Plant Soil (2010) 329:127–137 DOI 10.1007/s11104-009-0140-9
REGULAR ARTICLE
Molecular characterization of soil bacterial communities in contrasting zero tillage systems Javier A. Ceja-Navarro & Flor N. Rivera & Leonardo Patiño-Zúñiga & Bram Govaerts & Rodolfo Marsch & Antón Vila-Sanjurjo & Luc Dendooven
Received: 27 April 2009 / Accepted: 10 August 2009 / Published online: 2 September 2009 # Springer Science + Business Media B.V. 2009
Abstract It is well known that agricultural practices change the physical and chemical characteristics of soil. As a result, microbial populations can also be affected. The aim of this study was to analyze the effect on soil bacterial communities of zero tillage (ZT) under maize monoculture (MM) with crop residue removal (-R) (MM/-R treatment), compared to a ZT system under wheat monoculture (WW) with crop retention (+R) (WW/+R treatment). Phylogenetic analysis was used to characterize soil bacterial communities. Phylogenetic groups found exclusively in MM/-R were Caldilineales, Chromatiales, Oscillatoriales, Legionellales, Nitrosomonadales and unclassified ∂-Proteobacteria, while
Bacillales, Burkholderiales, Pseudomonadales and Rubrobacteriales were found only in WW/+R. Sequences of bacteria related to fluorescent Pseudomonas sp. were detected only in WW/+R. Acidobacteria, a largely unknown group of bacteria, were the dominant group in both treatments with a relative proportion of 0.703 and 0.517 for MM/+R and WW/-R respectively. It was found that zero tillage with removal of crop residue in soil cultivated with a monoculture of maize strongly reduced microbial diversity (H=3.30; D= 0.9040) compared to soil where crop residue was retained in a wheat zero tillage situation (H=4.15; D= 0.9848).
Responsible Editor: Petra Marschner.
Keywords Zero tillage . Residues management . 16S rDNA . Phylogenetic analysis
J. A. Ceja-Navarro (*) : F. N. Rivera : L. Patiño-Zúñiga : R. Marsch : L. Dendooven Department of Biotechnology and Bioengineering, Laboratory of Soil Ecology, Cinvestav, Av. Instituto Politécnico Nacional 2508, C.P. 07360 México, D.F., México e-mail:
[email protected] B. Govaerts International Maize and Wheat Improvement Center (CIMMYT), Apdo. Postal 6-641, 0600 México, D.F., México A. Vila-Sanjurjo Molecular and Cell Biology Department, University of California Berkeley, Berkeley, CA 94720, USA
Introduction Intensively cropped highlands in the world are prone to erosion and decreased soil quality (Hobbs et al. 2008). In the central highlands of Mexico, farmers apply continuous maize monoculture, use heavy mechanical tillage and remove crop residues for cattle feed. Although inorganic fertilizer inputs have increased, yields remain low due to a decline in soil fertility, soil structure deterioration and poor weed control (Fischer et al. 2002). It has been reported that practices such as reduced tillage and crop residues retention improve soil structure, increase soil micro-
128
bial biomass (SMB) content and soil fertility (Karlen et al. 1994; Alvear et al. 2005). Other studies have found that residue retention on the soil surface and reduced tillage increased soil microbial biomass and can mitigate CO2 and N2O emissions (Wardle 1992; Six et al. 2004). Despite these findings, our understanding of the composition of the soil microbial population and its biological functioning is still limited. To understand the roles played by microorganisms in the physical-chemical processes that shape agricultural soils, we first need to elucidate the composition of the soil microbial communities. To study the effects of agricultural practices on the soil, a long-term field experiment was started by International Maize and Wheat Improvement Center (CIMMYT) in which management practices, including crop rotation, tillage and residue management were compared. A total of sixteen possible combinations of soil management practices (conventional tillage, zero tillage, maize and wheat monoculture and rotations, residues removal and retention) were analyzed by Govaerts et al. (2005, 2006a, b). These authors reported that among all treatments, soils with zero tillage, crop residue retention, and continuous wheat monoculture (ZT/WW/+R) showed the largest crop yields, best soil quality, least amount of root rot, and lowest number of nematodes. In contrast, soils under zero tillage, with residue removal, and continuous maize monoculture (ZT/MM/-R) displayed the worst soil parameters of all sixteen treatments. The effect of these practices on micro-flora groups was also analyzed by plate count. From this study Govaerts et al. (2008) reported that among groups, fluorescent Pseudomonas were the most sensitive to management practices. To further characterize soil bacterial communities, WW/+R and MM/-R were subjected to phylogenetic analysis.
Plant Soil (2010) 329:127–137
typify the rainy season and evapotranspiration exceeds rainfall throughout the year, as total amount of yearly potential evapotranspiration is 1900 mm. The growing period at the El Batán experimental station has an average length of 152 days (FAO 1978). The El Batán research station is located near the former lake Texcoco. The soil is classified as a Cumulic Phaeozem in the World Reference Base system (IUSS Working Group WRB 2006) and as a fine, mixed, thermic Cumulic Haplustoll in the USDA Soil Taxonomy system (Soil Survey Staff 1998). The soil is characterized by good chemical and physical conditions for farming. The major limitations are periodical drought, periodical water excess and windand water erosion. The study described here is part of a long-term trial, which began in 1991. Details of this experiment can be found in Govaerts et al. (2008). Soil sampling Soil samples were taken from replicated plots at the end of the fallow period. Ten sub-samples (0–15 cm) were collected at random from each plot with a 2 cm diameter auger and mixed to yield one composite sample per treatment (Brons and van Elsas 2008). The soil was 5-mm sieved and the organic residues, e.g. roots and stones, removed. Soil samples were kept on ice during transport (1/S (Camargo 1992). Nucleotide sequence accession numbers The 16S rDNA gene sequences obtained in this study has been deposited in the GenBank database under accession numbers: EU192952-EU193040, EU193041-EU193123, EU276428-EU276578, EU449739-EU449769 and EU477104.
Community richness and composition analysis
Results
The distance matrixes (generated in PAUP* 4.0b10 software) were used to obtain the operational taxonomic units (OTUs) for each library. A 3% distance level between sequences was considered the cutoff among different OTUs. Rarefaction curve, richness estimator (bias-corrected Chao1 and Bootstrap), and diversity indices (Shannon’s (H) and Simpson’s (D)) were determined using DOTUR software version 1.51 (Schloss and Handelsman 2005; http://www.plantpath. wisc.edu/fac/joh/dotur.html) for each treatment. LIBSHUFF analysis (Singleton et al. 2001) was used to compute the differences between the clone libraries and estimated homologous and heterologous coverage within and between libraries from different treatment.
Initial characterization of the soil at the experimental site showed features similar to those found in agricultural soils of the central highlands of Mexico (Table 1), which were in good agreement with the results of Govaerts et al. (2008), so are the data obtained for the SMB, organic C and total N. The analysis of soil microbial communities by molecular procedures requires the extraction of high molecular size DNA, lysis of the microorganisms within the sample, and the extraction of DNA free from inhibitors. By using AlNH4(SO4)2 to removed humic and fulvic acids, we have been able to extract DNA with an average size of 20 Kb, with a total DNA yield of approximately 28.2 – 39.1 ng g−1 soil. After
Table 1 Physical and chemical soil characterization Treatment
WHC g kg−1 soil
Organic C
Total N
SMB (mg kg−1)
pH
EC (dS m−1)
MM/-R
474 (0.07)
12.2 (0.08)
1.27 (0.02)
376 (20)
6.1 (0.03)
1.08 (0.06)
WW/+R
439 (0.05)
12.9 (0.03)
1.60 (0.06)
402 (0.2)
6.2 (0.06)
0.58 (0.04)
Values between parentheses are standard errors of the estimates WHC Water holding capacity, SMB soil microbial biomass, EC electrolytic conductivity
Plant Soil (2010) 329:127–137
131
132
Plant Soil (2010) 329:127–137
Plant Soil (2010) 329:127–137
133
Fig. 2
Phylogenetic reconstruction based on maximum parsimony. Majority rule consensus tree of the 48 most parsimonious tree among clones of treatment WW/+R and GeneBank selected sequence of bacteria based on the 16S rDNA gene sequences (three length= 1633.27 steps; CI=0.3585; RI= 0.8607). Only bootstrap values greater than 50% are indicated (100 replicates). GenBank accession numbers of sequences of the most closely related bacteria are shown in parentheses. The scale bar represents the expected number of substitutions per nucleotide position. Desulfurobacterium thermolithotrophum was used as an out-group
PCR amplification of 16S rDNA, 172 clones were used for either phylogenetic reconstruction. Phylogenetic reconstructions were performed by distance (data not shown) and maximum parsimony methods, obtaining similar tree topologies. Figures 1 and 2 show the phylogenetic relationship between amplified 16S rDNA sequences from WW/+R and MM/-R with known 16S rDNA sequences from GeneBank. Phylogenetic groups were homogenized on the order level. Fourteen different bacterial groups were detected for MM/-R and twelve for WW/+R. After phylogenetic analysis of the MM/-R clones, 12 Table 2 Relative proportion of bacterial communities for treatments MM/-R and WW/+R basis on obtained phylogenetic groups
Phylogenetic Groups
Acidobacteriales
of them allowed characterization to the taxonomic level of genus, 22 to the level of family, and 138 to the level order. In contrast, after analysis of the clones obtained from WW/+R soil, 23 clones allowed characterization to the level of species, 18 to the level of genus, 21 to the level of family, and 110 the level of order. For both soil treatments, the most abundant groups were Acidobacteriales, with relative proportions of 0.703 and 0.512; Sphingomonadales with 0.070 – 0.122; and Xanthomonadales with 0.017 – 0.064, for MM/-R and WW/+R respectively (Table 2). Other groups found in both treatments were Actinomycetales, Gemmatimonadales, Myxococcales, Rhizobiales, and Rhodospirillales. Caldilineales, Chromatiales, Oscillatoriales, Legionellales, Nitrosomonadales, and unclassified ∂-Proteobacteria were exclusively found in MM/-R soil, whereas Bacillales, Burkholderiales, Pseudomonadales and Rubrobacteriales were found only in the WW/+R soil. The phylogenetic reconstruction of the Acidobacteriales, the most abundant group, was obtained once the global screening of all clones had been done. This reconstruction indicated that the Acidobacteriales MM/-R
WW/+R
Clones
pi
121
0.703
Clones 89
pi 0.517
Actinomycetales
8
0.047
2
0.012
Bacillales
–
–
2
0.012
Burkholderiales
–
–
3
0.017
Caldilinelaes
2
0.012
–
–
Chromatiales
4
0.023
–
–
Gemmatimonadales
4
0.023
8
0.047
Legionellales
1
0.006
–
–
Myxococcales
1
0.006
8
0.047
Nitrosomonadales
3
0.017
–
–
Oscillatoriales
2
0.012
–
–
Pseudomonadales
–
–
11
0.064
Rhizobiales
2
0.012
7
0.041
Rhodospirillales
8
0.047
5
0.029
Rubrobacteriales
–
–
5
0.029
Sphingomonadales
12
0.070
21
0.122
Xanthomonadales
3
0.017
11
0.064
Unclassified ∂ – Proteobacteria
1
0.006
–
Total Clones Groups (S) Frequency (1/S)
172
172
14
12
0.07
0.08
–
134
clones were distributed in seven of the eight subdivisions of the Acidobacteria phylum (Fig. 3a, b). Subdivisions 1, 3, 4, 5, 6 and 7 were well represented in WW/+R, whereas subdivisions 1, 4, 5 and 6 are the most abundant in MM/-R soils (Fig. 3). LIBSHUFF analysis of the sequence libraries generated for either treatment indicated that the Fig. 3 Phylogenetic reconstruction based on maximum parsimony for Acidobacteria groups. a Majority rule consensus tree of the 48 most parsimonious tree among Acidobacteria clones of treatment WW/+R based on the 16S rDNA gene sequences (three length=1633.27 steps; CI= 0.3585; RI=0.8607). b Majority rule consensus tree of the 8 most parsimonious tree among Acidobacteria clones of treatment MM/-R (three length=1287.25 steps; CI=0.4045; RI= 0.7859). Only bootstrap values greater than 50% are indicated. Brackets at the right of the tree indicate subdivisions and percentage distributions. GenBank accession numbers of sequences of the most closely related bacteria are shown in parentheses
Plant Soil (2010) 329:127–137
differences in their composition were due to the distinct makeup of the communities from which they were derived (P
