History

Above-ground biomass and productivity in a rain forest of eastern South America

Description
Above-ground biomass and productivity in a rain forest of eastern South America
Categories
Published
of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
   Journal of Tropical Ecology  (2008) 24 :355–366. Copyright © 2008 Cambridge University Pressdoi:10.1017/S0266467408005075 Printed in the United Kingdom  Above-ground biomass and productivity in a rain forest of easternSouth America J´erˆome Chave ∗ 1 , Jean Olivier  ∗ , Frans Bongers † , Patrick Chˆatelet ‡ , Pierre-Michel Forget § ,Peter van der Meer  † #, Natalia Norden ∗ , Bernard Ri´era §  and Pierre Charles-Dominique ‡ ∗ Laboratoire Evolution et Diversit´e Biologique, CNRS/UPS, Bˆatiment 4R3, 118 route de Narbonne, 31062 Toulouse, France † Forest Ecology and Forest Management Group, Centre for Ecosystem Studies, Wageningen University, P.O. Box 47, 6700 AH Wageningen, The Netherlands ‡ CNRS-Guyane UPS 2651, Station d’Etudes des Nouragues, French Guiana, France § D´epartement Ecologie et Gestion de la Biodiversit´e, UMR 5176 CNRS-MNHN, 4 av. du Petit Chˆateau, F-91800 Brunoy, France #Alterra, Centre for Ecosystem Studies, Wageningen University and Research Centre, P.O. Box 47, 6700 AH Wageningen, The Netherlands (  Accepted 10 April 2008  ) Abstract:  The dynamics of tropical forest woody plants was studied at the Nouragues Field Station, central FrenchGuiana. Stem density, basal area, above-ground biomass and above-ground net primary productivity, including thecontribution of litterfall, were estimated from two large permanent census plots of 12 and 10ha, established oncontrasting soil types, and censused twice, first in 1992–1994, then again in 2000–2002. Mean stem density was512 stems ha − 1 and basal area, 30m 2 ha − 1 . Stem mortality rate ranged between 1.51% and 2.06% y − 1 . In bothplots, stem density decreased over the study period. Using a correlation between wood density and wood hardnessdirectlymeasuredbyaPilodynwoodtester,wefoundthatthemeanwooddensitywas0.63gcm − 3 ,12%smallerthanthe mean of wood density estimated from the literature values for the species occurring in our plot. Above-groundbiomass ranged from 356 to 398Mgha − 1 (oven-dry mass), and it increased over the census period. Leaf biomasswas 6.47Mgha − 1 . Our total estimate of aboveground net primary productivity was 8.81MgCha − 1 y − 1 (in carbonunits),notaccountingforlosstoherbivory,branchfalls,orbiogenicvolatileorganiccompounds,whichmayaltogetheraccount for an additional 1MgCha − 1 y − 1 . Coarse wood productivity (stem growth plus recruitment) contributed to4.16MgCha − 1 y − 1 .Litterfallcontributedto4.65MgCha − 1 y − 1 with3.16MgCha − 1 y − 1 duetoleaves,1.10MgCha − 1 y − 1 to twigs, and 0.39MgCha − 1 y − 1 to fruits and flowers. The increase in above-ground biomass for both trees andlianasisconsistentwiththehypothesisofashiftinthefunctioningofAmazonianrainforestsdrivenbyenvironmentalchanges, although alternative hypotheses such as a recovery from past disturbances cannot be ruled out at our site,as suggested by the observed decrease in stem density. Key Words:  above-ground biomass, carbon, French Guiana, net primary productivity, tropical forest INTRODUCTION Terrestrial carbon cycling largely depends on thecontribution of tropical forests, and quantifying thiscontribution has proven challenging despite over 40years of active research (Golley & Lieth 1972). Estimatesof tropical forest carbon stocks vary widely amongstudies, and this variation contributes largely to theuncertainty in estimates of carbon flux. Recent estimatesof carbon pools in South America range from 150 to200MgCha − 1 for above-ground carbon in old-growth 1 Corresponding author. chave@cict.fr forests, and exceed 250 PgC for total carbon (Brown &Gaston 1995, Dixon  et al . 1994, Houghton  et al .2001, Malhi  et al . 2006, Saatchi  et al . 2007). All of these estimates made important assumptions in order toconvert permanent forest tree datasets into estimates of above-groundbiomass(henceforthAGB)andthencarbonstocks.Theseincludeaninadequatecoverageofsamplingsites, poorly validated methods for extrapolating site-based studies to continental-scale estimates, and roughassumptionsaboutthebelow-groundcontributiontothecarbon pools of tropical forests (Cairns  et al . 1997, Chave et al . 2004, Clark  et al . 2001a, Houghton  et al . 2001).The estimation of tropical forest net primary produc-tivity(NPP)isevenmoredifficultthanestimatingcarbon  356 J´ERˆOME CHAVE  ET AL. stocks.Currentestimatesbasedonstandardfieldprotocolssuggest values larger than 10MgCha − 1 y − 1 (Clark  et al .2001b), and tropical forests may contribute to up toa third of the net primary productivity of terrestrialecosystems (Field  et al . 1998). A large number of studieshave attempted to combine field measurements, remote-sensing data and ecophysiological models to produceworld maps of NPP, but these estimates remain proneto a considerable amount of uncertainty in the tropics(Field  et al . 1998, Schuur 2003, Turner  et al . 2005).Attheecosystemscale,NPPisbalancedbycarbonlossthrough heterotrophic respiration, Rh. If the differencebetween NPP and Rh, or net ecosystem exchange (NEE)is positive, then the carbon stocks of tropical forests areincreasing. It has been observed in several empiricalstudies that tropical forests are currently acting as anatmospheric carbon sink (Lugo & Brown 1992, Phillips et al .1998).Notonlywouldthisexplaintheatmosphericcarbon sink currently unaccounted for in global carbonmodels (Grace 2004), but this would also reinforce theconservation value of tropical forests, as they contributeto buffer current C emissions due to the burning of fossilfuels (Malhi  et al . 2008). Although a great deal of recentwork has been devoted to improve the quantificationof NEE in the tropics (Chave  et al . 2008, Grace 2004,Lewis  et al . 2004, Loescher  et al . 2003, Malhi  et al . 2008,Saleska  et al . 2003, Stephens  et al . 2007), Amazonianforests remain severely understudied (Baker  et al . 2004,Bonal  et al . 2008). For instance, the total area coveredbyforestcensusplotsisstillverylimitedintropicalSouthAmerica, with no more than 78ha of forest where treeswere measured at least twice (Baker  et al . 2004).In a previous study, we published the first AGBestimates for the forest surrounding the NouraguesResearch Station, central French Guiana, as deducedfrom a 22-ha network of permanent tree plots surveyedin 1992–1994 (Chave  et al . 2001). Here, we study thedynamics of this carbon pool, based on data from asecond census, and improved methods. We provide thefirst estimate of AGB change for a period of   c . 8 y, and thefirst estimate of above-ground net primary productivity(ANPP).DuetothedesignofourexperimentalsetupattheNouragues Research Station, the effect of the geologicalsubstrate on AGB estimates could also be appraised.Finally, we discuss our results in light of other study sitesin the region. MATERIALS AND METHODSStudy site OurstudywascarriedoutatNouraguesResearchStation(4 ◦ 05 ′ N, 52 ◦ 40 ′ W), located 120km South of Cayenne,in the lowland rain forest of French Guiana (Bongers et al.  2001, http://www.cnrs.fr/nouragues, Figure 1).This station was established in 1986, near an inselberg(granitic outcrop) that reaches 430m asl. The landscapeis a succession of small hills, between 60–120m asl.Rainfall is 2960mm y − 1 (average 1987–2001), witha dry season that averages 73 d, from late August toearlyNovember,andashorterdryseasoninMarch.Dailytemperature ranges between 20 ◦ C and 33 ◦ C (annualmean 27 ◦ C). Wind is never strong (maximum  < 14.2ms − 1 between 1999 and 2002, mean = 0.11 ± 0.07ms − 1 ). No hurricanes or cyclones reach French Guiana.Human activity is unlikely to have induced majordisturbancesintherecenthistory:theNouraguesIndiansare reported to have inhabited this area during the 18thcentury, but departed at least 200 y ago.The research station is located on the west bank of asmallriver,called‘criqueNouragues’,thatflowsonafaultseparating two geomorphological entities (Grimaldi &Ri´era 2001). The west bank has a weathered graniteparent material, with sandy soils of variable depth, onwhich a 400 × 300-m plot called Petit Plateau (PP) hasbeen established (van der Meer & Bongers 1996a). TheeastbankisonametavolcanicrockparentmaterialoftheParamacaformation,withclayeysoilsrichinferruginousnodules, typical of the decomposition of a laterite crust.Ontheeastbank,and c .500mfromPP,a1000 × 100-mplot,calledGrandPlateau(GP),hasbeenestablishedonauniform and gentle slope toward the creek. Permanentsampling plots are delineated by a grid of trails every100m on two plateaux along a compass bearing of 137 ◦ either sides of crique Nouragues.The forest around the station harbours a diverse treeand liana flora (Poncy  et al.  2001, Sabatier & Pr´evost1990),withover1700angiospermspeciesrecordedintheReserve (http://www.nouragues.cnrs.fr/plantspecies2.html). To the south of the GP, patches of forest withan overabundance of lianas of unknown origin areencountered (Chave  et al . 2001, Schnitzer & Bongers2002, Schnitzer  et al . 2006). Plot censuses Thetwoplotswerefirstestablishedandcensusedbetween1992and1994.InboththeGPandthePPplotsallwoodystems(treesandlianas) ≥ 10cmdbh(diameterat130cmabove ground, or  c . 50cm above buttresses, if present)were inventoried. All stems were mapped, tagged, andmeasured with a cloth tape (Chave  et al . 2001, Olivierunpub.data,vanderMeer&Bongers1996a,b).Thefirstcensus took place between March 1992 and November1994 in the GP plot, and in July–August 1992 in the PPplot.Between August 2000 and October 2002, we recen-sused a total of 12 630 trees and lianas with dbh  Biomass and productivity in a tropical forest  357 Figure1. Mapofthestudyarea(PP = petitplateauplot,GP = grandplateauplot).Thecoordinatesoftheinselbergtopare4 ◦ 05 ′ 31 ′′ N,52 ◦ 40 ′ 43 ′′ W. ≥ 10cm in the 22ha of the GP and PP plots. Plots weresubdivided into 10 × 100-m lines using 100-m stringstagged every 10m. For dead trees, the type of mortalitywasnoted(fallentrees,standingdead,snappedtrees).Wealsodocumentedmissingstems.Stemcircumferencewasmeasured to the nearest mm rounding down. We thenconverted this value into diameter assuming a circularstem.Insomecases,mostlyforbigtreeswithbuttressesorwithirregularboles,diameterwasnotdirectlymeasured,but estimated using a relascope technique (DeWalt &Chave 2004). Specifically, we took a digital photographof the stem and of a ruler positioned at a specified height,and estimated the diameter at that height using image-processingsoftwaresuchasPhotoshop(Adobe,SanJose,USA). The accuracy of this method for trunk diameterestimation was found to be of 5% (DeWalt & Chave2004). Data were digitized and checked carefully on acomputer while in the field, and obvious errors, such asanomalousdbhchangesormissingstems,wererecheckedthe following day.During the second census, we paid special attention toirregular-shaped stems for which an accurate measure-ment was difficult. We carefully estimated the stem basalarea at breast height, assuming that the stem crosssection has the shape of a polygon. Then, we convertedbasal area into a theoretical dbh value, as if the trunkwere cylindrical in shape. The stem diameter of thepalm species  Astrocaryum sciophilum  (Miq.) Pulle couldnot be measured with great precision because of thepresence of leaf scars on the trunk (Charles-Dominique et al . 2003). For multi-stemmed trees or forked stems,we counted genets as one individual, and calculated  358 J´ERˆOME CHAVE  ET AL. the total cross-sectional area of the ramets. The dbhwas estimated rather than directly measured for 7.7%of the stems. For the first census, we could not correctpossible errors of measurement on these stems, and if themeasurements appeared to be in error, we corrected itusing the procedure explained below.We discovered that the dbh of many of the large trees(typically, trees ≥ 70cm) had been overestimated in thefirst census. Since our second census was more accurate,weusedatrimmingmethodtocorrectforthisbias(Baker etal .2004,Chave etal .2003,2008;Sheil1995).Anydbhdecreasesof5mmy − 1 ormore,orincreasesof35mmy − 1 or more were assumed to be anomalous, if the trees didnotbelongtooneofthefamilieswithfast-growingspecies(Cecropiaceae,Vochysiaceae),andthefirstmeasurementwasdeclaredtobeincorrect.Wethencorrectedthevaluebyassumingthatthedbhgrowthwasequaltotheaveragedbh growth of trees in the same dbh class (Chave  et al .2003). Wood density measurement Wood specific gravity is an important variable in theestimation of tree above-ground biomass (Baker  et al .2004). In most previous studies, wood density wasestimatedforeachstemfromtheinformationofthespeciesbyassumingaspeciesmeanwooddensity.Inmanycases,however, the lack of reliable taxonomic identification forthe censused trees makes this procedure difficult. In ourplots,forinstance,only50%ofthestemswereidentifiedtothe species. Moreover, the procedure assumes that thereis no intraspecific variability in wood density, and thatwood density values from the literature are accurate. Wetherefore assessed its validity.Weobtainedaplot-averagedwoodspecificgravityfromavailablespeciesabundancesintwo1-hasubplotswhereall the stems had been identified to species (Poncy  et al .2001, Sabatier & Pr´evost 1990), combined with a largedatabase of mean specific gravity for neotropical treespecies (Chave  et al . 2006). We also tested for a potentialbias related to this assumption, namely, that large treesdonothavealowerwoodspecificgravitythansmallones,as it is the case in the BCI forest (Chave  et al . 2004). Wefound no difference in mean wood density between smalland large trees in the forest around the Nouragues FieldStation.We also developed a different strategy to estimate thewood specific gravity of the trees in our plots. We used awood tester, the Pilodyn 6J (Proceq USA, Aliquippa PA,USA), a tool commonly used to measure wood hardnessin plantation trees and construction wood. This deviceis pressed firmly onto the stem surface. The impact pinis shot into the wood by pressing a trigger, and thedepth of penetration can be read immediately on a scalemounted on the tester. We calibrated the instrumenton 144 trees (dbh range: 9–130cm), that were alsocored using a forestry wood increment borer (Suunto,Vantaa, Finland). For these trees, wood specific gravitymeasured from cored wood samples varied between 0.21and 0.96gcm − 3 (Figure 2). We found that wood densitycorrelated strongly with the wood hardness as measuredby the Pilodyn (Figure 2). Running a stepwise selectionlinearmodelonourdataset,wefoundthatthebestmodelpredictingwooddensity ρ  usedboththePilodynhardness h  and the stem dbh, as follows:ln( ρ ) = 1 . 01 + 0 . 77ln( h ) + 0 . 15ln( dbh )Here, h isdefinedasoneoverthepenetrationdepthofthepin into the wood (measured in mm). The coefficient of correlationisofthisregressionwasr 2 = 0.79.Weappliedthis model on all censused trees in two 1-ha subplots(n = 1044 trees), one in the GP plot, the other in thePP plot, for which we measured the Pilodyn-hardness.Finally,wecomparedournewestimateofplot-meanwooddensity with the one obtained through literature values.In most studies on the carbon sequestration of tropicalrain forests, it has been assumed that above-groundbiomass in live trees contains 50% carbon (Clark  et al .2001a). Although the wood carbon fraction may exhibitsome variation as faster-growing trees may have fewerof the more reduced and stable carbon compounds thandoslower-growingones(Elias&Potvin2003,Malhi et al .2004),wealsousedtheconventionof50%carbonindrybiomass here. Litterfall monitoring We measured litterfall separately for leaves, twigs(typically  < 1cm in diameter), and reproductive organs(flowersandfruits),usinganetworkoflitterfalltraps,each0.5m 2 in size. We initially installed 100 traps on the GPplot,and60trapsonthePPplots,followingarandomizedlocation procedure. Traps were made of square piecesof large-mesh polyethylene fabric tied by ropes to fourneighbouring live trees, at about 1.5m above ground toavoid disturbances by large mammals. When a trap wasdamaged by the fall of woody debris, the correspondingdata were discarded (0.7% of the measurements). Thecontentofthe160trapswascollectedtwiceamonthfromFebruary 2001 to July 2003. At this time, we analysedthe data and selected 15 representative traps in the PPplot, and 25 traps in the GP plot. Starting in January2004, we continued the same protocol with this reducedsamplingscheme.Thecontentofthetrapswasseparatedinto leaves, branches, fruits and flowers when wet, thenoven-dried at 70 ◦ C for up to 48 h, and weighed with anelectronicbalance(precision0.1g).Averageswerebased  Biomass and productivity in a tropical forest  359 Figure 2.  Calibration of the wood specific gravity estimation based on the Pilodyn. Wood specific gravity (gcm − 3 ), and Pilodyn hardness (definedas the inverse of the Pilodyn 6J reading, expressed in mm) were estimated for 144 trees, belonging to 98 different tree species. Sampled trees were9–130cm in dbh, and wood specific gravity ranged from 0.21gcm − 3 to 0.96gcm − 3 . on the data available from February 2001 to July 2007(77mo). Statistical analyses Mortality and recruitment were computed using anexponential model. For a cohort with  N  1  individualsduring the first census, with  N  S  survivors at the secondcensus, the formula is  N  S = N  1  exp(- m T  ), where  m  is themortality rate and  T   the census interval. Recruitmentrate  r   was estimated from the number  N  2  of individualspresent at census 2 and  N  S , assuming that  N  2 = N  S exp( r T  ). In the following, these quantities are reportedon an annual basis. Turnover is defined as the meanbetween recruitment and mortality (Phillips & Gentry1994). Because the census intervals had a similar span,we did not account for the fact that m should decrease asT increases (Sheil & May 1996).To estimate the stand-level AGB in trees, we used aregressionequationbasedonalargesamplesizeofdirectlyharvested trees. This equation relates the dbh  D  of a stemin cm and its oven-dry specific gravity  ρ  in gcm − 3 , to its AGB  in kg (n = 1804 trees, see Chave  et al . 2005): AGB = ρ × exp( − 1 . 499 + 2 . 148ln( D ) + 0 . 207ln( D ) 2 − 0 . 0281ln( D ) 3 )Lianas were also taken into account in the total AGBestimation,usingaformulabasedon424harvestedlianas(see Schnitzer  et al . 2006). AGB = exp( − 0 . 968 + 2 . 657ln( D ))To estimate the leaf biomass in our plots, we alsoused empirical regression methods. We used the dataset
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks