µ cm−1 ) Cl− (mg l−1 ) SO= 4 (mg l−1 ) Na+ (mg l−1 ) K+ (mg l−1 ) Mg2+ (mg l−1 ) Ca2+ (mg l−1 ) N-NH4 (mg l−1 ) N-NO3 (mg l−1 ) TN (mg l−1 ) P-PO4 (mg l−1 ) TP (mg l−1 )

ROL1

ROL2

ROL3

LSL

PSL

SP

13.8 5.6–21.7 51.5 10–95 0.05 nd–0.8 6.5 5.1–7.8 2.46 0.96–3.87 2049 320–3410 356.2 32.6–1058 211.4 15.8–646.1 511.7 143.9–1285.8 12.49 3.71–26.96 12.18 3.11–24.18 19.57 6.44–36.19 0.228 nd–1.201 0.341 0.014–2.871 2.294 0.780–3.582 0.561 0.163–1.084 0.705 0.280–1.187

14.5 6.1–23.5 171.3 25–359 0.02 nd–0.4 6.5 4.4–7.7 1.86 0.84–3.25 3012 558–5800 383.1 49.1–718.7 232.9 0.28–525.6 710.3 175.7–2146.3 14.45 2.14–26.57 21.43 4.37–51.85 32.76 10.46–60.32 0.188 0.024–0.709 0.355 0.023–2.137 2.074 0.867–3.266 0.609 0.095–1.792 0.610 0.241–1.000

14.6 6.3–23 76.5 17–203 0.2 nd–1.2 6.3 4.9–7.8 2.44 1.04–3.49 1837 429–2790 376.9 31.5–1340 250.2 18.3–1030.3 1114.5 152–9854.5 12.32 4.87–21.86 11.48 3.25–20.76 16.95 7.12–28.68 0.150 0.026–0.555 0.423 nd–4.853 2.621 1.588–4.351 0.370 nd–0.840 0.599 0.251–1.009

20.7 9.4–30.7 19.5 4–52 4.9 2–9.1 6.9 5.2–8.1 1.31 0.88–1.8 1845 470–3150 281.1 41.1–966.2 197.5 29.1–560.2 464.7 197.1–1177.3 11.98 4.58–33.76 10.29 3.27–23.42 17.31 7.92–35.68 0.096 nd–0.864 0.053 0.003–0.152 1.903 1.146–2.886 0.371 0.077–1.705 0.417 0.012–1.044

21.2 9.8–30.3 8.7 2–26 7.3 3.3–13.8 7.6 5.7–9.1 0.98 0.7–1.27 1327 510–2170 361.6 51.7–2595.2 135.1 35.7–386.8 392.0 192.2–879.7 9.86 5.13–15.49 8.61 3.88–14.08 14.27 8.14–27.19 0.032 nd–0.075 0.074 0.002–0.344 1.682 0.841–2.343 0.227 0.024–0.493 0.354 0.138–0.660

16.2 5.8–25.6 306.7 5–3493 1.5 nd–5.6 6.3 3.8–7.4 1.83 1.11–2.57 3527 697–7100 621.3 76.7–3106.8 727.5 39.6–4906.8 983.7 293.7–2021.8 13.54 8.01–28.45 20.25 4.94–61.61 28.11 6.67–57.07 1.220 0.023–7.168 0.093 0.013–0.390 2.975 1.054–9.743 0.501 nd–2.208 0.585 0.095–1.846

site followed a seasonal pattern, with minimum values in winter time and peaks in either spring or summer. Although no difference was observed between the variation range of algal densities between the ROLs and SL, a characteristic algal composition can be assigned to each one of these environments (Fig. 4). Cyanobacteria, represented by small Chroococcales (Synechococcus spp. and Synechocystis spp.) and an unidentified Oscillatorian taxa dominated the ROLs accompanied by some pennate benthic diatoms (Achnanthes hungarica, A. lanceol-

ata, Amphora veneta, Cocconeis placentula, Eunotia bilunaris, E. flexuosa, Nitzschia clausii, N. sigma, Neidium iridis). This flora constituted almost 80% of the average phytoplankton density during the study period. On the contrary, strictly planktonic taxa represented by small Chlorococcales (Didimocystis bicellularis, Monoraphidium caribeum, M. circinale, M. contortum) and several Cryptomonas species (C. curvata, C. erosa, C. marssonii, C. ovata) were the major components of the algal assemblages at SL and constituted more than 65% of the average phytoplankton dens-

31 Table 2. List of dominant species found in the studied floodplain. Numbers make reference tot the species used in the Canonical Correspondence Analysis Cyanobacteria Anabaena sphaerica Born. et Flah. Aphanocapsa delicatissima W. et G.S. West 1 Aphanocapsa grevillei (Berkeley) Rabenh. Aphanocapsa hyalina (Lingbye) Hansgirg 2 Aphanothece smithii Kom´arkova-Legnerov´a et Cronby Aphanothece stagnina (Sprengel) A. Braun in Rabenh. Chroococcus minutus (Kütz.) Nag. Chroococcus minimus (Keissler) Lemm. Coelosphaerium kuetzingianum Näg. Coelosphaerium minutissimum Lemm. Eucapsis starmachii Kom. et Hind. Leptolyngbya fragilis (Menegh.) Gomont Leptolyngbya frigida (Fritsch) Anag. et Kom. Merismopedia punctata Meyen Merismopedia tenuissima Lemm. 3 Microcystis firma (Kütz.) Schimidle 4 Oscillatoria sancta (Kütz.) Gomont. Oscillatoria subbrevis Schmidle Planktolyngbya subtilis (W. et G. S. West) Anag. et Kom. 5 Pseudoanabaena catenata Lauterb. Raphidiopsis spp. Romeria leopoliensis (Racib.) Koczw. 6 Synechoccocus spp. 7 Synechocystis spp. 8 Woronichinia elorantae Kom. et Kom.-Legn. Oscillatoriales nov. sp. 9 Bacillariophyceae Achnanthes cf. reversa Lange-Bertalot et Krammer Achnanthes biasolettiana Grun. Achnanthes exigua Grun. Achnanthes hungarica Grun. in Cl. et Grun. 10 Achnanthes lanceolata (Bréb.) Grun. Achnanthes lemmermanii Hust. Amphiprora alata Kütz. Amphora lybica Ehr. Amphora veneta Kütz. Anomoeneis sphaerophora (Ehr.) Pfitzer Aulacoseira granulata var. angustissima (Müll.) Sim. Aulacoseira granulata var. granulata (Ehr.) Sim. Aulacoseira italica (Ehr.) Sim. Cocconeis placentula Ehr. Cyclotella meneghiniana Kütz. 11 Cymbella mesiana Choln. Cymbella silesiaca Bleisch in Rabh. Diatoma mesodon (Ehr.) Kütz. Eunotia bilunaris (Ehr.) Mills Eunotia flexuosa Bréb. et Kütz. Eunotia monodon Ehr. Eunotia praerrupta Ehr. Fragilaria construens (Ehr.) Grun. Fragilaria ulna (Nitzsch) L.-B. Gomphonema augur Ehr. Gomphonema clavatum Ehr. Gomphonema gracile Ehr. Gomphonema parvulum Kütz. 12 Navicula capitatoradiata Germain Navicula cryptocephala Kütz. Navicula cuspidata Kütz.

(cont. Bacillariophyceae) Navicula halophila (Grun.) Cl. Navicula peregrina (Ehr.) Kütz. Navicula pupula var. nyassensis (Müll.) L.-B. Navicula pupula var. pupula Kütz. Neidium dubium (Ehr.) Cl. Neidium iridis (Ehr.) Cl. 13 Nitzschia acicularis (Kütz.) Smith Nitzschia acicularoides Hust. Nitzschia amphibia Grun. Nitzschia clausii Hantzsch Nitzschia frustulum (Kütz.) Grun. Nitzschia inconspicua Grun. Nitzschia linearis (Agardh) W. Smith Nitzschia palea (Kütz.) W. Smith Nitzschia sigma (Kütz.) W. Smith Pinnularia acrosphaeria W. Smith Pinnularia gibba var. gibba Ehr. Pinnularia gibba var. linearis Hust. Pinnularia interrupta W. Smith Pinnularia maior (Kütz.) Rabenh. Rhopalodia gibba (Ehr.) Müll. 14 Stauroneis phoenicenteron (Nitzsch.) Ehr. Chrysophyceae Chrysophyceae n.i. (flagelada) Mallomonas spp. Synura uvella Ehr. cf. Chroomonas caudata Geither Cryptophyceae Cryptomonas curvata Ehr. Cryptomonas erosa Ehr. 15 Cryptomonas marssonii Skuja 16 Cryptomonas ovata Ehr. 17 Chlorophyceae Chlamydomonas spp. Chlorella saccarophila (Krüg.) Mig. Chlorella vulgaris Beij. 18 Closterium acutum var. variabile (Lemm.) Krieger Didimocystis bicellularis (Chod.) Kom. 19 Monoraphidium arcuatum (Korsch.) Hind. 20 Monoraphidium caribeum Hind. Monoraphidium circinale (Nyg.) Nyg. 21 Monoraphidium contortum (Thur. in Bréb.) Kom.-Legn. 22 Monoraphidium griffithii (Berk.) Kom.-Legn. 23 Monoraphidium minutum (Nag.) Kom.-Legn. 24 Phacotus sp. Tetrastrum triangulare (Chod.) Kom. Euglenophyceae Euglena acus Ehr. Euglena variabilis Klebs Lepocinclis salina Fritsch Phacus sp. nov. Trachelomonas volvocinopsis Swir. Dinophyceae Peridinium sp.

32

Figure 4. Mean percentages of the algal classes at the sites located in the relictual oxbow lakes and in the shallow lake.

ity. Diatoms were quantitatively less important and were mainly represented by the centric Cyclotella meneghiniana. Moreover, the most frequent Cyanobacteria in the SL, Merismopedia tenuissima and Romeria leopoliensis, were not the same taxa registered as dominant in the ROLs. The proportion of the different algal groups at SP varied along the study period. These fluctuations were related to the variations of dissolved oxygen content. In general terms, during periods of low oxygen content and complete macrophyte cover the algal assemblages resembled that of the ROLs, whereas when oxygen conditions improved and macrophytes became less abundant phytoplankton composition was similar to the lake (Fig. 5). Regarding species richness, an increasing spatial gradient from ROL1 to the littoral area of the SL was observed (Fig. 3); the littoral zone exhibited the highest figures resulting from the contribution of both planktonic and tychoplanktonic taxa. Canonical correspondence analysis Among the variables analyzed, only nine were included in the model by forward selection (dissolved oxygen, total N, temperature, total P, ammonia, pH,

K+ , suspended solids and Ca++ ). The first two axes accounted for 56.2% of the variance (axis 1: 35.7%; axis 2: 20.5%). A Monte Carlo unrestricted permutation test on the first eigenvalue indicated that the abiotic factors were significantly correlated with the first axis (p < 0.01). The triplot of sampling sites, species and environmental variables according to the first two axes is shown in Fig. 6. The taxa illustrated were selected taking into account their abundance, frequency of occurrence along the study period, niche amplitude, and their fitness to the environmental variables included in the model. Samples from SL are plotted on the right side, clearly separated from those of the ROLs, while samples from SP are placed at an intermediate position between both groups. The first axis is mainly defined by a combination of dissolved oxygen, K+ , total N, total P and pH (intra-set correlation coefficients: 0.96; −0.80; −0.77; −0.73; 0.62, respectively). Axis 2 is mainly correlated with Ca++ and suspended solids (intra-set correlation coefficients: 0.78; 0.60). Small chlorococcaleans, several Cryptomonas species, Cyclotella meneghiniana, Romeria leopoliensis and Merismopedia tenuissima are ordinated towards

33

Figure 5. Annual variations of dissolved oxygen concentrations and of the percentage of the algal classes.

the right side of the figure, jointly with SL samples. Most of the Cyanobacteria (Synechococcus spp., Synechocystis spp., Microcystis firma, Planktolyngbya subtilis, Oscillatorian n.i.) and some diatoms (Achnanthes hungarica, Neidium iridis and Rhopalodia gibba), together with the ROLs samples, are plotted towards the left side.

Discussion The results obtained in this study are evidence of the important differences in both limnological features and algal assemblages across the wetland, from the relictual oxbow lakes to the shallow lake. Such differences may be associated to the dense and persistent

macrophyte cover of the ROLs and to its irregular occurrence in SP and SL. The abundant floating macrophyte cover is the cause of the severe light penetration reduction in the ROLs as well as of its almost anoxic conditions. Likewise, Hamilton et al. (1997) found severe depletions in oxygen content in floodplain environments with abundant emergent vegetation. Moreover, Pithart (1999) reported low oxygen contents for some floodplain pools with high levels of shading by surrounding vegetation and described an algal flora tolerant to low light intensities. In particular, Zalocar de Domitrovic (1993) asserted that the presence of macrophytes constitutes an important factor in the determination of phytoplankton assemblages in different systems of the Middle Paraná River.

34

Figure 6. Species triplot based on the Canonical Correspondence Analysis of the algal densities from the six sampling sites located in the study area. Only the selected species indicated with a number in Table 1 are displayed. Environmental variables are indicated by arrows.

On the other hand, the higher values of P, N and K registered in the ROLs can be attributed to the release of these elements from decomposing macrophytes. As it was reported by Junk (1997), a major pathway for sediment-bound nutrients into the lake water is via higher plants growing in the ‘aquatic terrestrial transitional zone’. The profuse aquatic vegetation also accounts for the high concentration of humic substances in these water bodies, which is one of the characteristics assigned by Hamilton et al. (1995) to such wetland systems. In this sense, even though the temporal variation of the pH within each sampling site follows the same pattern, the pH differences among the ROLs and SP with the SL are notorious. Even though in the basin there are evidences of salinization and/or alkalinization (INTA, 1990), the ROLs and SP presented lower pH due their distrophic characteristics. Multivariate analysis showed that the samples of the ROLs are clearly separated from those of SL (littoral and pelagial zones). The ROLs exhibit a high degree of independence from SL, being most of the time indirectly connected. Only when water levels were very high, the whole area was entirely flooded, but even during these events there was not a net flushing from the shallow lake to the ROLs, and still then their dense macrophyte cover persisted. The importance of

the degree of connection among floodplain water bodies on their dynamics has been repeatedly documented for other world regions. De Melo & Huszar (2000) reported that shallow lakes of the Amazonian floodplain can strongly differ in their metabolism as a consequence of their location on the floodplain. For the Middle Paraná River floodplain, García de Emiliani (1997) stressed the importance of the position and connection of the lakes regarding the main river, as one of the stirring factors of their phytoplankton structure and dynamics, and pointed out their strong dependence on water-retentive mechanisms. Moreover, Izaguirre et al. (2001a) and Unrein (2002), found a gradual change in the physico-chemical characteristics for a wetland of the Lower Paraná floodplain, together with a clear replacement of algal taxocenosis along a transitional system from the river to a permanently connected shallow lake. The extreme characteristics of the ROLs account for their very particular algal flora adapted to almost anoxic conditions and extremely low light penetration. Many species such as Synechococcus spp. and Synechocystis spp. found in the ROLs have been reported in the literature as mixotrophic (Izaguirre et al. 2001b) and are probably capable of anoxygenic photosynthesis (Sorokin, 1999). Moreover, several diatoms

35 as Achnanthes hungarica also very abundant in the ROLs, are mentioned in the literature to be capable to grow in darkness and tolerate very low oxygen contents (Van Dam et al., 1994; Tuchman, 1996). The affinity of the species belonging to these taxonomic classes to the extreme conditions encountered in the ROLs, is clearly revealed by the fact that between both groups more than 90% of the algal density is achieved. Phytoplankton in SL comprises mainly many small green algae, typically C-strategist (Reynolds, 1997), with a high growth rate and surface:volume ratio, and a rapid nutrient absorption. Most of them are typically autotrophic planktonic Chlorococcal species, such as Didimocystis bicellularis, Monoraphidium caribeum, M. circinale, and M. contortum. Studies carried out in other shallow lakes of the Paraná floodplain, as well as in lakes of other South American wetlands, report a similar algal flora, with abundant nanoplanktonic Chlorophyceae species, accompanied by many flagellates of the classes Cryptophyceae, Euglenophyceae and Dinophyceae (Zalocar de Domitrovic, 1990; García de Emiliani, 1993, 1997; Huszar and Reynolds, 1997; Train & Rodrigues, 1998; Unrein, 2002). Moreover, some species such as Cryptomonas spp. and many euglenoids reported by Jones (1994, 2000) as mixotrophic, were abundant and richly represented in SL. Nevertheless, the cause of their success in SL is probably different. Mixotrophy in the ROLs is probably related to very poor light conditions, whereas in SL N limitation could condition the selection of facultative mixotrophic species. In this sense, Unrein (2001) reported N limitation for this shallow lake during some periods of this study. Jansson et al. (1996) suggested that mixotrophic behaviour is a competitive advantage in absence of available inorganic nutrients. The floristic composition reveals that the species found mainly in the ROLs are typical components of the benthic and periphytic communities due to the profuse development of floating macrophytes that determines a tycoplanktonic profile of the community. On the other hand, the existence of areas free of vegetation in SL defines the dominance of euplanktonic taxa in the open waters. Vyverman (1996) pointed out that the development and success of individual species and communities in floodplain systems is probably regulated by a set of environmental factors, including the optical properties of the water, sediment load, nutrient concentrations, hydrology, lake morphology and stratification regime, together with biological interactions. Our results coincide with this view, but additionally stress the import-

ance of the macrophyte cover in the determination of such physico-chemical factors. In this way, we postulate that in wetlands where water bodies present major differences in floating vegetation, this variable must be considered as one of the main factors in the regulation of the algal assemblages composition and dynamics. Thus, the persistence of a dense macrophyte cover determines the existence of an adapted algal flora to these extreme environmental conditions.

Acknowledgements We thank the staff of the Otamendi Reserve (Parques Nacionales) for their assistance in the field. We are grateful to CONICET, UBA and ANCYPT for their financial support.

References APHA, 1975. Standard Methods for the Examination of Water and Wastewaters. American Public Health Association, Washington, 1193 pp. Chichizola, S. E., 1993. Las comunidades vegetales de la Reserva Natural Estricta Otamendi y sus relaciones con el ambiente. Parodiana 8 (2): 227–263. de Melo, S. & V. L. M. Huszar, 2000. Phytoplankton in an Amazonian flood-plain Lake (Lago Batata, Brazil): diel variation and species strategies. J. Plankton Res. 22: 63–76 Garcia de Emiliani, M. O., 1993. Seasonal succession of phytoplankton in a lake of the Paraná River floodplain, Argentina. Hydrobiologia 264: 101–114. Garcia de Emiliani, M. O., 1997. Effects of water level fluctuation in a river-floodplain lake system (Paraná River, Argentina). Hydrobiologia 357: 1–15. Hamilton, S. K., S. J. Sippel & J. M. Melack, 1995. Oxygen depletion and dioxide and methane production in waters of the Pantanal wetland of Brasil. Biogeochemistry 30: 115–141. Hamilton, S. K., S. J. Sippel, D. F. Calheiros & J. M. Melack, 1997. An anoxic event and other biogeochemical effects of the Pantanal wetland on the Paraguay River. Limnol. Oceanogr. 42: 257–272. Holland, M. M., 1988. SCOPE/MAB technical consultations on landscape boundaries: report of a SCOPE/MAB workshop on ecotones. Biology International, Special Issue 17: 47–106. Huszar, V. L. de M. & C. S. Reynolds, 1997. Phytoplankton periodicity and sequences of dominance in an Amazonian Floodplain lake (Lago Batata, Pará, Brazil): response to gradual environmental change. Hydrobiologia 346: 169–181. INTA, 1990. Atlas de suelo de la República Argentina. Escala 1: 500.000 y 1: 1.000.000. Secretaría de Agricultura, Ganadería y Pesca. Proyecto PNUA ARG. 85/019. INTA, Centro de Investigaciones de Recursos Naturales. Izaguirre, I. & I. O’Farrell, 1999. Temporal fluctuation of phytoplanktonic chlorophyll a and primary production in a turbid floodplain lake of the River Paraná (Argentina). Gayana Bot. 56 (1): 39–51. Izaguirre, I., I. O’Farrell & G. Tell, 2001a. Variation in phytoplankton composition and limnological features in a water-water

36 ecotone of the lower Paraná Basin (Argentina). Freshwat. Biol. 46: 63–74. Izaguirre I., R. Sinistro, I. O’Farrell, F. Unrein & G. Tell, 2001b. Algal assemblages in anoxic relictual oxbow lakes from the Lower Paraná floodplain (Argentina). Nova Hedwigia, 123: 95–106. Jansson, M., P. Blomqvist, A. Jonsson & A-K. Bregström, 1996. Nutrient limitation of bacterioplankton, autotrophic and mixotrophic phytoplankton, and heterotrophic nanoflagellates in Lake Örträsket. Limnol Oceanogr. 41(7): 1552–1559. Jones, R. I., 1994. Mixotrophy in planktonic protists as a spectrum of nutritional strategies. Marine Microbial Food Webs 8 (1–2): 87–96. Jones, R. I., 2000. Mixotrophy in planktonic protists: an overview. Freshwat. Biol. 45: 219–226. Junk, W. J., P. B. Bayley & R. E. Sparks, 1989. The flood pulse concept in River-floodplain systems. In Dodge, D. (ed.), Proceeding of the International Large River Symposium, Canadian Spec. Publ. Fisheries and Aquatic Science 106: 110–127. Junk, W. J., 1997. The Central Amazon Floodplain: Ecology of a Pulsing System. Springer-Verlag, Berlin, 525 pp. Koroleff, F., 1983. Simultaneous oxidation of nitrogen and phosphorus compounds by persulfate. In Grosshoff, K., M. Eberhodt & K. Kremling (eds), Methods of Seawater Analysis. Verlag Chemie, Wainheimer, Germany, pp. 168–169. Kronberg, L., 1999. Content of Humic Substances in Freshwater. In Keskitalo, J. & Eloranta (eds), Limnology of Humic Waters. P. Backhuys Publishers, Leiden, 2: 9–10. Naiman, J. & Décamps, H., 1990. The ecology and management of aquatic-terrestrial ecotones. Man and the Biosphere Series 4, UNESCO and Parthenon Publishing Group, Paris, 316 pp. Neiff J. J., M. H. Iriondo & R. Carignan, 1994. Large Tropical South American Wetlands: An Overview. In Link, G. L. & R. J. Naiman (eds), Proceedings of the International Workshop on The Ecology and Management of Aquatic–terrestrial ecotones. University of Washington, Seattle: 156–165. Pithart, D., 1999. Phytoplankton and water chemistry of several alluvial pools and oxbows after the flood event – a process of diversification. Algological Studies 95: 93–113. Reynolds, C. S., 1997. Vegetation Processes in the Pelagic: A Model for Ecosystem Theory. In Kinne, O. (ed.), Excellence in Ecology. Ecology Institute, Germany I, 339 pp.. Sorokin, Y. I., 1999. Aquatic Microbial Ecology. Backhuys Publishers, Leiden, 247 pp. ter Braak, C. J. F., 1988. CANOCO-aFORTRAN program for canonical community ordination by [partial] [detrended]

[canonical] correspondence analysis, principal components analysis and redundancy analysis (version 2.1). Report LWA-88-02. Agricultural Mathematics Group, Wageningen, 95 pp. Train, S. & L. C. Rodrigues, 1998. Temporal fluctuations of the phytoplankton community of the Baía River, in the Upper Paraná River floodplain, Mato Grosso do Sul, Brazil. Hydrobiologia 361: 125–134. Tuchman, N. C.,1996. The role of heterotrophy in algae. In Stevenson, R. J. et al. Algal Ecology. Freshwater Benthic Ecosystems. Academic Press, San Diego: 299–319. Unrein, F. & G. Tell, 1994. Respuesta del fitoplancton al enriquecimiento con fósforo y nitrógeno en aguas de una laguna de inundación del río Paraná. Rev. Hydrobiol. trop. 27 (4): 315–328. Unrein, F., 2001. Efecto de los nutrientes y el pH sobre el crecimiento y la estructura del fitoplancton en ambientes de la llanura aluvial del Paraná Inferior. Tesis de Doctorado, Universidad de Buenos Aires, 162 pp. Unrein, F., 2002. Changes in phytoplankton community along a transversal section of the Lower Paraná floodplain, Argentina. Hydrobiologia 468: 123–134. Utermöhl, M., 1958. Zur Vervollkommung der quantitativen Phytoplankton Methodik. MIH. Verh. int. Ver. Limnol. 9: 1–38. Van Dam, H., A. Mertens & J. Sinkeldam, 1994. A coded checklist and ecological indicator values of freshwater diatoms from the Netherlands. Netherlands J. Aquatic Ecol. 28: 117–133. Venrick, E. L., 1978. How many cells to count?. In: Sournia, A. (ed.), Phytoplankton Manual. UNESCO, Paris: 167–180. Vyverman, W., 1996. Phytoplankton diversity and abundance along environmental gradients in tropical floodplain lakes (SepikRamu Floodplain, Papua New Guinea). Perspectives in Tropical Limnology: 277–294. Welcomme, R. L., 1985. River fisheries. FAO Fish. Tech. Paper No. 262, Rome. Wilson, E. O., 1988. The current state of biological diversity. In Wilson, E. O. & F. M. Peter (eds), Biodiversity. National Academy Press, New York, 521 pp. Zalocar de Domitrovic, Y., 1990. Efecto de las fluctuaciones del nivel hidrométrico sobre el fitoplancton en tres lagunas isleñas en el área de confluencia de los ríos Paraná y Paraguay. Ecosur 16: 13–29. Zalocar de Domitrovic, Y., 1992. Fitoplancton de ambientes inundables del Río Paraná (Argentina). Revue d’ Hydrobiologie Tropicale 25: 177–188. Zalocar de Domitrovic, Y., 1993. Fitoplancton de una laguna vegetada por Eichhornia crassipes en el valle de inundación del río Paraná (Argentina). Ambiente Subtropical 3: 39–67.