Stratification and seasonal stability of diverse bacterial communities in a Pinus merkusii (pine) forest soil in central Java, Indonesia


13 pages

Please download to get full document.

View again

of 13
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.
Stratification and seasonal stability of diverse bacterial communities in a Pinus merkusii (pine) forest soil in central Java, Indonesia
   © 2002 Blackwell Science Ltd Stratification and seasonal stability of diverse bacterial communities in a Pinus merkusii  (pine) forest soil in central Java, Indonesia enced by drought. In the litter layer, the moisturecontent was significantly lower than in the fragmen-tation and mineral layers during the dry season. Aclone library was made from a litter sample takenduring the wet season. Partial sequencing of 74clones and linking the DGGE banding positions ofthese clones to bands in the DGGE profile of thesample from which the clone library was derivedshowed considerable bacterial diversity. Alpha-proteobacteria (40.5% of the clones, of which 57%belonged to the Rhizobium–Agrobacterium  group)and high-G+C content, Gram-positive bacteria(36.5%) dominated the clone library.Introduction Because of deforestation, primary tropical rain forest hasnearly disappeared from the island of Java, Indonesia. Anintensive reforestation programme was initiated in 1963,and pine plantations now comprise ª 50% of the total areaof the forest ecosystem. The pine forests consist mainlyof Pinus merkusii  , a native plant srcinating from northSumatra. Most plantations have been established in highelevation areas on marginal or nutrient-poor soils that do not support sustainable growth of crops (Bruijnzeel,1984). Abetter understanding of the factors governingnutrient cycling in tropical pine plantations will help toimprove management of these plantations, which are ofboth environmental (e.g. erosion prevention) and eco-nomical (e.g. logging) importance. As Java is subjected totropical weather conditions with relatively constant, hightemperatures (25–30 ∞ C) and high seasonal precipitation,results of studies on pine forests in temperate zonescannot be applied directly.Previous ecological research on pine forests on theisland of Java has shown that nitrogen is one of thegrowth-limiting nutrients (Bruijnzeel, 1984). Litterfall andlitter turnover are high (9–10 and 0.5 tonnesha - 1 year - 1 respectively) and provide about 60kg of Nha - 1 year - 1 tothe forest floor (Bruijnzeel, 1985; Gunadi and Verhoef,1993). The forest floor consists of three layers, from top tobottom, a litter layer (a 3- to 6-cm-thick layer of dead plantmaterial), a fragmentation layer (3–5cm thick and consist-ing of fragmented litter) and a mineral layer. Bacteria and fungi are directly responsible for the degradation of Environmental Microbiology (2002) 4 (6), 361–373 Agna S. Krave, 1,2 Bin Lin, 1 Martin Braster, 1 Anniet M. Laverman, 1 Nico M. van Straalen, 3 Wilfred F. M. Röling 1† and Henk W. van Verseveld 1 * 1 Section Molecular Microbial Ecology, Department of Molecular Cell Physiology, Faculty of Earth and Life Sciences, Research School SENSE, Vrije Universiteit,De Boelelaan 1087, NL-1081 HV Amsterdam, The Netherlands. 2 Universitas Kristen Satya Wacana, Faculty of Biology,Department of Microbiology, Jalan Diponegoro 52–60,50711 Salatiga, Indonesia. 3 Institute for Ecological Science, Faculty of Earth and Life Sciences, Vrije Universiteit, De Boelelaan 1087,NL-1081 HV Amsterdam, The Netherlands. SummaryIn Java, Indonesia, many nutrient-poor soils are inten-sively reforested with Pinus merkusii  (pine). Informa-tion on nutrient cycles and microorganisms involvedin these cycles will benefit the management of theseimportant forests. Here, seasonal effects on the stratification of bacterial community structure in thesoil profile of a tropical pine forest are described, anddifferences in bacterial communities are related tochemical and physical soil parameters. Culture-independent community profiles of litter, fragmentedlitter and mineral soil layers were made by denatur-ing gradient gel electrophoresis (DGGE) of 16S rDNA-specific polymerase chain reaction (PCR) fragments.The community profiles of the different soil layersclustered separately, correlating with significant dif-ferences in organic matter content between the threelayers. The bacterial communities appeared to bestable during the wet season of 1998. The drought in1997, caused by the El Niño  climatic effect, did notinfluence the bacterial communities in fragmentationand mineral soil, although moisture content and othersoil parameters were markedly lower than in the wetseason. However, communities in litter were influ- Received 18 February, 2002; revised 4 April, 2002; accepted 4 April,2002. *For correspondence. E-mail; Tel. (+31)20 444 7193; Fax (+31) 20 444 7229. † Present address: Fossil Fuelsand Environmental Geochemistry, University of Newcastle uponTyne, Newcastle upon Tyne NE1 7RU, UK.  organic materials present in the forest floor horizon.Gunadi and Verhoef (1993) and Gunadi etal  . (1998) foundthat, in the litter layer, a net immobilization of nitrogen tookplace, whereas net mobilization occurred in the fragmen-tation layer. Microbial activity, measured as dehydroge-nase activity, was positively correlated with N content. Thetotal decomposition rate in Javanese pine forest soils wasdriven mainly by seasonal moisture variations (Gunadi etal  ., 1998). The highest microbial activities occurred inthe litter layer during the tropical wet season (Gunadi etal  .,1998). Seasonal variation in the soil inorganic N contentand a correlation with the soil moisture content have beenwell established for many tropical forest ecosystems(Dubey, 1968; Vitousek etal  ., 1982; Luizao etal  ., 1992).Changes in the water regime of soil can alter its microbialcomposition (e.g. Bossio etal  ., 1998).Thus, microbial activity and nutrient availability areinfluenced by soil stratification and seasonal dynamics.We hypothesized that the microbial community structurediffers in the different soil layers and is influenced by sea-sonal fluctuations in moisture, pH and nutrient content. Totest these hypotheses, the different soil horizons in a pineforest were sampled monthly for one season, the mois-ture, soil pH and nutrient content were determined, andthe bacterial communities were analysed. Denaturing gradient gel electrophoresis (DGGE) of 16S rDNApoly-merase chain reaction (PCR) fragments (Muyzer etal  .,1993) was used to profile bacterial communities, as itallows rapid, culturing-independent analysis of largenumbers of samples. Numerical analysis of DGGE pat-terns (e.g. Eichner etal  ., 1999; Röling etal  ., 2000) wasused to compare the time series of community fingerprintsof the three soil layers. Community clustering fingerprintswere related to soil chemical variables. To obtain moredetailed insight into the phylogenetic composition of thebacterial communities, and as a measure of biodiversity,a clone library was constructed from a litter sample takenduring the rainy season when the highest activities occur.Seventy-four 16S rDNAclones were partially sequenced. Results Seasonal fluctuations in moisture and nutrient contents of different soil layers  Samples of litter, fragmented litter and mineral soil (first 10cm) were taken from the Pinus merkusii  forest at MountUngaran from September 1997 to August 1998. Samplesfrom the first 3months (September, October and Nov-ember 1997) were extremely dry because no rain fromMay until the first half of November 1997. From mid-November 1997 to August 1998, however, rain fell everymonth at a high level (data not shown) as a result of the La Niña  climatic effect. The onset of the rainy season wasclearly reflected in the increase in the moisture content ofthe three soil layers during December 1997 and January1998 to constant values, which were maintained from February until August 1998 (Fig.1A). To calculate aver-ages for the dry and wet seasons and to make statisticalanalyses, data for moisture, organic matter, ammoniumand nitrate contents and pH were pooled per soil layer forthe periods of September–November 1997 and February–August 1998 (Fig.1B–F). Mathematically significanteffects ( P  <0.01) of layer and season were observed. Thesole exception was the absence of a seasonal effect onorganic matter content (Fig.1C). The strongest seasonaleffects (increases in the moisture content, pH, ammoniumand nitrate concentrations in the wet season; Fig.1B andD–F) occurred in the litter layer, whereas the weakestchanges were noted for the mineral soil, in which the pHand ammonium content did not change significantly(Fig.1D and E). Overall, the litter and fragmentation layers were more comparable than the mineral soil,although a significant difference in the organic matter con-tents (Fig.1C) was recorded, and the moisture content ofthe litter layer was significantly lower than the fragmenta-tion layer during the dry season (Fig.1B). Bacterial community profiles  Reproducibility of DNAextraction, PCR and DGGE was extensively tested. As an example, Fig.2 shows theclustering of five DGGE tracks (coded ‘Lit0398’), whichcorrespond to the same litter sample (obtained in March1998), which was subjected to five independent DNAextractions and PCRs. The five PCRs were run on twoDGGE gels. Owing to the complexity of the fingerprints,which often showed more than 40 partially overlapping,peaks (see also Figs3 and 4), Pearson product-momentcorrelation of the densitometric curves (the intensity perpixel in the digitized DGGE tracks) was used for clusteranalysis throughout this study. This band-independent  © 2002 Blackwell Science Ltd , Environmental Microbiology  , 4 , 361–373 362 A. S. Krave et al . Fig.1. Seasonal effects on chemical parameters in the litter, fragmentation and mineral soil layers of a Pinus merkusii  (pine) forest at MountUngaran, central Java, Indonesia.A. Changes in moisture contents from September 1997 to August 1998 of the litter (filled circles), fragmentation (open triangles) and mineral(filled squares) soil layers. The error bars indicate the standard deviations, but often these were smaller than the size of the symbol.B–F. Average moisture contents (B), organic matter contents (C), pH (D), ammonium contents (E) and nitrate contents (F) in the litter,fragmentation and mineral layers during the dry (stacked bars; average of data from September–November 1997) and wet (open bars;average of data from February–August 1998) seasons. The error bars indicate standard deviation. The different letters at the top indicatesignificant differences ( P  <0.01).  Stratification and seasonal stability of highly diverse bacterial communities  363  © 2002 Blackwell Science Ltd , Environmental Microbiology  , 4 , 361–373  analysis gives objective results in contrast to methodsbased on band assignment (Rademaker etal  ., 1999).Numerical analysis using Pearson product-moment cor-relation (Fig.2) showed that the five tracks clustered at>88% similarity, with an average similarity between thetracks of 92% (standard deviation 4%). The referencemarkers, used in gel analysis and amplified from a standard mixture of 12 clones, clustered at >95% over different gels (data not shown). Asecond independentDGGE analysis of the soil samples shown in Fig.3 gavea comparable clustering (data not shown). Thus, themethods used for DNAextraction, PCR and DGGE analy-sis in this study gave reproducible results.DGGE patterns of the amplified 16S rDNAsequencesof the indigenous bacterial communities are shown inFig.3. Figure4B–D shows the densitometric curves of theDGGE tracks of the January 1998 litter, fragmented litterand mineral soil samples respectively. Adensitometriccurve of one of the marker lanes is shown in Fig.4A. Ascan be seen from the individual DGGE tracks in Figs3and 4, most variations in the profiles were observed forthe litter samples. Compared with these, bands disap-peared at lower denaturant concentrations in the profilesfrom the fragmented litter samples. The mineral soilsamples showed the lowest numbers of bands of ampli-fied 16S rDNAbut a higher percentage of bands at higherdenaturant concentrations. Also, two dominant bands(Figs2– 4, arrows) characterized the mineral soils. Thesebands were also pronounced in the fragmented litter, butwere much less prominent in the litter samples.Numerical analysis showed that the community profilesof the three horizons clustered separately (Fig.3). The fin-gerprints of the litter samples clearly separated from thecommunity profiles of the fragmented litter and mineralsoil samples (with a similarity of only 28%). The bacterialcommunity fingerprints of the fragmented litter andmineral soil clustered together with a similarly of only57%, and two groups were clearly distinguished. AllDGGE profiles of the fragmented litter samples groupedtogether with a similarity of 78%, whereas those ofmineral soil samples clustered with a similarity of 80%. Asindependent DNAextractions and PCRs from a singlesample showed 88% similarity (Fig.2), no obvious influ-ence of season on the bacterial community structure infragment litter and mineral soil was evident. In contrast, adistinct seasonal effect was recorded for the litter layer(Fig.3). DGGE profiles of bacterial communities in thelitter samples of October and November 1997, in whichthe moisture content was very low (Fig.1A), were com-pletely different (<10% similarity) from the other DGGEfingerprints of litter samples. The DGGE profile of the littersample taken early in the rainy season (December 1997),when the moisture content had not yet stabilized (Fig.1A),was also obviously different from the profiles from the wet season samples (February 1998 to August 1998).Although the January 1998 moisture content of the litterhad not stabilized, its bacterial community profile wasindistinguishable from the wet season profiles (Fig.3). Community composition  To obtain insight into the phylogenetic composition, 16SrDNAsequences in the DNAsample from the litter layerof March 1998 were amplified with the primer set 8f/1512r,which covers most of the 16S rRNAgene. This samplewas chosen because the highest microbial activities wereobserved during the wet season in the litter layer (Gunadi etal  ., 1998), which also showed increases in ammoniumand nitrate concentrations during the wet season (Fig.1Eand F).  © 2002 Blackwell Science Ltd , Environmental Microbiology  , 4 , 361–373 364 A. S. Krave et al . Fig.2. Negative image of ethidium bromide-stained normalized DGGE patterns of bacterial communities from litter of March 1998 from a Pinus merkusii  (pine) forest at Mount Ungaran, central Java, Indonesia. Clustering of the data was carried out with the GELCOMPAR softwareusing Pearson product-moment correlation and UPGMA clustering. Nest, nested PCR, first primer set F8/R1521, second primer set F341-GC/R518. The five ‘Lit0398’lanes were the result of independent DNAextractions and PCRs in order to test the reproducibility of the DNAextraction and DGGE methodology. The nested PCRs, ‘Lit0398 nest’, were performed to examine possible biases between the DGGE andcloning approaches. The two arrows indicate the positions of dominantly occurring bands in the soil layers and are discussed in the text.  Stratification and seasonal stability of highly diverse bacterial communities  365  © 2002 Blackwell Science Ltd , Environmental Microbiology  , 4 , 361–373 Fig.3. Negative image of ethidium bromide-stained normalized DGGE patterns of bacterial communities from Litter (Lit ‘month year’),fragmented litter (FragL‘month year’) and mineral soil (Mineral ‘month year’) of a Pinus merkusii  (pine) forest at Mount Ungaran, central Java,Indonesia. Clustering of the data was carried out with the GELCOMPAR software using Pearson product-moment correlation and UPGMA clustering. Denaturant concentration: 45% (left) to 65% (right). The two arrows indicate the positions of dominantly occurring bands in the soillayers and are discussed in the text.
Related Documents
View more...
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