Combined eco morphological functional groups are reliable indicators of colonisation

Ecological Indicators 64(2016)31–38Contents lists available at ScienceDirectEcologicalIndicatorsj o u r n a l h o m e p a g e :w w w.e l s e v i e r.c o m /l o c a t e /e c o l i ndCombined eco-morphological functional groups are reliable indicators of colonisation processes of benthic diatom assemblages in a lowland streamViktória B-Béres a ,∗,Áron Lukács b ,Péter Török c ,Zsuzsanna Kókai a ,Zoltán Novák b ,Enik ˝oT-Krasznai a ,Béla Tóthmérész c ,István Bácsi b aHajdú-Bihar County Government Office,Department of Environment and Conservation,Environmental Laboratory,Hatvan Street 16,Debrecen H-4025,Hungary bUniversity of Debrecen,Department of Hydrobiology,Egyetem Square 1,Debrecen H-4032,Hungary cMTA-DE Biodiversity and Ecosystem Services Research Group,Egyetem Square 1,Debrecen H-4032,Hungarya r t i c l ei n f oArticle history:Received 23August 2015Received in revised form 22November 2015Accepted 19December 2015Available online 7January 2016Keywords:Eco-morphological functional groups Diatom guildsCell size categories Colonisation Lowland streama b s t r a c tClassifying benthic diatom taxa based on ecological and morphological features became increasingly important in recent years due to the demand of understanding the dynamics and functioning of diatom assemblages.The great potential in using these functional classifications in diatom ecology involves further refinement of current classification.In our experimental study,colonisation processes of diatom assemblages were studied in a typical small lowland stream,using both diatom guilds and cell size categories.We also tested newly proposed combined eco-morphological functional groups (ecological guilds combined with cell size categories)in the study of the colonisation process in benthic diatom assemblages.We hypothesised that (i)there is a decrease in the proportion of low profile guild,while an increase in that of high profile and motile guilds in time with the decreasing rate of physical disturbance;(ii)the presence of small size categories will be pronounced at the beginning of the colonisation processes,while proportion of larger size categories will be higher in the latter phases of colonisation;and (iii)the relationship between taxa and environmental factors are better reflected by the use of combined eco-morphological functional groups than by the sole analyses of rough guilds or cell size categories.The first hypothesis was not confirmed,and our second hypothesis was only partially confirmed by the results.We found that the relationship between environmental factors and guilds,as well as cell size categories was not appropriate to reveal the relationship between abiotic factors and taxa composition.In contrast we found that compositional changes in colonisation were appropriately reflected by the newly defined combined eco-morphological functional groups.In the combined eco-morphological functional groups,such kind of taxonomical and ecological features can be prevailed which are hidden in guilds or cell size categories separately.Thus these combined eco-morphological functional groups could help to come one step closer to develop a widely used ecological classification in diatom researches.©2016Elsevier Ltd.All rights reserved.1.IntroductionThe ecological status assessment using species composition data have became increasingly debated by the authors of sev-eral comprehensive studies in aquatic biota (Reynolds et al.,2002;Padisák et al.,2009;Berthon et al.,2011).This is considered to be a rather static approach and especially problematic in assemblages with (i)high number of not-easy identifiable or taxonomically∗Corresponding author.Tel.:+36307006994.E-mail address:beres.viktoria@ (V.B-Béres).problematic taxa and in assemblages where (ii)highly stochastic processes frequently occur like in the phytoplankton or benthic diatom (Reynolds et al.,2002;Padisák et al.,2009;Berthon et al.,2011).While the functional classification of planktic algae and cyanobacteria has a rather long history (see Salmaso et al.,2014),classifying benthic diatoms (Chromista,Bacillariophyta –Guiry and Guiry,2015)into functional groups began only very recently (CSR strategy types –Biggs et al.,1998;ecological guilds –Passy,2007;Rimet and Bouchez,2012a,b ;biovolume –Berthon et al.,2011;Kókai et al.,2015).The link between the appearance,and abundance changes of diatom life forms (e.g.mobile,colonial,tube-forming,stalked,pioneer)and environmental factors are quite well/10.1016/j.ecolind.2015.12.0311470-160X/©2016Elsevier Ltd.All rights reserved.32V.B-Béres et al./Ecological Indicators64(2016)31–38studied(see for details for example Berthon et al.,2011).The com-position and dynamics of benthic diatom assemblages are primarily linked to present environmental conditions and their changes(Ács et al.,2000;Porter et al.,2013).The establishment of a respec-tive taxon in a benthic assemblage is primarily determined by its colonisation ability(Ács and Kiss,1993a),while the success of the establishment is affected by some specific morphological charac-teristics like(i)the cell size and biovolume(Ács and Kiss,1993a; Snoeijs et al.,2002;Kókai et al.,2015)or by ecological adaptation features like(ii)nutrient demand or(iii)tolerance against physical disturbances(Passy,2007;Berthon et al.,2011;Rimet and Bouchez, 2012a,b;Stenger-Kovács et al.,2013;B-Béres et al.,2014).Benthic diatom taxa can be classified intofive classes based on the cell size(S1:5–99␮m3,S2:100–299␮m3,S3:300–599␮m3, S4:600–1499␮m3,S5:≥1500␮m3;Berthon et al.,2011).Distri-bution of the different sized diatom taxa in an assemblage and considerable changes in cell size structure of the assemblages can provide vital information about changes in abiotic and biotic factors (i.e.eutrophication;climate change;organic pollution;secondary salinisation;macroalgal host effect–Snoeijs et al.,2002;Berthon et al.,2011;Kókai et al.,2015).It has to be emphasised also that the changes of the cell size categories in time or in connection with the changing abiotic factors depend strongly not only on the size of a respective diatom taxa,but also on ecological characteristics. Some taxa with the same size could belong to different ecological guilds,life forms,or strategies(e.g.Plenkovic-Moraj et al.,2008; Kókai et al.,2015).Based on ecological characteristics Passy(2007)classified diatom taxa into three functional groups(low profile,high pro-file and motile guilds).Classification of taxa into these functional groups(in the forthcoming called‘guilds’)was based on their resis-tance to nutrient enrichment and physical disturbances.Rimet and Bouchez(2012a)added to Passy’s original classification a fourth planktic guild,because planktic diatom taxa are also able to settle into the biofilm.The intensive study of the relationship between the diatom guilds and environmental factors started only in the last few years(Passy,2007;Berthon et al.,2011;Gottschalk and Kahlert, 2012;Rimet and Bouchez,2012a,b;Stenger-Kovács et al.,2013;B-Béres et al.,2014).Based on these studies it can be stated that there is a great potential in the use of diatom guilds for environmen-tal status assessment.However,it was also stressed,that further refinement of the robust guild classification is necessary(B-Béres et al.,2014).One possible solution for this refinement can be based on the combination of morphological(cell-size)and ecological features(guild affiliation),which jointly in an eco-morphological functional classification might provide enough high number of groups for afine-scale assessment of a given status and also com-positional and dynamical changes.Directional changes characterised also by highly stochasticfluc-tuations in composition,like colonisation processes in diatom assemblages could be displayed as robustly by the use of func-tional groups as by the conventional taxonomical approach.Despite of this fact,there is rather limited information about the changes of ecological guilds during colonisation(Passy and Larson,2011; Stenger-Kovács et al.,2013),and there are only indirect and sporadic data provided for changes of cell size categories dur-ing colonisation(but see Roberts et al.,2004;Sekar et al.,2004; Plenkovic-Moraj et al.,2008;Passy and Larson,2011).However, the development of a biofilm(colonisation)provides a new per-spective for evaluating the effect of environmental factors and their alterations on compositional changes of aquatic assem-blages(Stenger-Kovács et al.,2013).The retained information can play a decisive role in ecological status assessment or in watercourse typification.In spite of these facts,according to our present knowledge there is no information about the use of combined ecological groups assessing the changes of diatom assemblages in changing environments,and/or during colonisation processes.In this study we followed the colonisation processes in diatom assemblages using ecological guilds,cell size categories and newly created combined eco-morphological functional groups in a small lowland stream.We used the following study hypotheses directly linked to present literature:(i)Guild-hypothesis–There is a decrease in the proportion of lowprofile guild,while an increase in that of high profile and motile guilds with increasing time and/or decreasing rate of physical disturbance(Stenger-Kovács et al.,2013)(ii)Cell-size hypothesis–The proportion of small size categories will be high at the beginning of the colonisation process,while that of larger size categories will be higher in the latter phases of colonisation(Ács et al.,2000;Berthon et al.,2011).(iii)Combined eco-morphological functional group hypothesis–Colonisation processes and the relationship between taxa and environmental factors are better reflected by the use of com-bined eco-morphological functional groups than by the sole analyses of guilds or cell size categories(B-Béres et al.,2014).2.Materials and methods2.1.Sampling setup and environmental parametersSamples were collected at Debrecen-Józsa on the stream bed of Tócó(EOVX:254755;EOVY:839873,Fig.1).ColonisationFig.1.The sampling site on the Tócóstream in Trans-Tisza region of Hungary.Black lines:the main rivers in the region and the Tócóstream.Black mark:The sampling site on Tócóstream.V.B-Béres et al./Ecological Indicators64(2016)31–3833processes were studied from15.03.2014to06.06.2014.The sched-ule of sampling was the following:samples were collected more frequently at the beginning of the colonisation(hourly in thefirst 9h);daily in thefirst week;later twice a week(till the35th day) and weekly at the second part of the study period.Water depth (WD–cm)was measured by a measuring tape,water velocity(WV –m s−1)was measured by a SonTek FlowTracker Handheld-ADV (USA).Conductivity(COND–␮S cm−1),oxidation-reduction poten-tial(ORP–mV),pH,dissolved oxygen concentration(DO–mg L−1), oxygen saturation(OS–%)and water temperature(T–◦C)were measured with a portable-multiparameter digital meter(HQ30d, Germany)in thefield.Water samples of80mL were kept at4◦C in cooler bag during transportation to the laboratory for the measure-ments of the concentrations(mg L−1)of nitrite,nitrate,ammonium and soluble reactive phosphate(NO2−,NO3−,NH4+−and PO43−–spectrophotometric methods;MSZ ISO7150-1:1992;MSZ EN ISO 6878:2004;MSZ1484-13:2009).2.2.Sample collection and preparationWooden slides(length:6cm;width:2cm;height:0.5cm;with altogether a32cm2surface area of each)were used as colonisa-tion substrates.Three chains of wooden slides were put into the stream bed.All chains contained36pieces of wooden slides(one for each sampling time).Altogether three slides were collected at each sampling date(one from each chain).The slides were pre-served in Lugol’s solution until further processing.The biofilm was removed from the wooden slides by tooth brushing.Diatom valves were prepared by the hot hydrogen-peroxide method(EN 13946:2003).Naphrax synthetic resin was used for embedding. Leica DMRB microscope with1000–1600-fold magnification was used for identification of diatom taxa.At least400diatom valves were counted(EN14407:2004).We used Krammer and Lange-Bertalot(1997a,b),Krammer and Lange-Bertalot(2004a,b),and Potapova and Hamilton(2007)for diatom identification.2.3.Data processing and analysesClassifying diatom taxa into low profile,high profile and motile guilds was done according to Passy(2007)and Berthon et al.(2011). Planktic guild was also introduced according to Rimet and Bouchez (2012a).Diatom taxa were assigned tofive size classes according to Berthon et al.(2011–see the size classes in the introduction).We classified the detected species into20combined eco-morphological functional groups(four guilds×five cell size categories;see in Sup-plementary Material)for the assessment of the changes during colonisation processes.To analyse the relationship between envi-ronmental factors and the samples collected in each sampling date and to select crucial abiotic factors,principal component analyses (PCA)were calculated using CANOCO4.5software package(Lepˇs andˇSmilauer,2003).To analyse the relationship between some crucial environmental factors(verified by PCA)and the various characteristics of diatom assemblages(taxonomical composition, ecological guilds,cell size classes,or combined groups)canonical correspondence analyses(CCA)were calculated,where environ-mental factors were added by weighted averaging(ter Braak and ˇSmilauer,2002).3.Results3.1.Changes in environmental factorsAltogether13physical and chemical factors were included to PCA analysis(WD,WV,COND,ORP,pH,DO,OS,T,NO2−,NO3−, NH4+,PO43−,and Time).Thefirst two axes of PCA explained the 76.7%of variance.The factors showing the highest correlation with thefirst axis were time(−0.6635),DO(0.6112),pH(0.5338)and WD (0.5136),and with the second axis T(0.6972)and NO3−(−0.6308). Further analyses of the relationship between abiotic factors and species,furthermore environmental factors and ecological groups, six abiotic factors were performed by CCA based on the highest correlation with the two axes of the PCA(see above).The most important changes of the six environmental factors occurred after the rainy-period between the27th and35th days,and in the arid-period following these rainy days(Fig.2).Water depth decreased from18cm to4cm in the arid period(Fig.2a),water temperature increased from11.9◦C to16.9◦C from the27th day(Fig.2b).The pH,dissolved oxygen and nitrate concentration decreased the27th day,onwards(Fig.2c–e).The stream dried out to the91st day.3.2.Colonisation processes of diatom assemblages–ecological guildsThere were significant correlation between the selected six environmental factors and diatom taxa composition based on the results of CCA analysis(Fig.3).The eigenvalues of the CCA were 0.293and0.067,while the species and environmental correlation were68.4%and84%for thefirst and second axes,respectively.The Monte-Carlo permutation test indicated(N=499;p=0.002for the first and p=0.002for all canonical axes)for the pattern displayed by the CCA a significant difference from a random pattern.The abundance of some of the low profile guild taxa (Achnanthidium minutissimum, A.eutrophilum,Amphora pedicu-lus,Rhoicosphenia abbreviata)correlated strongly positively with colonisation time,and negatively with water depth,nitrate and dissolved oxygen concentrations;while some of that with other taxa showed a week(Planothidium lanceolatum),or even a nega-tive correlation(Meridion circulare,Planothidium frequentissimum, Fig.3).The abundance of some high profile guild taxa(Melosira varians, Fragilaria capucina s.l.,Gomphonema micropus)correlated positively with water depth and other abiotic factors(in exception of water temperature).In contrast,other taxa correlated positively with colonisation time or water temperature(Eunotia bilunaris,Gom-phonema clavatum,G.olivaceum,G.parvulum,Fig.3).Most of the motile guild taxa showed a strong positive corre-lation with time or water temperature(e.g.Navicula radiosa,Na. tripunctata,Na.veneta,Naviculadicta seminulum,Nitzschia dissipata, Ni.liearis,Surirella brebissonii).The correlations between motile guild taxa and physical and chemical factors were usually negative (NO3−,DO,pH,ORP,Fig.3).Although there were significant correlation found between the species composition and the environmental parameters,ecologi-cal guilds showed no significant correlations with the selected six environmental factors based on the results of CCA analysis.The eigenvalues of the CCA were0.037and0.017for thefirst and sec-ond axes,respectively.The Monte-Carlo permutation test did not indicate the significance of the CCA(N=499;p=0.3660for thefirst and p=0.2020for all canonical axes).3.3.Colonisation processes of diatom assemblages–cell size categoriesThe separation of the small(S1and S2)and large(S5)sized taxa was conspicuous along the abiotic factors(Fig.5).Taxa in S1 size class correlated negatively with water depth(e.g.Achnanthid-ium minutissimum,A.eutrophilum,Amphora pediculus,Naviculadicta seminulum),or nitrate content(F.capucina var.gracilis,Fig.3).The relation of some taxa of the S2size class to nitrate was positive (F.capucina,Nitzschia perminuta,P.frequentissimum),other S2taxa correlated positively with time(Navicula veneta,Figs.3and5).Taxa34V.B-Béres et al./Ecological Indicators 64(2016)31–38Fig.2.Changes of the most important physical and chemical parameters during the study (a)water depth (WD –cm);(b)temperature (T –◦C);(c)pH;(d)dissolved oxygen concentration (DO –mg L −1);(e)nitrate (NO 3−–mg L −1).of the S5size class correlated positively with water depth (e.g.M.varians ,Figs.3and 5).The separation of the taxa of S3or S4classes were not so obvi-ous than for the other size categories,there were opposite trends ofcorrelation of taxa belonging to the same size class (Figs.3and 5).While some taxa correlated positively with time (e.g.S3–Gom-phonema olivaceum ,Navicula cryptocephala ,R.abbreviata ;S4–E.bilunaris ,Navicula tripunctata ,Ni.dissipata )and water temperatureV.B-Béres et al./Ecological Indicators64(2016)31–3835Fig.3.Changes of the relation of species composition of diatom assemblages and the environmental variables displayed by CCA based on species abundances.The eigenvalues of the CCA were0.293and0.067,while the species and environmental correlation were68.4%and84%for thefirst and second axes,respectively.The most abundant30taxa were shown using four letter OMNIDIA codes.(S3–Gomphonema parvulum;S4–G.clavatum),other taxa rather correlated positively to nitrate or water depth or the other abi-otic factors(S3–G.micropus,Gomphonema sarcophagus,Navicula gregaria;S4–M.circulare,Figs.3and5).The correlation between the environmental variables and cell size categories were significant based on the results of CCA anal-ysis(Fig.5).The eigenvalues of the CCA were0.119and0.021, while the cell size categories and environmental correlation were 82.4%and97.1%for thefirst and second axes,respectively.The Monte-Carlo permutation test indicated the significance of the CCA(N=499;p=0.002for thefirst and p=0.002for all canonical axes).3.4.Colonisation processes of diatom assemblages–combinedeco-morphological functional groupsThe correlation between physical and chemical factors and the abundance of combined eco-morphological functional groups were significant(Fig.6).The Monte-Carlo permutation test indicated this significance of the CCA(N=499;p=0.002for thefirst and p=0.002 for all canonical axes).The eigenvalues of the CCA were0.219and 0.052,while the combined groups and environmental correlation were75.2%and92.9%for thefirst and second axes,respectively. The correlation between physical and chemical factors and taxa (Fig.3)were in correspondence with the correlation between envi-ronmental factors and the eco-morphological functional groups (Fig.6).Certain low profile and motile groups(LS1with Achnanthid-ium minutissimum,A.eutrophilum,Amphora pediculus;MS1with Naviculadicta seminulum;MS4with Navicula tripunctata,Ni.dis-sipata)correlated positively with time(Figs.3and6).Other low profile guild taxa with a different size(e.g.LS4–M.circulare,Fig.3) correlated positively with the measured physical and chemical parameters(NO3−,DO,pH,ORP,Fig.6).In contrast to the weak correlation between abiotic factors and high profile guild(Fig.4), combined high profile groups(HS1and HS4)correlated clearly pos-itively with time and water temperature,while the correlation of other combined groups(HS2and HS5)with these abiotic factors were negative(Fig.6).Fig.4.Changes of the relation of diatom ecological guilds and the environmental variables displayed by CCA based on guilds’abundances.The eigenvalues of the CCA were0.037and0.017,while the guilds and environmental correlation were70.9% and76.2%for thefirst and second axes.4.Discussion4.1.Colonisation processes of diatom assemblages–anecological-guild-based perspectiveDuring colonisation the benthic diatom assemblages are influ-enced by gradually changing stress factors:at the beginning,the most important factors are the nutrient content of the water the light availability and the water velocity(by causing shear effects).With increasing time,the taxonomical composition in the biofilm became affected by increasing interspecific competition for decreasing amounts of light and nutrients(Zhang et al.,1997;Zhang and Lee,1999;Bhagavathy et al.,2011;Bhagavathy and Sumathi, 2012;Stenger-Kovács et al.,2013).We hypothesised that the pro-portion of low profile guild will decrease in time because their establishment is strongly linked to open surfaces(Stenger-Kovács et al.,2013).This hypothesis was not confirmed by the results –although the abundance of some respective low profile guild taxa decreased with time(Fig.3),there was no overall temporal decrease in the abundance of the low profile guild taxa(Fig.4). Those low profile guild taxa(M.circulare,P.frequentissimum,P. lanceolatum)appeared in high number in the samples at the begin-ning of colonisation process,which were abundant in the matured benthic diatom assemblages of the studied stream.These taxa are generally characterised with low migration ability but with a high settling rate(Ács and Kiss,1993a).It is also important to emphasise that small low profile taxa, like Achnanthidium minutissimum and Amphora pediculus corre-lated even positively with the time of colonisation.This interesting phenomenon was supposedly linked with the initiation of a new colonisation-phase after a heavy rainy period,which caused harsh physical disturbances between the27th and35th days(opened some new colonisation patches in the biofilm).The above listed small-sized low profile taxa are able to re-establish after heavy dis-turbances(Stenger-Kovács et al.,2013),primarily because of their short generation-time and high resistance to physical disturbances (Passy,2007;Stenger-Kovács et al.,2013;B-Béres et al.,2014).The36V.B-Béres et al./Ecological Indicators64(2016)31–38re-establishment of these small low profile taxa after the35th day resulted the observed lack of decrease of the overall abundance of the low profile guild in time.Increase in abundance of high profile guild taxa is primarily influenced by high nutrient supply and low intensity of physical disturbances(i.e.shear effects caused by water velocity–Passy, 2007;Stenger-Kovács et al.,2013).Thus,we hypothesised that the abundance of high profile guild will increase in time.This was not confirmed by our results:the abundance of high profile guild was even slightly decreased with time(Fig.4),because some high pro-file guild taxa(F.capucina,M.varians)were characteristic in the assemblages from thefirst day and they correlated negatively with time(Fig.3).They are constant member of the matured biofilm in Tócóstream.These high profile guild taxa can be characterised by a fast settling ability;therefore,they might appear in the early stage of the colonisation process(Ács and Kiss,1993a,1993b;Ács et al., 2000).Their high abundance in thefirst hours was primarily caused by the breakaway from matured biofilm from the surroundings and settling to the experimental substrate.It is important to note that the relative abundances of these taxa decreased drastically during the new colonisation phase(from31st day)resulted the lack of increase of the overall abundance of the high profile guild during the colonisation process.Motile guild taxa are able to change their position and choose the most suitable microhabitat(Johnson et al.,1997;Passy,2007). These taxa usually avoid physically disturbed places(Stenger-Kovács et al.,2013)and they are successful competitors for nutrients(Pringle,1990;Van der Grinten et al.,2004).Thus, we hypothesised that their abundance correlates positively with time.Although our assumption was not fully confirmed,the rel-ative abundance of motile guild increased until the rainy-period (between the27th and35th days).After a high decline,the rela-tive abundance of this guild started to increase again during a new colonisation phase(to the70th day).But the further succession of biofilm was not possible to study because of the drying out of the stream(to the91st day).These opposite changes(increases and decreases in relative abundance)caused the lack of clear positive correlation of motile guild with time.4.2.Colonisation processes in diatom assemblages–cell-size-categories-based perspectiveThere are contrasting opinions of the changes of cell size cat-egory abundances with the changes of physical and chemical parameters in time(Ács et al.,2000;Berthon et al.,2011).Some authors consider the small cell sized taxa(Achnanthidium minutis-simum,Amphora,A.pediculus and Am.inariensis)as pioneer,first colonising ones,because of their high resistance to physical dis-turbances(Sekar et al.,2004;Passy,2007;Berthon et al.,2011; Gottschalk and Kahlert,2012;Rimet and Bouchez,2012b;B-Béres et al.,2014).Some others reported that the relatively large cell-sized taxa(Diatoma vulgaris,M.varians,Ulnaria ulna)can also be considered in the initial period of colonisation asfirst settlers due to their rapid settling ability on various substrates(Ács and Kiss, 1993a;Ács et al.,2000).Taking into consideration these contrasting views,we hypothesised that the abundance of small taxa will be high at the beginning of the colonisation processes,while the pro-portion of larger size categories will be higher in the latter phase of colonisation.Our results confirmed this assumption only partially (Fig.5):Taxa belong to S2size class(e.g.F.capucina,P.frequen-tissimum)correlated negatively with time(Fig.3).The abundance of small pioneer taxa(Achnanthidium minutissimum and Amphora and A.pediculus)increased only from the beginning of the new colonisation phase from the27th day(see above)resulting positive correlation of S1category withtime.Fig.5.Changes of the relation of cell size classes and the environmental variables displayed by CCA based on the classes’abundances.The eigenvalues of the CCA were 0.119and0.021,while the cell size categories and environmental correlation were 82.4%and97.1%for thefirst and second axes.Our assumptions of increasing abundance of the large cell-size class in time were not confirmed.The relation of large taxa(mem-bers of S5class:M.varians,Gomphonema species)and time was negative.These taxa are constant member of the matured biofilm in Tócóstream(see above).The high initial number of the large taxa was probably due to their breakaway from matured biofilm from surroundings and fast settling to the experimental substrate. These results might indicate also the high settling-ability of these taxa from the disturbed areas of shallow water,but further data are needed to prove the high colonisation ability of the large sizeclass,Fig.6.Changes of the relation of combined groups and the environmental variables displayed by CCA based on groups’abundances.The eigenvalues of the CCA were 0.219and0.052,while the combined groups and environmental correlation were 75.2%and92.9%for thefirst and second axes.。