Biologia, Bratislava, 61/Suppl. 20: S467—S476, 2006 Section Zoology DOI: 10.2478/s11756-007-0063-y
Structure of chironomid assemblages along environmental and geographical gradients in the Bohemian Forest lakes (Central Europe): An exploratory analysis Peter Bitušík1,2 & Marek Svitok3 1 Faculty of Sciences and 2 Science and Research Institute, Matej Bel University, Tajovského 40, SK–97401 Banská Bystrica, Slovakia; e-mail:
[email protected] 3 Department of Biology and General Ecology, Faculty of Ecology and Environmental Sciences, Technical University in Zvolen, T. G. Masaryka 24, SK–96053 Zvolen, Slovakia
Abstract: Eight glacial lakes of the Bohemian Forest (Czech Republic and Germany) were characterised by the distribution of chironomids collected as pupal exuviae. Twenty-eight taxa were identified, including some faunistically interesting species of the region. Two-way indicator species analysis (TWINSPAN) was used to classify lakes according to their taxonomic composition. Canonical correspondence analysis (CCA) and multiple regression were used to relate the chironomid assemblages to two sets of explanatory variables: (i) local environmental variables, and (ii) broad-scale spatial variables. The TWINSPAN classified the lakes into four groups, whereas presence/absence of three taxa was indicative for this classification. The CCA of assemblage composition on environmental variables showed that chironomids respond significantly to altitude and alkalinity. The ordination of composition data on geographical variables revealed strong longitudinal gradient in chironomid distributions. Altitude and alkalinity accounted for 36.2% of the total variation, while the geographic gradient explained 20.5%. As revealed by the variation partitioning procedure, the significant effect of these variables was, in large part, independent of each other. Overall taxonomic richness appeared to be governed by altitude only. Causal ecological and historical factors underlying these results are discussed. This paper may provide a basis for hypothesis testing in future research of the Bohemian Forest lakes. Key words: Chironomidae, pupal exuviae, mountain lakes, classification, variation partitioning, spatial analysis, the Bohemian Forest.
Introduction For more than 130 years, limnological survey has been performed on eight small glacial lakes situated along the Czech-Bavarian border in the Bohemian Forest (Šumava, B¨ohmerwald) (Vrba et al., 2003), and this is particularly so for five of them that are the only natural lakes in the Czech Republic. The research has been intensified during the last three decades as a result of changes in lake water chemistry and the reduction of biodiversity due to atmospheric acidification (Veselý, 1994). Recently, the reversal of the lake water chemistry from acidity and biological recovery are a focus of interest in limnological research of the Bohemian Forest lakes (Vrba et al., 2003). Compared with zooplankton, benthic macroinvertebrate fauna still remains poorly understood (see Vrba et al., 2000 for review), with the exception of available long-term data on mayflies and stoneflies (Soldán et al., 1998, 1999). In the amount of literature qualitative data on the other groups of benthic macroinc 2006 Institute of Zoology, Slovak Academy of Sciences
vertebrates are scattered (Papáček & Soldán, 1995; Soldán et al., 1996). Chironomid fauna of the Bohemian Forest lakes has remained virtually unknown until the recent palaeolimnological analyses of chironomid remains (Bitušík & Kubovčík, 2000), and identification of some pupal exuviae material collected from the lakes (Bitušík, 2006). In most of the freshwater benthic communities, larval chironomids predominate with respect to number of species, and very often, to abundance, as well. Species composition and community structure closely reflect the conditions of various freshwater habitats, and they have traditionally been used as key benthic organisms for studying lake metabolism and assessing water quality (e.g., Seminara et al., 1990). Chironomids are useful environmental indicators because they are sensitive to changes in oxygen concentration, food quality and quantity, and other factors linked to lake trophic status (Sæther, 1979; Heinis & Davids, 1993; Brodersen et al.,2004). Due to the severe atmospheric depo-
S468
P. Bitušík & M. Svitok
Fig. 1. The map of location of the Bohemian Forest lakes (their codes used in Table 1 and Figures are in parentheses). Abbreviations of lake names: PL – Plešné jezero (Plešné Lake), CT – Čertovo jezero (Čertovo Lake), CN – Černé jezero (Černé Lake), PR – Prášilské jezero (Prášilské Lake), LA – Laka (Laka Lake), GA – Grosser Arbersee, KA – Kleiner Arbersee, and RA – Rachelsee.
sition in catchments of the Bohemian Forest lakes (cf. Kopáček et al., 2001), it is reasonable to expect that low pH can affect chironomid assemblages. In conditions of increasing acidity, the changes in species composition and chironomid assemblage structure can be observed (Lindegaard, 1995). However, interpretation of the acidification effect may be confused by the change in the trophic level as a typical feature of acidified lakes, and the lake trophic level may be a more decisive factor structuring chironomid assemblages than lake pH itself (Brodin, 1990). Other interpretation constraints may arise from the nature of ecological data. Most ecological phenomena, investigated by sampling geographic space, are structured by forces that have spatial components. In general, the species distributions result from the combined action of several forces; some of them are external (environmental factors), and others are intrinsic to the community, such as population dynamics and biotic interactions, respectively (Legendre, 1993; Legendre & Legendre, 1998). Therefore, the spatial structuring of natural communities poses the problem of the relative contribution of different factors whose interaction often results in an overlaid effect in space, and may confound the interpretation of correlative analyses (Borcard et al., 1992; King et al., 2005). The quantification of spatial patterns may provide clues to the identity of important processes (Liebhold & Gurevich, 2002). Since spatial autocorrelation is a very general property of ecological data, it is always important to get a measure of the amount of spatial structuring in species data (Borcard et al., 1992; Borcard & Legendre, 1994, 2002; Lichstein et al., 2002). Moreover, pilot studies, such as this one, should provide information about the nature of the underlying spatial structure of the variables as a basis for subsequent studies (Legendre et al., 2002).
The objectives of this exploratory study were: (i) to report the first comprehensive chironomid list of the Bohemian Forest lakes, (ii) to classify lakes on the basis of the recent chironomid assemblages, (iii) to find the environmental and geographical determinants structuring those assemblages. Study area and sampling sites The investigated lakes are situated in the Bohemian Forest in the Czech Republic (5 lakes) and the Bavarian Forest (Bayerischer Wald) in Germany (3 lakes) at altitudes between 918 and 1087 m a.s.l. (Fig. 1). All 8 lakes are reported as the Bohemian Forest lakes further in the text. Their catchments are relatively small (65–279 ha), with bedrock composed mainly of gneiss, mica-schist and/or granite, and covered predominantly by Norway spruce (Picea abies (L.) Karst.), and more sparsely by beech (Fagus sylvatica L.) and fir (Abies alba Miller). The lake areas range from 2.6 to 18.8 ha, the maximum depths from 3 to 40 m, and volumes from 0.05 to 2.92 106 m3 (Tab. 1). More details on the observed lakes are given by VESELÝ (1994) and VRBA et al. (2000).
Methods Field and laboratory The lakes were surveyed once in early September, 2003. The sampling time agreed with the date of water chemistry survey and collection of other biological material. Pupal exuviae were collected along the lake shore from sites where floating debris had accumulated. A fine-meshed (200 µm) hand net attached to an extension pole was used. Sampled material was placed into polythene bottles, labelled, and preserved with 4% formalin. In the laboratory, all of the pupal exuviae in the sample were removed using a low-powered stereomicroscope (× 7–40). The exuviae were mounted in groups on slides using Berlese fluid, and identified under
Chironomids from the Bohemian Forest lakes
S469
Table 1. Location, morphological and physicochemical characteristics of the Bohemian Forest lakes. Lake Characteristic
Location Latitude Longitude Morphological Altitude Lake area Average depth Lake volume Physicochemical Alkalinity Conductivity (25 ◦C) Dissolved O2 Chlorophyll-a pH Total P Transparency
CN
CT
GA
KA
LA
PL
PR
RA
49◦ 11 13◦ 11
49◦ 10 13◦ 12
49◦ 06 13◦ 07
49◦ 08 13◦ 09
49◦ 07 13◦ 20
48◦ 47 13◦ 52
49◦ 05 13◦ 24
48◦ 58 13◦ 24
m a.s.l. ha m 106 m3
1008 18.8 16 2.92
1027 10.7 17 1.86
935 7.7 6 0.45
918 9.4 3 0.25
1085 2.6 2 0.05
1087 7.6 8 0.61
1079 4.2 8 0.35
1071 5.7 3 0.18
mmol L−1 µS cm−1 mg L−1 µg L−1
–7 20.3 3.45 0.83 5.01 1.6 9.8
–17 21.6 0.37 2.83 4.68 3.1 5.3
34 14.1 1.06 5.18 6.11 5.6 4
23 13.3 1.85 17.88 5.81 10.7 1.7
58 15.6 5.64 6.3 6.22 5.1 2.5
0 14.1 0.49 14.28 5.26 8.5 1
5 11.3 0.42 4.23 5.37 7 4.5
–6 21.5 3.31 0.64 5.14 2.9 8.5
N E
µg L−1 m
Explanations: Values of physicochemical characteristics with the exception of dissolved O2 were based on surface samples collected in September 2003. Amount of dissolved oxygen was measured near bottom during the same sampling dates. Abbreviations of lake names are given in Fig. 1.
high magnification (× 400) to species-level, if possible. Identification was greatly aided by the key by LANGTON (1991) and LANGTON & VISSER (2003). Data analyses The data were summarised into three matrices: faunistic data matrix, local environmental variables matrix, and matrix of broad-scale spatial variables. The faunistic data matrix was comprised of the presence/absence of chironomid taxa across eight lakes. The local environmental data matrix consisted of 11 variables, hypothesised to determine the variation of chironomids (cf. BRODERSEN & LINDEGAARD, 1999; RUSE, 2002a; NYMAN & KORHOLA, 2005; BITUŠÍK et al., 2006), that were measured for each lake (Tab. 1). To eliminate redundant, strongly correlated variables, which may cause multicollinearity problems (QUINN & KEOUGH, 2002), the environmental data were screened prior to analysis. Collinear variables were deleted sequentially until all of the remaining variables had product moment correlation coefficients less than 0.75. In total, five local environmental variables were deleted: volume (positively correlated with depth and area), pH (positively correlated with alkalinity), water transparency, conductivity, chlorophyll-a (all highly correlated with total phosphorus and with each other). The broad-scale spatial matrix was comprised of the geographical coordinates of sampling sites and their product monomials in the form of second order trend surface equation (see LEGENDRE, 1990). The addition of higher order and interaction terms allowed for more complex spatial features to be analysed. Geographical coordinates were centred to their respective means to avoid the presence of collinearity in the matrix (LEGENDRE & LEGENDRE, 1998). Two-way indicator species analysis (TWINSPAN; HILL, 1979) was used to classify lakes according to their taxonomic composition, i.e., to delimit assemblage types. The data were analysed using the computer program WinTWINS 2.3 (HILL & ŠMILAUER, 2005) with downweighting of rare taxa (see below) and strict convergence criteria for classification (OKSANEN & MINCHIN, 1997). This method has been criticised for failing to identify secondary gradients
besides one strong dominant gradient, and for rather arbitrary cutting of the major gradient axis, which may lead to the separation of sites with relatively similar species composition (BELBIN & MCDONALD, 1993). Therefore, an alternative clustering method based on the UPGMA algorithm with S¨ orensen’s similarity coefficient was employed to ensure the robustness of TWINSPAN results. The final division level was chosen to maximise ecological interpretability of groupings in terms of major environmental gradients and taxonomic similarity. The method was applied with the aim to use it in the future survey of the lakes, and to make it cost effective. The distributions of chironomid taxa along an environmental gradient appeared to be non-linear as judged by the relatively long gradient (gradient length > 3 SD units), extracted by detrended correspondence analysis with detrending by segments (TER BRAAK & ŠMILAUER, 2002). Thus, canonical correspondence analysis (CCA; TER BRAAK, 1986), with the Monte Carlo permutation test (999 permutations) was used to identify which variables could directly account for the variation observed in the chironomid data matrix. The results of the correspondence analysis and its derivatives CCA and TWINSPAN are sensitive to the presence of rare species (TER BRAAK, 1995). For that reason, an unduly large influence of rare taxa was reduced by downweighting the rare species according to an algorithm available in CANOCO software (TER BRAAK & ŠMILAUER, 2002). The same species weights as in CCA were applied in TWINSPAN. Stepwise CCA with forward selection procedure was used to select the parsimonious model with the combination of environmental variables that best explained the variation in the faunistic matrix. Only variables with the conditional effect significant at P < 0.1 were entered into the model. The relatively liberal significance level was adopted to prevent the elimination of potentially important variables (BOWERMAN & O’CONNELL, 1990). Forward selection procedure is also useful in preventing the artificial increase of the explained variation by mere chance (BORCARD et al., 1992). Intra-set correlations were examined to
S470
P. Bitušík & M. Svitok
Fig. 2. Two way ordered table produced by TWINSPAN based on absence/presence of chironomids in eight Bohemian Forest lakes. Dendrograms demonstrated successive divisions of lakes and species, respectively. Abbreviations of lake names are given in Fig. 1.
estimate the relative contributions of the individual environmental variables to the ordination axes (TER BRAAK, 1995). In addition, a variation partitioning procedure consisting of a series of CCAs was applied to distinguish the importance of environmental conditions and broad-scale spatial structure, respectively, to the chironomid assemblages (BORCARD et al., 1992). Total variation in the chironomid data matrix was then partitioned into four components: (a) a non-spatially structured component explained by the environmental variables, (b) a component explained by spatially structured environmental variables, (c) a component of spatially structured variation which is not explained by the environmental variables in the model, and (d) unexplained variation (BORCARD & LEGENDRE, 1994). Prior to variation partitioning, both the environmental matrix and spatial matrix were submitted to forward selection procedure as mentioned above. An analogous approach via stepwise multiple regressions (STATSOFT, 2001) was used to provide an insight into processes that affect chironomid taxonomic richness (LEGENDRE & LEGENDRE, 1998). Since sample sizes varied greatly among the lakes, rarefaction analysis (HURLBERT, 1971; SIMBERLOFF, 1978) was used to adjust species richness for confounding the effect of different sample size. Monte Carlo randomisation procedure (1000 randomisations) was used to estimate the expected taxonomic richness for a given number of individuals drawn randomly from each sample (GOTELLI & ENTSMINGER, 2006). Samples were rarefied down to the smallest sample size (n = 29). The average expected number of taxa in the rarefied samples was then used in the regression analysis. All variables con-
formed to a normal distribution, where Shapiro-Wilks tests were all non-significant at P < 0.05.
Results A total of 974 pupal exuviae from eight samples were identified to 28 taxa (Fig. 2). One species only (Tanytarsus buchonius) was found in all the eight lakes. In only one lake twelve taxa were recorded, and on average, there were nine taxa per lake. The chironomid fauna was composed largely of lentic taxa. The TWINSPAN procedure classified the Bohemian Forest lakes into four groups based on their chironomid assemblages (Fig. 2). The classification generated by UPGMA revealed a very similar pattern, thus the results of TWINSPAN will be regarded as stable. The first division of TWINSPAN classification separated Plešné Lake and Prášilské Lake from the other lakes. Plešné and Prášilské lakes were characterised by the indicator species Zavrelimyia melanura/barbatipes and preferred by Psectrocladius sordidellus and P. oligosetus. Further division separated Laka Lake from the remaining five lakes, which were indicated by the presence of Procladius (H.) sp. and preferred by Ablabesmyia monilis, Heterotrissocladius grimshawi, and other species. The last dichotomy differentiated species-poor Rachelsee from the group of four lakes: Černé Lake, Čertovo Lake, Grosser Arbersee, and Kleiner Arbersee. Heterotrissocladius marcidus was
Chironomids from the Bohemian Forest lakes
S471
Fig. 3. CCA ordination biplot of chironomid presence/absence with respect to altitude and alkalinity (A); ordination triplot showing significant relationship among local and spatial variables and chironomid assemblages of Bohemian Forest lakes (B). Arrows represent vectors of environmental and spatial variables, respectively. Triangles and circles represent species and site centroids. Only species with the highest fit (> 30%) to the model are displayed. Scaling focused on inter-species distances. Abbreviations of lake names are given in Fig. 1. For abbreviations of taxa names see Fig. 2.
present in all the lakes in this group. Those lakes were preferred by Psectrocladius bisetus, Phaenopsectra flavipes and Corynoneura scutellata and other taxa. In CCA, altitude was the first variable among the local environmental variables entering the model, followed by alkalinity. Chironomid assemblages responded strongly to altitudinal gradient where altitude accounted for 21.9% in species variation (P < 0.05). When alkalinity was included, local environmental variables explained 41.5% of the variation in the species matrix (P < 0.01). The species-environmental relationship is shown in Fig. 3A. The first canonical axis was significant (P < 0.01), and reflected altitudinal variation (intra-set correlation, r = 0.81). The second axis was correlated strongly with alkalinity (r = 0.98), however was not significant (P = 0.17). Ordination showed that Z. melanura/ barbatipes and P. sordidellus occurred at higher altitudes in water with a somewhat lower alkalinity. In contrast, Psectrocladius psilopterus, Tanytarsus cf. smolandicus, Phenopsectra Pe f. Bala, and other species situated in the upper left quadrant of the ordination space preferred lower situated lakes with higher alkalinity. CCA showed a strong spatial structure in the Chironomidae data set. Out of the five geographical variables considered in the analysis, longitude was retained by the forward selection procedure as the significant predictor of the faunistic variation. It accounted significantly for 25.8% of the variation in the chironomid data set (P < 0.001). The overall model of chironomid response to significant spatial and environmental variables is shown in Fig. 3B. Displayed canonical axes were statistically significant (P < 0.001 and P < 0.05, respectively) and explained 26.1% and
22.1% of the total chironomid variation, respectively. The first axis reflected mainly a spatial change and was highly correlated with longitude (r = 0.97), and somewhat weakly correlated with altitude (r = 0.80). The second axis represented environmental change attributable to acidification and was correlated with alkalinity (r = 0.64). Plešné Lake was located somewhat asymmetrically and appeared to be an outlier as detected by leverage diagnostics in the CCA. When the outlier was omitted, the overall model remained significant. The results of variation partitioning showed that 36.2% of Chironomidae data variation can be significantly explained by the local effect of environmental variables (P < 0.05), after the effect of longitude has been removed (Fig. 4A). A large amount of variation (20.5%) was significantly accounted for by lake position along geographical gradient (P < 0.05), and cannot be related to the measured environmental variables. A small portion of shared variation (5.3%) showed that the effect of environmental variables was relatively independent for large-scale spatial gradients. When the effects of altitude and alkalinity were assessed separately (Fig. 4B), altitude uniquely accounted for a significant fraction of variation (19.0%, P < 0.05) whereas alkalinity did not (19.9%, P = 0.051). The locally acting environmental variables jointly explained the negative portion of variation (–2.7%), indicating that they explained chironomid variation better together, than the sum of their individual effects and the effect of altitude probably counteracts the effect of alkalinity (see Legendre & Legendre, 1998; Yamasaki et al., 1999 for details).
S472
P. Bitušík & M. Svitok
Fig. 4. Schematic representation of variation partitioning of chironomid presence/absence data matrix among environmental and spatial variables (A); results of refined variation partitioning in which altitude and alkalinity are analysed separately (B).
The observed taxonomic richness in the Bohemian lakes ranged from 5 to 15 with the highest richness in Grosser Arbersee Lake and the lowest in Laka Lake and Rachelsee Lake, respectively (Fig. 5). In the stepwise multiple regressions on local environmental variables, altitude was the single best predictor, explaining 69.4% of the variation in rarefied Chironomidae taxonomic richness (F1,6 = 13.58, P < 0.05). The richness decreased significantly as the altitude increased (Fig. 6). In the analysis of broad-scale spatial variables, none of the variables were included in the final model. Discussion
Fig. 5. Individual-based rarefaction curves for chironomids in Bohemian Forest lakes. Abbreviations of lake names are given in Fig. 1.
Fig. 6. Response of expected chironomid taxonomic richness (rarefied to 29 individuals) to altitudinal gradient in Bohemian Forest lakes. Abbreviations of lake names are given in Fig. 1.
The analysis of the pupal exuviae material collected in September 2003 reveals chironomid assemblages with a rather high proportion of Orthocladiinae members related to Chironominae. Besides the taxa/species generally distributed in lakes and ponds (e.g., Ablabesmyia spp., Prodiamesa olivacea, Psectrocladius sordidellus, P. psilopterus, Endochironomus albipennis, Phaenopsectra flavipes, Tanytarsus signatus), some taxa are considered to be cold stenothermic (e.g., Zavrelimyia spp., C. cf. magus, Heterotrissocladius spp., Psectrocladius barbatipes, P. bisetus, P. oligosetus, Phaenopsectra Pe f. Bala, T. cf. smolandicus), and some prefer dystrophic conditions (e.g., Macropelopia adaucta, P. bisetus, P. oligosetus, G. paripes, T. buchonius) (Fittkau, 1962; Dowling & Murray, 1981; Wiederholm, 1983, 1986; Langton & Visser, 2003; Ekrem, 2004). Five species were recorded from the Czech Republic for the first time (Bitušík, 2006) and finding of other faunistically interesting species is highly probable. Moreover, it could be supposed that the Bohemian Forest lakes provide the only habitats for some species (e.g., H. grimshawi) in the Czech Republic. Chironomid pupal exuviae technique (CPET), as a method for river monitoring and assessment (Wilson & Bright, 1973), is applicable to nearly all types of surface freshwaters (Wilson & Ruse, 2005), includ-
Chironomids from the Bohemian Forest lakes ing lakes (e.g., Ruse, 2002a, b; Bitušík et al., 2006). Moreover, the pupal exuviae technique provides a rapid system for qualitative data collecting due to easier and more reliable identification comparing with larvae analyses. Pupal exuviae samples contain a mixture of species arising from different lake habitats. Most chironomid species live in the littoral zone which provides much more heterogeneous conditions than the relatively stable and uniform profundal (Brodersen & Lindegaard, 1999). Without data on the distribution of living larvae in the lake, the interpretation of pupal exuviae data must be made carefully, using all autecological information. All the lakes with exception of Laka Lake used to be thermally stratified with a clinograde distribution of dissolved oxygen and a more or less pronounced bottom oxygen deficit (Vrba et al., 2000). Both the TWINSPAN and CCA analyses distinctly separated the shallow, well-oxygenated Laka Lake from the other lakes. Chironomid assemblage of the lake consists of a low number of taxa. However, the absence of Procladius (H.) sp. as the indicative characteristic of this lake is probably spurious. Features of Procladius larvae biology contribute to their success in inhabiting a wide range of environmental conditions (Brown & Oldham, 1984), and it is no reason for the absence of them in this site. The exceptionality of the lake might be rather underlined by the presence of P. barbatipes that was found to be one of the indicative species for subalpine lakes in the Tatra Mountains (Bitušík et al., 2006). The TWINSPAN group comprising of Plešné Lake and Prášilské Lake is separated due to presence of Z. melanura/barbatipes. Both the lakes are the most productive systems among the investigated lakes (Nedbalová et al., 2006). Chironomid assemblage from Plešné Lake is remarkable because of the presence of three species of the genus Zavrelimyia, Parakiefferiella sp. and Chironomus sp. While most of the recorded species are indicative of cold, well oxygenated conditions in the littoral, Chironomus larvae could be dwellers of soft sediments under lower oxygen concentrations. CCA supports the exceptionality of Plešné Lake by illustrating the position of the lake apart from the other lakes. Lastly, the assemblage composition of the TWINSPAN group of five lakes with the presence of H. marcidus and H. grimshawi infers oligotrophic conditions (Sæther, 1979). This is especially true for Černé Lake, čertovo Lake and Rachelsee, which are considered to be the most oligotrophic (Vrba et al., 2000; Nedbalová et al., 2006). The overall model explaining the structure of chironomid assemblages, which included altitude, alkalinity and longitude, accounts for 62.0% of the total variation (Fig. 4A). The relatively high amount of explained variation demonstrates that important factors were included in the model.
S473 Altitudinal gradient is closely connected with a major climatic gradient in the mountains and many studies revealed significant relationship between faunistic changes in aquatic organisms of mountain lakes, and altitude-temperature gradient (e.g., Lotter et al., 1997; Larocque et al., 2001; Bitušík et al., 2006; Krno et al., 2006). Temperature has a strong influence, both directly (growth, metamorphosis, behaviour) and indirectly (dissolved oxygen, food sources, etc.), on chironomid distributions. In this case, altitude, not surprisingly, shows a strong correlation with both the chironomid faunistic composition and the taxonomic richness. Moreover, the lakes located at lower altitude, on the south slopes of the mountain range (Kleiner Arbersee and Grosser Arbersee), are inhabited by the most diverse chironomid fauna, whereas the uppermost situated lakes (Laka Lake, Rachelsee, Prášilské Lake), are the species-poorest. A relatively high number of taxa sampled from Plešné Lake has an apparent connection with the mesotrophic status of the lake. Furthermore alkalinity shows a conditional effect in conjunction with altitude. The variability of chironomid data is explained better by both of the variables than summation of their individual effects (Fig. 4B). The Bohemian Forest was exposed to strong influences of atmospheric acidification during the last century (Kopáček et al., 2002). The lake ecosystems responded to the acidification stress differently according to their catchments characteristics and buffer capacity. This means that Nedbalová et al. (2006) recognised two categories of lakes following their acidification status in 2003. Chironomid data are correlated with alkalinity as a surrogate of pH, but it should be stressed that it is difficult to discern the effects of anthropogenic acidification in these lakes with humic and naturally acid water. Palaeolimnological analyses of the sediment cores from Černé Lake and Prášilské Lake (Bitušík & Kubovčík, 2000) revealed the changes in chironomid thanatocoenoses during the acidification period. The acidification process was connected with a decline in sub-fossil remains and with the disappearance of certain taxa (Microtendipes pedellus group, Polypedilum spp., Pagastiella orophila, Paratanytarsus spp.) in the most recent sediment layers. It is worth mentioning that these taxa were absent in the pupal euxviae collections, as well. Because some of them can be found under acid conditions (e.g., Microtendipes, Pagastiella, Wiederholm & Eriksson, 1977; Schnell & Willassen, 1996), it is reasonable to suppose the negative influence of some other factors connected with low pH (e.g., high concentrations of total reactive Al, Fott et al., 1994). In general, chironomids do not reflect the effects of the acidification stress in the investigated lakes in such a dramatic way as Ephemeroptera and Plecoptera (Vrba et al., 2003). Some studies (e.g., Meriläinen & Hynynen, 1990; Walker et al., 1991; Olander et al., 1997) did not find a significant relationship between
S474 chironomid fauna and lake water pH, especially when pH gradient was small. The ordination analysis showed the great importance of geographical gradient (from west to east) in structuring the assemblages (Fig. 3B). Spatialstructuring processes explain almost one quarter of the variation in chironomid data (Fig. 4). Spatial patterns of response variables may originate in two different ways (Legendre et al., 2002); (i) spatial dependence, i.e. the response variable is spatially structured because it depends upon explanatory variables that are themselves spatially structured by their own generating processes, (ii) spatial autocorrelation, i.e. the value of the response variable at one site depends on the values of the response variable at neighbouring sites. Thus, the spatial matrix may act as a synthetic descriptor of unmeasured underlying processes such as external causes or biotic factors (Borcard et al., 1992; Borcard & Legendre, 1994). However, real-case field observations often result from a combination of spatial dependence and autocorrelation (Legendre et al., 2002). In our case, some evidence should allow for explanations of the geographical pattern. The longitudinal gradient in geology of the catchments seems to play a crucial role. The easternmost Plešné Lake is situated on granitic bedrock, whilst the more westward situated Prášilské Lake is partly on granit and mica schists. Mica schist or gneiss dominates in the other lakes (see Tab. 1 in Vrba et al., 2000). The difference in geology results in a different soil composition, and consequently leads to high terrestrial P input into the lake (see Kopáček et al., 2006 for details). Plešné Lake differs from the others because it has the highest level of productivity, heterotrophic microbial and plankton biomasses. Furthermore there are differences in some aspects of chemistry, and structure of phytoplankton (Nedbalová et al., 2006). However, the ordination model was not changed even when Plešné Lake was removed from the analysis. The second reason why chironomid fauna is significantly related to geographical gradient should be the distances between the lakes. Chironomid assemblages of Kleine Arbersee, Grosser Arbersee, čertovo and Černé lakes consist of some common species (Fig. 2) despite their different chemistry, and other characteristics (cf. Nedbalová et al., 2006). While the distance between čertovo and Černé lakes is about 2 km, and Kleine Arbersee is situated ca 4 km to Grosser Arbersee, Plešné Lake is more than 60 km away from this lake group. Although flight ability of chironomid adults is limited, the wind can be the cause of passive longdistance dispersal both after eclosion, during swarming, and female oviposition flight (Armitage, 1995). Chironomids achieve dispersal using a variety of mechanisms, and some of them are very particular (see, e.g., Hoffrichter, 1973; Block et al., 1984). Green & Sánchez (2006) recently presented the first evidence that passive internal transport within water birds may be relevant for chironomids. Distance of water bodies
P. Bitušík & M. Svitok is the main factor explaining the spatial distribution of adult chironomids (Delettre & Morvan, 2000). It should be supposed that close vicinity of the lakes enables easier inter-watershed dispersal of species. Finally, we may hypothesise that chironomid assemblage composition could reflect the historical human influence on the lakes which displayed the west-east gradient. Whereas the human impact upon the westernmost (Bavarian) lakes started in the 8th century and has increased manifold since the Middle Ages, and various human activities in the basins of some Czech lakes have been documented since the 16th century. Easternmost Plešné Lake and its forested catchment remained less affected until the 20th century (Veselý, 1994). This effect is not presumably correlated with an explicit environmental variable, and consequently might be displayed as a part of spatial variation. We are conscious that the “snap-shot” survey cannot collect the total chironomid species richness in the investigated lakes; thus, some relationships between chironomid data and environmental variables have not been entirely explained. Despite this, we believe that the chironomid-based classification presented in this study can be instrumental in determining the ecological conditions in the Bohemian Forest lakes and a basis for further long-term research.
Acknowledgements We are very grateful to J. KOPÁČEK for collecting pupal exuviae material and together with J. VRBA providing environmental data from the lakes. Thanks are also due to both and J. MATĚNA for useful suggestions to the first version of the paper. We also want to thank J. BLESSING and M. BITUŠÍK for linguistic correction of the manuscript.
References ARMITAGE, P.D. 1995. Behaviour and ecology of adults, pp. 194–224. In: ARMITAGE, P.D., CRANSTON, P.S. & PINDER, L.C.V. (eds) The Chironomidae: Biology and ecology of nonbiting midges, Chapman & Hall, London. BELBIN, L. & MCDONALD, C. 1993. Comparing three classification strategies for use in ecology. J. Veg. Sci. 4: 341–348. BITUŠÍK, P. 2006. Faunistic records: Chironomidae. Dipterologica Bohemoslovaca 13, Acta Univ. Carol., Biol. (in press). BITUŠÍK, P. & KUBOVČÍK, V. 2000. Sub-fossil chironomid assemblages (Diptera: Chironomidae) from the Černé lake and Prášilské lake (Bohemian Forest, Czech Republic). Silva Gabreta 4: 253–258. BITUŠÍK, P., SVITOK, M., KOLOŠTA, P. & HUBKOVÁ, M. 2006. Classification of the Tatra Mountains lakes (Slovakia) using chironomids (Diptera, Chironomidae). Biologia, Bratislava 61, Suppl. 18: S191–S201. BLOCK, W., BURN, A.J. & RICHARD, K.J. 1984. An insect introduction to the maritime Antarctic. Biol. J. Linn. Soc. 23: 33–39. BORCARD, D. & LEGENDRE, P. 1994. Environmental control and spatial structure in ecological communities: an example using oribatid mites (Acari, Oribatei). Environ. Ecol. Stat. 1: 37– 61.
Chironomids from the Bohemian Forest lakes BORCARD, D. & LEGENDRE, P. 2002. All-scale spatial analysis of ecological data by means of principal coordinates of neighbour matrices. Ecol. Model. 153: 51–68. BORCARD, D.P., LEGENDRE, P. & DRAPEAU, P. 1992. Partialling out the spatial component of ecological variation. Ecology 73: 1045–1055. BOWERMAN, B.L. & O’CONNELL, R. 1990. Linear statistical models: An applied approach, 2nd edition, Duxbury Press, Belmont, California, 1040 pp. BRODERSEN, K.P. & LINDEGAARD, C. 1999. Classification, assessment and trophic reconstruction of Danish lakes using chironomids. Freshwater Biol. 42: 143–157. BRODERSEN, K.P., PEDERSEN, O., LINDEGAARD, C. & HAMBURGER, K. 2004. Chironomids (Diptera) and oxy-regulatory capacity: An experimental approach to paleolimnological interpretation. Limnol. Oceanogr. 49: 1549–1559. BRODIN, Y.-W. 1990. Midge fauna development in acidified lakes in northern Europe. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 327: 295–298. BROWN, A.E. & OLDHAM, R.S. 1984. Chironomidae (Diptera) of Rutland Water. Arch. Hydrobiol. Suppl. 69: 199–227. DELETTRE, Y.R. & MORVAN, N. 2000. Dispersal of adult aquatic Chironomidae (Diptera) in agricultural landscapes. Freshwater Biol. 44: 399–411. DOWLING, C. & MURRAY, D.A. 1981. The distribution of the Chironomidae (Diptera) in two Irish blanket bogs. Proc. R. Irish Acad. 81: 53–61. EKREM, T. 2004. Immature stages of Europaean Tanytarsus species I. The eminulus-, gregarius-, lugens- and mendax species groups (Diptera, Chironomidae). Mitt. Mus. Natkd. Berl., Dtsch. Entomol. Z. 51: 97–146. FITTKAU, E.J. 1962. Die Tanypodinae (Diptera, Chironomidae). Abh. Larvalsyst. Insekten 6: 1–453. FOTT, J., PRAŽÁKOVÁ, M., STUCHLÍK, E. & STUCHLÍKOVÁ, Z. 1994. Acidification of lakes in Šumava (Bohemia) and in the High Tatra Mountains (Slovakia). Hydrobiologia 274: 37–47. GOTELLI, N.J. & ENTSMINGER, G.L. 2006. EcoSim: Null models software for ecology. Version 7. Acquired Intelligence Inc. & Kesey – Bear. Jericho, VT 05465. http://garyentsminger.com/ ecosim.htm. 21.06.2006. GREEN, A.J. & SÁNCHEZ, M.I. 2006. Passive internal dispersal of insect larvae by migratory birds. Biology Letters 2: 55–57. HEINIS, F. & DAVIDS, C. 1993. Factors governing the spatial and temporal distribution of chironomid larvae in the Maarseveen Lakes with special emphasis on the role of oxygen conditions. Neth. J. Aquat. Ecol. 27: 21–34. HILL, M.O. 1979. TWINSPAN – a FORTRAN program for arranging multivariate data in an ordered two-way table by classification of the individuals and attributes. Cornell University, Ithaca, New York. HILL, M.O. & ŠMILAUER, P. 2005. TWINSPAN for Windows version 2.3. Centre for Ecology and Hydrology & University of South Bohemia, Huntingdon & České Budejovice, 29 pp. HOFFRICHTER, O. 1973. On the role of traffic as possible means of dispersal in chironomids. Chironomus 1: 102–104. HURLBERT, S.H. 1971. The nonconcept of species diversity: a critique and alternative parameters. Ecology 52: 577–586. KING, R.S, BAKER, M.E., WHIGHAM, D.F., WELLER, D.E., JORDAN, T.E., KAZYAK, P.F. & HURD, M.K. 2005. Spatial considerations for linking watershed land cover to ecological indicators in streams. Ecol. Appl. 15: 137–153. KOPÁČEK, J., STUCHLÍK, E., VESELÝ, J., SCHAUMBURG, J., ANDERSON, I.C., FOTT, J., HEJZLAR, J. & VRBA, J. 2002. Hysteresis in reversal of Central European mountain lakes from atmospheric acidification. Water Air Soil Pollut.: Focus 2: 91–114. KOPÁČEK, J., TUREK, J., HEJZLAR, J., KAŇA, J. & PORCAL, P. 2006. Element fluxes in watershed-lake ecosystems recovering from acidification: Plešné Lake, the Bohemian Forest, 2001– 2005. Biologia, Bratislava 61, Suppl. 20: S427–S440. KOPÁČEK, J., VESELÝ, J., & STUCHLÍK, E. 2001. Sulphur and nitrogen fluxes and budgets in the Bohemian Forest and Tatra
S475 Mountains during the industrial revolution (1850–2000). Hydrol. Earth Syst. Sci. 5: 391–405. KRNO, I., ŠPORKA, F., GALAS, J., HAMERLÍK, L., ZAŤOVIČOVÁ, Z. & BITUŠÍK, P. 2006. Litttoral benthic macroinvertebrates of mountain lakes in the Tatra Mountains (Slovakia, Poland). Biologia, Bratislava 61, Suppl. 18: S147–S166. LANGTON, P.H. 1991. A key to pupal exuviae of West Palaearctic Chironomidae. Privately published by P.H. Langton, 3 St. Felix Road, Ramsay Forty Foot, Cambridgeshire, England, PE 17 1YH, 386 pp. LANGTON, P.H. & VISSER, H. 2003. Chironomidae exuviae. A key to pupal exuviae of the West Palaearctic Region. Interactive Identification System for the Europaean Limnofauna (IISEL), World Biodiversity Database, CD-ROM Series. LAROCQUE, I., HALL, R.I. & GRAHN, E. 2001. Chironomids as indicators of climate change: a 100–lake training set from a subarctic region of northern Sweden (Lapland). J. Paleolimnol. 26: 307–322. LEGENDRE, P. 1990. Quantitative methods and biogeographic analysis, pp. 9–34. In: GARBARY, D.J. & SOUTH, R.R. (eds) Evolutionary biogeography of the marine algae of the North Atlantic, NATO ASI Series vol. G 22, Springer-Verlag, Berlin, Germany. LEGENDRE, P. 1993. Spatial autocorrelation: trouble or new paradigm? Ecology 74: 1659–1673. LEGENDRE, P., DALE, M.R.T., FORTIN, M.-J., GUREVITCH, J., HOHN, M. & MYERS, D. 2002. The consequences of spatial structure for the design and analysis of ecological field surveys. Ecography 25: 601–615. LEGENDRE, P. & LEGENDRE, L. 1998. Numerical ecology, 2nd edition. Developments in Environmental Modelling 20. Elsevier Science B.V., Amsterdam, 853 pp. LICHSTEIN, J.W., SIMONS, T.R., SHRINER, S.A. & FRANZREB, K.E. 2002. Spatial autocorrelation and autoregressive models in ecology. Ecol. Monogr. 72: 445–463. LIEBHOLD, A.M. & GUREVICH, J. 2002. Integrating statistical analysis of spatial data in ecology. Ecography 25: 553–557. LINDEGAARD, C. 1995. Classification of water-bodies and pollution, pp. 385–404. In: ARMITAGE, P.D., CRANSTON, P.S. & PINDER, L.C.V. (eds) The Chironomidae: Biology and ecology of non-biting midges, Chapman & Hall, London. LOTTER, A.F., BIRKS, H.J.B., HOFMANN, W. & MARCHETTO, A. 1997. Modern diatom, cladocera, chironomid, and chrysophyte cyst assemblages as quantitative indicators for the reconstruction of past environmental conditions in the Alps. I. Climate. J. Paleolimnol. 18: 395–420. MERILÄINEN, J.J. & HYNYNEN, J. 1990. Benthic invertebrates in relation to acidity in Finnish forest lakes, pp. 1029–1049. In: KAUPPI, P., ANTTILA, P. & KENTTÄMIES, K. (eds) Acidification in Finland, Springer-Verlag, Berlin, Heidelberg. NEDBALOVÁ, L., FOTT, J., KOHOUT, L., KOPÁČEK, J., MACEK, M., SOLDÁN, T. & VRBA, J. 2006. Biological recovery of the Bohemian Forest lakes from acidification. Biologia, Bratislava 61, Suppl. 20: S453–S465. NYMAN, M.T. & KORHOLA, A.A. 2005. Chironomid-based classification of lakes in western Finnish Lapland. Boreal Environ. Res. 10: 239–254. OKSANEN, J. & MINCHIN, P.R. 1997. Instability of ordination results under changes in input data order: explanations and remedies. J. Veg. Sci. 8: 447–454. OLANDER, H., KORHOLA, A. & BLOM, T. 1997. Surface sediment Chironomidae (Insecta: Diptera) distributions along a ecotonal transect in subarctic Fennoscandia: developing a tool for paleotemperature reconstructions. J. Paleolimnol. 18: 211–230. PAPÁČEK, M. & SOLDÁN, T. 1995. Biogeograficky významné druhy vodního hmyzu (Ephemeroptera, Odonata, Plecoptera, Heteroptera: Nepomorpha) v oblasti Šumavy. Klapalekiana 31: 41–51. QUINN, G.P. & KEOUGH, M.J. 2002. Experimental design and data analysis for biologists. Cambridge University Press, Cambridge, UK, 520 pp.
S476 RUSE, L. 2002a. Chironomid pupal exuviae as indicators of lake status. Arch. Hydrobiol. 153: 367–390. RUSE, L. 2002b. Colonisation of gravel lakes by Chironomidae. Arch. Hydrobiol. 153: 391–407. SÆTHER, O.A. 1979. Chironomid communities as water quality indicators. Holarct. Ecol. 2: 65–74. SCHNELL, O.A. & WILLASSEN, E. 1996. The chironomid (Diptera) communities in two sediment cores from Store Hovvatn, S. Norway, an acidified lake. Ann. Limnol. 32: 45–61. SEMINARA, M., BAZZANTI, M. & TAMORRI, C. 1990. Sublittoral and profundal chironomid (Diptera) communities of Lake Vico (Central Italy): relationship to the trophic level. Ann. Limnol. 26: 183–193. SIMBERLOFF, D. 1978. Use of rarefaction and related methods in ecology, pp. 150–165. In: DICKSON, K.L., CAIRNS, J. JR, & LIVINGSTON, R.J. (eds) Biological data in water pollution assessment: quantitative and statistical analysis, ASTM STP 652, American Society for Testing and Materials, Philadelphia, Pennsylvania. SOLDÁN, T., LANDA, V. & ZAHRÁDKOVÁ, S. 1999. Long-term changes of diversity of mayflies (Ephemeroptera) in the Křemelná river basin (Šumava Mts, Czech Republic). Silva Gabreta 3: 95–114. SOLDÁN, T., PAPÁČEK, M., NOVÁK, K. & ZELENÝ, J. 1996. The Šumava Mountains: a unique biocentre of aquatic insects (Ephemeroptera, Odonata, Plecoptera, Megaloptera and Heteroptera – Nepomorpha). Silva Gabreta 1: 179–186. SOLDÁN, T., ZAHRÁDKOVÁ, S., HELEŠIC, J., DUŠEK, V. & LANDA, V. 1998. Distributional and quantitative patterns of Ephemeroptera and Plecoptera in the Czech Republic: a possibility of detection of long-term environmental changes of aquatic biotopes. Folia Fac. Sci. Nat. Univ. Masaryk. Brun., Biol. 98: 1–306. STATSOFT, INC. 2001. STATISTICA (data analysis software system), Version 6. TER BRAAK, C.J.F. 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology 67: 1167–1179. TER BRAAK, C.J.F. 1995. Ordination, pp. 91–173. In: JONGMAN, R.H.G., TER BRAAK, C.J.F. & VAN TONGEREN, O.E.R. (eds) Data analysis in community and landscape ecology, Cambridge University Press, Cambridge, UK.
P. Bitušík & M. Svitok BRAAK, C.J.F. & ŠMILAUER, P. 2002. CANOCO reference manual and CanoDraw for Windows user’s guide: software for canonical community ordination (Version 4.5). Microcomputer Power, Ithaca, New York, USA, 500 pp. VESELÝ, J. 1994. Investigation of the nature of the Šumava lakes: a review. Čas. Nár. Mus. Praha, Řada Přírodověd. 163: 103– 120. VRBA, J., FOTT, J. & KOPÁČEK, J. 2000. Long-term limnological research of the Bohemian Forest lakes and their recent status. Silva Gabreta 4: 7–28. VRBA, J., KOPÁČEK, J., FOTT, J., KOHOUT, L., NEDBALOVÁ, L., PRAŽÁKOVÁ, M., SOLDÁN, T. & SCHAUMBURG, J. 2003. Long-term studies (1871–2000) on acidification and recovery of lakes in the Bohemian Forest (central Europe). Sci. Total Environ. 310: 73–85. WALKER, I.R., SMOL, J.P., ENGSTROM, D.R. & BIRKS, H.J.B. 1991. An assessment of Chironomidae as quantitative indicators of past climatic change. Can. J. Fish. Aquat. Sci. 48: 975–987. WIEDERHOLM, T. (ed.) 1983. Chironomidae of the Holarctic region. Part 1. Larvae. Entomol. Scand., Suppl. 19: 1–457. WIEDERHOLM, T. (ed.) 1986. Chironomidae of the Holarctic region. Part 1. Pupae. Entomol. Scand., Suppl. 28: 1–482. WIEDERHOLM, T. & ERIKSSON, L. 1977. Benthos of an acid lake. Oikos 29: 261–267. WILSON, R.S. & BRIGHT, P.L. 1973. The use of chironomid pupal exuviae for characterizing streams. Freshwater Biol. 3: 283– 302. WILSON, R.S. & RUSE, L.P. 2005. A guide to the identification of genera of chironomid pupal exuviae and their use in monitoring lotic and lentic fresh waters. Freshwater Biol. Assoc., 164 pp. YAMASAKI, S.H., PELLETIER, B., DUTILLEUL, P. & FYLES, J.W. 1999. Sign, magnitude, and elements of ecological interpretation of shared variation among predictor variables in multiple regression: I. The univariate case, pp. 147–162. In: YAMASAKI, S.H. (ed.) Influences of Kalmia angustifolia on black spruce in eastern Canada’s boreal forest, National Library of Canada, Ottawa, Canada. . TER
Received June 22, 2006 Accepted November 15, 2006
