full text
Transcription
full text
COMPARATIVE ANALYSIS OF'SUCCESSIONALSTAGESOF' SANDY VEGETATION - A CASE STUDY K. Marq6czi MaryiczL K (1993): Comparati'e a allsk of successionalnas6 af sandy wgetatian - 4 case study. - Tiscia 27, 3-8. different stagesof pimary Abstrrct. The vegetationof sevenselectedsites, represenring soheme,and to comparethe suooession, is aMlyzedin orderto esiablisha feasiblesuccessional palhway by the multivariate methods parhways rhe suggesled succ€ssional and traditional a nore or lesslinear successional relev6ssuggests The ordinationof micro-scalecoenological wiih thetsditional theory. s€quenc€ ofstudiedstandsconespoDding Five g,?ical patchesw€re identifiedin eachstandby olassificaiionin order to repres€ntthe Ordinalionof 6ese lypical patcheson the basisof species intemalvariabilityof the vegetation. frequ€ncies shows,thatthe primarydynanic ofthe sandvegetationis moreooinploxThe linoar suooessional sequenceof studi€dstandsis uolikety, differeni kansitionsbetwoenstagesare equiprobable, andthe graphstructureofseriesis ratherreticular. lype Threegroupsofstandswere identifiedby diversityorderingi(i) an initial, oPengrassland with mediumdiversity;the forestherblayeralsobelongsto with low diversity;(ii) a grassland andthe shrubbyhabitat.DiversityProfiles this oategory;(iii) the mostdiverse,olosedgrassland, of thesegroupswerecleariyseparated. The m€thodsemployeddiffelenliateih€ vegetationof the studjedstandsaocordjngto their sequence doesnot necessarily diversityandcomplexity,but the €stablished speciescomposition, reflectto therealhistorioaldevelop eutofthe standsin quesiion Kewottlsi sancl! wgetatiotl, successioLntultituriate ethah, di|e,sity ordering Pi 659,H-6701Hungary K. Marg6czi,DeparmenlofEcology,JATE Univ€rsity,Szeged, Introductlon On€ofthe typicalvegetationandflorl typesof the HungarianPlain can be found on smdy areas between the Danube and Tisza riv€|s. This vegetationhas been shrdied by many authors, amongthen such pioneersas Kerner(1863) and Rapaics (1918). The successionalstases were describedby Hargjtai (1940), zsolt (1943) and Magyar (1960). Besides this olassical works, Pr€cs€nyi(1981) studled the diversiry chaDses dudns suocessionand there are severalpublications phenomena of elucidatingcedainecophysiological suchvegetation(e.9.Tubq 1984and Feketeet ai., 1988). The spatial pattem as weli as the niche belongingto the gnssland r€lationsof somesp€cies commlnity Festucetu raginatae lr.rc studied by Molnir andNosek(1979)andFeketeet a1.(i980). RecentlyFekete(1992)publisheda nelv conwich is differentfrom oept of primary succession, TISCIA 27 the traditional interpfetationby the Hungarian phytosociologists. He pointedoui that the primaD/ dynamicsof sandyvegetationis very complexrthe vegetaiion-soil evolution is non-patali6l; ihe suocession is often deteminedby stoohastictran" sitions between stages; the successionalgraph structul€of the sand s€re is ratherrcticular than The linear;origin of climaxstageis hetercgeneous. pioneersnssland do€s not accumulatesuffioient humus fof the €srablislmentof steppe-m€adow species,so the steppem€adow does not fit the lin€. The standsof oak pioneergmssiand-shrub wood (the climax connuniq, in the kaditional concept)arehabitat-dependent. Although the classical phytocoenological of sand)!egetation successron is given oescr,prion in the above meniioned papers, no detailed nultivariate analysishas been made to describe anongthe stnges. vegetation differ€nces Classifications and ordinationsarc suitablefor analysis of succession,but the effectiveness of djfferent methodsdep€ndson ihe natureof analyzed dara(Mazzoleni,1991). Unequivocal successional sequencesare not necessarily self evident to critical observer (Andersoq 1986). Ior example,computerbased classification and ordination tecbdques d€fined a successionalgradient that differed fion a theoretical or assumedsuccessionalclassification1nthe case of a big sagebrusl/grasscomnuniry (Tueller ald Platou,1991). The aim of this studyis to analyz€the v€gelation of seleot€dhabitatsin a Hungariansandyarea using multivariate methods and diversity comparisons, andto evaluatethe differert concepts ofsuccession accordingto th€r€sults. The art comnuniry compositiona.]d epigeic faunaofthe samestudyplotsanalysedin lhis paper hasalsobeenstudied(Jlirdlinet al., 1993) nesophilousand nesophilousspeciesbesidesthe xerotolerant ones.The phytocoenologicai siatusof thisstandis uncedain. Plot 6. Shrubbyhabitatwith 50-60% covemge of Crataegus nanagya, Juhipen$ connu,,i\. Berbe s vulAaris andLigtsttun vulgarc sht\bs Plot 7. Closedpoplar(Pop&lrsa/ba)forestwith sama Robi ia pseudaacacia .Jees. Crutaegus nonagtna and Junipetw cannunit give a sparse The areasof the studyplotswereabout400 m2. 2. Field sanplins A sinilar samplingprocedurex,asappliedthatr desoLibed by Szolletand Bartha(1991)with sone modjfications.Long tnnseot of 200 contagious smallplots, eachof 20x20 cm size, wereusedfol Study area and Methods samplingin eachstand.The total lransectof 40 m lenghtwasbroken4-5 times,resultinga zig'zagline L Site desctiption and absenoe "netling"the wholeplot. The presence The field studieswer€ canied out in a naiure ofspeoi€irootingin the subplotsv,ererecorded.hl leseNe in fhe southemparl of Hungary,betwe€n the shrubbyhabitaiandin the poplarforestonly the of detectable lhe rivers, Duna and Tisza near !o the village herblayerwassampled.The presenoe w€rerecord€dandanalysed K6leshalomrn June 1991.The study siLeis a lichenandmossspecies complexof wind-blolvnsand dunes,SeveLalsuc- togetherwith thebiSherpLants. cessionalsiagesof sandy vegetationoccur here fiom bare sand to popl forcst For the preseni 3. Data ahalysis The sevenstudysiteswere ordinatedaccording study 7 plots were s€lect€d,representig these ro rbe pooledfrequencyof occuningspeciesusing stagesl principal coordinateanalysis with Czekanowski Plot 1. Almosi baresandwith sorneplant spe" index by the program package NuCoSA cios belonging to the associationFr,ttrcalr,, (T6thlneresz,1991, 1993a).Virrgh (1986) havs yaginatae.The bare sudacewithin the plot nlay found the sarneordinationrnerhodand similarity index to be usetul for deiecting vegetation derivefrofi an abandoned sand-mine. differences in a similarscal€study. In orderto represenr the intemal variabilityof Plot 2. Openperennialgrassland with 30-40% plant cov€rage,predonir\ ting by Festuca r.lgi ata the standsfive most abundantoharacterisiic patch (Festucetwn |aginatae danubiale). The ]|r,ossnnd types (10 contagious20x20 cm subplotof each) were seleotedin eachstandby classificationof aU lichenlayerwasalsoconsiderable. 200subplotsacoording to theirspeci€scomposition, similariq,indexby the NCLAS2 Plot 3. Similar to plot 2, but predominated by usingthe Sorensen Stipa borysthenica. (Fesh.tceh.t,nvaEinatae stipe- program of SYN-TAX lll prograF package (Podani, 1988). These 20x100 cmz subplots, tasutnborysthenicae) represenlirglypical patches,were classificat€dby Plot 4. Openperennialgrassland predoninated NUCoSAusins Czekanowskisimiladty index and by the species belonging to the Festucetun singl€ linkage soiting algorithm. Principal laginatae comm]rr,ity ^ad Popuhts alba fomls a coordinate analyses (NuCoSA)ofthe samesubplols shrubstoreyhere.Heightofit is s0 cn, about2s % were peLfomedusins asain Czekanowskiindex, andthe resriltsof clusteringmethodand ordinalion PIot 5. Closed grasslud with sone xero, TISCIA 27 Analysis ol the diversity conditions were performed by dive'siry o de iog rTodmeras? 1993a,b).This evaiuationdifferentiatesih€ diveBiry sequedce of studied commurities based on dominant lercus lare species. The Hill diversity value is seisitive on rare species at low scale parameter value and it is sensitive on dominant speciesat high paranet€r value. Ta$le 1- T[cpoolcd foqr...y values otth sldi€s in tlc 200 2ox2o 3n quadntctet shrdyplol Dah of22 rarc $ccies, whoscloolcd f@qu;cy valucav.r. tclow l0 ae not indicaiod.Th€ uscd nonetrclahnc h aftor siron ( 1992). SPECIES Conplolh.cirn lfi8eta r 2 0 0 0 0 0 0 0 0 0 12 0 5 0 3 1 o 0 0 0 24 9 3 0 0 Equisetun rcn*istimrn l Euphorbia ctpaissi$ 1 3 0 59 2 12 r24 10t t1 5 3 0 5 0 0 0 9 STUDY ?LOTS 4 5 6 7 0 1 2 26 0 0 21 0 3 6 o Ilt 0 0 1 2 t21 0 o o 17 4 5a l 0 0 0 0 0 0 0 5 0 29 0 63 0 3 2 1 r 3 0 a 2 0 0 2 3 0 0 0 18 o 24 55 0 2 t61 22 0 1 J 3 0 0 12 t3 3 PLllochdr. sqrd,rdd 5 0 2 3 3 9 a 3 6 53 0 0 2 0 0 2 o 0 0 l l 0 3 ? t6 23 2 a 0 0 0 1 0 '1 A 12 51 5 l3 2 0 23 2 2 7 1 132 0 0 3 l5 NSCIA 27 t9 l5 33 1 6 2 5 23 R€sult Ordi ation and cfassification of st dy "ai The ordinationolthe sevenstandsaccordingto the pooledfrequencyof occuring speciesdefineda -...\ successional sequence, rhat did not differ \, considembly from a theoreiical or assun€d successionalorderj if the arch etrecl of the ordina' tion methodis laken into coffideralion;or y the plot 4 doesnot fit well into thesequence (Fig. 1.) The selected representaiive subplots of the standswere sepantedrather well on the PCoA scatiergran(Fig. 2.), and the classification resulis confirmthe togethemess of thes€subplots(Fi8. 3.). Thereare very few overlapsb€lweenthe diffefent stands, consequentlybetween ihe successionaL subplotr stages.Ohly the subplotsof dre shrubbyhab(al r18.2.Pnncrpdcod'lDarcM(lyss ol Llie'eDresentative o f r h cr J d . c d , k n d , L . i n c se d c i ' c lrch c! u b p l o G bcloncrns (plot 6) form l1o separalecluster and they are to lhc ,omochstcr,indiortcdby rta6 ir Fig,L Th€ d!,hcd positionedfa. from eachotheron the cenlralpa|l of linessisr lossibl. orddrsof rho samplcsco.sid.dng th. rrchoffc.tof llr.ordin iotrmcLhod. the scattergram. The vegeiationofthis standis very heierogeneous. The identified patch types mny rcpres€nidifferen!successional stages,and can be area.Insidetb€ denoledclusterthe two siudy sites (Fig 3.). orderedin a feasiblesuccessional sequence that is areseparated parallelwith the sequence (Fig. of tbe otherstands Diversit, 2.t. lig. L P ,cipalcoordlrar.nalysn oftlc shdy plolsaccodtug ro thc poolcdfrc'$ancic,oftho occnirinsspccica. ror tho a.srirtion ofthc atudylioh indicatcd by nuhlcs sc. thc "Sludy d.a md Mctlods . Altogether76 specieswere recordedin the 7 siands.The frequencies of speciesare presented on Table 1. (Datfl of 22 rare specjes,whosesummed fiequencyvalueswerebelow10 arenot indicaled.) Three groups of standswefe identifLedby diversityorderingr(i) an initial, opeograssland type with low diversity;(ii) a grasslandwith ftediuln diversitytthe forestherb layer also belongsto this cAtegory;(iii) the nrosl djverce,closedgrassland, andthe shrubbyhabitat.Div€rcityprcfiles of these groupswere.clearly separaled. In the first group the diversity of plot 1 is unambiguouslylower, than that of the plot 4. Howevernr the nediurn diveNily groupthe profile ofplots 2,3 and7 crosseachother,this meanstha! thesethreeco nuniiiescannotbe orderedsimply by th€irdiversity.At lowerscalepalameter,ihat is, regardingtherarespeci€sthe diversityofthe plot 7 is the highest,but the oppositeis right ai the higher valueof scaleparameter- iegading the donr;nant Discusslon Some subplots do not join to any group; ihey represent unicalpatchrypes.On Fig. 2. the subplots According to the tradilionai inte{pretations ofplot 1 andplot 4. are encircledtogether,tlis lwo (Zsalr, 1943 and Magyar, 1960) the feasible plots represent the initial stageofsuccession in this successional sequenceof siudied stands is the 6 TISC]A 27 communiry,an amual grassland, which is regarded by thefaditional conceptls the firsr stageof sandy successioncannotbe found in K6leshalomsite. Festucdwgi ata is rhe rl]ost frequentspeciesm the very open,pr€s'rmablypioneerstages(plot I and plot 4). It is remarkable,that poplar can colonize even in the very initial forn ol Festrcetut vagrrdla?oommunitythe (plot4). Fekete(1992)presenleda differentconoeptof primary succession on sandi the startingpoint is Dsually Festucetun wgitlatae community. The succession terminatesar the Juniperc- Papuletun along nany lines; this comnunity corresponds to the real forestsin this sere.The sleppemeadow (closed grassland)does no! fit to the proneer grassland'' shrub line and this makesthe graph reticulaied in thisxerosefies The fesultof analysisreflectingto the intemal Fis.3. DcndrogBboflh. Elrcscnhliv. sulploh oftho studied variabilityof the stands(Fig. 2.) suppo( rhis con$ande.Tlc a6 indi.rtc thc &lcvst clust rs,subflorsof ceptfathef thanthe traditional,lin€ar successional whicl oro cnciicl.dby lir$ in th. ordin ion diasmmD sequefice.The starling point is undoubt€dlythe complexoflhe plot I andplot 4; plot 2, 3, 5 and 7 e dpornis of successrol. followirgalmosi baresand(plat 1) "- Ferhrenml could be nlrernarrv€ yaginatae danubiale (plot 2) - Fasfircetn Besides this ar alternntiveexplanationof the \)aginatae ltipetosrtl sabulosae (plar 3) - open scailergra is alsopossibler(I-a) - 3 - 2 - 5 - 7 plot couldbe rccognizedaloflg an arch in the grassland with small poplartrees(plot 4) - closed sequence Fig.2. gLassland (plot 5) - shrubbyhabital Glol 6) Tbe subplotsof plo! 6 may rcpresen!intermepoplarforest(plot 7). On the PCoAscattergram of standsthis soquenoe oaDbe mofe ol lessrecogujzed diate stngesof altemativepathwaysleadingfrom (plot 1,4)to the closedgrassland if consideLing!h€ arch effect of the ordination the opeflgrasslnnd nethod (Fig,1.).The arch effeci is strongerwhen Glot 5) andpoplarfol€st(plot 7). At th€ samerime lypical patchesof plot 6 the ordinaredsampleshavefew speciesir conrnlon the subploisrepres€nling (Mazzoleniet al., 1991) The Brcneho tcctonutt are ordered along aD arch, parallel with the nlenlioned(1-zl)-. 3 " 2' 5 - 7 plot sequ€nce in the scattergran.That is, aherfiativeinteryretations are possiblewhefl t0,fu9 to identiry the successronaL sequence on the bas€of ihe ordinationresult.The used methods are suitable to arange rhe communitiesaccordhg to their complexjry,and .n speclescompositionbul it doesnot mean,that the identifiedsequence exactlycorresponds to ihe r€al ,t historical de\,elopment questior, ofthe stands in = ln a hilly areatherearc considerable diff€rences in the physico-clremical condiiionof the soil (i.e., humusandnitrogencontert,watercontentand soil granulesize)ai ihe top ofthe sandhills, in the wind groovesbetweenthem and at the relatively flat afeis (K6rm6czi, 1983). Such environnenral - - - - r l l ! o.o r.z ,.a r.6 610 differences mry cause the development of ai altemativeerdpoinisofsuccession, accordingto the concept of Drake (1990), who resard rh€ Iis 4. Divc$iry ord€nngof thc ,tudy piots.lccoiins lo llic environmentrlgrrdientsas | filter definug which poolcd fEqnency valu$ of th. .ccunis sp€cies.Ior setofspeciesis pemissibleto colonjze. d€scriptionoftn. shdy plots indicaled!y dunbc.s scclxe .studyd.a md M.tlods '. The specles diversityjs the highestin the closed grassland(plot 5) and in the shrubbyhabitat(plot T]SCIA27 6), but presumablebecauseof ditrerent reasons ln and L. call6 providedhelptul revjew of an earlier the oaseof closedgrasslandconmunity $e nore venion of ihe manuscript.The work was founded favourable environmental conditions of the slight by the OTKA V3 grantNo. 160?. wind srove habitat (higher soil humidity, lower of wind effectandinsolation)allow the coexisrence The shrubbyhabitatshaved€veloped more species. - In: Kikkava, J. succession. in the upper slope of a sand hil1, and lhe sbrtbs Andes6n,D.r. (1986):Ecolosical D.J.(.ds.):CommunityE olosy,P.tic Md md Andcsn, stand separatelyproviding a wide varieiy of Proccs,Blrckw.ll sci. hbl., cdton, Austmlir,pp. 269' microhabitats from the open, dry patchesbetween 235. the shrubsto the shadedsitesunder their canopy, Drakc,J-A..(1990):Thc mechrnicsaf coDmuity asshbly ud $cc.$ion - J.Th.or.Biol.l,/2,213-233. were ihe soil is coveredby litter. Here the habrtat Iokete,6. (1992):Thc I'olislicvi€w of$cccssionrconsidcred.causes highernumberof speoies. heierogeneity P acsenyr rckcld,c.,Trba,Z, ald Mclk6,!, (1933):Backsroundlrocescs Studyirgsimilar sandyvegeLalion at tho populrtionl.vcl durinBsucc.ssionin Snssl ds o. (1981) found that the speciesdiversiE,was the sand.- Vcgclalio//,33 .ll. is iD ihe of successiot, that lowestin the last stage Fckere,C., Tnba,z. ud Picsanti, I. (1930):Niche studicsoD forest.In presentcasethe diversityof foLestberb sono phut sp.ciosol r srNslandconNnity. VU - Acla B.L tl\ng, 26, 2At-297, layer is high if resardingthe rare sPecies,bu! lower if regardinglhe dominan!sPe- Hnrsitai,z, (1940)i NogykdrijsDijvaDyviligaIL A honoki oonsiderably (ll r lifc of NagykiJrais. IL Plst niiv6nyeiivclk@ctck, cies (Fig. 4.). This communitydoes not evolve coodrnitioson sand),- Bot$ikai. Ki'21.n6nyek.J7,205slage bur aulomaticailylrom a pre"iousgrassland 2,44 the light shotage(theshadingeffeotofthe growing I5rd6n,C,,, G.ll{ L, Md Marg6cziK. (19t3): An! conmunity qospositiotrin r Husr on $cc.$ion!l sanddunc uca, of treecanopy)driv€sits developmenlSo,because the considerable change ill environnenta] Komor, A.M. (1863): Pn$zorloboD d( Doraul.ndar' conditionsa different speciesPool is allowed !o b.t*.on tha zonaiionof,Ndy K6rn6czi,L, (r9S3)rCotrclarion colorize(of. Drake,1990) gRsElNdaDdthr physi.o-chcfiical conditionoflhoi! soilin The vegetaiionof plot 4 providesan eviderce Bugrc,-Acir Biol. SzcAcd29, I l7-127 possrble is of opengrassland rharlhe leforestation ^ l)I d l J L i ! . I i , , - ^ k r d { r n . r . K . d d 6 . B u t { e r l M0srrr,P.Ll96O but canno! be found cLear tfansitional slages MMzolctri,s., FLcnch,DD Nd Milos,J. (1991)rCo'nrralivc sludyof chssification!trd odin ion mclhodson succet and folest herb liyer betweenthe open grass)and riond d!t!, - Coonosc! d,9l"l0l. amongthe studiedPlots. Molhir, E, N, lud Nosok,t,N,r Sidtial !6ce$c6 it r s6slmd of Pfesentpaper ref€rsto field investigations cunnunit,l, -Aota Bol,Acad.sci. H!tr8,?J,339-348, one single time Point. Direct evidecesabout the Podahi,J. (1988)rSYN"TAX lll User's h.nurl - ,{bstractr L Bot.niq /r, Supplomont pathwayscan be drawnuP oDly aftel successional tuacs,hyi,l, (1981)iChdrsosin t[e divcEity of l[e legclalior severalyears study period on Pe|m rehl plots dldrg srcco$iotr-ach Bot,Acad.sci. Hung.27,139r93, Resultof sucha long teffi sludycaMedout on the Ropaier,R.(ler8), Az Alliild ndvantffjldDjzij.lhm.. - Edesz€ti vegetationof Hungariansandyarcashavenot beeD l. (1992):A nrsy.loszasi .dany.s n6h hatiroz6F, of the Sinon, publjshedyet. Althoughbetterunderstanding (D.t ihinlliotr book of tb. Hnhgdridr nsculd n66), successionis very importnnt from the natue T.nkdnyvkiad6, Budap.ri consewatiorpoint of wiev as well, becausethe szollit, Cy, ud Brdl'a, S. (lr9l)r Paltetuft.ly* or dolonit. sr.ssltrd coho!tritics u,inc nforhation thcory nodch ' of consewation activiryhaveio aim at naintainance Abstrlct!Botrni.! /J, 4r-60. stagesclose10 each other if it is 'L6tl1nfr6sz,a, (1993n):N!CoSA l.oi Nr'nbcr ctu"clri lor all succ€sionftl possible.BLtt ih€ nininal spalial scale of the co'ntunitt !tudics and oll$r ecologicalarulicahons ' AbstactaBot ica I/,233-237. rsno'known) errs dell successionalprocesses T6rllnaraszB. (19936.):Divoid 1,50:A ptogLaBfor diw6ity Acknowledgem€nt was I an thanttul to L. Call6:this investigation b) hin. r,d carnedoul on rn€snrdysi'esselecred joint io his projecton communityorganization dufS. Bafha on sand.B. T6thm6r€sz, ing succession orderina- Tisoia27,33-44. Txellcr,P.T.lnd rlabu, K.A. (l9el):aphr succcsioDsrad'odt 9r,5?-63. nr.6ts s.g.b&sl/gF$ ecosyst€n.Vegetatio viiish, K. (1936): Th. cficct of hcrbicideson v.g.t tior a Brltivarialcstudt- AbstractaBottrica 10,31? dynamics: 340. (Thc vcscZsoll,J. (1943):A ecnlendroi szisetn6v6nytrkamja. tation oa szhtcrdrc Islmd). Indcx ltoni Bolarici Univ. T]SCIA 27