Table Of Content

© Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at The role of leaf anatomy and tannins in litter decay in a tropical stream El rol de la anatomia foliar y de los tanninos en la descomposición de la hojarasca en un arroyo tropical Astrid RIEMERTH, Maria GUSENLEITNER, Gerhard DRAXLER, Wolfgang WANEK & Fritz SCHIEMER Abstract:LeaflitterofninetreespeciesoccurringintheriparianforestofQuebradaNegra,afirst-ordertropicallowlandstream inCostaRica,wasanalysedbyplant-anatomicalandphytochemicalmethodsinordertotestthehypothesisthatr-strategistsin- vestlessindefenceagainstherbivoresandpathogensthanK-strategistsandthereforehavefasteraquaticdecayrates.Inaparal- lelstudyonleaflitterdecomposition,detritusformationandcolonisationbymacroinvertebrates,TSCHELAUT (2008b)showed cleardifferencesbetweenplantmaterialoriginatingfrom4speciesoftreescommonintheriparianforestofthestream.The4 speciesrepresenteddifferentlifehistorystrategiesoftrees,fastgrowingpioneerspecies(forsimplicitycalledr-strategists)(Aca- lyphadiversifolia–Euphorbiaceae,Cecropiaobtusifolia–Cecropiaceae,Tetrathylaciummacrophyllum–Flacourtiaceae)andaslow growingclimaxspecies(K-strategists)(Sloaneamedusula–Elaeocarpaceae).Inadditiontothese4plants,5additionalspecieswere analysedinthepresentstudyinordertoobtainabroaderpresentationofdefencemechanisms.Theplantsdifferintheiranatom- icalstructuresandcontentoftanninsoftheleaves.Anatomicaldifferenceswerequantifiedby13parameters,e.g.proportionof sclerenchymaandcollenchyma,onanatomicalsectionsandcombinedina“toughnessindex”.Theresultsaresetrelationtothe observeddecayrates.Thecomparisonshowsthattheamountoftanninsinfluencesthelitterdecayduringthefirstdayswhereas anatomicalstructuresplaythemainroleduringthewholedecompositionprocess. Keywords:litterdecomposition,tannins,plantanatomy,toughness. Resumen:Lahojarascade9especiesdeárbolespresentesenelbosqueribereñodeQuebradaNegra,unarroyodeprimerorden enlostrópicosbajosdeCostaRica,fueronanalizadaspormétodosanatómicosyfitoquímicos,paraprobarlahipótesisqueloses- trategasrinviertenmenosenladefensacontraherbívorosypatógenosquelosestrategask,porlotanto,tienentasasdedescom- posiciónmásrápidasenelagua.Enunestudioparalelosobredescomposicióndehojarasca,formacióndedetritusycolonización pormacroinvertebrados,TSCHELAUTetal.(2008b),mostróclarasdiferenciasentreelmaterialvegetalprovenientede4especies frecuentesdeárbolesenelbosqueribereñodelarroyo.Las4especiesrepresentabandiferentesestrategiasdehistoriadevida,es- peciespionerasderápidocrecimiento(porsimplicidadllamadasestrategasr),(Acalyphadiversifolia–Euphorbiaceae,Cecropiaob- tusifolia–Cecropiaceae,Tetrathylaciummacrophyllum–Flacourtiaceae),yespeciesconunclimaxdelentocrecimiento(estrategas K)(Sloaneamedusula–Elaeocarpaceae).Ademásdeestas4plantas,5especiesadicionalesfueronanalizadasenelpresenteestu- dio,paraobtenerunarepresentaciónmásampliadelosmecanismosdedefensa.Lasplantasdifierenensuestructuraanatómicay contenidodetaninosenlashojas.Lasdiferenciasanatómicasfueroncuantificadaspor13parámetros,porejemplo,proporciónde esclerénquimaycolénquima,encortesanatómicosycombinadosenun“Indicededureza”(ToughnessIndex).Losresultadoses- tablecenrelacionesparalastasasdedescomposiciónobservadas.Lacomparaciónmuestraquelacantidaddetaninosafectalades- composicióndelahojarascadurantelosprimerosdías,noobstantelasestructurasanatómicasjueganunrolprincipalduranteto- doelprocesodedescomposición. Palabrasclave:descomposicióndelahojarasca,taninos,anatomíavegetal,dureza. Introduction fungiwhichformsamajorfoodsourceforinvertebrates and detritivorous fish. The shredding of plant material Inloworderstreams,allochthonousmaterialgener- by certain macrozoobenthic species transforms coarse Stapfia88, ally plays a major role as the energetic basis of the particulateorganicmaterial(CPOM)tofineparticulate zugleichKatalogeder oberösterreichischen stream ecosystem. Allochthonous material, mainly in organic material (FPOM) and enhances the microbial Landesmuseen NeueSerie80(2008): theformofleaflitter,developsabiofilmofbacteriaand breakdown and decomposition processes (VANNOTE et 467-484 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at of the leaves (combined to a “toughness index”) and theirtannincontent.Overallanatomicalandchemical defencemechanismsrelatedthisfindingstotheleaflit- terdecompositionrateswereobtainedbyTSCHELAUTet al. (2008b). Material and Methods Studysiteandselectionofspecies The plant material was collected at the Biological Field Station La Gamba, Costa Rica, located at the Piedras Blancas National Park (8°27‘–8°41‘N and 83°15‘–83°45‘W), especially in the riparian forest of Fig.1:Positioninthelifehistorystrategycontinuumoftheninespecies arrangedaccordingtoaselectedrangeofcriteriaproposedbyWHITMORE QuebradaNegra,afirst-orderlowlandrainforeststream. (1993); Ithasitssourceintheprimaryforest(~180mabovesea Table1:Valuesofdailylitterdecompositionrate(k)fromfourdifferenttypes level) of the Esquinas rainforest (TSCHELAUT et al. ofleafpacks(Acalyphadiversifolia,Cecropiaobtusifolia,Tetrathylacium 2008a). macrophyllum,Sloaneamedusula)exposedintheQ.Negrafor28days, obtainedfromtheslopesofregressionequations,betweenlogW andt InadditiontofourplantsusedbyTSCHELAUTetal. t (TSCHELAUT2005). (2008b)intheirlitterdecompositionexperiments,five morespecieswereanalysedinordertoobtainabroader taxon W W t k 0 t representationofthetreevegetationinthegalleryfor- Acalyphadiversifolia(Euphorbiaceae) 3,16 1,045 28 0,0395 Cecropiaobtusifolia(Cecropiaceae) 3,295 1,71 28 0,0234 est along the river Quebrada Negra. The selected nine Tetrathylaciummacrophyllum(Flacourtiaceae) 3,175 0,88 28 0,0458 specieswereordinatedalongalifehistorystrategycon- Sloaneamedusula(Elaeocarpaceae) 4,67 3,73 24 0,0093 tinuumdefinedaccordingtocriteriaproposedbyWHIT- MORE (1993), e.g. growth rate, reproduction, spread of al.1980).Tanninsandanatomicalleafstructuresareef- fruits and seeds, quality of wood, requirement of light fective defence mechanisms against herbivores. It can and lifetime. The continuum ranges from fast-growing be assumed that leaf palatability and litter decomposi- pioneerplantstoslowgrowingclimaxspecies.Forsim- tionarestronglycorrelated(e.g.SCHÄDLERetal.2003). plicity the terms r- and K-strategists are used Leaflitterdecayratesofdifferenttreesfromtheri- (MACARTHUR&WILSON1967). parianforestofQuebradaNegraselectedaftertheirlife Figure1outlinesthepositionoftheselectedspecies history strategy were analysed by TSCHELAUT et al. inalifehistorystrategycontinuum.Acalyphadiversifolia, (2008b). The four exposed plant species can be either Cecropia obtusifolia, Guatteria chiriquiensis, Myriocarpa put into the “fast” or “slow” decomposition group ac- longipes and Tetrathylacium macrophyllum are termed r- cording to the classification of PETERSEN & CUMMINS strategists, while Apeiba tibourbou, Luehea seemannii, (1974).Acalyphadiversifolia(withadecayrateof0.040), SloaneamedusulaandVirolakoschnyiareK-strategists,al- Cecropia obtusifolia (0.023) and Tetrathylacium macro- thoughthedistinctionisnotsharp. phyllum (0.046) can be assigned to the “fast” group. Leaf material of the nine species was collected in Sloaneamedusula(0.009)canbeplacedintothe“slow” February 2005. Attention was given to collect leaves group.Duringtheexperimentdurationof28daysA.di- neartoabscission.IncaseofVirolakoschnyi,whichcasts versifolia and T. macrophyllum lost about 70% of initial its leaves in dry seasons, only relatively young leaves dryweightandC.obtusifoliaabout45%.S.medusulalost couldbeharvested. lessthan20%ofinitialdryweightduring24daysofex- posureintheriver(Table1). Leafanatomy Inordertotestthehypothesisthatfast-growingpi- Material was fixed in 30% ethanol. Five regions of oneer species make a smaller investment in defence the leaf were processed by anatomical sectioning tech- against herbivores and pathogens than slow-growing niquesandthesectionswerestainedforfurtheranalysis. plants (climax species) and therefore pioneer species Thesafranin-astrabluecolorationandtheauramin-mu- havefasterdecayrates,fivemorespecies(Apeibatibour- cicarmincolorationareteststodistinguishbetweenlig- bou – Tiliaceae, Guatteria chiriquiensis – Annonaceae, nified and non-lignified cell structures (BRAUNE et al. Luehea seemannii – Tiliaceae, Myriocarpa longipes – Ur- 1999;FärbevorschriftenFirmaCHROMA1962).Further- ticaceae,Virolakoschnyi–Myristicaceae)wereanalysed morethelignin-specificphloroglucin-HCltestandthe inthepresentstudywithrespecttoanatomicalfeatures lipid-specific Pearse reaction were applied. All plant 468 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at anatomical analysis were carried out at the Depart- Results mentofEcophysiologyandFunctionalAnatomyof Leafanatomy Plants,UniversityofVienna. C.obtusifoliashowlongunicellularhairsontheun- Withthestainedcross-sectionpreparations,theat- dersideoftheleafbladeandcone-shapedandglandular tempt was made to quantify and summarise the plant hairs on the upper side. The cross sections of the mid- anatomical parameters considered to represent defence vein vascular bundles show dispersed lignified struc- strategies for herbivore predation. The nine following tures.Sclerenchymacellsandcrystalsarerare,secretion measurable parameters were combined in a “toughness canalsfrequent.Theside-veinsintheleafbladearesur- index”(TI): roundedbyabundlesheath,whichprotrudetoepider- •meanrelativeleafdryweight malcells(seePlate1). •areaofthemid-vein M.longipes(Urticaceae)exhibitsgreatvariationin •sizeofepidermalcells the leaf size between sun and shade leaves. The upper •cuticleformationandcuticleridges surfaceoftheleafbladeisglabrousbutwarty.Oneofthe •amountofside-veins typical characteristics of Urticaceae are cystoliths as •lignificationsinthemid-vein supplements of the epidermal cell wall incrusted with •lignificationsintheleafblade calciumcarbonate.Sclerenchymacellsareabsent.The •amountofsclerenchymainthemid-vein vascular bundles occur in form of a ring. In the •amountofcollenchymainthemid-vein parenchyma,secretioncanalsandcrystalsarefrequent. Inordertocalculatethe“toughnessindex”theindi- Side-veins have a prominent bundle sheath without vidualresultswereconvertedintorelativevaluesfrom0 sclerenchymacells(seePlate2). (lowest toughness value) to 100 (highest toughness T.macrophyllum(Flacourtiaceae)areglabrousonthe value). The scale finally was divided into three equal top.Theundersideiscoveredwithunicellularhairs.The sections and allocated a factor from one to three. The mid-veinshowspronouncedlignifiedstructuresthatoc- indexisthesumoftheresultsfromeachsectioninor- cupy almost half of the mid-vein area. Sclerenchyma dertoobtainasinglerobustvalue. cellsformacircleseveralcellswidearoundthevascular Measurementsoftannins bundles. The spongy parenchyma in the leaf blade is looseandcontainslargegasspaces(seePlate3). To quantify the amount of tannins in plant tissue, twocommonmethodswereapplied.Totalphenolicson G.chiriquiensis(Annonaceae)leavesarecoveredby freshanddriedleafsamplesweremeasuredbytheFolin hairs on both sides. Especially on the veins the hairs Denis Method, first described by SWAIN & HILLIS grow very densely. The cross section of the mid-vein (1959) and modified by MARTIN & MARTIN (1982), shows a ring-shaped sclerenchyma around the vascular HAGERMAN(1988),COLEY&BARONE(1996)andHÄT- bundles.Thespeciescontainscrystalsinlargeepidermal TENSCHWILER et al. (2003). The Folin Denis Assay de- cellswhichisatypicalfeaturefortheAnnonaceae.We tectshydrolysableandcondensedtanninsandalsonon- foundsomespecialcellswithlipidstores.Theside-veins tanninphenolicsandotherreadilyoxidisablematerials have prominent bundle sheets of sclerenchyma cells such as ascorbate (HAGERMAN & BUTLER 1991). The thatdonotreachtheepidermis(seePlate4). determinationofcondensedtanninswascarriedoutby A.diversifolia(Euphorbiaceae)haslonghairsalong thebutanolHClmethod(PORTERetal.1986)ondried the veins on the underside of the leaves. Older leaves leafmaterial.Forbothassays,aqueousacetonewasused are heavily covered with epiphylls. Some solitary cells as extraction solvent. Acetone is an effective solvent ofsclerenchymalieoutsidethevascularbundles.Crys- whichinhibitsinteractionbetweentanninandproteins talsareveryfrequentinepidermalcells,atypicalattrib- andthuspreventstanninfrombindingtoleafproteins ute for many species of the Euphorbiaceae, although duringhomogenisation(discussedinHAGERMAN1988). theyoccuronlyrarelyintheparenchymaandphloem. Fresh samples were processed at the Biological Field Largecrystalsformadistinctivefeatureoftheleafblade Station La Gamba, Costa Rica. Dry samples were (seePlate5). analysed at the Department of Chemical Ecology and Ecosystem Science, University of Vienna. At A. tibourbou (Tiliaceae) leaves are barely covered leastfivesampleswereanalysedforeachparameter.The withhairsontheupperside,whereasontheunderside, statisticalanalysisofdatawasperformedwiththesoft- dense stellate hairs without ramification occur. The wareStatgraphicsPlusforWindows5.0.Statisticalsig- crosssectionsofthemid-veinsshowaring-shapedvas- nificancewastestedwithone-wayANOVAandScheffe cular bundle with a prominent bundle sheet of scle- orLSDtest(p<0,05). renchymacells.Intheparenchymaoutsidethevascular 469 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at bundle, secretion canals and crystals were found. Be- tals.AsdiscussedinleavesofA.tibourboubigidioblasts neath the epidermal cells of the leaf blade, large id- beneath the epidermal cells of the leaf blade are fre- ioblastsarefrequent(seePlate6). quent(seePlate8). V. koschnyi (Myristicaceae) shows a significant S.medusula(Elaeocarpaceae)haslargeleaveswhich heart-shapedleafnode.Onbothsidesaswellasonthe carryhairsontheveinsonbothsides.Theuppersideof veins,star-likehairscanbefound.Anatomicalsections theleafbladefeelsfurry.Anatomicalcutsshowthere- of the mid-veins show a ring-shaped xylem. Phloem is lation between the extremely prominent mid vein and arranged in small groups surrounded by parenchyma theblade.Vascularbundlesformaringwiththreehor- cells. In the outer parenchyma, a disconnected ring of izontalstrataofxylemandphloeminthecentreofthe sclerenchyma cells occurs. Further sclerenchyma sur- ring. The upper layer is inversely orientated. A domi- rounded by phloem can be found in the inner nant element of stability is the extensive development parenchyma.Cellswallsoftheparenchymaarepartial- of collenchyma. In the collenchyma and parenchyma ly lignified as shown by fluorescence microscopy. Epi- cells,cubicandcolumnarcrystalsoccur.Inthexylemof dermal cells of the leaf blade are homogeneous in di- thehorizontalbundles,secretioncanalsofvarioussizes mensionandform.Cellsofthepalisademesophyllhave occur. In the leaf blade, the high number of side-veins an isodiametric form. Among the palisade mesophyll, withbundlesheetsthatreachuptotheupperepidermis larger cells with crystals occur. The spongy mesophyll is conspicuous. Analyses by fluorescence microscopy ofleavesofV.koschnyiisrelativelydense.Betweenpal- showed that stomatal cells of S. medusula are lignified. isade and spongy mesophyll idioblasts which have a Sometimessecretioncanalsareembeddedinthemeso- positive reaction with gentian violet are conspicuous phylloftheleafblade(seePlate9). (seePlate7). Toughnessindex L. seemannii (Tiliaceae) leaves are clearly charac- Theproportionsofthevariousmechanicalstabilis- terisedbytheirbrownundersidewhichiscoveredbya ing elements in the mid-vein of the leaves of the nine tomentum of hairs whereas the upper side is rarely pu- speciescomparedaregivenintable2. bescent.Thevascularbundleofthemid-veinsurround- ed by a prominent ring of sclerenchyma is made up of Thecharacteristicsoftoughnessforeachspeciesare twosemicirclesofwhichthesmalleruppersemicircleis showninfig.2and3. inverselyorientated.Crystalsareabundantinparenchy- Thetoughnessindex(TI)valuescalculatedforthe macellsneartheprominentsclerenchyma.Sectionsof nine species ranged from 18 to 36. The highest values theleafbladeshowthetomentumofmulti-cellularhairs amongtheK-strategistswerefoundforL.seemanniiand onthebottomside.Lignifiedstructuresoccurinvascu- S. medusula (36) followed by A. tibourbou (27) and V. lar bundles of side-veins. Cell walls of mesophyll cells koschnyi(21).Amongther-strategiststherangewas17 arepartiallylignified.Bleachedanduncolouredsections to23(T.macrophyllum:23,C.obtusifolia:22,A.diversi- show that the mesophyll of the leaf blade is extremely folia:21,M.longipes:18,G.chiriquiensis:17).Thedif- densely packed. The cross section of leaf blade has a ference between the two groups is statistically signifi- characteristicundulatingform.Attheconstrictions,lig- cant. nified structures occur which reach from the upper to A correlation coefficient between toughness index the lower epidermis. Above the lignified structures, andlitterdecayratesofA.diversifolia,C.obtusifolia,S. largeidioblastscanbefoundwhichcontaincubiccrys- medusulaandT.macrophyllumisR²=0,632. Table2:Summaryoflignifiedstructures,amountofsclerenchymaand Contentoftannins collenchymainpercentofthemid-veinarea. TheFolinDenismethodforoveralltanninsonfresh species lignifiedstructures amountof amountof leaves of the nine species indicated concentrations of sclerenchyma collenchyma 1.9% to 45.9% tannic acid equivalents on dry weight C.obtusifolia 13,1% 0,2% 12,5% M.longipes 5,2% 0% 16,3% (TAE). Among the K- strategists the highest amounts A.diversifolia 18,6% 2,9% 8,9% weremeasuredforV.koschnyi withvaluesup to45.9% G.chiriquiensis 26,2% 12,9% 0% TAE (s = 10.6, n = 10), whereas the lowest level was T.macrophyllum 43,6% 16,6% 8,2% foundinleavesofL.seemanniiwith13.7%TAE(s=1.9, A.tibourbou 29,3% 8,5% 10,9% n=10).Thelowestlevelsforr-strategistswerefoundin V.koschnyi 42,4% 6,5% 2,3% G.chiriquiensiswith1.88%TAE(s=0.01,n=15)and L.seemannii 43,6% 21,3% 6,9% M.longipeswith2.30%TAE(s=2.4,n=10).Among S.medusula 45,2% 8,3% 16,5% ther-strategists,amaximumleveloftanninswasfound 470 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at Fig.2:Summaryofthecharacteristicanatomicalparametersforthefastgrowingspecies.Allmeasuredvalues wereassignedfromceroto100.Scaleisdividedintothreeequalrangesforcalculationofthe“ToughnessIndex”. inleavesofC.obtusifoliawith20.6%TAE(s=2.5,n= 51.2%TAE(s=12.0,n=10)followedbyS.medusula 10). The difference between the two groups is statisti- with 36.0% TAE (s = 17.9, n = 10). Similar to the callysignificant. analysisonfreshmaterialM.longipesandG.chiriquien- In comparison to measurements on fresh material sisshowedthelowestlevels(1.6%TAE,s=1.7,n=10 the same method was carried out on dry material. The and 4.9% TAE, s = 1.9, n = 5, respectively). Again, a amounts of TAE range between 1.6% and 51.2%. The differencebetweenthetwolifehistorygroupsisstatisti- highest level was found in leaves of V. koschnyi with callysignificant. 471 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at Fig.3:Summaryofthecharacteristicanatomicalparametersfortheslowgrowingspecies.Allmeasuredvalues wereassignedfromceroto100.Scaleisdividedintothreeequalrangesforcalculationofthe“ToughnessIndex”. Adifferentpatternwasobtainedwhenusingthebu- tanol HCl method with dry leaf material. The highest concentrations on condensed tannins were found in Table3:Resultsoftanninmeasurements;Allovertannins(%TAE)measured leaves of V. koschnyi with 13.3% cyanidin equivalents bytheFolinDenismethodondryandfreshsamplesancondensedtannins ondryweight(CE)(s=1.5,n=10)closelyfollowedby (%CE)bytheButanolHClmethodondrymaterial. C.obtusifoliawith11.3%CE(s=2.3,n=10).Thenext TAEfresh TAEdry CEdry range is formed by S. medusula (7.4% CE, s = 0.7, n = C.obtusifolia 20,6± 2,5 17,7± 3,9 11,3±2,3 10) and L. seemannii (5.9% CE, s = 0.2, n = 10). The M.longipes 2,3± 2,4 1,6± 1,7 0,5±0,8 amounts of condensed tannins in A. diversifolia, G. A.diversifolia 10,6± 2,5 27,8± 9,4 1,9±0,3 chiriquiensisandM.longipesareclearlylower.Inthecase G.chiriquiensis 1,9± 0,6 4,9± 1,3 0,6±0,3 ofT.macrophyllumandA.tibourbou,singlevalueswere T.macrophyllum 8,4± 1,9 7,3± 2,1 0,1±0,1 belowdetectionlevel(Table3). A.tibourbou 19,6± 6,5 27,4±12,7 0,0±0,0 V.koschnyi 45,9±10,6 51,2±12,0 13,3±1,5 AcomparisonbetweendecayratesofA.diversifolia, L.seemannii 13,7± 1,9 9,6± 3,4 5,9±0,2 C.obtusifolia,S.medusulaandT.macrophyllumandthe S.medusula 25,8± 4,4 36,0±17,9 7,4±0,7 overalltanninlevelsmeasuredbyFolinDenisonfresh 472 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at material shows a very high correlation (R² = 0.988), whereas the condensed tannins measured by buthanol HClmethodhaveaweakercorrelation(R²=0.614). Analysesofthetanninlevelsduringdecomposition showthatoveralltanninsaswellascondensedtannins areleachedordecomposedafteratleast25daysofex- posure in the river. Tannins of S. medusula leached or decomposedmoreslowlythanthoseofA.diversifolia,C. obtusifolia and T. macrophyllum, in which the tannins werealreadygoneafter14daysofexposureintheriver. Discussion Anatomicalparametersaredifficulttoquantifyand compare, because morphology and leaf anatomy are highlydiverse.Thetoughnessofleaveshasbeenanissue of many studies but there is no consistent definition of Fig.4:Correlationbetween„ToughnessIndex“anddecayrate. thistermandmethodforassessments.CHOONG(1996) used the epidermis, thick walled cells beneath and the cuticle,WRIGHT&WESTOBY(2002)theconcentration of cell wall, vascular tissue, fibre or sclerenchyma in a leafasparametersfortoughness.Itwasoneoftheobjec- tivesofourstudytoquantifyandcombineabroadrange ofplantanatomicalcharacteristicsthatarethoughttobe important in leaf litter decay in an “toughness index” (TI).Ourindexshowsastatisticalsignificantdifference betweenplantsofdifferentlifehistorystrategiesandcor- relateswellwithaquaticdecayrates. Tannin levels after the Folin Denis assay showed a clear(R²=0.828)correlationbetweenthevaluesmeas- uredondryandfreshmaterial.Thismeansthatarough estimateontanninlevelscanalsobeobtainedformdry material extractions. Tannin levels measured by the Folin Denis assay (TAE) on fresh material vary from 1.9% to 45.9% TAE. BALDWIN & SCHULTZ (1988) Fig.5:Correlationbetweenallovertanninsanddecayrate. foundvaluesofoveralltanninswiththesameassayun- der17%,whereasCOLEY&BARONE(1996)citeanav- erage value of 6.9% from a literature search. Few cita- tions describehigher overall tannin values,e.g. CAMP- BELL & FUCHSHUBER (1995) up to 35% and SWAIN (1965)upto60%.Thesedifferencesmaybecausedby variations in extraction and purification of extracts. A statisticaloutlierisV.koschnyi,whichshowsthehighest overalltanninvalue.Thismaybecausedbytherelative young leaves of this species used for the measurements (seeCOLEY&AIDE1991). Condensed tannins (CE) are assumed to be the mainfactorregulatingdecompositionprocessesinrivers and streams (CAMPBELL & FUCHSHUBER 1995). COLEY &BARONE(1996)foundvaluesofcondensedtanninsin mature leafs of tropical species at about 5.5% CE and HÄTTENSCHWILERetal.(2003)valuesupto14%.Inthis studylevelsofCErangedfrom0%to13.3%.V.koschnyi Fig.6:Correlationbetweencondensedtanninsanddecayrate. 473 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at again shows the highest levels, closely followed by C. COLEYP.D.&T.A.KURSAR(1996):Causesandconsequencesofepi- obtusifolia.Leavesofther-strategistC.obtusifoliaareof- phyll colonization. — In: MULKEY S.S., CHAZDON R. & A.P. SMITH(eds),TropicalForestPlantEcophysiology,.Chapman ten directly exposed to the sun. COLEY & KURSAR andHall,NY.:337-362. (1996) demonstrate that carbon based defence mecha- CHOONGCM.F.(1996):Whatmakesaleaftoughandhowthisaf- nismsliketanninscanbeincreasedbysunlightindif- fectsthepatternofCastanopsisfissaleafconsumptionby ferent tropical plants. For T. macrophyllum and A. ti- caterpillars.—FunctionalEcology10:668-674. bourbou,condensedtanninscouldnotbedetected,pos- FIELDJ.A.&G.LETTINGA(1992):Toxicityoftannincompoundsto siblyasaresultofthedryingprocess.Despitethesein- microorganisms.—In:HEMMINGWAYR.W.&P.E.LAKS(eds), consistencies all three methods, Folin Denis assay on PlantPhenols,PlenumPress,NewYork:673-692. dryandfreshmaterialandbuthanolHClmethod,show HAGERMANA.E.(1988):Extractionoftanninfromfreshandpre- a statistical significant difference between the two dis- servedleaves.—JournalofChemicalEcology14:453-461. tinguishedlifehistorystrategygroups.Inourresults,the HAGERMAN A.E. & L.G. BUTLER (1991): Tannins and Lignins. Pp. correlation between decay rates and overall tannins is 355-388—In:ROSENTHALG.A.&M.R.BERENBAUM(eds).,Her- bivores: Their interaction with secondary plant metabo- stronger than with condensed tannins. This is in con- lites;Volume1;Thechemicalparticipants;AcademicPress. trast to results of STOUT (1989), FIELD & LETTINGA HÄTTENSCHWILER S., HAGERMAN A.E. & P.M. VITOUSEK (2003): (1992)andCAMPBELL&FUCHSHUBER(1995). Polyphenolsinlitterfromtropicalmontaneforestsacrossa widerangeinsoilfertility.—Biogeochemistry64:129-148. Insummary,astatisticallysignificantdifferencebe- tween life history strategy groups of plants are found MACARTHURR.H.&E.O.WILSON(1967):TheTheoryofIslandBio- geography.—PrincetonUniv.Press,Princeton,NJ. bothintanninlevelsandinanatomicalparametersand thus a classification according to life history strategies MARTINJ.S.&M.M.MARTIN(1982):TanninAssaysinEcological Studies.—Oecologia54:205-211. canbeusedasapredictorforaquaticdecompositionof NAPP-ZINNK.(1973):AnatomiedesBlattesII,Blattanatomieder aquaticleaflitterdecayintropicalrivers.Ourresultsare Angiospermen.―GebrüderBorntraeger,Berlin,Stuttgart. consistent with the observation of SMITH et al. (1998) PETERSENR.C.&K.W.CUMMINS(1974):Leafprocessinginawood- that the first stages of leaf litter decay are mainly con- landstream.—FreshwaterBiology4:343-348. trolled by chemical compounds whereas later on me- PORTERL.J.,HRSTISCHL.N.&B.G.CHAN(1986):Theconversionof chanicalstructuresbecomemoreimportant. procyanidinsandprodelphinidinstocyanidinanddelphini- din.—Phytochemistry25:223-230. Acknowledgements SCHÄDLERM.,JUNGG.,AUGEH.&R.BRANDL(2003):Palatability,de- compositionandinsectherbivory:patternsinasuccession- We thank T. Weissenhofer for the logistic support alold-fieldplantcommunity.—Oikos103:121-132. duringthecollectionofplantmaterialinLaGamba. SMITH C.K., GHOLZ H.L. & DE ASSIS OLIVEIRA F. (1998): Fine litter chemistry,earlystagedecay,andnitrogendynamicsunder plantations and primary forest in lowland Amazonia. — References SoilBiol.Biochem.30:2159-2169. BALDWINI.T.&J.C.SCHULTZ(1988):Phylogenyandthepatternsof STOUTR.J.(1989):Effectsofcondensedtanninsonleafprocess- inginmid-latitudeandtropicalstreams:atheoreticalap- leafphenolicsingap-andforest-adaptedPiperandMico- proach.—Can.JournalFish.Aquat.Sci.46:1097-1106. niaunderstoryshrubs.—Oecologia75:105-109. SWAINT.(1965):Thedistributionoftanninsandtheirfunction. BRAUNEW.,LEMANA.&H.TAUBERT(1999):Pflanzenanatomisches —In:BONNERJ.&J.E.VARNER,PlantBiochemistry;Academic PraktikumI.—SpektrumAkademischerVerlag,Berlin,8. Press,NewYorkandLondon:574-580. Auflage. SWAINT.&W.E.HILLIS(1959):Thequantitativeanalysisofpheno- CAMPBELL I.C. & L. FUCHSHUBER (1995): Polyphenols, condensed licconstituents.—JournaloftheScienceofFoodAgricul- tannins, and processing rates of tropical and temperate ture10:63-68. leavesinanAustralianstream.—J.N.Am.Benthol.Soc. 14:174-182. TSCHELAUTJ.(2005):Leaflitterdecompositionandmacroinverte- bratesinaneotropicallowlandstream,QuebradaNegra, CHROMA-Gesellschaft (1962): Ausgewählte Färbemethoden für CostaRica.―Diplomarbeit,UniversitätWien. Botanik, Parasitologie, Zoologie. — Schmid & Co., Stutt- TSCHELAUTJ.,WEISSENHOFERA.&F.SCHIEMER(2008a):Macroinverte- gart,Untertürkheim. brates and Leaf Litter Decomposition in a Neotropical COLEY P.D. & AIDE T.M. (1991): Comparison of hervivory and Lowland Stream, Quebrada Negra, Costa Rica. — In this plant defenses in temperate and tropical broad-leaved volume. forests.Pp.25-49—In:PRICEP.W.,LEWINSONT.M.,FERNANDES TSCHELAUTJ.,PICHLERC.,WEISSENHOFERA.&F.SCHIEMER(2008b):The G.W.&W.W.BENSON(eds),Plant-AnimalInteractions:Evo- RiverNetworkofthePiedrasBlancasNationalPark,Costa lution Ecology in Tropical and Temperate Regions. John Rica.—Inthisvolume. WileyandSons;N.Y. VANNOTER.L.,MINSHALLG.W.,CUMMINSK.W,SEDELLJ.R.&CUSHING, COLEYP.D.&J.A.BARONE(1996):Herbivoryandplantdefensesin C.E. (1980): The River Continuum Concept. — Can. J. of tropicalforests.—Ann.Rev.Ecol.Syst.27:305-335. Fish.Aquat.Sci.37:130-137. 474 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at WEBERA.,HUBERW.,WEISSENHOFERA.,ZAMORAN.&G.ZIMMERMANN (2001):Anintroductoryfieldguidetothefloweringplants oftheGolfoDulcerainforests,CostaRica.―Stapfia78:1- 462. WHITMORE T.C. (1993): Tropische Regenwälder – Eine Einfüh- rung.—SpektrumAkademischerVerlag,Heidelberg,Ber- lin,NewYork. WRIGHTI.J.&M.WESTOBY(2002):Leavesatlowversushighrain- fall:coordinationofstructure,lifespanandphysiology.— NewPhytologist155:403-416. Addressesofauthors: AstridRIEMERTH MariaGUSENLEITNER FritzSCHIEMER DepartmentofLimnologyandHydrobotany UniversityofVienna Althanstraße14 A-1090Vienna,Austria E-mail:[email protected] [email protected] [email protected] Dr.GerhardDRAXLER DepartmentofEcophysiology andFunctionalAnatomyofPlants UniversityofVienna Althanstraße14 A-1090Vienna,Austria E-mail:[email protected] WolfgangWANEK DepartmentofChemicalEcology andEcosystemsResearch UniversityofVienna Althanstraße14 A-1090Vienna,Austria E-mail:[email protected] 475 © Biologiezentrum Linz/Austria; download unter www.biologiezentrum.at Plate1: Cecropiaobtusifolia 476