Bacteria, Fungi and the Breakdown of Leaf Litter in a Large River Author(s): Virginie Baldy, Mark O. Gessner, Eric Chauvet Source: Oikos, Vol. 74, No. 1, (Oct., 1995), pp. 93-102 Published by: Blackwell Publishing on behalf of Nordic Society Oikos Stable URL: http://www.jstor.org/stable/3545678 Accessed: 11/04/2008 04:46 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=black. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.
JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We enable the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact
[email protected].
http://www.jstor.org
OIKOS 74: 93-102. Copenhagen1995
Bacteria, fungi and the breakdown of leaf litter in a large river Virginie Baldy, Mark O. Gessner and Eric Chauvet
Baldy,V., Gessner,M. 0. andChauvet,E. 1995.Bacteria,fungiandthebreakdownof leaf litterin a largeriver.- Oikos 74: 93-102. We examinedthe dynamicsof leaf mass loss and microbialbiomassassociatedwith decomposingleaf litter in a seventhorderriver.This was in an attemptto test the to thebreakdownprocessis less importantin major hypothesisthatfungalcontribution riversthanwas previouslyfoundfor headwaterstreams.Bacterialbiomasswas estimatedfrom directcell counts coupledwith determinations of bacterialbiovolumes. Fungal biomass was estimatedon the basis of ergosterolmeasurements,speciesspecificconversionfactors,andthe relativeabundanceof the dominantfungalspecies. Sporulationrates of aquatic hyphomyceteswere determinedby counting conidia releasedfrom leaf litterduringbrief laboratoryincubations.Comparedto low-order streams,the breakdownof willow, poplarandplaneleaves was slow withexponential decay coefficientsk rangingfrom 0.0045 to 0.0091 d-~. Numbersof bacteriafirst increasedexponentiallyon all leaf speciesbutreacheda plateauof almost108cells per mg AFDMafter4-8 weeks of leaf submergence.This correspondsto a peakbacterial biomassof 0.3-0.5% of detritalcarbon.Fungalbiomassattainedpeaksof 5-10% of detritalcarbonafter4-8 weeksandgreatlyexceededbacterialbiomassat anyinstance. On average,fungi accountedfor 96% of the total microbial(fungalplus bacterial) biomassin leaf litter.Dynamicsof sporulationrates of aquatichyphomyceteswere characterizedby early peaks of 1.2-1.4 conidiapg-1 AFDM d-~,followed by sharp declinesto about0.2 pg-~d-1.Peaksoccurredbeforethe corresponding peaksin fungal biomass.Roughorganicmatterbudgetssuggestthatfungi assimilateda minimumof 16-23%of the initialleaf carbon,andaccountedfor42-65% of the overallcarbonloss from leaves duringperiodsof highestfungalactivity.Takentogether,these findings indicatethatfungiplay an eminentlyimportantrolein the biologicaltransformation of leaf littereven in majorrivers.Bacterialcontributionis likely to be smallin spite of increasesin biomassat advancedstagesof breakdown.Withregardto leaf decomposition, largefluvial systemswouldthus appearto behavelike theirheadwatercounterparts,suggestingthatthe presentresultscan be generalizedfor lotic ecosystems. V Baldy and E. Chauvet, Centre d'Ecologie des Systemes Fluviaux, Centre National de la Recherche Scientifique, 29 rue Jeanne Marvig, F-31055 Toulouse Cedex, France. M. O. Gessner, Centre for Ecosystem Research, Univ. of Kiel, Schauenburgerstrafie 112, D-24118 Kiel, Germany.
Streams and rivers that flow through forested catchments receive substantial amounts of leaf litter from the riparian vegetation, and the decomposition of this allochthonous material has been recognized as a vital process in the functioning of running waters (Webster and Benfield 1986, Boulton and Boon 1991, Maltby 1992a). Emphasis in past research on leaf decomposition in streams has been primarily laid on the determination of breakdown
rates as a function of leaf species (Petersen and Cummins 1974, Hill et al. 1992, Gessner and Chauvet 1994), the chemical transformations that accompany leaf breakdown (Suberkroppet al. 1976, Rosset et al. 1982, Chauvet 1987, Gessner 1991) and the pattern of macroinvertebrate colonization on decomposing leaf material (Anderson and Sedell 1979, Cuffney et al. 1990, Chauvet et al. 1993).
Accepted22 February1995 Copyright? OIKOS1995 ISSN 0030-1299 Printedin Denmark- all rightsreserved OIKOS 74:1 (1995)
93
dicate in fact that peak fungal biomass in leaf litter is normally greater than 5% of detrital mass and may even exceed 10% of detrital mass (Suberkropp 1992b, clearly E Gessner and Chauvet 1994). 200 a) Those studies that have taken microorganisms into account, were generally restricted to either the fungal (Barlocher 1992a, Chergui and Pattee 1993) or the bacterial (Cuffney et al. 1990, Leff and McArthur 1990) comco 0O cn ' ' or measured bulk parameters of microbial bioponent, ..' 0 mass and activity such as ATP content and respiration (Petersen and Cummins 1974, Mulholland et al. 1984, Cuffney et al. 1990, Hill et al. 1992). A separate quantitao 15 tive consideration of the two groups of microorganisms has rarely been attempted (but see Padgett 1976, Findlay and Arsuffi 1989). However, owing to the fundamentally different life forms of fungi (generally filamentous) and bacteria (apart from Actinomycetes typically singlea) 5 celled), these two groups probably exert distinct impacts 0 on decomposing plant material. This idea ties in with the E observation that fungi appear to dominate microbial as-0 semblages in leaves initially as long as the tissue is more D F M or less intact, while bacteria tend to increase when leaves of the become partially broken down (Suberkropp and Klug Fig. 1. Daily meansof dischargeand watertemperature GaronneRiver during the field experimentfrom November 1976). If so, a clear distinction between fungal and bacte1991 to April 1992. rial activity would be paramount to understanding the leaf decomposition process in running waters. In this paper, we report an investigation into the breakMost of these stum dies were carried out in small head- down of three leaf species in a large river and the dynamwaters, because the c:ontributionof allochthonous organic ics of both the prokaryotic (bacteria) and eukaryotic matter to stream me:tabolism tends to diminish with in- (fungi) component of the microbial assemblage associcreasing channel width (Conners and Naiman 1984). ated with these leaves. Our specific intention was (1) to However, even in large rivers, inputs of organic matter test whether leaf-associated fungal biomass would be as derived from the terrestrial environment may be signif- important in the seventh order Garonne River as was icant. Chauvet and Jlean-Louis (1988) estimated, for ex- previously found in two low order streams (Suberkropp ample, that the sev enth order Garonne River receives et al. 1993, Gessner and Chauvet 1994), (2) to determine about 10% of the coarse particulate organic matter the relative importance of bacterial and fungal biomass in (CPOM) typically entering fully canopied low order this river, and (3) to test the general hypothesis that streams. Additional material may be imported from up- microbial assemblages on decomposing leaf litter in runstream reaches and tthe floodplain (Conners and Naiman ning waters are initially dominated by fungi and, with 1984, Cuffney 1988)). Thus, the decomposition of CPOM progressing leaf breakdown, gradually shift to assemand particularly leaff litter would appear to constitute a blages in which bacteria assume greater importance (Sucentral process not c)nly in low order streams but also in berkropp and Klug 1976). major rivers (cf. Thorp and Delong 1994). Before the fundamental alterations of channel morphology and riparian vegetation by man, the importance of the process in whole system functioning would have been even grea- Materials and methods ter (Naiman et al. 1988). site Leaf breakdown in aquatic ecosystems is mainly a Study biological process involving three types of organisms: The study was conducted in the Garonne River, a seventh large particle detritivores commonly referred to as shred- order hardwater river draining large portions of southders, fungi and bacteria. Given the considerable attention western France. We chose a site downstream from the devoted to invertebrates and their role in the leaf break- confluence with the Ariege River some 7 km upstream down process, surprisingly little work has been accom- from the city of Toulouse. At this site, the river channel is plished on the microbial assemblages associated with 200 m wide. Dominant tree species of the fragmented decomposing leaves (e.g., Kaushik and Hynes 1971, Su- floodplain forest are willow (mainly Salix alba L.) and berkropp and Klug 1976, Chamier 1987) although the poplar (Populus nigra L. and various hybrids), but genercritical role of microorganisms is clearly established (Su- ally only willows grow immediately adjacent to the river berkropp 1992a, Maltby 1992b). Recent estimates in- channel. River discharge was recorded continuously at a
-? CO
94
300
OIKOS 74:1 (1995)
characteristics of theGaronne River 1-mm-meshscreen. Subsamplesof about250 mg were Table1. Physico-chemical fromNovember 1991to April1992(N ashedfor 3 h in a muffle furnaceat 550?Cto determine duringfieldexperiments = 7). ash free dry mass (AFDM).Carboncontentwas determinedby a conductimetric analysisof CO2releasedafter Parameter Mean Range combustionof 250-mg subsamples(Carmhograph 8 anaWosthoff oHG, Germany). lyser, 8.5 7.9-9.1 pH 158 127-185 Conductivity(IS/cm, 20?C) 1.5 1-2 (mmol/1) Alkalinity [P-PO4] (lg/1)
[N-NO3](mg/l) [N-NH4](mg/1)
8
1.8 0.08
3-16
1.2-2.8 0.05-0.10
gaugingstationnearthe experimentalsite. The long-term annualmean is 192 m3/s. During the study period, it rangedfrom 61 to 270 m3/s (Fig. 1). Watertemperature was measuredonce daily and variedbetween 5?C and 14?C.Dischargeand temperature datawere kindlycommunicatedby local authorities.Additionalphysical and chemicalcharacteristics of the GaronneRiverweredeterminedaccordingto standardproceduresand are summarized in Table 1.
Fungal community structure and sporulation rate Three willow leaves, one poplarleaf or a quarterof a plane leaf were arbitrarilytaken from each litter bag, placedin Erlenmeyerflaskscontaining100 ml of filtered riverwater(WhatmanGF/D, approximatepore size 2.7 ,lm), andincubatedon a rotaryshakerfor 3 d at 10?Cso as to inducesporulationof aquatichyphomycetes.At the end of incubation,250 pl of a 0.5% TritonX-100 (Prolabo, France)solutionwere addedto ensure a uniform distributionof conidia, and 2 ml of the supensionwas passedthrougha Whatmanmembranefilter (poresize 5 pm). The conidiatrappedon the filterwere stainedwith 0.01% trypanblue in lactic acid, and identified and countedundera Zeiss Axioplanmicroscopeat a magnificationof 200. Two filtersper samplewere preparedand at least 100 conidia per filter were counted. Conidial numberswere convertedto conidial carbon based on publishedvalues of biovolumesof the identifiedspecies (Barlocherand Schweizer1983) and assumingan average carbondensityof 250 fg C/pm3 (FindlayandArsuffi 1989, Newell 1992). A particle-platingmethodwas used to identify leafcolonizing fungi other than aquatichyphomycetes.Six leaf discs (4 mmdiameter)werecut fromleaf materialof each replicatelitterbag using a cork borer.Discs were placedon 0.1%maltextractagar(Merck)and incubated at 10?C.Coloniesgrowingout of the discs' edges were transferredto fresh plates and selected strainsof each colonytypeidentifiedto genusor species.Isolatedstrains showing macroscopicallyidenticalcharacteristicswere assumedto belong to the same taxon.
Experimental set up and field procedures Freshlyfallenleaves of willow, poplarandLondonplane (PlatanushybridaBrot.)were collectedat the studysite on 14 November1991.Londonplaneis a speciesuncommon to riparianforests of westernEuropebut was includedin this studybecauseit is frequentlyplantedalong watercoursesin Franceand elsewherein Europe,while being more recalcitrantto decompositionthan the two other naturalspecies. The collected leaves were stored overnight,thenweighedintoportionsof 2-7 g freshmass (dependingon leaf species),andplacedin meshbags (15 cm sides) made from rigid plastic nettingwith a mesh size of 2 mm (Chauvet1987). Fourreplicatebags were set aside to determineinitial leaf dry mass, fungal and bacterialbiomass. The remainingleaf bags were hotsealedandsecuredin the riverby meansof metallicrings slippedover steel rods,whichwere anchoredin the river bottom.Waterdepthat base flow was 30-50 cm at the placewhereleaf packswereexposed.After2, 4, 8, 12, 16 and 20 weeks of immersion,four bags of each leaf spe- Fungal biomass cies were retrievedand returnedto the laboratoryin a Fungal biomass was estimated as ergosterol content. bucketcontainingriverwater. Analyses were performedwith six willow leaves, two poplarleaves or one half of a planeleaf fromeach litter bag. Ergosterolwas extractedfromlyophilizedleaf samLeaf mass loss ples by 30 minrefluxingin alcoholicbase andquantified In the laboratory,the leaves wereremovedfromthe litter by measuringUV-absorbance afterpurificationby HPLC bagsandindividuallyrinsedwithdeionizedwaterso as to (Gessneret al. 1991). The chromatographic systemconremove the greaterpart of extraneoussediments and sisted of a Kontronpump, autoinjectorand a variable adheringinvertebrates.Subsamplesof the leaf material wavelengthdetectorset at 282 nm. The columnwas a 25 were taken for the determinationof microbialbiomass cm x 4.6 mm RP18 Lichrospher(5 pumparticlesize). and fungal communitystructure.The remainingleaves Columntemperature was maintainedat 28?C in a water were driedat 105?Cfor at least 24 h, cooled in a des- bath.The systemwas runisocraticallywith HPLC-grade iccator,weighedto the nearest0.1 g, andgroundto pass a methanolat a flow rateof 1.5 ml/min.Ergosteroleluted OIKOS 74:1 (1995)
95
Table2. Exponentialbreakdowncoefficients(mean+ asymp- were projected on a screen, and normally 150-200 bactetotic standarderror)of leaf litterdecomposingin the Garonne ria were sized and assigned to one of 34 classes defined River as determinedwith non-linearregressionanalysis(N = on the basis of bacterial size and shape. The distribution 21-23). of bacteria among classes was assumed to be repreLeaf species Breakdowncoefficient Breakdowncoefficient sentative of the bacterial assemblage on a given leaf (d-1) (degree-d-1) species at a given date. To account for size-dependent differences in the dry mass density of bacterial cells, we 0.0091 0.0013 Salix 0.00112+0.00007 converted bacterial biovolume (V in um3)to carbon (C in 0.00086?0.00004 0.0070+0.0006 Populus fg) according to the empirically determined relationship Platanus 0.0045?0.0004 0.00054?0.00002 C = 89.6.V059 (Simon and Azam 1989). This equation yields carbon densities ranging from 130 to 446 fg/pm3 for the usually encountered bacterial size spectrum of after 9 min; peak identity was checked on the basis of 0.02 to 0.40 uim3per cell. Occasionally, we found large retention times of commercial ergosterol purchased from cells measuring up to 1.1 im3, corresponding to carbon Fluka (> 98% purity). Ergosterol contents were converted densities of 95 fg/,um3.To calculate the total amount of to fungal biomass using species-specific factors estab- bacterial carbon associated with leaves, the information lished for the eight dominant fungal species sporulating on carbon densities was used in combination with total on the leaves, coupled with estimates of their relative cell counts and the size-frequency structure of bacterial abundance in the fungal community (Gessner and Chauassemblages. vet 1993). Fungal carbon content was determined for the three dominant species of aquatic hyphomycetes identified in the present study, thus permitting us to express fungal biomass on a carbon rather than on a dry mass Statistical analysis basis. To obtain mycelia for these analyses, fungal strains Using non-linear regression analysis, mass loss data were were isolated from single spores, and grown in shake adjusted to the simple exponential model mt = m-e-kt, culture on a glucose mineral-salt solution (Gessner and where mt is the leaf mass remaining at time t, mo is the Chauvet 1993). They were harvested during late growth initial leaf mass, and k is the exponential breakdown and analyzed in the same way as leaf material. coefficient. One-way and repeated-measures analysis of variance were used to test for differences among leaf species in terms of bacterial numbers, ergosterol concenBacterial numbers and biomass trations, and sporulation rates of aquatic hyphomycetes. At each sampling date, 15 discs (7 mm diameter) were Differences were considered significant when P < 0.05. cut from 2-4 leaves of each litter bag with a cork borer, All calculations were made with the statistical package preserved in 2% filter-sterilised formalin, and stored until SYSTAT (Wilkinson 1990). analysis. Five ml of filtered autoclaved water was added before homogenizing the leaf discs for 3 min with an Ultra Turraxblender. Samples were diluted with 20 ml of 100; water and vigorously vortexed. After allowing the homogenate to settle for 1 min (Suberkropp and Klug 0) 1976), a 0.1-ml subsample was taken about 5 mm below 75 the surface and completed to a final volume of 10 ml. cCZ DAPI (4',6-diamidino-2-phenylindole) was added to a E final concentration of 5 mg/l before the suspension was Q) L,. a black filter passed through polycarbonate (Nuclepore, 50 pore size 0.22 pm). A cellulose nitrate filter (Whatman or Millipore, pore size 0.45 pm) was placed under the black filter in order to spread the vacuum evenly, ensuring an LL 0 25 even distribution of the particles over the entire filter surface (Fry 1988). After filtration, the filter was rinsed with 10 ml of filtered autoclaved water in order to remove excess staining solution (Fry 1988). Three filters were 0 prepared for each replicate leaf pack and the bacteria 4 12 0 8 16 20 retained on the filter were counted at 1000x by using epifluorescence microscopy with 10 or 20 microscopic Time (week) fields being viewed per filter. Photographs were taken from filter portions from one 2. Lossin ash-freedrymass(AFDM)fromleavesof willow of the four replicate leaf packs per species per sampling Fig. (0), poplar (-) and Londonplane (A) decomposingin the date. Photographic slides (Ektachrome ISO 400/27?) GaronneRiver.Verticalbarsindicate? 1 SE. 96
OIKOS 74:1 (1995)
their initial AFDM (Fig. 2). After 20 weeks, 60% of the initial AFDM remained in leaf packs with London plane while less than half of the initial leaf mass remained in poplar and willow at that time. Sporulation rates of aquatic hyphomycetes on leaf litter reached peaks of 1.2 (willow), 1.4 (poplar) and 1.3 (London plane) conidia per ,g AFDM per day after 2 or 4 weeks of leaf submersion in the river (Fig. 3). Subsequently, sporulationrates declined to 0.09-0.57 conidia ug-1AFDM d-'. Differences in sporulation rates between leaf species were not significant. At the two-week sam-
Salix - 2.5
0.8
-2.0
0.6
-1.5 0.4 - 1.0 0.2
- 0.5
0.8
0
-b
-2.5
LL
-9
Salix
9-
8LL
0.6 b)
-1.5
0.4 - 1.0 0) o
-8
-2.0
0.2-
-0.5
-.
Platanus
A
I/
-6
r0 c-
0
2
'o 01
- 2.5 - 2.0
0.6-
-7
6-
a)
0 LL
0.8-
---
?':-
I
o0 (-.
0L
0O
p
7-
- 1.0
,
,
,
,
,
5 cm
Populus
9-
-9
E cn 0
.0 .0 0
E aC) C,D
o O
8-
-8
7-
-7
ca L_ a)
-6
-o
0) CJ o
- 1.5 0.4-
5
0
0
6
0 o
;
o
0.2-
- 0.5
I
v
0
.
.
4
8
.
12 12
16 16
n1
20 20
Time (week)
'-a) 0 v
a)
5
I
I
5
I
-9
E .3
E zI
z
o
Fig. 3. Dynamicsof ergosterolconcentrations(in leaves) and sporulationratesof aquatichyphomycetesassociatedwith leaf litterdecomposingin theGaronneRiver.Verticalbarsindicate? 1 SE.
Results Exponential breakdown coefficients ranged from 0.0044 to 0.0091 per day (Table 2) and appeared to be different among leaf species, as judged from the visual comparison of confidence limits around means of AFDM remaining at each sampling date. However, much of the difference between willow and poplar was due to the first two weeks of breakdown when willow leaves lost nearly 40% of OIKOS 74:1 (1995)
0
4
8
12
16
20
Time(week) Fig. 4. Dynamicsof bacterialnumbersassociatedwithleaf litter decomposingin the GaronneRiver. Verticalbars indicate+ 1 SE. 97
Vries,species of the generaFusarium,Pythium,and Saprolegnia,and sterilemycelia.Quantitativecomparisons of these fungi with aquatichyphomyceteswere not pos10 sible owing to the differencesin methodsused to detect the differentfungalgroups. Concentrations of ergosterolin leaf litterwere minute (