Combined influence of river discharge and wind on littoral nematode communities of a river mouth area of

A. Wittho¨ ft-Mu¨ hlmann Æ W. Traunspurger Æ K. O. Rothhaupt

Abstract The littoral nematode community species diversity and the occurrence of chewers adjacent to a river mouth (River Schussen) on were influenced mainly by river discharge. The Lake Constance () was studied from impact of both these stress factors was modified February 1999 to January 2000 in order to by a third variable, water level. Moderate and determine the influence of stress resulting from high levels of combined habitat stress led to fluctuations in river discharge on local nematode significant changes in community structure. Un- assemblages. Additionally, the influence of wind der conditions of calm weather and low discharge, as a second important stress factor was consid- reduced species diversity and an increased pre- ered. Six sample sites were chosen, reflecting a dominance of deposit feeders were observed. In gradient of river influence within the broader most cases, species diversity was found to be river mouth area. Nematode communities, vary- higher under moderate stress conditions, an ing in a mean range from 121 to 165 ind/10 cm2, observation that offers support for Connel’s were found to differ significantly in terms of Intermediate Disturbance Hypothesis. abundance, feeding type composition and species diversity. Deposit feeders were most abundant at Keywords Habitat stress Á Feeding types Á all sites followed by chewers. Deposit feeders Diversity Á Spatial pattern Á Temporal pattern Á were affected mainly by wind events, while Disturbance

A. Witthoft Muhlmann (&) Institute for Lake Research, Argenweg 50/1, 88085 Introduction Langenargen, Germany e mail: [email protected] Nematodes are one of the most abundant taxa in freshwater benthic communities (Traunspurger W. Traunspurger Animal Ecology, University of Bielefeld, 33615 2000). Four of five multicellular animals are 5 Bielefeld, Germany nematodes with estimated 10 species (Coomans e mail: Traunspurger@uni bielefeld.de 2000). Freshwater nematode densities can reach up to 3.5 · 106 individuals per m2 (Traunspurger A. Witthoft Muhlmann Á K. O. Rothhaupt Institute Limnology, University of Konstanz, 78457 2002). They play a major role in ecosystem Konstanz, Germany processes like grazing on periphyton assemblages K. O. Rothhaupt (bacteria, protests, algae, Montagna 1995). Work- e mail: Karl.Rothhaupt@uni konstanz.de ing in terrestrial environments, Bongers (1990) developed an environmental monitoring system composition and species diversity (Petraitis et al. by classifying nematode communities into colo- 1989; Death and Winterbourn 1995; Taniguchi nizers and persisters. Today this Maturity Index and Tokeshi 2004). In the present study, six (MI) is also used as bioindicator in aquatic sample sites were investigated on a monthly basis benthic assessments (Bongers and Ferris 1999). over a one year-period within a river–lake-eco- Summing up, due to their abundance, species tone, thus covering a broad potential range of richness, ecology, functional morphology (with disturbance stress. We investigated the functional regard to feeding type), and short life cycles they diversity of the nematode community in relation are able to respond quickly to environmental to the pattern of local disturbance and distin- change, making them ideal subject organisms for guished differences in feeding type compositions ecological studies (Platt and Warwick 1980; Giere and species diversity. The following objectives 1993; Hakenkamp et al. 2002). were addressed in detail: (1) Does disturbance Seasonal and spatial distribution patterns of caused by wind and/or river discharge alter nem- nematodes are well described for several lakes atode communities of the river mouth ecotone? (Biro´ 1973; Holopainen and Paasivirta 1977; Prejs (2) Do ‘‘unstable’’ habitat conditions favour 1977b; Bretschko 1984; Strayer 1985; Prejs and colonizers or affect feeding type composition? Prejs 1992; Raspopov et al. 1996; Traunspurger (3) Is species diversity increased under moderate 1996; Bergtold and Traunspurger 2004; Wu et al. habitat stress as predicted by the Intermediate 2004; Michiels and Traunspurger 2005b). How- Disturbance Hypothesis (Connell 1978)? ever, studies dealing with the nematode commu- nities of river mouth areas are rare (Nalepa and Quigley 1983). The area used for the current study Material and methods can be described a lenticular aquatic-terrestrial ecotone, sensu Pieczynska (1990), coupling a Study area water body and adjacent riparian system. Ecoton- al habitats exhibit great heterogeneity with Lake Constance is the second largest lake in respect to abiotic and biotic factors due to the Central Europe (surface area: 571.5 km2; mean import of organic and inorganic components into depth: 91 m, max. depth: 254 m, catchment area: the transition zone and offer opportunities to test 11,487 km2, IGKB 1994) and is located on the important ecological concepts. River mouth areas northern fringe of the Alps (4739¢ N). This study represent zones of transition between rivers and covers the river mouth of the River Schussen and pelagic habitats and function as river–lake inter- the adjacent littoral area. The Schussen (dis- faces. Among external physical factors influencing charge: 15.5 ± 0.15 m3/s (mean ± SE; Hydrologi- the littoral benthic community, wind forces are cal Services Agency Baden Wu¨ rttemberg); considered to be of strong importance due to catchment area: 790 km2, IGKB 1994) enters generation of waves and consequent resuspension the lake on the northern shore and the river of sediment. In river–lake ecotones, the river mouth is characterized by an extended littoral discharge is presumed to play a significant further area (width: 1.2 km, slope: <0.2%). The inflow role in shaping the benthic community. In a pattern of the river depends largely on the wind previous study, we showed that wind and river direction, wind speed, and discharge rate. Since discharge have a combined impact on the meio- the mean wind direction is south-westerly (Ger- faunal community of a river mouth ecotone, man Weather Agency DWD), the inflow however the strength of both components was angle of the Schussen River is toward the south- highly variable (Wittho¨ ft-Mu¨ hlmann et al. 2005a). east. Based on this prevailing pattern, six sample In addition to causing direct mortality and relo- sites were chosen, representing the area directly cation of organisms, disturbance (sensu Connel in contact with river discharge, an intermediate 1978) by these factors can alter habitat stability, zone, and a far distant zone (Fig. 1). Further food supply and inter- and intra-specific competi- details are given in Wittho¨ ft-Mu¨ hlmann et al. tion, and thus ultimately affect community (2005a). Table 1 Definition of habitat stress resulting from wind strength and river discharge, classified for the month pre ceding each sampling day Force Habitat stress class Low Moderate High

Wind speed ‡ 7 m/s £5 events 5 10 events >10 events and duration ‡ 8h River discharge £10 m3/s 10 20 m3/s >20 m3/s

speeds and consequent higher waves that can cause the resuspension of sediments, and/or enhanced discharge. High habitat stress is typical of storms and/or high river discharge is charac- terized by strong sediment erosion and high deposition of river-borne seston. Habitat stress was calculated and classified for the month preceding each sampling day.

Sampling and processing

Sediment cores (surface area: 26.4 cm2) were taken from the study sites using a hand-held Fig. 1 The littoral zone influenced by the Schussen River corer at monthly intervals from February 1999 at Lake Constance (Germany). Characteristic features, through January 2000. Sampling of site LB prevailing annual wind direction, and sample sites are commenced one month later. At each site, four shown. Different zones are represented schematically. replicate samples were taken randomly within a Sample sites E1 E3 are located within the transition zone of the river mouth, which is adjacent to the littoral area rough circle of diameter 8 m. In the laboratory, containing samples site C, which presumably experiences water overlying the sediment cores was carefully low influence of the river. Sample site LB lies on the siphoned off, leaving the sediment surface undis- border of the littoral and pelagic zones and sample site R is turbed. Subsequently, cores were sliced in half at the river mouth. The aerial view shows the ripples generated by wind action (1), river borne ditches (2), and longitudinally and photographed in order to patches of river borne dissolved and particulate matter (4). document the various sediment layers. One half- Dynamics of wind and river can lead to resuspension of the core was then used for meiobenthic processing, upper sediment layer (3). Two further important aspects while the second half-core was set aside for are interactions (e.g., sediment import/export) with the pelagic zone (5) and seasonally changing water level (6). microbial and chemical analyses. Each half-core For further details, see Table 1 and the text. The aerial was divided into four layers at: 0–13, 14–23, 24– picture was reproduced with kind permission of the 53, and 54–103 mm depth. Prior to chemical and German Aerospace Centre microbial analysis, corresponding depth layers of all four replicate half-cores were pooled. Habitat stress was assumed to be dependent on Meiobenthic fauna were extracted by means of the strength of the river discharge and wind a modified gravity gradient centrifugation tech- conditions. Both forces were categorized into nique after Pfannkuche and Thiel (1988) using classes of low (1), moderate (2), and high stress Ludox AM 40TM (DuPont, Inc., Sigma-Aldrich (3) (Table 1). Low habitat stress is characterized Chemical Co.; colloidal silica, specific density: by reduced wave action and/or low river dis- 1.21 g/cm3) and preserved in formaldehyde (4% charge. Moderate conditions mean higher wind final concentration). For counting purposes, each processed sample was stained with 1% Rose- present models (Smyth and Verbyla 1996). Bengal. Nematodes were classified into four Whenever model results were significant, Tukey’s feeding types: epistrate feeder, deposit feeder, HSD post hoc test was applied to identify suction feeders and chewers (Yeates et al. 1993; differing means (overall p-level 0.05). To fit Traunspurger 1997). assumptions for ANOVA, data were log + 1 transformed. DIVERSE is a routine of the Microbial, chemical, and granulometric PRIMER software package V5.2.9 (Primer-E analyses Ltd.). All other procedures were computed with JMP V5.0.1 (SAS Institute Inc.). A detailed description of the methods applied is given in Wittho¨ ft-Mu¨ hlmann et al. (2005a). Briefly, algal pigment composition was analysed Results using a modified HPLC technique after Schmid and Stich (1995). One gram fresh sediment was Study sites conditions suspended in 3 ml filtered lake-water (0.45 lm, Schleicher & Schuell OE 67), again diluted 1:100 River discharge and lake levels fluctuated con- and then aliquots used for analyses. Bacteria were siderably over the study period, and the site fixed in 4% formaldehyde, stained with diamidin- experienced a series of storms with wind speeds ophenylindole (DAPI), and then counted directly >7 m/s (Fig. 2). Increased river discharges often using epifluorescence microscopy. Total carbon coincided with stormy periods in spring and was determined with an infrared analyzer (Leco winter, but not during summer and autumn. The CS 125) using 200–300 mg dry sediment. Organic study year had a long period of calm conditions carbon was measured after acidification with HCl. from mid-July to mid-November, whereas the Inorganic carbon was calculated as the difference remaining months were marked by alternating between total and organic carbon. Granulometry periods of moderate and high habitat stress. Lake of the <63-lm fraction was conducted via Laser- levels followed a seasonal pattern with strong particle sizing (Galai CIS-01; 100–200 mg DW); seasonal fluctuations in water level at the sample the >63-lm fraction was obtained by sequential sites (maximal range: 2.30 m; Table 2). At the site sieving of the sample (6–10 g DW) through mesh sizes of >500-, 250–500-, 126–250-, and 64–125- lm. Discharge MHM HMML L L M H Wind HLMMHLMLLMH Schussen 550 120 survey days s Data analysis 100 500 lake level m( e m( 80 450 l) v mc( v mc(

g/) 400 60 l Univariate diversity indices were calculated with e a

h³ 350 40 ke

the DIVERSE routine: species number S, species sr iD

300 L 20 ae richness d (Margalef’s index (Clifford and Ste- c 250 0 phenson 1975)), Shannon-Wiener-Index H¢ (log ) 14 e s/m 12 wind events 10 and Evenness J¢. The disturbance/enrichment 8 related MI was calculated according to Bongers FMAMJ J ASOND J (Bongers 1990; Bongers and Ferris 1999). ANO- Febr. '99 - Jan. '00 VA was used to test for differences in diversity index values and feeding type abundance between Fig. 2 Schussen River discharge, lake level, and wind events from February 1999 until January 2000. Black locations and with habitat stress parameters (wind line = lake level, grey shading = discharge, open dot direction and river discharge). Details are given in s = hours with wind speed >7 m/s, black triangles = sam Table 3. Water level was included as a factor pling days. L = low, M = moderate, and H = high affecting the strength of habitat stress. Season was discharge and wind strength respectively. Sources wind, lake level, and discharge data: German Weather Agency considered as random effect using the restricted (DWD) & Environmental Protection Agency Baden residual maximum likelihood method to fit the Wurttemberg (LUBW) 9 nearest to the river mouth (site R) the mean 10

· organic carbon content was 3.8- to 7-fold (3.47%) and the mean bacteria concentration 1.9- to 2.7- 9 ± 1 38

9 fold higher (4.6 · 10 cells/g DW; p < 0.0001,

10 Tukey’s HSD) than those recorded at all other · sites. Organic carbon levels varied significantly 1 2 40 ± 0 17 7 4 ± 1 5 1 53 ± 0 22 0 77 ± 0 06 33 91 ± 8 1 23 ± 0 61 0 88 ± 0 18 165 ± 141 2130 43 1 ± 12 6 2 48 between sites (Table 3 a, b). Mean algal pigment 8

10 concentration was 47–126% higher at the remote · site C (43.1 lg/g DW). All sites are characterized as fine-sandy according to mean grain size. In ± 7 75

9 general, the range and variability of all parame-

10 ters was highest at site R (Table 2). · 94 2 25 ± 0 17 6 9 ± 2 1 1 31 ± 0 41 0 70 ± 0 14 31 163 ± 10 2 40 ± 0 63 0 56 ± 0 14 138 ± 102 1120 22 5 ± 9 9 1 66 Nematode community pattern 8 10 · .

1 Mean annual abundance of nematodes ranged from 121 ind/10 cm2 (site E1) to 165 ind/10 cm2 ± 9 57

9 (site C; Table 2) and nematodes were the pre- 10

· vailing taxon of the meiobenthic community

33 (49.2%, site C to 71.4%; site LB). Altogether 2 43 ± 0 23 7 9 ± 2 8 1 52 ± 0 32 0 76 ± 0 09 43 121 ± 13 1 47 ± 0 65 0 92 ± 0 38 163 ± 68 1000 30 8 ± 19 2 2 23 Less frequent Strong flood event Scarce 106 species were identified and recorded. Deposit 9

10 feeders dominated at all sites in terms of abun- · dance (58–81%) followed by chewers (12–35%) and epistrate feeders (3–11%). Suction feeders ± 1 14 9 were rare (1–4%). The differences between sites 10

· were significant with respect to total nematode

34 abundance and for abundance of epistrate feeders 2 40 ± 0 20 8 0 ± 2 8 1 51 ± 0 35 0 75 ± 0 10 46 166 ± 20 1 47 ± 0 63 0 66 ± 0 30 133 ± 76 865 20 5 ± 8 4 1 64 Flood event

8 and chewers, but not for deposit feeders 10 (Table 3). No site differed completely from the · others. Table 4 summarizes the dominant species

± 8 60 categorized according to feeding type. 9

10 Mean annual values for univariate diversity

· measures varied significantly between sites with 34 2 35 ± 0 20 8 1 ± 2 5 1 51 ± 0 29 0 74 ± 0 08 52 183 ± 106 1 51 ± 0 65 0 62 ± 0 24 121 ± 81 930 19 2 ± 16 2 1 78 Frequent exception of Margalef’s species richness d (Ta-

9 ble 3). The Shannon-Wiener diversity index H¢ 10

· was significantly lower at river mouth site R (H¢ = 1.17) than at sites E1–E3 and the remote

± 4 28 site C (H¢ = 1.51–1.53). Species number S was 9

10 highest at site R, but the community was charac- · terized by just a few dominant species (Evenness 54 8 2 ± 6 1 1 17 ± 0 82 0 56 ± 0 24 71 170 ± 162 0 88 ± 0 71 100 Sites R E1E2E3LBC J¢ = 0.56). Species richness and MI varied within a narrow range at all sites (d = 6.8 – 8.2; MI = 2.2– ) 128 ± 156 2 2.4). Seasonal variation in species diversity indi- ¢ g/g DW) 30 6 ± 26 6

H ces was strong with higher values in spring and d S

)3 lower values in autumn, with exception of site C n m) l ¢

J where seasonality was less pronounced (Fig. 3). Main features of sample sites during the investigation period Margalef’s species richness and the Shannon- Wiener-Index increased substantially in spring at mouth (m) Sample size ( Shannon-Wiener Evenness Maturity Index, MI 2 20 ± 0 22 Species richness Species number Nematodes (Ind /10 cm Grain size ( Bacteria (cells/g DW) 4 60 Depth (m) Organic carbon (%) 3 47 ± 2 88 Distance to river Table 2 River water contact Continuous Algal pigments ( l Values are presented as mean ± standard deviation. Forthe location of sample sites R, E1–E3, river LB & C see Fig. mouth site R. No clear seasonality was Table 3 ANOVA, model results; H¢ = Shannon Wiener Index, J¢ = Evenness, S = species number, d = species richness, MI = Maturity Index, Corg = organic carbon Depen dent Global effect Effect tests variable Site Water Wind Discharge Water Water level level · wind level · discharge

Nema todes r2 = 0.33; F = 6.68; P = 0.0001 P = 0.059 P = 0.013 P = 0.59 P = 0.053 P = 0.011 P < 0.0001 Epistrate r2 = 0.41; F = 9.29; P £ 0.0001 P = 0.46 P = 0.28 P = 0.078 P = 0.88 P = 0.74 feeders P < 0.0001 Deposit r2 = 0.42; F = 9.69; P = 0.17 P = 0.09 P = 0.005 P = 0.055 P = 0.011 P = 0.009 feeders P < 0.0001 Chewer r2 = 0.43; F = 10.04; P £ 0.0001 P = 0.66 P = 0.39 P = 0.044 P = 0.38 P = 0.36 P < 0.0001 H¢ r2 = 0.50; F = 13.60; P £ 0.0001 P = 0.40 P = 0.44 P = 0.03 P = 0.99 P = 0.66 P < 0.0001 J¢ r2 = 0.41; F = 9.21; P £ 0.0001 P = 0.024 P = 0.41 P = 0.22 P = 0.33 P = 0.73 P < 0.0001 Sr2 = 0.48; F = 12.22; P = 0.025 P = 0.11 P = 0.32 P = 0.028 P = 0.34 P = 0.84 P < 0.0001 dr2 = 0.42; F = 9.45; P = 0.09 P = 0.3 P = 0.5 P = 0.054 P = 0.04 P = 0.15 P < 0.0001 MI r2 = 0.55; F = 16.25; P £ 0.0001 P = 0.98 P = 0.42 P £ 0.0001 P = 0.032 P = 0.21 P < 0.0001 C 2 org r = 0.66; F = 27.16; P £ 0.0001 P = 0.005 P £ 0.0001 P = 0.011 P = 0.067 P = 0.06 P < 0.0001 Italicized values indicate significant difference; n = 243 (number of samples)

Table 4 The relative Feeding type Dominant species S Relative density (% ± SD) abundance of the three most dominant species Epistrate feeders Ethmolaimus pratensis 13 5.2 ± 7.7 categorized according to Prodesmodora circulata 0.35 ± 1.15 feeding type, respectively Chromadorita leuckarti 0.16 ± 0.77 Deposit feeders Monhystera stagnalis/paludicola 49 32.3 ± 21.8 Daptonema dubium 28.4 ± 20.6 Eumonhystera filiformis 1.17 ± 2.73 Chewers Tobrilus medius 17 13.5 ± 12.0 Tobrilus gracilis 5.1 ± 7.6 Tobrius stefanskii 4.4 ± 5.0 Suction feeders Dorylaimus stagnalis 27 0.55 ± 1.69 S = number of found Mesodorylaimus spec. 0.12 ± 0.57 species, SD = standard Aporcelaimus spec. 0.03 ± 0.23 deviation evident in the MI, which decreased continuously events was high (nematodes: low = 144 ± 95.9, at all sites until autumn, followed by a slight moderate = 165.3 ± 123, high = 114.7 ± 101.7 increase during winter. ind./10 cm2 ± SD; deposit feeders: low = 109.9 ± 77.7, moderate = 111.6 ± 115.1, high = Influence of water level, wind field and river 67.3 ± 61.2 ind./10 cm2 ± SD) and increased with discharge on the nematode communities rising water level. Furthermore, water level had a significant influence on wind and discharge Wind events and water level affected the abun- effects (Table 3). Conversely increasing river dance of nematodes in general and of deposit discharge increased the abundance of chewers feeders in particular. Abundances of both groups (low = 24.7 ± 22.8, moderate = 36.9 ± 28.6, were lowest when habitat stress caused by wind high = 46.2 ± 48.8 ind./10 cm2 ± SD) but water 100 (15; 44; 28) 100 (14;45; 50)

90 90

() 80 80 c%

adn 70 E1 70 R

ue 60 27 3,0 MI 60 27 d 3,0 MI 24 d 24 an 21 2,5 21 2,5 15 vb 50 15 50 2,0 2,0 12 12 40 1,5 40 1,5 9 9 H' 1,0 1,0 6 6 Cumulati e 30 H' 30 0,5 0,5 3 3 20 20 1610 011060

100 (13; 34; 38) 100 (15; 22; 20)

90 90

80 80

adn 70 E2 70 LB

uce(%) 27 27 60 Sd 3,0 MI 60 d 3,0 MI an 24 24 21 2,5 21 2,5

vb 50 15 50 15 ti e 2,0 2,0 12 12 1,5 40 1,5 40 9 9 1,0 1,0 Cumula 30 6 H' 30 6 H' 0,5 0,5 3 3 20 20 1610 011060

100 (12; 32; 33) 100 (17; 25; 16)

90 90

() 80 80 c%

adn 70 E3 70 C

ue 60 27 d 3,0 MI 60 27 d 3,0 MI 24 24 an 21 2,5 21 2,5

vb 50 15 50 15 2,0 2,0

ltie 12 12 1,5 1,5 40 9 40 9 1,0 1,0

Cumu a 6 6 H' 30 H' 30 0,5 0,5 3 3 20 20 1610 011060 Species rank Species rank Fig. 3 Site specific diversity patterns. Cumulative relative change in species richness (d), Shannon Wiener Index abundance curves in dependence of low (circles), (H¢) and Maturity Index (MI). Numbers in parentheses ate (triangles) and high discharge (squares; comp. Fig. 2). refer to number of species with increasing discharge (L; M; X axis = log10 scale. Embedded graphs illustrate seasonal H). Bars = standard error (n =4) level or its interaction with habitat stress was exposed to low river discharge and calm wind not significant. The abundance of both feeding conditions or high stress resulting from high types differed significantly between low and discharge and strong wind effects. The relative moderate stress. The relative abundance of abundance of chewers increased when river dis- deposit feeders was highest when sites were charge was high and wind events were moderate. Table 5 Mean values of relative abundance of feeding types and Shannon Wiener diversity index (H¢) differentiated according to increasing wind and discharge stress

Epistrate feeders (%) Deposit feeders (%) Chewers (%) H’

H / 10.0 4.0 / 50.2 71.5 / 37.2 23.9 / 1.64 1.24

tre ss M 3.3 7.3 6.2 78.9 60.4 57.4 17.8 29.1 33.5 1.12 1.70 1.82

L 1.6 2.6 13.5 84.2 76.4 42.5 13.8 20.7 41.2 1.13 1.09 1.86 Wind s L M H L M H L M H L M H discharge stress

L = low, M = moderate, and H = high stress (comp. Table 1). Grey gradient: 0 5% (white), 5 10%, 10 15%, 15 25%, 25 35%, 35 45%, 45 65%, >65% relative abundance; / = combination did not occur

Epistrate feeders showed a similar pattern to Discussion chewers (Table 5). Diversity indices and MI differed significantly The study revealed that (1) the abundance of between sites and were strongly affected by river nematodes in general and deposit feeders in discharge but not by wind events. Species diversity particular were influenced by the strength of increased with increasing river discharge (H¢: wind events, (2) the overall diversity and abun- low = 1.13 ± 0.34, moderate = 1.54 ± 0.37, high = dance of chewers are mainly affected by discharge 1.63 ± 0.46; S: low = 5.9 ± 1.7, moderate = 8.2 ± stress, (3) the benthic organic matrix, represented 2.8, high = 9.0 ± 4.0; MI: low = 2.18 ± 0.11, mod- by organic carbon content, was influenced by both erate = 2.41 ± 0,21, high = 2.43 ± 0.20). Evenness forces and (4) changing water levels modified the was significantly reduced by rising water levels impact of both wind events and river discharge on when discharge was low. Interaction between the community. water level and habitat stress played a minor role. Annual variation of Shannon-Wiener species General composition of nematode diversity differed between sites. H¢ altered greatly communities at site R and at the remote site C least (Fig. 3). Cumulative dominance plots differed in depen- With 106 species identified, the study area is one dence of discharge stress and revealed three of the most species rich freshwater habitats groups of sites. Dominance was strongest at site studied thus far. Michiels and Traunspurger R, where under low stress only one species (2005a) described 152 species in the littoral area accounted for 88% of the entire assemblage. This of Lake , but in most other studies of dominance fell to 45–50% under conditions of littoral habitats the number of species recorded is moderate and high stress. Dominance was weakest below 50 (Table 6). Species number was highest at the remote site C, where the influence of habitat at the most heterogeneous site R, followed by stress was marginal. The remaining sites E1–E3 sites E1–E3 and lower at the remote site C and and LB represent an intermediate group, in which the deeper site LB. Hence, species numbers a few species were dominant under low stress decrease with decreasing influence of the river. conditions, but less so under moderate and high Thus our sample sites are seen to provide an stress (Fig. 3). In all cases, species number was accurate representation the transition zone from lowest when discharge was low and highest under the river to the lake. However, this pattern was moderate discharge. A combination of intermedi- not repeated by the pattern of Shannon-Wiener- ate wind and discharge stress yielded significantly species diversity, which indicated the greatest higher species diversity (H¢)(p < 0.0001, Tukey’s diversity in sites E1–E3 and the remote site C. HSD; Table 5). The dominance of three species (Daptonema Table 6 Compilation of diversity indices in shallow zones of lakes and estuaries with different trophy Lake/Estuary Trophy Depth (m) SH¢ MI Authors

Char Oligotrophic <3 21 2.4 2.7 Prejsa Zielony Gasienicowy Oligotrophic 6 9 7 9 1.7 2.0 Prejsb Konigssee Oligotrophic 1 5 48 51 1.0 2.1 Traunspurgerc 11 alpine lakes Oligotrophic <2 24 75 1.7 3.6 Michielsh 3 swedish lakes Oligotrophic <2 14 22 1.5 2.0 Petersj Lake Constance Oligo /mesotrophic 0.5 4.2 31 71, total: 106 1.2 1.5 2.2 2.4 This study Zarnowieckie Mesotrophic <3 16 2.8 Prejsa 8 swedish lakes Mesotrophic <2 11 33 1.2 2.1 Petersj Neusiedlersee Thalassohyalin, <3 26 0.1 2.0 Schiemerd meso/eutrophic Obersee Eutrophic 1 152 Michielsi Czarna Kuta Eutrophic 2.5 3.0 3 0.9 1.2 Prejsb Donghu Eutrophic 2.2 7 20 1.0 2.1 2.6 3.3 Wue Mikolajskie Eutrophic <3 52 2.3 2.4 Prejsa Smolak Eutrophic 4.5 2 0.30 Prejsb 6 swedish lakes Eutrophic <2 8 34 1.0 2.9 Petersj Ems estuary Eutrophic Estuary 1.4 2.5 Essinkf Gironde Estuary Eutrophic Estuary 1.6 Soetartg a Prejs (1977a), b Prejs (1977b), c Traunspurger (1991), d Schiemer (1978), e Wu et al. (2004), f Essink and Keidel (1998), g Soetart et al. (1995), h Michiels and Traunspurger (2005a), i Michiels and Traunspurger (2004), j Peters (2005) S, Species number; H¢, Shannon Wiener Index, MI, Maturity Index. All H¢ values calculated as log base e. Values given in the studies of Schiemer and Peters have been transformed dubium, Monhystera stagnalis/paludicola,and resuspension, resettlement and relocation of abi- Tobrilus medius) resulted in low evenness scores. otic matter and organisms under the influence of The combination of high species numbers and wind events (Wittho¨ ft-Mu¨ hlmann et al. 2005a). dominance of only few species has previously The littoral area can switch within days or even been described as characteristic of lotic habitats hours from one set of more or less stable physical (Beier and Traunspurger 2003). The MI, calcu- habitat conditions to another. The import of lotic lated as a measure of pollution was >2.0 at all species and relocation of littoral nematodes by sites, indicating comparatively low levels of pol- the river may result in increased inter- and intra- lution (Beier and Traunspurger 2003). The study specific competition between members of the in revealed significant differences in feeding type situ community and relocated organisms. The composition between all sites with regard to same mechanism can also influence resource epistrate feeders and chewers, but not to deposit allocation within the entire river mouth area, as feeders, which were well represented at all sites. observed in the case of organic carbon. However, This outcome may be indicative of the role of we were unable to estimate the degree to which deposit feeders as colonizers, able to recover passive recruitment contributes to the species quickly after disturbance events in dynamic envi- diversity observed. ronments (Death and Winterbourn 1995). The results of this study have refined our understanding of river–lake interfaces. Wind Influence of wind effects and river discharge events were important with regard to abundance on nematode community structure of nematodes and in particular to deposit feeders, whereas species measures and the occurrence of In a previous study at the same investigation site, chewers were influenced mainly by river dis- we were able to show that deposition of river charge. Water level influenced the benthic assem- borne benthic particulate matter (rBPM) is a blages generally. The littoral area undergoes large function of primary settlement of matter sus- diurnal and seasonal changes in temperature. pended in by river discharge, and subsequent However, previous analyses indicated that this parameter played only a minor role as stress maximum species richness at sites of greatest factor (Wittho¨ ft-Mu¨ hlmann et al. 2005a). stability, and maximum evenness at sites of Disturbance is known to alter species diversity intermediate stability. They stressed that high in benthic communities (Peterson 1996; Schratz- species diversity can be maintained by habitat berger and Warwick 1999; Tolonen et al. 2001). patchiness. The results of mescosm experiments Decreasing evenness resulting from rising water of Austen et al. (1998) were consistent with levels led to a shift in the nematode community Intermediate Disturbance Hypothesis (Connell towards few deposit feeders like Monhystera 1978) and the authors suggested that dominance stagnalis/paludicola and Daptonema dubium, of few species, as it is the case in our study, is both of which have relatively short life cycles. important in maintaining regional diversity. Our Annual mean nematode abundances were rela- results agree with this study, in that species tively similar at all sites investigated. However, diversity (H¢) and species number (S) were periods of low habitat stress, with calm wind reduced when wind stress and discharge stress conditions and low discharge, were linked with were either both low and both high. Conversely, greater nematode abundance, a strong dominance any other combination of stresses may enhance of deposit feeders and low species diversity. habitat patchiness. Urban (2004) showed in a Deposit feeders comprise both selective and freshwater-pond study that coping with local non-selective species, feeding on bacteria and environmental heterogeneity requires trade offs unicellular eucaryotes. In this group, reproduction between adaptations that increase environmental rates are high. The importance of habitat stability tolerance and those that promote fitness within and, indirectly, of water level, was demonstrated intense biotic interactions. Such trade offs will by the results from the littoral border site LB. This often regulate metacommunity dynamics. Our site is located 1 m deeper than all the others, and results provided hints that these trade offs might was thus less prone to resuspension of sediment by vary along a disturbance gradient such as that wave action. Here, deposit feeders remained very within a river–lake interface. abundant under any stress condition. Stable con- ditions may enhance reproductive fitness of deposit feeders and inter-specific competition Conclusion may be increased, reducing species diversity and evenness. Conversely, species diversity and the The results obtained describe the broader river proportions of chewers and epistrate feeders all mouth area as a very dynamic, relatively unpol- increased under conditions of moderate and high luted benthic habitat, representing a transition stress, with river discharge as the major driving from riverine to fully littoral habitat with no river factor. In a previous microcosm experiment using influence. Within this ecotone, the composition of river-borne benthic particulate matter of the nematode feeding types differed significantly Schussen, we could show that increased rBPM between sample sites with the most pronounced can have a suppressive effect, especially on chew- differences recorded for the site closest to the ers (Wittho¨ ft-Mu¨ hlmann et al. 2005b). Concen- river mouth. Consideration of habitat stress clas- tration of rBPM is higher when discharge is low. ses has improved understanding of the processes These field results support our previous experi- involved in regulating the composition of nema- mental findings. tode communities. Disturbance or stress resulting According to Allan (1995) and Tokeshi (1993), from wind and river discharge regulated habitat abiotic factors and physical disturbance can patchiness and thereby led to changes in commu- weaken biological interactions. However, the four nity composition. The impacts of wind and month-long calm period encountered during the discharge are in turn regulated by water level. investigation may have allowed the communities Low stress led to greater nematode abundance, to equilibrate and given colonizers like deposit reduced species diversity, and an increased dom- feeders an advantage over persisters such as inance of deposit feeders, thus colonizers were in chewers. Death and Winterbourn (1995) found favour under more stable habitat conditions. The results support the Intermediate Disturbance Giere O (1993) Meiobenthology the microscopic fauna Hypothesis. in aquatic sediments. Springer, Berlin, p 328 Hakenkamp CC, Morin A, Strayer DL (2002) The functional importance of freshwater meiofauna. In: Acknowledgements We thank Simone Eckenfels and Rundle S, Robertson AL, Schmid Araya MJ (eds) Petra Heim for help with microbial analyses and sample Freshwater Meiofauna: biology and ecology. Back processing, Susanne Fitz and Viola Burkhard Gehbauer huys Publishers, Leiden, The Netherlands, pp 321 335 for help with water chemistry analyses, Barbara Haibel for Holopainen IJ, Paasivirta L (1977) Abundance and determining phytobenthos, Nicola Reiff for preparing biomass of the meiozoobenthos in the oligotrophic nematodes, and Alexander Brinker for statistical advice. and mesohumic Lake Paarjavi, southern Finland. The manuscript has benefited from the comments of Annl Zool Fennici 14:124 134 Alexander Brinker, Hans Gude, and Tom Moens. This IGKB (1994) Limnologischer Zustand des Bodensees, research is part of a dissertation to be submitted to the 1 60 University of Constance by A. Witthoft Muhlmann, Michiels IC, Traunspurger W (2004) A three year study of funded by the German Research Foundation (Special seasonal dynamics of a zoobenthos community in a Collaborative Research Programme 454 ‘‘Littoral of Lake eutrophic lake. Nematology 6:655 669 Constance’’). Michiels IC, Traunspurger W (2005a) Benthic community patterns and the composition of feeding types and reproductive modes in freshwater nematodes. Nema References tology 7:21 36 Michiels IC, Traunspurger W (2005b) Seasonal variation Allan JD (1995) Structure and function of runnings waters. of biodiversity and assemblage structure in freshwater Chapman & Hall, London, p 388 nematodes. Arch Hydrobiol 163:183 194 Austen M, Widdicombe S, Villano Pittaco N (1998) Montagna PA (1995) Rates of metazoan meiofaunal Effects of biological disturbance on diversity and microbivory: a review. Vie Milieu 45:1 9 structure of meiobenthic nematode communities. Mar Nalepa TF, Quigley MA (1983) Abundance and biomass Ecol Prog Ser 174:233 246 of the meiobenthos in nearshore Lake Michigan with Beier S, Traunspurger W (2003) Seasonal distribution of comparisons to the macrobenthos. J Great Lakes Res free living nematodes in the Krahenbach, a fine 9:530 547 grained submountain carbonate stream in southwest Peters L (2005) Species distribution of free living nema Germany. Nematology 5:113 136 todes and other meiofauna in littoral periphyton Bergtold M, Traunspurger W (2004) The benthic commu communities of lakes. Nematology 7:267 280 nity in the profundal of Lake Brunnsee: seasonal and Peterson CG (1996) Response of benthic algal communi spatial patterns. Arch Hydrobiol 160:527 554 ties to natural physical disturbance. In: Stevenson RJ, Biro´ K (1973) Nematodes of Lake Balaton. IV. Seasonal Bothwell MI, Lowe RL (eds) Algal ecology: freshwa qualitative and quantitative changes. Annl Inst Biolog ter benthic ecosystems. Academic press, San Diego, ici (Tihany) Hungaricae Acad Scientiarum 40:135 158 pp 375 402 Bongers T (1990) The maturity index: an ecological Petraitis PS, Latham RE, Niesenbaum RA (1989) The measure of environmental disturbance based on maintenance of species diversity by disturbance. nematode species composition. Oecologia 83:14 19 Quart Rev Biol 69:393 418 Bongers T, Ferris H (1999) Nematode community struc Pfannkuche O, Thiel H (1988) Sample processing. In: ture as a bioindicator in environmental monitoring. Higgins RP, Thiel H (eds) Introduction to the study of Environmental Toxicology and Chemistry 14:224 228 meiofauna. Smithsonian Institution Press, Washing Bretschko G (1984) Free living nematodes of a high ton D.C., pp 134 145 mountain lake (Vorderer Finstertaler See, Tyrol, Pieczynska E (1990) Lentic aquatic terrestrial ecotones: Austria, 2237 m asl). I. Monhystera cf. stagnalis and their structure, functions, and importance. In: Naiman Ethmolaimus pratensis. Arch Hydrobiol 101:39 72 RJ, De´camps H (eds) Ecology and management of Clifford HT, Stephenson W (1975) An introduction to aquatic terrestrial ecotones. UNESCO, Paris, pp 103 numerical classification. Academic Press, New York, 140 xii + 229 p Platt HM, Warwick RM (1980) The significance of free Connell JH (1978) Diversity in tropical rain forests and living nematodes to the littoral ecosystem. In: Price coral reefs. Science 199:1302 1310 JH, Irvine DEG, Farnham WF (eds) The shore Coomans A (2000) Nematode systematics: past, present environment, vol. 2, Ecosystems. Academic Press, and future. Nematology 2:3 7 London and New York, pp 729 759 Death RG, Winterbourn MJ (1995) Diversity patterns in Prejs K (1977a) The littoral and profundal benthic nem stream benthic invertebrate communities: the influ atodes of lakes with different trophy. Ekologia polska ence of habitat stability. Ecology 76:1446 1460 25:21 30 Essink K, Keidel H (1998) Changes in estuarine nematode Prejs K (1977b) The species diversity, numbers and communities following a decrease of organic pollu biomass of benthic nematodes in central part of lakes tion. Aquat Ecol 32:195 202 with different trophy. Ekologia polska 25:31 44 Prejs K, Prejs A (1992) Importance of predation in Traunspurger W (1991) Fischbiologie des Konigssees. Das regulating density of meio and macrofauna in sea Meiobenthos des Konigssees. Systematische und sonal tropical waters. Hydrobiologia 242:77 86 okologische Untersuchungen unter besonderer Raspopov IM, Andronikova IN, Dotsenko ON, Kurashov Berucksichtigung der Nematoda. Nationalpark Ber EA, Letanskaya GI, Panov VE, Rychkova MA, chtesgaden, Berchtesgarden, Germany, p 152 Telesh IV, Tchernykh OA, Vorontsov FF (1996) Traunspurger W (1996) Distribution of benthic nematodes Littoral zone of Lake Ladoga: ecological state eval in the littoral of an oligotrophic lake (Konigssee, uation. Hydrobiologia 322:39 47 National Park Berchtesgaden, FRG). Arch Hydrobiol Schiemer F (1978) Verteilung und Systematik der freile 135:393 412 benden Nematoden des Neusiedlersees. Hydrobiolo Traunspurger W (1997) Bathymetric, seasonal and vertical gia 58:167 194 distribution of feeding types of nematodes in an Schmid H, Stich HB (1995) HPLC analysis of algal oligotrophic lake. Vie et Milieu 47:1 7 pigments: comparisons of columns, column properties Traunspurger W (2000) The biology and ecology of lotic and eluents. J Appl Phycol 7:487 494 nematodes. Freshw Biol 44:29 45 Schratzberger M, Warwick RM (1999) Differential effects Traunspurger W (2002) Nematoda. In: Rundle SD, Rob of various types of disturbances on the structure of ertson AL, Schmid Araya MJ (eds) Freshwater Mei nematode assemblages: an experimental approach. ofauna: biology and ecology. Backhuys Publishers, Mar Ecol Prog Ser 181:227 236 Leiden, The Netherlands, pp 63 104 Smyth GK, Verbyla AP (1996) A conditional likelihood Urban MC (2004) Disturbance heterogeneity determines approach to REML in generalized linear models. freshwater metacommunity structure. Ecology J Roy Statist Soc Ser B Methodol 58:565 572 85:2971 2978 Soetart K, Vincx M, Wittoeck J, Tulkens M (1995) Witthoft Muhlmann A, Traunspurger W, Rothhaupt KO Meiobenthic distribution and nematode community (2005a) Influence of river discharge and wind field on structure in five European estuaries. Hydrobiologia the freshwater meiofauna of a dynamic river mouth 311:185 206 area. Arch Hydrobiol 163:307 327 Strayer D (1985) The benthic micrometazoans of Mirror Witthoft Muhlmann A, Traunspurger W, Rothhaupt KO Lake, New Hampshire. Archiv fur Hydrobiol Suppl (2005b) Meiobenthic response to river borne benthic 72:287 426 particulate matter a microcosm experiment. Freshw Taniguchi H, Tokeshi M (2004) Effects of habitat com Biol 50:1548 1559 plexity on benthic assemblages in a variable environ Wu J, Fu C, Liang Y, Chen J (2004) Distribution of the ment. Freshw Biol 49:1164 1178 meiofaunal community in a eutriphic shallow lake of Tokeshi M (1993) Species abundance patterns and com China. Arch Hydrobiol 159:555 575 munity structure. Adv Ecol Res 24:111 186 Yeates GW, Bongers T, De Goede RGM, Freckman DW, Tolonen KT, Hamalainen H, Holopainen IJ, Karjalainen J Georgieva SS (1993) Feedings habits in soil nematode (2001) Influences of habitat type and environmental families and genera an outline for soil ecologists. variables on littoral macroinvertebrate communities J Nematol 25:315 331 in a large lake system. Arch Hydrobiol 152:39 67