Morpho-Functional Groups and phytoplankton development in two deep lakes (Lake Garda, Italy and Lake Stechlin, Germany)
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Hydrobiologia (2007) 578:97–112 DOI 10.1007/s10750-006-0437-0 P H Y T O P L A N K T O N W O RK S H O P Morpho-Functional Groups and phytoplankton development in two deep lakes (Lake Garda, Italy and Lake Stechlin, Germany) Nico Salmaso Æ Judit Padisák Springer Science+Business Media B.V. 2007 Abstract Phylogenetic classifications of plants Lake Garda, southern Alps, zmax = 350 m; bien- often do not reflect their ecological functions. In nium 2002–2003) and Lake Stechlin (north-east fact, the functional mechanisms of biological com- Germany, zmax = 67 m; 1995, 1998 and 2001). In munities may be better understood if species are both lakes, the temporal evolution of the phyto- pooled into groups having similar characteristics. plankton communities within individual years fol- The objective of this work is to evaluate, with the use lowed a regular annual cycle, with the exception of of multivariate methods, classifications based on the Lake Stechlin in 1998, when an irregular phyto- morphological and functional characteristics (size plankton pattern was caused by a sudden mass and form, mobility, potential mixotrophy, nutrient appearance of Planktothrix rubescens in the spring requirements, presence of gelatinous envelopes) of and summer months, resulting in a collapse of the cyanobacteria and eukaryotic algae to explain the whole community in autumn. Overall, the temporal seasonal dynamic of the phytoplankton community. developments of the phytoplankton communities The analyses involve data from two deep lakes: obtained on the basis of patterns of the morpho- functional groups appeared highly comparable with those obtained, in the single years, on the Guest editors: M. Albay, J. Padisák & L. Naselli-Flores basis of the original phytoplankton species matri- Morphological plasticity of phytoplankton under different ces. The comparison of the morpho-functional environmental constraints groups of the lakes Garda and Stechlin showed N. Salmaso (&) important differences in the abundance and sea- Natural Resources Department, Limnology and Fish sonality of the dominant phytoplankton types. Research Unit, IASMA Research Center, The results obtained in this study underline that Via E. Mach, 1, I-38010 S. Michele a/Adige (Trento), Italy the use of classifications based on the adaptive e-mail: nico.salmaso@iasma.it strategies of the single species may represent a useful tool to investigate the community evolution J. Padisák and to compare phytoplankton assemblages of Department of Limnology, University of Veszprém, Egyetem u. 10, Veszprem H-8200, Hungary different lakes, overcoming problems related to e-mail: padisak@almos.vein.hu possible differences of taxonomic accuracy and identification. J. Padisák Leibniz-Institut für Gewässerökologie und Binnenfischerei, Alte Fischerhütte 2, D-16775 Keywords Phytoplankton Morpho-Functional Stechlin-Neuglobsow, Germany groups Seasonal cycles Deep lakes 123
98 Hydrobiologia (2007) 578:97–112 Introduction ciations, with each comprising species that coexist and increase or decrease in numbers simulta- The need to understand relationships between neously. The species of a particular association structural and functional properties of the eco- share common ecological attributes so that they systems has initiated a number of classifications of may potentially dominate or co-dominate. Suc- aquatic biota on basis of their taxonomy or cessively, this approach was refined and expanded functional and structural characteristics. Such (Reynolds, 1997; Padisák & Reynolds, 1998). At classifications are of practical use for generalisa- present, the system comprises 31 associations tions across species and represent necessary tools accomodated on the basis of expert judgement for scientific communication and water-body and experience (Reynolds et al., 2002). Subse- analysis (Körner, 1993). quent studies defined three further associations The taxonomic classification of the living (Padisák et al., 2003a, 2006). organisms recently has benefited from the use Reynolds (1988) defined another system of of genetic and molecular techniques. However, classification based on the functional properties of despite the support of these recent methods and the phytoplankton species. Adapting the C–R–S the refinement of the phylogenetic classifications, concept formulated by Grime (1977) for the ecologists have often been dissatisfied with the terrestrial plants, he classified the phytoplankton taxonomic assemblages because they do not species into three groups based on their suscep- always and necessarily reflect the perceived tibility to disturbance, stress and utilisation of ecological functions (Solbrig, 1993). A taxo- limited resources. In this model, the C-strategists nomic unit may be composed, especially at (competitive species) dominate in lakes with low higher taxonomic levels (e.g., classes), by species intensity of disturbance and stress. The S-strate- possessing very different structural and func- gists (stress-tolerant species) develop in situations tional properties. Considering phytoplankton, we of low disturbance and high stress, whereas the R- may refer to examples from Chlorophyceae strategists (‘‘ruderals’’) dominate at high distur- which include types characterised by very differ- bance intensity and low stress. ent structure and life strategies, ranging from The different groups defined in these classifi- single picoplanktonic cells (Stockner & Antia, cations generally include species with common 1986) to long filaments and large multicellular features (Reynolds et al., 2002), indicating that colonies. To overcome these problems, ecologists the algae belonging to a specific group have have tried to group organisms with similar common morphological characteristics which are structural and functional characteristics with the powerful predictors of optimum dynamic perfor- aim to obtain a better understanding, and mance (Reynolds & Irish, 1997). In fact, the possibly generalisation, of the functioning of different strategies of different phytoplankton the ecosystems. The functional groups may be organisms are strongly related to differences in defined utilising elements that bear a certain set geometrical dimensions and shapes. Morpholog- of common structural and/or functional features. ical characteristics are essential features that These include quality criteria (size/form, physi- influence sinking properties (Padisák et al., ological and life strategy characteristics), tempo- 2003b, d; Salmaso, 2003), growth rates (Sommer, ral appearance and distributional characteristics, 1981; Tang, 1995), efficiency of nutrients and light whereas, at the species level, the functional utilisation (Sommer, 1984; Tilzer, 1984), and groups may include taxonomic units as well susceptibility to grazing (Lehman, 1988). These (Körner, 1993). elements, together with the ability to regulate One of the first formal attempts to define a buoyancy, the requirement for specific resources system of classification based on the functional (e.g., silica) and the ability to obtain alternative properties of the phytoplankton was proposed by carbon and nutrient sources, represent strong Reynolds (1980, 1984). Utilising long series of selective factors that are able to select the best phytoplankton observations from a group of lakes competitors under different environmental con- in Northwest England, he distinguished 14 asso- straints (Weithoff, 2003). 123
Hydrobiologia (2007) 578:97–112 99 In this work we evaluate, with the use of Materials and methods multivariate methods, classifications based on the morphological and functional characteristics of Methods in the field and laboratory cyanobacteria and eukaryotic algae to capture much of the seasonal phytoplankton variations in In Lake Garda, samples were collected every two deep lakes located at the southern border of 4 weeks between January 2002 and December the Alps (Lake Garda) and in north-east Ger- 2003 at the deepest zone of the lake (out of many (Lake Stechlin). Brenzone; zmax = 350 m); sampling was not car- ried out in January 2003. The average values of the chemical variables and phytoplankton abun- Study sites dance in the upper 20 m were estimated from samples collected at the integrated depths of 0– Detailed hydrological, morphological as well as 2 m, 9.5–11.5 m and 19–21 m. Phytoplankton limnological descriptions of these large and deep samples were fixed with acetic Lugol’s solution water bodies can be found in Salmaso (2003, and counting was performed with inverted micro- 2005) and Koschel & Adams (2003). scopes following the criteria reported by Lund et Concentrations of total phosphorus (TP) in the al. (1958). Biovolumes were calculated from euphotic layer (0–20 m) of Lake Garda show recorded abundances and specific biovolumes typical seasonal fluctuations from around 5 lg l–1 approximated to simple geometric solids (Rott, (summer) to 20 lg l–1 (spring overturn); corre- 1981). Zooplankton was collected by vertical tows sponding fluctuations of Soluble Reactive Phos- from 0 to 50 m using nets with mesh size of 80 lm phorus (SRP) range from below detection limit (Salmaso & Naselli-Flores, 1999). Zooplankton up to 10 lg l–1 (Salmaso, 2005). During summer, biomass was computed according to de Bernardi the nitrate concentrations in the productive layers & Canale (1995). Euphotic depths (zeu) were may decrease to very low levels (down to estimated from direct measurements made with a 100 lg l–1). NH4–N concentrations are generally radiometer LiCor, model Li-192SA. A detailed below 20 lg l–1. Thermal stratification begins in description of the analytical methods used in the May (10–12C at the surface), and strong thermal laboratory was reported by Salmaso (2002, 2005). stratification persists from June to September The phytoplankton patterns selected for this (20–25C). Winter euphotic depths (zeu) are study from Lake Stechlin include 3 years: 1995, around 25–40 m, whereas, during summer, the 1998 and 2001. Phytoplankton abundances were lower limit of zeu is between 15 and 20 m estimated from pooled samples collected from six (Salmaso, 2003). depths evenly distributed within the 0–25 m layer. The annual average TP concentrations in Lake Sampling frequency was weekly between March Stechlin vary between 12 and 21 lg l–1. SRP in and October, and biweekly throughout the rest of the euphotic layer is below the limit of detection the year. The statistical analyses were carried out (1–2.5 lg l–1) with maximum values (4–5 lg l–1) using monthly averages of the phytoplankton after the autumnal overturn. Dissolved inorganic biovolumes. Phytoplankton samples were treated, nitrogen (DIN: NOx–N + NH4–N) in the eupho- counted, and biomass was estimated as described tic zone is rather low, varying from the detection for Lake Garda. To make the data comparable to limit to around 100 lg l–1 (with occasional peaks those from Lake Garda, picophytoplankton was up to 150 lg l–1). The DIN/SRP ratios in the removed from the data-set. Samples for chemical euphotic zone vary between 2 and 159 (average: analyses were collected from 0 to 25 m and 30). Low values occur typically in the period July– analysed according to the OECD standards and October. Thermal stratification begins in May standards of Deutsche Einheitsverfahren. Crusta- (10–12C at the surface) and begins to be strongly cean zooplankton was collected by vertical tows pronounced in July, lasting to middle September from 0 to 22 m using nets with mesh size of (20–25C). Euphotic depth during the stratified 90 lm. Biomass of the single individuals was period extends to the upper 25 m. estimated following Bottrell et al. (1976) and 123
100 Hydrobiologia (2007) 578:97–112 Kasprzak (1984). Euphotic depths were estimated light and nutrient conditions in heterogeneous from zs, the Secchi disk depths (zeu = zs · 2.7). environments (Reynolds, 1997). Within flagel- Detailed description of the analytical methods lates, the second division separates the large was reported by Padisák et al. (2003c). The group of the Phytomonadina from the other chemical and trophic characteristics of the two (large and small) flagellates; this division corre- lakes during the examined periods are summar- sponds roughly to the separation of the potential ised in Table 1. mixotrophs from mostly autotrophic algae (Isaks- In both lakes, relative thermal resistance son, 1998; Jones, 2000), though, at present, no (RTR) was computed by the density difference complete information exists about the importance in the examined strata (0–20 m, Garda and 0– of mixotrophy for the single species. As for the 25 m, Stechlin) compared to the density differ- other non-flagellated organisms, a further division ence between 4 and 5C (Kalff, 2002); water separates the cyanobacteria and diatoms from the densities were estimated from temperature val- remaining groups. Cyanobacteria are prokaryotes ues. and exhibit defined morphological and physiolog- ical differences compared with the eukaryotic Identification of the Morpho-Functional phytoplankton. Among these, the ability to reg- groups ulate the vertical position through the formation of gas-vescicles (aerotopes) and the ability to fix The list of the Morpho-Functional Groups atmospheric nitrogen constitute a set of unique (MFG) is reported in Table 2. The criteria features (Whitton & Potts, 2000). Similarly, the adopted to discriminate the groups include motil- siliceous walls of the diatoms distinguishes these ity, the potential capacity to obtain carbon and organisms from other taxa and causes severe nutrients by mixotrophy, specific nutrient require- survival problems in stratifyed pelagic environ- ments, size and shape, and presence of gelatinous ments due to the enhanced susceptibility to envelopes (Weithoff, 2003). These criteria, along sinking losses of the heavier cells (Padisák et al., with the separation of the Cyanobacteria from the 2003d). Further subdivisions (3rd column of other algae and a further subdivision based Table 2) were based on size and shape. In this considering different life strategies, resulted in work, the separation of the large and small taxa differentiation of 31 groups. has been based on the susceptibility to grazing. The first division is based on the presence or However, a clear-cut delimitation of the graze- absence of flagella. Despite their small size, many able fraction for the entire phytoplankton com- flagellates can move vertically within the water munity is not possible. On practical grounds, column for tens of centimeters to several meters single cells and small colonies with linear dimen- daily (Wetzel, 2001). Motility offers the ability to sions
Hydrobiologia (2007) 578:97–112 101 Table 2 Morpho Functional groups (MFG; see explanation in the text) Flagellates Potential 1 Large (colonial or 1a Large Chrysophytes/Haptophytes 1a- mixotrophs unicellular) LargeChry 1b Large Dinophytes 1b- LargeDino 1c Large Euglenophytes 1c-LargeEugl 2 Small (unicellular) 2a Small Chrysophytes/Haptophytes 2a- SmallChry1 2b Small Dinophytes 2b-SmallDino 2c Small Euglenophytes 2c-SmallEugl 2d Cryptophytes 2d-Crypto Mostly 3 Phytomonadina 3a Unicellular Phytomonadina 3a-UnicPhyto autotrophs 3b Colonial Phytomonadina 3b-ColoPhyto Without Cyanobacteria 4 Unicellular 4 Unicellular cyanobacteria 4-UnicCyano flagella 5 Colonies 5a Thin filaments (Oscillatoriales) 5a-FilaCyano 5b Large vacuolated Chroococcales 5b- LargeVacC 5c Other large colonies, mostly non-vacuolated 5c- Chroococcales OtherChroo 5d Small colonies, Chroococcales 5d- SmallChroo 5e Nostocales 5e-Nostocales Diatoms 6 Large 6a Large Centrics 6a-LargeCent 6b Large Pennates 6b-LargePenn 7 Small 7a Small Centrics 7a-SmallCent 7b Small Pennates 7b-SmallPenn Others— 8 Large 8a Large unicells—Unicellular 8a- Unicellular Conjugatophytes/Chlorophytes LargeCoCh 8b Large unicells—Other groups 8b-LargeUnic 9 Small 9a Small unicells—Conjugatophytes 9a-SmallConj 9b Small unicells—Chlorococcales 9b-SmallChlor 9c Small Chrysophytes 9c- SmallChry2 9d Small unicells—Other groups 9d-SmallUnic Others— 10 Filaments 10a Filaments—Chlorophytes 10a- Colonial FilaChlorp 10b Filaments—Conjugatophytes 10b-FilaConj 10c Filaments—Xanthophytes 10c-FilaXant 11 Non filament. 11a Chlorococcales—Naked colonies 11a- colonies NakeChlor 11b Chlorococcales—Gelatinous colonies 11b- GelaChlor 11c Other colonies 11c-OtherCol Besides grazing, size and shape are important for in the dynamic evolution of different taxonomic influencing losses from suspension (Padisák et al., groups sharing common morphological features. 2003b). The final list of MFG in Table 2 inte- grates other two discriminant elements, i.e. the Data analysis presence of large gelatinous envelopes, with the separation of the naked and gelatinous Chloro- Phytoplankton matrices (taxa · sampling dates) coccales, and taxonomy. As for this last point, the were analysed considering the original data (the subdivisions were carried out considering the biomass of the single species) and two synthetic existence of different life strategies; in turn, this matrices obtained by considering the Morpho gives the opportunity to identify real differences Functional Groups and the phylogenetic algal 123
102 Hydrobiologia (2007) 578:97–112 classes. The three matrices were ordinated sepa- using the program CANOCO 4.5 with focus on rately, for every lake and year, by applying Non interspecies distances and biplot scaling (ter Metric Multidimensional Scaling (NMDS). The Braak & Šmilauer, 2002). Before the computa- ordinations were applied to Bray & Curtis’ tion, the biovolumes of the single MFG were log- dissimilarity matrices (Podani, 2000). Before transformed to reduce the weight of the most computation, the data were transformed by dou- abundant groups. The significance of the axes ble square root to reduce the weight of the more obtained by the CCA analysis was tested by abundant taxa/groups (Salmaso, 1996). The re- means of Monte Carlo permutations (ter Braak & sults of the ordinations obtained from the MFG Šmilauer, 2002). and the algal classes were compared with those obtained utilising the phytoplankton data. As in NMDS the orientation of the axes is arbitrary, Results before the comparison the configurations ob- tained from the MFG and the classes were Dominant classes and phytoplankton rotated and transposed by orthogonal Procrustes rotation to maximise their fit with the corre- The temporal development of the algal classes in sponding phytoplankton ordinations (Podani, the lakes Garda and Stechlin is reported in Fig. 1. 2000). Axes I and II from the phytoplankton, The dominant classes in Lake Garda were the MFG and classes ordinations were compared by cyanobacteria, diatoms and chlorophytes. These parametric and non-parametric correlations. three classes were dominated by Planktothrix The comparison of the phytoplankton commu- rubescens, Fragilaria crotonensis and Mougeotia nities of the lakes Garda and Stechlin was carried sp., respectively. The cyanobacteria showed their out considering the differences in the structure of maximum development in summer and/or au- their respective MFG. The common matrix, tumn (Fig. 1a). Besides Planktothrix, other including the biovolumes of the MFG of the two important cyanobacteria were large colonies of lakes, was analysed by Principal Components Aphanothece/Aphanocapsa. Also, in the 2 years Analysis calculated from the correlation matrix, studied Anabena lemmermannii developed epi- followed by orthogonal rotation (varimax meth- sodic oligotrophic blooms (sensu Salmaso, 2000) od). Before computation, the data were log- strictly localised to the upper few centimetres of transformed, Yi = log(Xi + 1), to reduce the the water column and with low overall biovo- weight of the most abundant taxa. The ordination lumes (
Hydrobiologia (2007) 578:97–112 103 spp. In the successive year, despite the lack of a teria were between 8 and 23 mm3 m–3 in all discernible spring peak, the diatoms developed 3 years. Apart from the cryptophytes (developing seasonally with the same pool of species. The irregularly with Cryptomonas erosa/ovata, Rho- chlorophytes showed a comparable development domonas lens and R. minuta), the remaining in the 2 years studied, with a small spring classes developed substantial biovolumes and increase, followed by large early and mid-summer showed repetitive patterns in the three studied growth. Though with low biovolumes, besides years. Chlorophytes were mainly present from Mougeotia other chlorophytes showed a notice- early spring to early autumn; the most abundant able development from early summer to late were Neocystis policocca, Pseudosphaerocystis summer/early autumn (Ankyra judayi, Coela- lacustris and Scenedesmus costato-granulatus. strum spp., other coccal greens and Chlorococ- The contribution of dinophytes was mainly due cales). The remaining classes developed lower to Ceratium hirundinella and Gymnodinium hel- biovolumes. The xanthophytes (Tribonema sp.) veticum. In the 3 years considered here, the showed their maximum development from April biovolume of the chrysophytes and haptophytes to the end of June, whereas the chrysophytes was mainly provided by Chrysochromulina parva, mainly occurred in late spring (Dinobryon diver- Dinobryon sociale and Ochromonas sp. Xantho- gens, D. sociale and Ochromonadaceae). The phyceae and Euglenophyceae showed low occur- dinophytes showed a greater biovolume peak rences and low biovolumes. from late summer to early autumn (Ceratium hirundinella); Gymnodinium helveticum was Phytoplankton ordinations—single lakes detectable also during the spring months. The cryptophytes (mainly Plagioselmis nannoplanctica The configurations of the phytoplankton samples and Rhodomonas minuta) showed a more irreg- of Lake Garda obtained by mean of NMDS are ular distribution over the 2 years. reported in Fig. 2. For the 2 years considered, The temporal development of the phytoplank- the configurations obtained utilising the phyto- ton in Lake Stechlin showed a different pattern plankton classes and the MFG were superim- (Fig. 1b). The two more abundant classes in the posed to the corresponding configurations examined period were the diatoms and the obtained utilising the phytoplankton species. cyanobacteria. However, the diatoms peaked high The stress values (Kruskal & Wish, 1978) of in 1995 and 2001, whereas the cyanobacteria, the six NMDS configurations ranged between largely Planktothrix rubescens, showed a conspic- 0.05 and 0.13. The configurations obtained with uous development exclusively in 1998, from the phytoplankton species showed an ordered January to September; in October Planktothrix and cyclic development. In 2002, the annual underwent a complete breakdown and the com- development based on the phytoplankton classes munity was composed exclusively of a few followed a different pattern compared with the diatoms and chlorophytes. The spring peaks of ordination based on the phytoplankton samples; the diatoms in 1995 and 2001 were due to the differences were marked along the first axis, different species, namely Cyclotella spp., and whereas the ordination of the samples along the Aulacoseira islandica with a minor contribution second axis showed a comparable ranking of Cyclotella spp., respectively. Moreover, the between the two configurations (Fig. 2a; Ta- annual development of the diatoms in 2001 was ble 3). Though less evident, the differences also characterised by Asterionella formosa in between the two ordinations were detected also spring and Fragilaria crotonensis in early autumn. in 2003, but with the major discrepancies being Excluding Planktothrix, the contribution of the present along the second axis (Fig. 2b; Table 3). other cyanobacteria was always very low. Anaba- In contrast, the ordinations obtained on the basis ena lemmermannii was detected with low abun- of the Morpho-Functional Groups (MFG) dances (annual peaks of ca. 20–100 mm3 m–3) matched those achieved from the complete between June/July and August/October. The phytoplankton matrices very closely (Fig. 2c, d; annual biovolume peaks of the other cyanobac- Table 3). 123
104 Hydrobiologia (2007) 578:97–112 Fig. 2 Lake Garda. Phytoplankton species Ordination of Classes phytoplankton samples in 5 the two-dimensional Non 1 7 1 Metric Multidimensional Scaling configurations; the 4 6 7 5 NMDS Axis 2 arabic numbers indicate 7 9 the month of sampling. 0 3 9 4 a, c: 2002; b, d: 2003 10 6 1 0 10 2 8 8 1 11 -1 12 11 3 (a) (b) 2 12 -1 -1 0 1 -2 -1 0 1 Phytoplankton species MFG 1 1 NMDS Axis 2 0 0 -1 (c) (d) -1 -1 0 1 -2 -1 0 1 NMDS Axis 1 NMDS Axis 1 The NMDS ordinations of the Lake Stechlin distribution of the data) showed a better agree- samples are reported in Fig. 3. In 1995 and 2001, ment between the phytoplankton species and the temporal succession of the samples based on MFG configurations. the phytoplankton data followed a clear cyclic pattern (Fig. 3a, c; black circles), whereas in 1998 Phytoplankton ordinations—joint analysis the seasonal evolution was strongly typified by the separation of the October sample from the The results of the ordination of the samples of the remaining ones (Fig. 3b). This sample, as dis- lakes Garda and Stechlin by PCA are reported in cussed above, was characterised by a sudden Fig. 4 and Table 4. The analysis does not include collapse of Planktothrix rubescens and other the data recorded in Lake Stechlin in 1998, in dominant seasonal phytoplankton species. As in order to simplify the interpretation of the ordi- the case of the Lake Garda, in 1995 and 2001 the nation diagram. A two factor solution explaining ordinations of the MFG agreed more closely with 40% of the total variance was computed. (Ta- those for the phytoplankton species as did those ble 4). In addition to the significant proportion of based on the algal classes; however, the algal the explained variance, this solution was chosen classes performed better than for the Lake Garda because it yielded the most straightforward inter- samples, as evidenced by the high significance of pretation without loss of significant information. the relationships between the couples of axes The relative position of the single samples in (Table 3). As for the 1998 configurations, in Fig. 4 was confirmed by the application of the which the parametric correlations were strongly NMDS ordination (figure not shown); after determined by the October sample, the Spearman orthogonal Procrustes rotation of the NMDS correlations (which are independent from the configuration, the first and second axis of the 123
Hydrobiologia (2007) 578:97–112 105 Table 3 Correlation coefficients between the first two species (Phyto) and the corresponding NMDS axes axes of the Non Metric Multidimensional Scaling (NMDS) obtained with the algal classes and Morpho-Functional configurations obtained on the basis of the phytoplankton Groups A B Lake Garda Axes classes Axes MFG Axes classes Axes MFG 2002 Axis 1 Phyto 0.35 0.99 0.48 0.99 Axis 2 Phyto 0.96 0.99 0.97 0.99 2003 Axis 1 Phyto 0.87 0.99 0.81 0.99 Axis 2 Phyto 0.41 0.87 0.30 0.83 C D Lake Stechlin Axes classes Axes MFG Axes classes Axes MFG 1995 Axis 1 Phyto 0.83 0.98 0.67 0.97 Axis 2 Phyto 0.90 0.95 0.90 0.94 1998 Axis 1 Phyto 0.98 0.99 0.55 0.64 Axis 2 Phyto 0.77 0.90 0.50 0.80 2001 Axis 1 Phyto 0.93 0.97 0.92 0.94 Axis 2 Phyto 0.80 0.87 0.82 0.85 For example, in the first column, the entries 0.35 and 0.96 refer to the correlations between Axis 1 Phyo and Axis 1 Classes, and Axis 2 Phyo and Axis 2 Classes, respectively. A,C: Pearson correlations; B,D: Spearman correlations. The significant correlations are reported in bold (P < 0.01) and italics (P < 0.05) Phytoplankton species 1 1 5 6 Classes 4 4-9 1 6 3 4 3 7 10 7 8 5 NMDS Axis 2 NMDS Axis 2 NMDS Axis 2 0 0 2 3 0 2 1 9 2 8 1 1 10 12 9 12 11 10 11 -1 -1 11 -1 (a) (b) 12 (c) -2 -1 0 1 2 -1 0 1 2 3 -2 -1 0 1 Phytoplankton species 1 1 MFG 1 NMDS Axis 2 0 0 0 -1 -1 -1 (d) (e) (f) -2 -1 0 1 2 -1 0 1 2 3 -2 -1 0 1 NMDS Axis 1 NMDS Axis 1 NMDS Axis 1 Fig. 3 Lake Stechlin. Ordination of phytoplankton samples in the two-dimensional Non Metric Multidimensional Scaling configurations; the arabic numbers indicate the month of sampling. a, d: 1995; b, e: 1998; c, f: 2001 PCA ordination showed a strong and significant to Lake Stechlin (right panel). In both cases, the (P < 0.01) correlation with the corresponding annual development of the samples followed a NMDS axes (r = 0.93 and r = 0.88, respectively). cyclical pattern, although less evident compared Two main differences became apparent in the to those reported in the analyses of the single spatial arrangement of the samples of the two years (Figs. 2, 3). The cyclical patterns of the two lakes (Fig. 4). The samples formed two distinct lakes followed a common direction only along the groups belonging to Lake Garda (left panels) and second axis, and an opposite direction along the 123
106 Hydrobiologia (2007) 578:97–112 Table 4 Principal components analysis with varimax 3 Garda 2002 rotation: percentage of explained variance and correla- Garda 2003 tions between the first two factors and the input variables Stechlin 1995 2 Stechlin 2001 PCA Axis I II Variance explained 23.2% 16.7% 1 9d—SmallUnic 0.84 –0.11 Axis 2 7b—SmallPenn 0.76 –0.10 2b—SmallDino 0.74 0.27 0 2a—SmallChry1 0.63 0.32 10a—FilaChlorp 0.56 0.13 7a—SmallCent 0.52 0.25 -1 11a—NakeChlor 0.48 0.57 5a—FilaCyano –0.84 –0.07 10b—FilaConj –0.74 –0.14 -2 3a—UnicPhyto –0.71 0.22 -2 -1 0 1 2 8a—LargeCoCh –0.61 0.11 Axis 1 10c—FilaXant –0.60 –0.10 6b—LargePenn –0.49 –0.12 Fig. 4 Ordination of the samples by principal components 5e—Nostocales –0.09 0.86 analysis with varimax rotation. The arrows indicate the 5c—OtherChroo –0.07 0.66 direction of the temporal sequences. Succession of the 1b—LargeDino –0.27 0.61 months as in Figs. 2 and 3 11c—OtherCol 0.22 0.60 9c—SmallChry2 0.36 0.59 11b—GelaChlor 0.34 0.56 Garda stood out for the presence of large func- 9a—SmallConj 0.14 0.54 tional units, both filamentous or unicellular. The 1a—LargeChry –0.09 0.52 3b—ColoPhyto –0.06 0.48 second axis showed a strong and positive corre- 6a—LargeCent 0.19 –0.58 lation with 11 MFG; these groups assumed a 9b—SmallChlor –0.36 0.25 progressive and greater importance from the 5b—LargeVacC –0.30 0.15 winter to the summer samples. In contrast, the 2c—SmallEugl 0.06 0.31 5d—SmallChroo 0.09 0.24 large centric chain-forming diatoms showed a 2d—Crypto 0.24 –0.23 negative correlation with the second axis (Ta- The significant correlations are reported in bold (P < 0.01) ble 4). This second set of MFG included phyto- and italics (P < 0.05). Variables are abbreviated as in plankton types with more comparable temporal Table 2. Underlining of Morpho-Functional Groups patterns in the two lakes. indicates high abundance (see text) The more abundant MFG in Table 4 are underlined. These groups reached biovolume first axis. This shows that the spring and summer peaks greater or equal than 100 mm3 m–3 at least samples were characterised by the greater differ- in one lake and in 1 year. With this criterion it ences, with a major comparability of the phyto- was possible to discriminate the eleven groups plankton features during the autumn and early represented in Fig. 5. Among the ‘‘discriminat- winter months. ing’’ MFG, the small centric diatoms showed a The analysis of the factor loadings (Table 4) different seasonal development in the two lakes, allowed to interpret in terms of MFG composi- whereas the filamentous cyanobacteria, conju- tion the different patterns observed in the two gatophytes and xanthophytes, and the large pen- lakes. The first axis showed a strong correlation nate diatoms were typically present in the with thirteen MFG; among these, only the naked plankton of Lake Garda. However, as for the Chlorococcales showed a positive correlation also filamentous cyanobacteria, it is important to with the second axis. This first set of MFG emphasise that this analysis does not include the constituted the discriminant phytoplankton types contribution of the 1998 data for Lake Stechlin, in the two lakes. In particular, the MFG typical of which is characterised by a huge development of the Lake Stechlin were represented by small and Planktothrix rubescens. The remaining groups in mostly unicellular species, whereas the Lake Fig. 5 are represented by ‘‘common’’ MFG; 123
Hydrobiologia (2007) 578:97–112 107 300 600 7a-SmallCent 5a-FilaCyano 3000 10b-FilaConj -3 200 400 mm m 2000 3 100 200 1000 0 0 0 400 10c-FilaXant 1500 6b-LargePenn 100 5c-OtherChroo 300 -3 mm m 1000 75 200 3 50 100 500 25 0 0 0 300 1b-LargeDino 400 11b-GelaChlor 1a-LargeChry 300 -3 mm m 200 300 200 3 100 50 100 0 0 0 J F MA MJ J A SOND 1200 6a-LargeCent 250 2d-Crypto 200 Garda 2002 1000 -3 mm m 150 Garda 2003 800 3 150 100 Stechlin 1995 100 50 Stechlin 2005 50 0 J F MA MJ J A S ON D J F MA MJ J A SOND Fig. 5 Seasonal development of the dominant MFG (abbreviations as in Table 2) in the lakes Garda and Stechlin among these, some differences were observed in to 0.19 and 0.08, which contributed a percentage the gelatinous chlorococcales, which showed a of explained variance of 28.5% and 11.2%, more ample temporal development in Lake respectively. The application of the Montecarlo Stechlin (spring–early autumn) compared with tests to evaluate the significance of the extracted Lake Garda (summer–early autumn). axes yielded highly significant results for the whole set of axes and for the first axis MFG and environmental variables (P < 0.01). Excluding RTR (linked with the second axis) and, partly, SRP, the first axis was Figure 6 reports the simultaneous ordination of strongly linked to the whole set of variables. The the samples and MFG of the lakes Garda (Fig. 6a, first axis distinguishes the late winter and spring b) and Stechlin (Fig. 6c, d) in relation to the samples from the summer and early autumn principal environmental variables and zooplank- samples; the second axis contributes to separate ton biomass obtained by the application of the late spring and summer samples from the late Canonical Correspondence Analysis. As in the autumn ones (Fig. 6a). During the annual cycle, case of PCA, the CCA analysis applied to Lake from winter to late summer the samples were Stechlin does not include the data recorded in characterised by a progressive decrease in the 1998. Moreover, the group 2c (small eugleno- nutrient and light availability, and by a progres- phytes) was excluded, because it was present only sive increase of thermal stability and grazing in two occasions and with very low biovolumes pressure. Figure 6b shows a contraposition be- (
108 Hydrobiologia (2007) 578:97–112 1.0 1.0 6 RTR RTR 7 7 10c-FilaXant 1a-LargeChry 6 5 8 Temp. Temp. 11 7 5d-SmallChroo 12 2a-SmallChry1 11b-GelaChlor 5b-LargeVacC 5 10b-FilaConj 9b-SmallChlor 10 5e-Nostocales 4 1 Biom. Copepoda 8a-LargeCoCh 3b-ColoPhyto 0.0 0.0 Biom. Copepoda DIN 3 7 DIN 6b-LargePenn 11 9 8 Reactive-Si 1b-LargeDino 5c-OtherChroo Reactive-Si Zeu Biom. Cladocera Zeu Biom. Cladocera 4 6a-LargeCent 2b-SmallDino 11a-NakeChlor 2 SRP 2 10 9 SRP 3 (b) -1.0 (a) -1.0 12 -1.0 0.0 1.0 -1.0 0.0 1.0 1.0 1.0 7 5 8 4 1a-LargeChry Biom. Copepoda Biom. Copepoda 4 RTR 9d-SmallUnic RTR 3 9c-SmallChry2 SRP 7b-SmallPenn 5d-SmallChroo Temp. 8a-LargeCoCh Temp. SRP 11b-GelaChlor 3a-UnicPhyto DIN 3 6 7 DIN 0.0 0.0 6b-LargePenn 1 Biom. Cladocera 6a-LargeCent 7a-SmallCent Biom. Cladocera 12 6 5e-Nostocales 2 1 9 9 1b- LargeChry 2d-Crypto 10b-FilaConj Reactive-Si 10 9b-SmallChlor 5 Reactive-Si 2 8 Zeu 5c-OtherChroo 11c-OtherCol Zeu 11 5a-FilaCyano 9a-SmallConj 11 3b-ColoPhyto 12 -1.0 -1.0 (c) 10 (d) -1.0 0.0 1.0 -1.0 0.0 1.0 Fig. 6 Ordination of samples (left panels) and Morpho- indicate the month of sampling. The four MFG without Functional Groups (right panels) of (a, b) Lake Garda and labels around the origin in (b) are: 3a-UnicPhyto, 5a- (c, d) Lake Stechlin in the CCA plane defined by the first FilaCyano, 7a-SmallCent and 2d-Cryp. The four MFG two axes. Samples refer to (a) 2002 and 2003 (shaded and around the origin in (d) are: 2a-SmallChry1, 2b-SmallDi- empty circles, respectively), and (c) 1995 and 2001 (shaded no, 10a-FilaChlorp and 11a-NakeChlor. MFG are abbre- and empty circles). The arabic numbers in (a) and (c) viated as in Table 2 As for Lake Stechlin (Figure 6c, d), the eigen- (which were associated more with the second values associated to the first two axes were equal axis), the other variables showed a closer link to 0.16 and 0.03, which contributes a percentage with the first axis. An important difference in the of explained variance of 33.7% and 7.0%, respec- ordination of the MFG of the two lakes is tively. The Montecarlo tests to evaluate the represented by the position of the filamentous significance of the extracted axes yielded highly conjugatophytes (10b) in the two ordination significant results for the whole set of axes and for diagrams (Fig. 6b, d); however, in the Lake the first axis (P < 0.01). With a few differences, Stechlin this group was always identified with the ordination of the samples (Fig. 6c) showed a very low quantities and only in one occasion with pattern comparable with the results obtained for moderate biovolumes (June 2001;
Hydrobiologia (2007) 578:97–112 109 location of the common MFG (i.e. the groups if they increased or decreased in abundance at the both present in Fig. 6b, d) along the first axes of same time and at the same place. With the the two configurations followed a similar ranking, deductive approach (see Gitay et al., 1999), with Pearson and Spearman correlations between functional classifications are derived from state- the two set of coordinates equal to 0.57 (P = 0.01) ments established a priori on the basis of the and 0.54 (P < 0.05), respectively. importance of specific processes (e.g., physiology) or properties (e.g., morphology). The C–R–S classification of phytoplankton based on morpho- Discussion logical descriptors of the specific taxa proposed by Reynolds (1988: p. 396: Fig. 10-2) conforms to Functional groups were used in different fields of this procedure. Moreover, successive refinements ecological research, including vegetation studies of the classifications based on the recognition of (e.g. Leishman & Westoby, 1992; Pillar, 1999), associations (Reynolds et al., 2002) integrate prediction of effects of global climate change some of the elements of the deductive approach. (Gitay et al., 1999), conservation biology (Gitay Recognising that the separation of the phyto- et al., 1999; Pressey et al., 1993), studies of plankton species on the basis of their morpho- microorganisms (Meyer, 1993), fungi (Oberwin- logies coincides with the distributions of the same kler, 1993), macrophytes (Shipley et al., 1989) and species in different habitats with different light macroinvertebrates (Usseglio-Polatera et al., and nutrient resources, these authors proposed to 2000). Irrespective of the type of organisms employ the morphological properties to fit hith- studied, a common goal was to find classifications erto functionally unclassified taxa into existing useful to define and understand the dynamic functional categories. The results presented here behaviour of groups of species in relation to appear to support Reynolds’ approach (see also environmental variation. This approach assumes below). The different sets of characters may also that the characteristics of a community can be be processed with multivariate techniques to better understood and managed if species are group species with similar properties in clusters grouped into classes that possess similar charac- (‘‘data-driven approach’’; Gitay et al., 1999). teristics or behave similarly (Solbrig, 1993). In the work presented here, the deductive However, these general statements, though sim- approach has been used to identify the morpho- ple to understand and accept on an intuitive basis, functional groups on the basis of predetermined prove quite difficult in their practical implemen- relevant traits that influence essential functional tation. In fact, the structural and functional processes and ecological characteristics of the classifications does not produce exclusive and phytoplankton organisms. These traits have been ultimate solutions because the species may be chosen considering their relevance as proxies for grouped utilising different criteria. population performances and their facility of In the context of phytoplankton ecology, it is measurement or definition, discarding other char- convenient and simple to distinguish two ap- acters more difficult to determine or extract from proaches to obtain functional classifications: literature (e.g., growth rates; see Weithoff, 2003). dynamic and deductive. The most straightforward Some final subdivisions, based on taxonomy and method of establishing functional groups is based different life strategies, reflect the actual spectra on the temporal development of the species: of algae present in the lakes Garda and Stechlin, identifying groups that show similar dynamic so that it could not be exhaustive for lakes with behaviour; this method is based on the implicit different characteristics; this would require some notion of groups of species responding similarly minor integrations and arrangement of the final to a set of particular environmental conditions groups for other specific studies. (e.g., Fabbro & Duivenvoorden, 2000; Salmaso, The analyses carried out on the lakes Garda 2003). The phytoplankton associations originally and Stechlin demonstrate a correspondence defined by Reynolds (1980, 1984) are essentially between the seasonal development of the com- based on this approach. Species were put together munities based on the MFG and phytoplankton 123
110 Hydrobiologia (2007) 578:97–112 species. This suggests that the differentiation are needed to substantiate the reliability of this among phytoplankton on the basis of specific approach. adaptations, requirements and morphological In both lakes, the temporal development of characters determines the formation of groups phytoplankton within individual years followed a of species sharing more or less similar ecological regular annual cycle, with the exception of Lake characteristics. This might be, in part, confirmed Stechlin in 1998, due to a sudden mass appearance by the temporal dynamic of the MFG, which is of Planktothrix rubescens in the spring and sum- characterised by delimited temporal develop- mer months. Reasons for development to Plank- ments. However, these aspects will require tothrix dominance have not been clear since water further in-depth studies regarding specifically chemical variables did not indicate any conspicu- the degree of comparability of the ecological ous foregoing change; the only unusual event prior requirements of single species belonging to the dominance of this species was the long lasting the same MFG in different phytoplankton ice in the winter of 1995/1996 (Padisák, 2003a). communities. Excluding the 1998 case documented in Lake The CCA analysis showed that some groups— Stechlin, the regular cycles observed in the other such as the cyanobacteria (with the exclusion of years in the two lakes were determined only in Oscillatoriales), colonial Chlorococcales and Phy- part by the development of the same dominant tomonadina, and the large chrysophytes—appeared MFG. This suggests that if aquatic habitats con- more adapted to low nutrient concentrations and, strained by particular combinations of environ- due to their size, higher grazing pressure. These mental variables will support more than one likely groups showed a different sensitivity to the water MFG with appropriate adaptations, this does not column stability; the MFG with the higher scores necessarily imply that these groups will be present along the RTR gradient showed various adaptations with similar relative biomasses (or that, in some against sinking, i.e. the presence of flagella, gas cases, that they will be there; cf. also Reynolds et vescicles and large gelatinous envelopes. The late al., 2002). This aspect raises further questions and winter and spring MFG were composed also by perspectives about the identification of alternative large and potentially rapidly sinking colonies, i.e. causal factors able to explain the dominance of organisms that require low stability of the water different MFG in these two lakes. column. Moreover, the winter and spring groups In essence, the identification and use of MFG developed with conditions of high nutrient avail- may constitute a useful tool for the synoptic ability and euphotic depth. analysis of phytoplankton communities, particu- The ability of the MFG to mimic the seasonal larly comparing different lakes. However, the development of the phytoplankton community has usefulness of this approach in predicting the important practical consequences and opens inter- development of probable phytoplankton types esting perspectives. MFG may represent a useful as a result of different combinations of environ- tool to investigate concurrently the community mental factors needs to be analysed in detail evolution of different lakes based on the adaptive considering other lakes with different physio- strategies of the single species, overcoming graphic and trophic characteristics. problems related to the comparison of different taxa and to the existence of possible differences Acknowledgements The limnological research in Lake Garda was partially funded by the Veneto Region and of taxonomic accuracy and identification. More- ARPAV (Veneto Region Environment Protection over, MFG could contribute to predicting the Agency). We are grateful to Dr Giorgio Franzini and Dr development of most probable phytoplankton Fabio Decet (ARPAV) for their logistic support in the types as a result of different combinations of field and laboratory. We thank the technical staff of the IGB and especially Mrs Johanna Dalchow, Mrs Elke environmental factors. In this regard, the com- Mach, Mrs Monika Papke, Mrs Uta Mallok and Mrs parable responses of the MFG common to the Adelheid Scheffler for their careful technical assistance. lakes Garda and Stechlin to the same set of We wish also to thank Dr Ingrid Chorus and Dr Luigi physical, chemical and biological variables con- Naselli-Flores for helpful comments on the first draft of the manuscript. stitute promising elements, though more studies 123
Hydrobiologia (2007) 578:97–112 111 References Oberwinkler, F., 1993. Evoultion of Functional Groups in Basidiomycetes. (Fungi). In Schulze, E. D. & Bottrell, H. H, A. Duncan, Z. M. Gliwicz, E. Grygierek, H. A. Mooney (eds), Biodiversity and Ecosystem A. Herzig, A. Hillbricht-Illkowska, H. Kurasawa, Function. Ecological Studies. Springer-Verlag, Berlin, P. Larsson & T. Wenglenska, 1976. A review of some 143–164. problems in zooplankton production studies. Norwe- Padisák, J. & C. S. Reynolds, 1998. Selection of phyto- gian Journal of Zoology 24: 419–456. plankton associations in Lake Balaton, Hungary, in de Bernardi, R. & C. Canale, 1995. Ricerche pluriennali response to eutrophication and restoration measures, (1948–1992) sull’ecologia dello zooplancton del Lago with special reference to the cyanoprokaryotes. Hyd- Maggiore. Documenta Istituto Italiano di Idrobiolo- robiologia 384: 41–53. gia 55: 1–68. Padisák, J., G. Borics, G. Fehér, I. Grigorszky, I. Oldal, Fabbro, L. D. & L. J. Duivenvoorden, 2000. A two-part A. Schmidt & Z. Zámbóné-Doma, 2003a. Dominant model linking multidimensional environmental gradi- species and frequency of equilibrium phases in late ents and seasonal succession of phytoplankton assem- summer phytoplankton assemblages in Hungarian blages. Hydrobiologia 438: 13–24. small shallow lakes. Hydrobiologia 502: 157–168. Padisák, J., E. _ Soróczki-Pintér & Z. Rezner, 2003b. Gitay, H., I. R. Noble & J. H. Connell, 1999. Deriving functional types for rain-forest trees. Journal of Sinking properties of some phytoplankton shapes Vegetation Science 10: 641–650. and the relation of form resistance to morphological Grime, J. P., 1977. Evidence for the existence of three diversity of plankton—an experimental study. Hyd- primary strategies in plants and its relevance to robiologia 500: 243–257. ecological and evolutionary theory. American Natu- Padisák, J., W. Scheffler, P. Kasprzak, R. Koschel & L. ralist 111: 1169–1194. Krienitz, 2003c. Interannual variability in the phyto- Isaksson, A., 1998. Phagotrophic phytoflagellates in plankton composition of Lake Stechlin (1994–2000). lakes—a literature review. Archiv für Hydrobiolo- Archiv für Hydrobiologie/Advances in Limnology 58: gie/Advances in Limnologie 51: 63–90. 101–133. Jones, R. I., 2000. Mixotrophy in planktonic protists: an Padisák, J., W. Scheffler, C. Sı́pos, P. Kasprzak, R. overview. Freshwater Biology 45: 219–226. Koschel & L. Krienitz, 2003d. Spatial and temporal Kalff, J., 2002. Limnology. Prentice Hall, Prentice Hall, pattern of development and decline of the spring Upper Saddle River, NJ. diatom populations in Lake Stechlin in 1999. Archiv Kasprzak, P., 1984. Bestimmung des Körperkoohlenstoffs für Hydrobiologie/Advances in Limnology 58: 135– von Planktoncrustaceen. Limnologica 15:191–194. 155. Körner, C., 1993. Scaling from Species to Vegetation: The Padisák, J., I. Grigorszky, G. Borics & É. Soróczki-Pintér, Usefulness Of Functional Groups. In Schulze, E. D. & 2006. Use of phytoplankton assemblages for monitor- H. A. Mooney (eds), Biodiversity and Ecosystem ing ecological status of lakes within the Water Function. Ecological Studies. Springer-Verlag, Berlin, Framework Directive: the assemblage index. Hydro- 117–140. biologia 553: 1–14. Koschel, R., D. D. Adams (eds), 2003. Lake Stechlin—An Pillar, V. D., 1999. On the identification of optimal plant approach to Understanding an Oligotrophic Lowland functional types. Journal of Vegetation Science 10: Lake. Archiv für Hydrobiologie/Advances in Limnol- 631–640. ogy 58: 1–311. Podani, J., 2000. Introduction to the Exploration of Kruskal, J. B. & M. Wish, 1978. Multidimensional Scaling. Multivariate Biological Data. Backhuys, Leiden. Sage Publications, Beverly Hills and London. Pressey, R. L., C. J. Humpheries, C. R. Margules, R. I. Lehman, J. T., 1988. Selective Herbivory and its Role in Vane-Wright & P. H. Williams, 1993. Beyond oppor- the Evolution of Phytoplankton Growth Strategies. In tunism: key principles for systematic reserve selection. Sandgren, C. D. (ed.), Growth and reproductive Trends in Ecology and Evolution 8: 124–128. strategies of freshwater phytoplankton. Cambridge Reynolds, C. S., 1980. Phytoplankton assemblages and University Press, Cambridge, 369–387. their periodicity in stratifying lake systems. Holarctic Leishman, M. R. & M. Westoby, 1992. Classifying plants Ecology 3: 141–159. into groups on the basis of associations of individual Reynolds, C. S., 1984. The Ecology of Freshwater Phyto- traits—evidence from Australian semi-arid wood- plankton. Cambridge University Press, Cambridge, lands. Journal of Ecology 80: 417–424. 384 pp. Lund, J. W. G., C. Kipling & E. D. Le Cren, 1958. The Reynolds, C. S., 1988. Functional morphology and the inverted microscope method of estimating algal num- adaptive strategies of freshwater phytoplankton. In bers and the statistical basis of estimations by count- Sandgren, C. D. (ed.), Growth and Reproductive ing. Hydrobiologia 11: 143–170. Strategies of Freshwater Phytoplankton. Cambridge Meyer, O., 1993. Functional Groups of Microorganisms. In University Press, Cambridge, 388–433. Schulze, E. D. & H. A. Mooney (eds), Biodiversity Reynolds, C. S., 1997. Vegetation Processes in the Pelagic: and Ecosystem Function. Ecological Studies. Spring- A Model for Ecosystem Theory. Ecology Institute, er-Verlag, Berlin: 67–96. Oldendorf, 371 pp. 123
112 Hydrobiologia (2007) 578:97–112 Reynolds, C. S. & A. E. Irish, 1997. Modelling phyto- H. A. Mooney (eds), Biodiversity and Ecosystem plankton dynamics in lakes and reservoirs: the Function. Ecological Studies. Springer-Verlag, Berlin: problem of in-situ growth rates. Hydrobiologia 349: 97–116. 5–17. Sommer, U., 1981. The role of r- and K-selection in the Reynolds, C. S., V. Huszar, C. Kruk, L. Naselli-Flores & S. succession of phytoplankton in Lake Constance. Acta Melo, 2002. Towards a functional classification of the Oecologia Generale 2: 327–342. freshwater phytoplankton. Journal of Plankton Re- Sommer, U., 1984. The paradox of the plankton: fluctua- search 24: 417–428. tions of phosphorus availability maitain diversity of Rott, E., 1981. Some results from phytoplankton counting phytoplankton in flow-through cultures. Limnology intercalibrations. Swiss Journal of Hydrology 43: 34–62. and Oceanography 29: 633–636. Salmaso, N., 1996. Seasonal variation in the composition Stockner, J. G. & N. J. Antia, 1986. Algal picoplankton and rate of change of the phytoplankton community from marine and freshwater ecosystems: a multidisci- in a deep subalpine lake (Lake Garda, Northern plinary perspective. Canadian Journal of Fisheries Italy). An application of nonmetric multidimensional And Aquatic Sciences 43: 2472–2503. scaling and cluster analysis. Hydrobiologia 337: 49–68. Tang, E. P., 1995. The allometry of algal growth rates. Salmaso, N. & L. Naselli-Flores, 1999. Studies on the Journal of Plankton Research 17: 1325–1335. zooplankton of the deep subalpine Lake Garda. ter Braak, C. J. F. & P. Šmilauer, 2002. CANOCO Journal of Limnology 58: 66–76. reference manual and CanoDraw for Windows user’s Salmaso, N., 2000. Factors affecting the seasonality and guide: Software for canonical community ordination distribution of cyanobacteria and chlorophytes: a case (version 4.5). Microcomputer Power (Ithaca, NY, study from the large lakes south of the Alps, with USA). special reference to Lake Garda. Hydrobiologia 438: Tilzer, M. M., 1984. A quantum yield as a fundamental 43–63. parameter controlling vetical photosynthetic profiles Salmaso, N., 2002. Ecological patterns of phytoplankton of phytoplankton in Lake Constance. Archiv für assemblages in Lake Garda: seasonal, spatial and Hydrobiologie (Supplement) 69: 169–198. historical features. Journal of Limnology 61: 95–115. Usseglio-Polatera, P., M. Bournaud, P. Richoux & H. Salmaso, N., 2003. Life strategies, dominance patterns and Tachet, 2000. Biological and ecological traits of mechanisms promoting species coexistence in phyto- benthic freshwater macroinvertebrates: relationships plankton communities along complex environmental and definition of groups with similar traits. Freshwater gradients. Hydrobiologia 502: 13–36. Biology 43: 175–205. Salmaso, N., 2005. Effects of climatic fluctuations and Weithoff, G., 2003. The concepts of ‘‘plant functional vertical mixing on the interannual trophic variability types’’ and ‘‘functional diversity’’ in lake phytoplank- of Lake Garda, Italy. Limnology and Oceanography ton—a new understanding of phytoplankton ecology? 50: 553–565. Freshwater Biology 48: 1669–1675. Shipley, B., P. A. Keddy, D. R. J. Moore & K. Lemky, Wetzel, R. G., 2001, Limnology. Lake and River Ecosys- 1989. Regeneration and establishment strategies of tems, 3rd edn. Academic Press, San Diego, California. emergent macrophytes. Journal of Ecology 77: 1093– Whitton, B. A. & M. Potts (eds), (2000). The Ecology of 1110. Cyanobacteria, their Diversity in Time and Space. Solbrig, O. T., 1993. Plant traits and adaptive strategies: Kluwer Academic Publishers, Dordrecht, The Neth- their role in ecosystem function. In Schulze, E. D. & erlands. 123
You can also read