Community structures and substrate utilization of bacteria in soils from organic and conventional farming systems of the DOK long-term field experiment

Abstract

Preservation or improvement of soil quality and productivity is of major importance for sustainable agriculture. Microorganisms strongly influence these soil characteristics as they are involved in nutrient cycling, transformation processes and soil aggregate formation, as well as in plant pathology or plant growth promotion. A profound understanding of structure, dynamics and functions of soil microbial populations represents one key to the understanding and description of soil quality. Therefore, we analyzed long-term effects of three farmyard manure (FYM)-based farming systems, i.e. bio-dynamic (BIODYN), bio-organic (BIOORG) and conventional (CONFYM), on microbiological soil characteristics and compared them to long-term effects of minerally fertilized (CONMIN) and unfertilized (NOFERT) control systems. Furthermore, we compared these long-term effects of farming systems to short-term effects of the crops winter wheat and grass-clover ley. The DOK field experiment in Therwil, Switzerland, which was established in 1978, represents in a unique long-term comparison, allowing to approach these questions. Effects on microbiological soil characteristics were assessed with a polyphasic approach by analyzing soil microbial biomass, soil DNA content, colony forming unit (CFU) counts, community level substrate utilization (CLSU) patterns with Biolog™ EcoPlates, and terminal restriction fragment length polymorphism (T-RFLP) profiles of bacterial 16S rRNA genes. The soil biomass parameters, i.e. microbial biomass, DNA content and CFU, were all strongly influenced by the farming systems, whereas only CFUs were significantly affected by the two crops analyzed. Differences among the FYM-based farming systems BIODYN, BIOORG and CONFYM were only significant for microbial biomass and DNA content. CLSU and T-RFLP profiling, on the other hand, allowed for consistent differentiation of soil bacterial community structure in relation to the influence of farming systems and crops. The analyses revealed that the main and highly significant effect on microbiological soil characteristics was related to FYM applications. Less strong but significant effects were caused by the two crops, i.e. winter wheat and grass-clover. Effects of the farming systems BIODYN, BIOORG and CONFYM on soil bacterial community structure were relatively weak and not significant. These results suggest that for successful soil quality management fertilization regime and crop rotation are of major importance and that polyphasic approaches are needed to describe and assess microbiological soil characteristics.

Full text

Page 1

Community structures and substrate utilization of bacteria
in soils from organic and conventional farming systems
of the DOK long-term field experiment
Franco Widmera,*, Frank Raschea,b,1, Martin Hartmanna, Andreas Fliessbachb
aMolecular Ecology, Agroscope FAL Reckenholz, Swiss Federal Research Station for Agroecology
and Agriculture, Reckenholzstrasse 191, CH-8046 Zu ¨rich, Switzerland
bResearch Institute of Organic Agriculture (FiBL), Ackerstrasse, CH-5070 Frick, Switzerland
Received 12 April 2005; received in revised form 23 September 2005; accepted 27 September 2005
Abstract
Preservation orimprovementofsoil qualityand productivityis ofmajorimportanceforsustainableagriculture. Microorganisms
strongly influence these soil characteristics as they are involved in nutrient cycling, transformation processes and soil aggregate
formation, as well as in plant pathology or plant growth promotion. A profound understanding of structure, dynamics and functions
of soil microbial populations represents one key to the understanding and description of soil quality. Therefore, we analyzed long-
term effects of three farmyard manure (FYM)-based farming systems, i.e. bio-dynamic (BIODYN), bio-organic (BIOORG) and
conventional (CONFYM), on microbiological soil characteristics and compared them to long-term effects of minerally fertilized
(CONMIN) and unfertilized (NOFERT) control systems. Furthermore, we compared these long-term effects of farming systems to
short-term effects of the crops winter wheat and grass-clover ley. The DOK field experiment in Therwil, Switzerland, which was
established in 1978, represents in a unique long-term comparison, allowing to approach these questions. Effects on microbiological
soil characteristics were assessed with a polyphasic approach by analyzing soil microbial biomass, soil DNA content, colony
forming unit (CFU) counts, community level substrate utilization (CLSU) patterns with BiologTMEcoPlates, and terminal
restriction fragment length polymorphism (T-RFLP) profiles of bacterial 16S rRNA genes. The soil biomass parameters, i.e.
microbial biomass, DNA content and CFU, were all strongly influenced by the farming systems, whereas only CFUs were
significantly affected by the two crops analyzed. Differences among the FYM-based farming systems BIODYN, BIOORG and
CONFYM were only significant for microbial biomass and DNA content. CLSU and T-RFLP profiling, on the other hand, allowed
for consistent differentiation of soil bacterial community structure in relation to the influence of farming systems and crops. The
analysesrevealedthatthemainandhighlysignificanteffectonmicrobiologicalsoilcharacteristicswasrelatedtoFYMapplications.
Less strong but significant effects were caused by the two crops, i.e. winter wheat and grass-clover. Effects of the farming systems
BIODYN, BIOORG and CONFYM on soil bacterial community structure were relatively weak and not significant. These results
suggest that for successful soil quality management fertilization regime and crop rotation are of major importance and that
polyphasic approaches are needed to describe and assess microbiological soil characteristics.
# 2005 Elsevier B.V. All rights reserved.
Keywords: Organic agriculture; Conventional agriculture; Soil biomass; Soil bacterial community structure; Effect study
1. Introduction
In agriculturally managed ecosystems preservation
or improvement of soil quality and productivity is of
major importance. Soil quality definitions include
www.elsevier.com/locate/apsoil
Applied Soil Ecology 33 (2006) 294–307
* Corresponding author. Tel.: +41 44 377 73 76;
fax: +41 44 377 72 01.
E-mail address: franco.widmer@fal.admin.ch (F. Widmer).
1Present address: ARC Seibersdorf Research GmbH, Department
of Bioresources, Microbiology, A-2444 Seibersdorf, Austria.
0929-1393/$ – see front matter # 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.apsoil.2005.09.007

Page 2

physical,chemicalandbiologicalsoilcharacteristicsthat
are often closely interrelated (Doran and Zeiss, 2000).
Soil microbiota play an important role in these soil
characteristics since many of them are involved in
nutrient cycling, transformation processes and soil
aggregate formation, as well as in plant pathology or
plant growth promotion (Kennedy, 1999; Buckley and
Schmidt, 2001). Understanding structure, dynamics and
functions of soil microbial communities represents one
key to the understanding of soil fertility and soil quality
(Kennedy,1999;BuckleyandSchmidt,2003).However,
due to technical limitations, it is currently difficult to
describe dynamics of microbial communities and to
assesstheirroleinecosystemfunctions(DoranandZeiss,
2000;BuckleyandSchmidt,2003).Therefore,oneofthe
current challenges is to assess changes or differences in
microbial diversity in soils, with respect to community
composition and species distribution (Kennedy, 1999;
Hilletal.,2000;Kirketal.,2004).Ifaffectedpopulations
can be identified, future research then may focus on
affiliation of specific populations and their functions in
order to assess functional consequences or functional
redundancy in soil (Torsvik and Ovreas, 2002).
Microbial community structures in agriculturally
managed soils need to be reproducibly detectable and
the identity and possible functions of detected popula-
tions need to be described. Then, agricultural manage-
ment-dependentinfluences onsoilmicrobialcommunity
structuresmaybeassessedandindicativecomponentsbe
used for biological soil quality diagnosis (Kennedy,
1999).Inordertoachieve thesetasks,effectsofdifferent
agricultural management systems on microbiological
soil characteristics in well documented and designed
field experiments are necessary.
Changes in soil quality may develop slowly and may
adjust to a new long-term steady state after a change of
managementorconversiontoadifferentfarmingsystem.
Therefore, long-term agricultural field experiments are
particularly valuable to detect changes that would not be
detectable in short-term studies (Powlson and Johnston,
1994). The DOK long-term field experiment in Therwil,
Switzerland (Ma ¨der et al., 2002) has been established in
1978 and initially focused on the feasibility and
agricultural productivity of organic farming systems.
In recent years more emphasis has been given to the
effects of farming systems on soil quality (Ma ¨der et al.,
2002), which may be reflected in various biological soil
characteristics like soil microbial biomass, community
structures, functions and activities. Therefore, the
approach of assessing these characteristics in the DOK
long-term field experiment represents a relevant and
attractive approach. Currently, various methods are
available for the analysis of microbiological soil
characteristics, each with its advantages and disadvan-
tages (Hill et al., 2000; Kirk et al., 2004).
Determining community level substrate utilization
(CLSU) patterns is one approach for the characteriza-
tion of microbial communities and is based on specific
substrate utilizing functions performed by aerobic
heterotrophic bacteria. CLSU analyses with the
BiologTMsystem (Garland and Mills, 1991) have been
applied in many studies in order to gain information on
differences of microbial communities in various soil
systems (Grayston et al., 1998; Gomez et al., 2000;
O’Donnell et al., 2001; Larkin, 2003; Fliessbach and
Ma ¨der, 2004). Although the BiologTMtechnique is
restricted to culturable microorganisms and does not
provide a precise assessment of the functional proper-
ties of microbial communities in soil (Preston-Mafham
et al., 2002), it has been successfully applied in
comparative soil analyses and has shown to be a
powerful and sensitive low cost analytical tool for
demonstrating differences or changes in soil micro-
biological characteristics (Fliessbach and Ma ¨der, 1997;
Gomez et al., 2000; Widmer et al., 2001).
Only a small portion of the whole microbial diversity
of ecosystems has been isolated or adequately char-
acterized(Tiedjeetal.,1999;AmannandLudwig,2000).
Important progress towards cultivation-independent
characterization of soil microbial community structures
was achieved by direct extraction of microbial commu-
nity DNA from soils (Torsvik et al., 1990; Bu ¨rgmann
et al., 2001) and development of PCR-based, group-
specific detection protocols for microbial phyla, gen-
erally based on ribosomal RNA genes (rDNA) as
phylogenetic markers (Amann and Ludwig, 2000; Hill
et al., 2000; Theron and Cloete, 2000). Genetic profiling
of PCR-amplified small subunit (SSU) rDNA, such as
restriction fragment length polymorphism (RFLP)
analyses (Widmer et al., 2001), denaturing or tempera-
ture gradient gel electrophoresis (DGGE/TGGE; Muy-
zer, 1999) or single strand conformation polymorphism
(SSCP; Schwieger and Tebbe, 1998) have successfully
been applied for assessing microbial community
structures in various soil systems. Terminal restriction
fragment length polymorphism (T-RFLP) analysis (Liu
et al., 1997) or ribosomal intergenic spacer analysis
(RISA; Fisher and Triplett, 1999) with capillary
electrophoresis enable to characterize highly diverse
soil microbial communities and has considerably higher
resolution than common gel electrophoretic systems.
Depending on the PCR detection primers applied,
different microbial groups and populations as well as
different marker genes can be detected (Braker et al.,
F. Widmer et al./Applied Soil Ecology 33 (2006) 294–307295

Page 3

2001; Brodie et al., 2003; Wolsing and Prieme, 2004).
T-RFLPanalysesofSSUrDNAcanprovideresolutionat
the genus or even at the species level (Dunbar et al.,
2001).TheT-RFLPanalysisyieldsnumericdatathatcan
be further evaluated with standard statistic analyses and
results may be compared with data stored in sequence
databases,e.g.theRibosomalDatabaseProject(RDP)or
with data from other studies.
Organicfarmingpracticesareconsideredecologically
more sustainable than conventional farming practices
and to support soil microbial diversity and functions
(Ma ¨deretal.,2002).Therefore,itwasourobjectivetouse
soils from the DOK long-term field experiment, in order
to assess differences in microbiological soil character-
istics, and to relate them to effects of long-term bio-
dynamic,bio-organicandconventionalfarming.Inorder
to rank the magnitude of detected effects, we compared
them to long-term effects of minerally and unfertilized
controls as well as to short-term effects of two crops.
Soils from the DOK field experiment allowed to
approach these questions in a unique long-term field
experiment with a well designed split–split plot design
and three temporally shifted crop rotation parallels.
Effects on microbiological soil characteristics were
assessed with a polyphasic approach by analyzing soil
microbial biomass, soil DNA contents, colony forming
unit (CFU) counts, CLSU patterns with BiologTM
EcoPlates and T-RFLP of bacterial SSU rDNA.
2. Materials and methods
The DOK long-term field experiment was estab-
lished in 1978 in Therwil, Switzerland, on a haplic
luvisol on deep deposits of alluvial loess (Ma ¨der et al.,
2002).The experiment has been designed for evaluation
of agronomic and ecological effects of bio-dynamic
(BIODYN), bio-organic (BIOORG) and conventional
(CONFYM) farming systems. The two organic systems
have been maintained according to the regulations of
the respective organic producer organizations (Eidg.
Volkswirtschaftsdepartement, 1997), while the conven-
tional system is managed since 1992 according to the
Swiss guidelines for integrated farming (Eidg. Volks-
wirtschaftsdepartement, 1998). The farming systems
mainly differ in fertilization practice and plant
protection strategy (Fig. 1). The three main systems
(BIODYN, BIOORG and CONFYM) are fertilized with
system-specificfarmyardmanure(FYM)corresponding
to 1.4 livestock units ha?1, which represents approxi-
mately 2 metric tonnes organic carbon per hectare and
year (Ma ¨der et al., 2002). Fertilization in the CONFYM
system was supplemented with mineral fertilizers (N, P
and K) according to official recommendations. The
system CONMIN mimics a conventional system with-
out livestock and is fertilized with mineral fertilizers
only, but was left unfertilized during the first 7 years of
the study. NOFERT represents the unfertilized control,
F. Widmer et al./Applied Soil Ecology 33 (2006) 294–307296
Fig. 1. Farming system-specific differences in fertilization and plant protection in the DOK long-term field experiment. For details on the
experiment refer to Besson and Niggli (1991) and Ma ¨der et al. (2002). Farmyard manure (FYM) application of 1.4 livestock units ha?1year?1was
performed with aerobically treated composted FYM (C/N = 8), slightly aerobically treated rotted FYM (C/N = 11); anaerobically treated stacked
FYM (C/N = 12). CuSO4was used for plant protection in BIOORG potato until 1991.

Page 4

which is only treated with bio-dynamic field prepara-
tions. The 7-year crop rotation was designed according
toscientific and practical needsand hasbeen adjusted in
1999 to (1) potato, (2) winter wheat 1, (3) soybean, (4)
maize and (5) winter wheat 2, followed by 2 years of a
grass-clover ley. The first five crops represent the arable
phase of the crop rotation, whereas the 2 years of
permanent grass-clover ley without tillage represent the
recoveryphase.Croprotationandsoiltillagerepresenta
compromise of the onestypicallyapplied in organicand
conventional agricultural practice and were identical in
all farming systems. Crop rotation was repeated in three
temporally shifted parallels in the field. All farming
systems and crops were replicated four times in a split–
split plot design.
Soil samples were recovered on March 14th 2000
from the field plots planted with winter wheat 1 and a
2nd year grass-clover ley. Winter wheat represented the
arable phase of the crop rotation, whereas grass-clover
represented the end of the soil recovery phase after 2
years of permanent plant cover without soil tillage.
From each of the four replicate plots of the systems
BIODYN, BIOORG, CONFYM,
NOFERT, 14 single cores of 3 cm diameter were taken
from the plough layer (0–20 cm), pooled, and main-
tained at 4 8C. In the laboratory soil samples were
carefully dried at room temperature to reach approxi-
mately 25% water content, which represents 45% of
maximum water holding capacity. Soil water content
was determined gravimetrically by drying soils for 24 h
at 105 8C. Soils were then sieved (2 mm) and soil pH
was determined in a soil suspension diluted 1:10 (w/v)
with CaCl2(25 mM). Soils were stored for up to 1 week
at 4 8C prior to use.
Soilmicrobialbiomass(Cmic)wasestimatedusingthe
chloroform-fumigation-extraction(CFE)method(Vance
et al., 1987). Soils were equilibrated for 7 days at 20 8C
and triplicate sub-samples of 20 g (dry weight equiva-
lent) were fumigated with CHCl3 for 24 h at room
temperature. Fumigated and control soil samples were
extracted with 80 ml 0.5 M K2SO4(90 min at 300 rpm),
filtered (Macherey Nagel 615) and total organic carbon
(TOC) was determined by infrared spectrometry after
combustionat850 8C(DIMA-TOC100,Dimatec,Essen,
Germany). Cmicwas calculated from the difference in
extractable carbon of fumigated and unfumigated
samples using a kECfactor of 0.45 (Joergensen, 1996).
Community level substrate utilization (CLSU)
analysis was performed with sieved soil pre-incubated
for 6 days at 20 8C. Three replicate sub-samples of 10 g
(dry weight equivalent) were suspended (30 min at
300 rpm) in 90 ml sterile saline solution (0.8% NaCl).
CONMIN and
Soil suspensions were allowed to settle for 10 min
before the supernatant was diluted 10-fold to obtain a
final dilution of 10?2(Fliessbach and Ma ¨der, 1997).
Each well of a BiologTM-EcoPlate (Biolog Inc.
Hayward, CA, USA) was filled with 125 ml of the
final dilution (Garland and Mills, 1991). Inoculation
density was determined by counts of colony forming
units on a glucose minimal medium (Pochon and
Tardieux, 1962). From each soil sample three replicate
plates, with three replicate substrate sets were used
(n = 9). Including the four replicate field plots, a total of
36 replicate data sets were prepared for each treatment.
Plates were incubated at 20 8C and optical density at
600 nm (OD600) was read periodically in a microplate
reader (MRX, Dynex Technologies, Inc., Chantilly,
USA) at 12 predetermined time points between 24 and
96 h of incubation. Individual absorbance values of the
31 single substrates were corrected by subtraction of the
blank control value (raw difference; RD). Negative RD-
values were set to zero. To minimize effects of different
inoculum densities, data were normalized by dividing
the RD values by their respective average well colour
development (AWCD) values.
Soil DNA extraction was performed on sieved soils
usingaslightlymodifiedbeadbeatingprotocoldeveloped
by (Bu ¨rgmann et al., 2001). Approximately 0.5 g fresh
soil and 0.75 g silica beads (diameter 0.10–0.11 mm;
Braun Biotech International GmbH, Melsungen, Ger-
many) were suspended in 1.3 ml extraction buffer (0.2%
hexadecyltrimethylammonium-bromide (CTAB), 1 mM
dithiotreitol (DTT), 0.2 M sodium phosphate, 0.1 M
sodium chloride, 50 mM EDTA, pH 8.0) and processed
for 40 s using a FP 120 bead beater (Savant Instruments,
Inc., Holbrook, NY) at setting 5.5. Samples were
centrifuged (14,000 ? g, room temperature, 5 min) and
supernatantsweretransferredtofreshtubes.Eachsample
was extracted two additional times with 1 ml extraction
bufferandallthreecorrespondingextractswerepooledin
the same tube, yielding a total extract of approximately
3 ml.Followingextractionwith3 mlchloroform,nucleic
acids were precipitated with 3 ml precipitation solution
(20% polyethylenglycol 6000, 2.5 M NaCl) for 1 h at
37 8C followed by centrifugation (14,000 ? g, room
temperature 15 min). Pellets were washed with 800 ml
70% ethanol, air dried and resuspended in 1 ml TE
(10 mM Tris–HCl, 1 mM EDTA, pH 8.0) per gram
extractedsoil(dryweightequivalent).DNAwasstoredat
?20 8C until further processing.
Quantification of DNA yield was performed with
PicoGreen (Molecular Probes, Eugene, OR, USA)
according to (Bu ¨rgmann et al., 2001). Two microlitres
of PicoGreen, 2 ml DNA-extract, and 396 ml TE-buffer
F. Widmer et al./Applied Soil Ecology 33 (2006) 294–307 297

Page 5

were mixed and maintained at room temperature until
fluorometric quantification at 480 nm excitation and
520 nm emission (Luminescence Spectrometer LS 50
B, Perkin-Elmer, Rotkreuz, Switzerland). Bacterioph-
age l DNA (Promega, Madison, WI, USA) was used as
DNA concentration standard. DNA-yield was presented
as mg extracted DNA g?1soil (dry weight equivalent).
PCR-amplification of bacterial SSU rDNA was
performed by using primers 27F (50-AGAGTTTGATC-
MTGGCTCAG-30)and1378R
CAAGGCCCGGGAACG-30) (Heuer et al., 1997).
Primer 27F was 50-labeled with carboxyfluorescein
(FAM6, Microsynth, Balgach, Switzerland). Reaction
mixes of 50 ml contained 3 ng soil DNA, 1? PCR
reaction buffer (Qiagen, Hilden, Germany), 1.5 mM
MgCl2, 0.2 mM of each primer, 0.2 mM of each
desoxynucleoside triphosphate,
(Sigma, Buchs, Switzerland) and 1 U HotStar DNA
polymerase (Qiagen). PCR conditions were: initial
denaturation for 15 min at 95 8C followed by 37 cycles
consisting of denaturation for 45 s at 94 8C, primer
annealing for 30 s at48 8C andpolymerization for 2 min
at72 8Cfollowedbya finalextensionfor5 minat72 8C.
Quality of PCR products was inspected by electrophor-
esis of 5 ml PCR product in 1% (w/v) agarose gels (Life
Technologies, Paisley, Scotland) containing ethidium
bromide (0.5 mg ml?1).
For terminal restriction fragment length polymorph-
ismanalysisPCRproducts(45 ml)wereprecipitatedwith
45 ml isopropanol for 1 h at ?20 8C and centrifuged at
10,000 ? gfor15 minatroomtemperature,followedbya
wash-step with 100 ml ethanol (70%). Air-dried pellets
wereresuspendedin20 mlrestriction-mix(2 Urestriction
enzyme MspI in 1? reaction buffer B; Promega
Corporation,Madison,WI,USA)anddigestedovernight
at 37 8C. Quality of digests was inspected by gel-
electrophoresisof7 mldigestsin3.0%(w/v)MetaPhor1
gels (FMC, BioProducts, Rockland, ME, USA) contain-
ing ethidium bromide (0.5 mg ml?1). One microlitre of
MspI-digestedPCRproductswasmixedwith12 mlHiDi
formamide (AppliedBiosystems,FosterCity,CA,USA)
and 0.2 ml internal size standard (2500 TAMRATMSize
Standard,AppliedBiosystems),followedbydenaturation
at 92 8C for 2 min. T-RFLPs were analyzed by capillary
electrophoresis on an automated sequencer (ABI 310
Genetic Analyzer, Applied Biosystems) equipped with a
47 cm capillary and POP-4TMpolymer (Applied Bio-
systems) and the GeneScan software V3.1 (Applied
Biosystems). The baseline threshold for signal detection
was set to 50 fluorescence intensity units. Electropher-
ograms obtained were transformed into numeric data of
individual peak heights using the Genotyper 3.6 NT
(50-CGGTGTGTA-
1.2 mg ml?1
BSA
software(AppliedBiosystems).Manualpeakcallingwas
performedforpeakswhoseheightscouldunambiguously
bequantifiedinallsamples.Thevaluesofallscoredpeaks
were compiled in a data matrix. To minimize effects of
different PCR product quantities, data were normalized
bydividingthevalueofeachT-RFbytheaveragevalueof
all T-RFs from the corresponding sample.
Various descriptive and discriminative statistical
analyses were applied for comparison and evaluation
of data. Soil microbial biomass parameters, i.e. Cmic,
CFU and soil DNA content, were correlated by applying
the Pearson Product-Moment correlation coefficient
(JMP software; SAS Institute Inc., Cary, NC, USA).
Soil parameters were tested for significant differences in
relationtoallfactors,i.e.farmingsystemsandcrops,orin
relation to single factors, i.e. farming system or crop, as
well as for interaction between farming system and crop
using two-way analysis of variance (ANOVA) and post
hoc Tukey tests (JMP software). Explorative statistical
analyses of mean-transformed CLSU- and T-RFLP-data
were based on cluster analysis with Euclidean distances
andWardclustering(StatisticaVersion6.1;StatSoftInc.,
Tulsa, OK, USA) according to Blackwood et al. (2003).
Significant correlations of CLSU or T-RFLP finger-
print data with effects of farming systems or crops were
determined by applying Monte Carlo permutation
testing with a linear model (CANOCO for Windows
4.5; Microcomputer Power, Ithaca, NY, USA) according
to ter Braak and Smilauer (2002). Permutation tests
were conducted on all canonical axes with 1000
permutations. If only subsets of the data were analyzed
the Bonferroni correction of significance levels was
applied (Shaffer, 1995).
Influence of farming systems and crops on total
variance of the CLSU and T-RFLP data sets were
determined by partitioning the variance (CANOCO for
Windows 4.5) using redundancy analysis of a linear
model according to Borcard et al. (1992).
Two-way ANOVA (JMP software) was used to
determine significant effects of farming systems and
crops on each T-RF or CLSU value, as well as
interaction between farming system and crop.
3. Results
3.1. Soil microbial biomass (Cmic) and colony
forming units
Farming systems receiving farmyard manure, i.e.
BIODYN, BIOORG and CONFYM, showed signifi-
cantly higher (average increase 37%; p < 0.001) soil
microbial biomass (Cmic) contents, when compared to
F. Widmer et al./Applied Soil Ecology 33 (2006) 294–307 298

Page 6

systems without FYM application, i.e. CONMIN and
NOFERT (Tables 1 and 2). These differences were also
significant (p < 0.01) when analyzed separately for
winter wheat and grass-clover. Cmic-content of soils
from CONFYM averaged at 15% lower values when
compared to organic systems BIODYN and BIOORG,
but differences were not significant (p ? 0.05).
Differences between CONMIN and NOFERT as well
as between BIODYN and BIOORG were also not
significant (p ? 0.05). Soil Cmic-content varied slightly
between winter wheat and grass-clover, but revealed no
significant differences (p ? 0.05). Two-way ANOVA
showed significant (p < 0.001) changes of microbial
biomass among the farming systems, but no significant
(p ? 0.05) changes were found in relation to the crops
(Table 2). Interaction between farming systems and
crops was also not significant (p ? 0.05).
Colony forming units of bacteria did neither show
significant (p ? 0.05) changes in relation to FYM
application, nor within the FYM treated systems. CFU
valueswere58%(p < 0.001)higherinthewinterwheat
plots when compared to grass-clover plots (Table 1).
Two-way ANOVA revealed significant (p < 0.001)
changes of CFU-counts in relation to farming system
and crops, butno interaction between these factors were
found (Table 2). There was no significant correlation
(p ? 0.05) between Cmicand CFU values.
3.2. Total soil DNA content
Farming systems receiving FYM revealed higher soil
DNA contents (average increase 32%, p < 0.01) when
comparedtosystemswithoutFYMapplication(Tables1
and 2). These differences were only significant
(p < 0.05) in grass-clover plots, but not in winter wheat
plots.BIODYNshowedattheaverage18and26%higher
DNA contents when compared to BIOORG and
CONFYM, butdifferences
(p ? 0.05).
NOFERT were also not significant (p ? 0.05). DNA
contents between winter wheat and grass-clover plots
revealed no significant (p ? 0.05) differences.Two-way
ANOVA showed slightly not significant (p = 0.052)
changes of DNA content in relation to farming systems,
whereas correlation to crop was clearly not significant
(p ? 0.05, Table 2). Interaction between farming
werenot
CONMIN
significant
Differences betweenand
F. Widmer et al./Applied Soil Ecology 33 (2006) 294–307 299
Table 1
Soil biomass parameters determined for two crops planted in the five farming systems of the DOK long-term field experiment
Crop Farming systema
Soil Cmiccontent
(mg g?1soil dry wt)
Soil DNA content
(mg g?1soil dry wt)
CFUb
(?106g?1soil dry wt)
4.30 ? 0.51
7.20 ? 0.59
7.30 ? 0.49
5.20 ? 0.71
6.58 ? 0.44
2.83 ? 0.39
5.13 ? 0.79
3.82 ? 0.16
3.14 ? 0.18
4.47 ? 0.13
Winter wheatNOFERT
CONMIN
BIODYN
BIOORG
CONFYM
211 ? 21
196 ? 35
395 ? 35
290 ? 29
304 ? 17
243 ? 49
211 ? 26
369 ? 55
390 ? 57
308 ? 21
18 ? 3
21 ? 3
33 ? 4
22 ? 5
23 ? 3
20 ? 6
18 ? 3
33 ? 9
32 ? 7
26 ? 4
Grass-clover NOFERT
CONMIN
BIODYN
BIOORG
CONFYM
Plots planted with winter wheat or grass-clover, which represent two positions in the crop rotation, were analyzed. Data represent mean values and
standard deviations of four field replications.
aNOFERT, no fertilization; CONMIN, conventional exclusively with mineral fertilizer; BIODYN, bio-dynamic; BIOORG, bio-organic;
CONFYM, conventional with farmyard manure.
bCFU, colony forming units.
Table 2
Significance of effects of farming system-specific factors of the DOK
long-term field experiment on soil biomass parameters determined
with two-way ANOVA
Significance levelsa
Cmic
DNA CFU
All five farming systems
Farming system
Crop
Farming system ? crop
FYMbapplication
FYM treatment
Crop
FYM treatment ? crop
Two different models were used for the calculation of significant
influences of either all five farming systems individually or of farm-
yard manure application.
a(***) p < 0.001; (**) p < 0.01; (*) p < 0.05; (–) p ? 0.05.
bFYM, farmyard manure.
***
–
–
–
–
–
***
***
–
***
–
–
**
–
–
–
***
–

Page 7

systems and crops was also not significant (p ? 0.05).
However, linear correlation of soil DNA contents to soil
microbial biomass was highly significant (r = 0.75,
p < 0.01).
3.3. Community level substrate utilization
As determined by counts of bacterial CFU on a
glucose minimal medium, 7644 ? 1365 CFU were
inoculated to each well of the BiologTMEcoPlates for
the winter wheat soils and 4846 ? 923 CFU for the
grass-clover soils. Twelve time points were predeter-
mined for CLSU data collection from the 10 different
sample types. Substrate utilization of all 31 substrates
was monitored at these time points revealing clear
differences among CLSU-fingerprints from the differ-
ent farming systems and crops (data not shown). Since
average well colour development (AWCD) in the
BiologTMEcoPlates was more rapid for winter wheat
soil samples than for grass-clover soil samples, six pair-
wise sets of winter wheat and grass-clover data were
selected, which displayed most similar AWCD values
(Table 3). Analysis of this time course revealed that
CLSU patterns developed towards a clear differentia-
tion of certain treatment groups. Cluster analysis of
average values from data set six (Table 3 and Fig. 2)
exemplified and supported results obtained with Monte
Carlo permutation testing performed on the CLSU data
sets from all four replicate samples (Table 3). Average
CLSUvalues ofthe 31 substrates mostclearly separated
NOFERT from all other treatments, i.e. BIODYN,
BIOORG, CONFYM and CONMIN (Fig. 2, clusters I
and II). This separation was highly significant for all six
data sets (Table 3). On the second branching level of the
dendrogram a strong influence of the two crops was
evident (Fig. 2, clusters IIa and IIb). Monte Carlo
permutation testing revealed that significance levels of
crop effects increased from data sets two to six
(Table 3). Cluster analysis also revealed that CONMIN
(Fig. 2, branches IIa1and IIa2) associated with the FYM
treated winter wheat samples but had a tendency to
separate from BIODYN, BIOORG and CONFYM
(Fig. 2, cluster IIa3). Monte Carlo permutation testing
revealed significant differences between CONMIN and
FYM-treated plots in data sets four to six (Table 3).
Differences among the FYM-treatments, i.e. BIODYN,
BIOORG and CONFYM, were not significant for both
crops (Fig. 2, clusters IIb and IIa3).
The fraction of variance in the CLSU data related to
the five farming systems and two crops decreased over
thetimecoursefrom43.7%indataset1to36.6%indata
set six (Table 3). Crop and farming system effects
accountedfor8.6and27.9%ofthevariance,respectively,
while a ‘FYM-application’-effect accounted for 2.7%
and the ‘no FYM-application’-effect accounted for
11.5% of the variance. Two-way ANOVA of individual
substrate utilization intensities revealed significant
effects (p < 0.05) of the two crops in 13 of the 31
substrates and also 13 of the 31 substrates displayed a
system effect (Table 4). Eight of 31 substrates showed
significant effects for both, crop and system, whereas
cross effects of the two factors were significant for two
substrates(Table4).Thirteensubstratesdidnotshowany
significant changes related to farming system or crop. In
F. Widmer et al./Applied Soil Ecology 33 (2006) 294–307 300
Table 3
Time course of data determined with BiologTMEcoPlates
SetWinter wheatGrass-cloverSignificance levelsa
Varb(%)
Timec(h)AWCDTimec(h)AWCD Fertilizationd
Fertilizer typee
Cropsf
1
2
3
4
5
6
44
48
52
56
60
64
0.30
0.41
0.51
0.60
0.68
0.76
48
52
60
64
68
72
0.27
0.35
0.52
0.60
0.68
0.74
***
**
***
***
***
***
–
–
–
*
*
*
–
*
*
**
**
**
43.7
41.1
39.2
38.1
36.6
36.6
Average well colour development (AWCD) values were determined at twelve predetermined time points during plate incubation and six sets with
most similar AWCD values were composed. Significant effects related to overall fertilization, mineral or organic fertilizer types and crops were
determined. The percentage of total variance explained by the whole model was determined for each of the six data sets.
a(–) p ? 0.05; (*) p < 0.05; (**) p < 0.01; (***) p < 0.001 as determined by Monte Carlo permutation testing.
bPercent variance related to the seven factors WW, GC, NOFERT, CONMIN, BIODYN, BIOORG and CONFYM as determined by partitioning
the variance based on redundancy analysis.
cIncubation time of BiologTMEcoPlates.
dNOFERT vs. CONMIN, BIODYN, BIOORG and CONFYM.
eCONMIN vs. BIODYN, BIOORG and CONFYM.
fWinter wheat vs. grass-clover.

Page 8

general, the two crops had significant effects on the
CLSU patterns, while for the five systems significant
effects were exclusively related to differences between
the systems with FYM and those that did not receive
FYM. No statistically significant differences were found
betweenBIODYN,BIOORGandCONFYM,evenwhen
excluding CONMIN and NOFERT from the calculation
(data not shown).
3.4. Terminal restriction fragment length
polymorphisms analyses
Further analysis of bacterial community structures
was performed by RFLP analysis of PCR-amplified
bacterial SSU rDNA. Visual inspection by agarose gel
electrophoresis of RFLP patterns obtained from pooled
samples containing equal amounts of DNA from all four
fieldreplicatesrevealednonoticeabledifferencesamong
the different farming systems and between winter wheat
and grass-clover (data not shown). T-RFLP analysis
based on 32 terminal restriction fragments (T-RFs) that
couldunambiguouslybeidentifiedandquantifiedamong
all 40 samples, allowed for statistically supported
distinction of genetic profiles.
Clusteranalysis oftheaveragevaluesofthe 32 T-RFs
mostclearlyseparatedfieldplotsoilsfertilizedwithFYM
from those not receiving FYM (Fig. 3, clusters I and II).
Furthermore,thedendrogramrevealedastronginfluence
ofthetwocropsintheFYMtreatedplots,whichresulted
incrop-dependentsub-clustering (Fig. 3, clustersIIaand
IIb). On the other hand, the CONMIN and the NOFERT
treatments dominated effects of the two crops (Fig. 3,
clusters Ia and Ib). Monte Carlo permutation testing
performed on the T-RF data set with all 40 samples
revealed that the separation of clusters I and II and of
clusters IIa and IIb were highly significant with
p < 0.001 and p < 0.01, respectively. Separation of all
other clusters was not significant (p ? 0.05) (Fig. 3).
The five farming systems and the two crops
accounted for 43% of the variance in the T-RFLP data
set. Crop and farming system effects explained 7.4 and
35.6% of the variance, respectively, while a ‘FYM-
application’-effect accounted for 3.6% and the ‘no
FYM-application’ accounted for 6.4% of the variance.
Two-way ANOVA of individual T-RF values revealed
significant effects (p < 0.05) of the two crops for 8 of
the 32 T-RFs while 22 T-RFs displayed system effects.
Seven of the 32 T-RFs showed significant effects for
both, crop and system, whereas cross effects of the two
factors were significant for 6 T-RFs (Table 5). Nine of
the 32 T-RFs revealed no significant changes related to
farming system or crop. No statistically significant
F. Widmer et al./Applied Soil Ecology 33 (2006) 294–307301
Fig. 2. Cluster analysis of CLSU data derived from the DOK long-term field experiment based on meanvalues offour replicates of the five farming
systems and two crops. The Ward dendrogram was determined based on Euclidean distances calculated from all 31 blank- and AWCD-corrected
BiologTMEcoPlate substrate utilization values. Monte Carlo permutation testing on all four field replications was used to determine significant
branching in the dendrogram. GC, grass-clover; WW, winter wheat; BIODYN, bio-dynamic; BIOORG, bio-organic; CONFYM, conventional;
CONMIN, mineral fertilizer; NOFERT, unfertilized;***p < 0.001;**p < 0.01. Labels on specific branches refer to information specified in the text.

Page 9

differences were found between BIODYN, BIOORG
and CONFYM, even when excluding CONMIN and
NOFERT from the calculation (data not shown).
4. Discussion
In agricultural soils microbial diversity may be
decreased as compared to natural soils, which in turn
might lead to reduced or less robust soil functionality
(Torsvik et al., 2002). Currently, knowledge about the
relation of soil functionality and soil microbial diversity
is scarce as both characteristics are difficult to assess.
However, it is important to know, whether agricultural
management practices have an impact on soil microbial
characteristics, and which of the agricultural manage-
mentfactors,like fertilization orchoice ofcrops,induce
strongest effects on soil microbial communities. The
DOK field experiment was established in 1978 and has
been operated according to guidelines for conventional
and organic farming (Ma ¨der et al., 2002). It was
designed as a split–split plot with three temporally
shiftedcroprotationparallelsandfourfieldreplications.
Therefore, effects of farming systems and crops on
microbial soil characteristics could be investigated in
F. Widmer et al./Applied Soil Ecology 33 (2006) 294–307302
Table 4
Significances of correlations betweenthe 31 CLSU values of BiologTM
EcoPlates and the two factors farming system and crop as well as
interactions between the two factors
Substratesa
Significance levelsb
Cropsc
Systemsd
Interactione
b-Methyl-D-glucoside
D-Galactonic acid g-lactone
L-Arginine
Pyrovic acid methyl ester
D-Xylose
D-Galacturonic acid
L-Asparigine
Tween 40
I-erythritol
22-Hydroxy benzoic acid
L-Phenylalanine
Tween 80
D-Mannitol
4-Hydroxy benzoic acid
L-Serine
a-Cyclo-dextrin
N-acetyl-D-glucosamine
g-Hydroxy-butyric acid
L-Threonine
Glycogen
D-Glucosaminic acid
Itaconic acid
Glycyl-L-glutamic acid
D-Cellubiose
Glucose-1-phosphate
a-Keto butyric acid
Phenylethylamine
a-D-Lactose
D,L-a-Glycerol phosphate
D-Malic acid
Putrescine
–
–
**
**
**
***
**
*
–
–
**
***
–
–
–
***
–
–
–
–
–
*
–
***
–
***
–
–
–
–
*
–
***
–
*
*
*
–
–
–
**
**
***
–
–
–
***
–
–
*
***
–
–
–
*
–
–
***
–
–
–
***
–
–
***
–
–
–
–
–
–
–
–
–
–
–
–
–
*
–
–
–
–
–
–
–
–
–
–
–
–
–
–
aAll 31 specific substrates on a BiologTMEcoPlate.
b(–) p ? 0.05; (*) p < 0.05; (**) p < 0.01; (***) p < 0.001 (as
determined by ANOVA).
cWinter wheat and grass-clover.
dBIODYN, BIOORG, CONFYM, CONMIN and NOFERT.
eInteraction between crops and systems.
Table 5
Significances of correlations between the 32 terminal restriction
fragment (T-RF) peak height values and the two factors farming
system and crop as well as interactions between the two factors
T-RF sizesa
(rmu)
Significance levelsb
Cropsc
Systemsd
Interactione
62
65
74
82
85
91
92
94
–
–
–
–
*
**
–
–
–
–
*
–
***
–
–
–
–
–
–
–
*
**
*
*
–
–
–
–
–
–
–
–
–
*
**
**
***
**
*
***
***
*
*
–
–
–
*
**
***
–
–
–
**
***
***
***
***
***
***
**
–
–
–
*
–
*
–
**
–
–
–
*
–
–
–
–
***
–
*
–
–
*
–
–
–
–
–
–
–
–
–
–
–
–
–
–
125
126
127
136
138
142
147
149
155
158
159
165
168
197
265
277
280
290
295
437
450
494
513
539
aSizesin relativemigrationunits (rmu) ofall 32 terminal restriction
fragments (T-RF) scored.
b(–) p ? 0.05; (*) p < 0.05; (**) p < 0.01; (***) p < 0.001 (as
determined by ANOVA).
cWinter wheat and grass-clover.
dBIODYN, BIOORG, CONFYM, CONMIN and NOFERT.
eInteraction between crops and systems.

Page 10

this well designed field experiment. The polyphasic
approach, based on soil biomass parameters and
microbial community profiling techniques, allowed to
successfully assessing these effects.
4.1. Soil biomass parameters
The five farming systems of the DOK experiment
significantly affected soil microbial biomass Cmic
(Table 2) while the influence on soil DNA content
was only nearly significant (p = 0.052). In general, the
FYM treated soils in the farming systems BIODYN,
BIOORG and CONFYM, revealed significantly higher
values for microbial biomass and DNAwhen compared
to those not fertilized with FYM, i.e. CONMIN and
NOFERT. Among the systems with FYM treatment,
BIODYN tended to higher Cmicvalues for both crops,
when compared to the conventional system. These
results are supported by other studies (Bossio et al.,
1998; Carpenter-Boggs et al., 2000; Peacock et al.,
2001), indicating that the DOK system is a stable and
representative system for assessing agricultural man-
agement effects on biological soil characteristics.
Assessment of crop effects on soil Cmicand DNA
content revealed less pronounced differences, which
were not significant (Table 2), and not consistent
between the two crops. In contrast, the CFU values
revealed significant correlations with farming systems
and crops. In various studies moderate effects of plants
on biological soil characteristics were reported (Bard-
gett et al., 1999; Johnson et al., 2003; Kennedy et al.,
2004). The magnitude of effects strongly depended on
plant species, composition of plant populations, as well
asonwhethersampleswerederivedfromrhizosphereor
bulk soil. One distinct effect of crops in our study was a
higher content of Cmicand DNA in soils from grass-
clover plots of the BIOORG system, when compared to
the corresponding winter wheat plots. This observation,
however, may mainly be attributed to the system-
specific fertilization regime of BIOORG, where a larger
quantity of slurry was applied to the grass-clover plots
than in the other farming systems. Over-all, strongest
effects on soil biomass parameters were induced by
application of FYM, which may be explained by the
addition of nutrients as well as microbial biomass
contained in FYM. Crop effects were only significant
for the culturable fraction of soil bacteria detected with
plate counts. These significant effects of agricultural
management factors indicated distinct effects on
microbiological soil characteristics.
If soil microbial community structures are resolved
based on SSU rDNA gene analyses it is important that
F. Widmer et al./Applied Soil Ecology 33 (2006) 294–307303
Fig.3. ClusteranalysisofT-RFLPdataderivedfromtheDOKlong-termfieldexperimentbasedonmeanvaluesoffourreplicatesofthefivefarming
systems and two crops. The Ward dendrogram was determined based on Euclidean distances calculated from all 32 normalized T-RF peak height
values. Monte Carlo permutation testing on all four field replications was used to determine significant branching in the dendrogram. GC, grass-
clover; WW, winter wheat; BIODYN, bio-dynamic; BIOORG, bio-organic; CONFYM, conventional; CONMIN, mineral fertilizer; NOFERT,
unfertilized;***p < 0.001;**p < 0.01. Labeling of specific branches refers to information specified in the text.

Page 11

soil DNA represents total soil microbial biomass. Non-
representative DNA extracts may lead to biased
description of soil microbial communities (Miller
et al., 1999; Roose-Amsaleg et al., 2001). A strong
correlation between soil Cmicand soil DNA contents,
ascertained the representativeness of soil DNA extracts
(Marstorp et al., 2000; Bundt et al., 2001; Blagodats-
kaya et al., 2003; Hartmann et al., 2005). The high
correlation between soil Cmicand soil DNA contents
presented here, i.e. r = 0.75 (p < 0.001), is supporting
this concept and the quality of the DNA extraction and
quantification protocol used (Bu ¨rgmann et al., 2001).
4.2. Soil bacterial community structures
The influence of the farming systems explained 28%
of the total CLSU data variability while 42% of the
BiologTM
EcoPlate substrates revealed significant
changes due to the farming system with no crop
interaction (Table 4). T-RFLP analyses revealed that
36% of total variance was specifically related to the
farming systems, which also significantly influenced
56% of all detected soil bacterial T-RFs with no crop
interaction (Table 5). These statistical analyses of
CLSU and T-RFLP data indicated a good comparability
of the two approaches applied even though they
revealed a slightly better differentiation for the T-RFLP
data set (Tables 4 and 5).
Comparing the cluster analyses of the CLSU and the
T-RFLP data revealed one clear difference between the
two approaches. Cluster analysis of T-RFLP data
significantly separated the two farming systems without
FYM amendment, i.e. CONMIN and NOFERT, from
the FYM-based farming systems (Fig. 3), while the
CLSU-derived dendrogram significantly (p < 0.001)
separated the unfertilized controls, i.e. NOFERT, from
the other four farming systems and clustered the
CONMIN samples to FYM amended winter wheat
samples (Fig. 2). However, as substantiated by time
course and Monte Carlo permutation analyses, also the
CLSU analysis tended to separate CONMIN from FYM
treated systems by revealing significant (p < 0.05)
separation between CONMIN and FYM treated soils in
the last three time points of the time course (Table 3).
Possibly, NOFERT was separated so significantly from
the other farming systems because of the decrease in
soil pH or because crop yield was strongly reduced due
to the lack of nutrients after 22 years without fertilizer
input. These specific soil conditions may well be
represented inmicrobial
Enhanced effects induced by reduced crop or root
biomass are likely and will have to be studied in greater
community structure.
detail in the future. For T-RFLP analysis of CONMIN
this explanation may also apply since the corresponding
soils tended to lower soil Cmic contents. In CLSU
analysis all CONMIN samples clustered with the FYM
amended winter wheat plot samples, which represented
the arable phase in the crop rotation. Even though
CONMIN produces normal crop yields, these soils
tended to lower biomass contents, and therefore may
have displayed similar substrate utilization patterns as
the soils from winter wheat plots in the arable phase of
the crop rotation.
In an earlier study on soils from winter wheat plots
from the DOK field experiment, principal component
analysis of CLSU profiles detected seasonal variations,
while differences between the farming systems were
only observed in spring but variation among field
replications was relatively high (Fliessbach and Ma ¨der,
1997). O’Donnell et al. (2001) also reported only small
differences of CLSU data among FYM and minerally
fertilized soils. Possibly, the relatively low variability
among a high number of replicate BiologTMmicro-
plates, the reduced number of substrates in BiologTM
EcoPlates as compared to the GN-plates used earlier,
and the time course analysis applied in the present study
may explain the higher resolution and significances
found.Denaturing gradient
(DGGE)-based comparison of organic and inorganic
fertilizer types for their long-term effects on bacterial
community structures revealed
between organic and inorganic fertilization as deter-
mined with redundancy discriminate analysis (Marsch-
ner et al., 2003). T-RFLP analysis targeting the
functional genes nirK and nirS, which are specific for
denitrifying bacteria, revealed clear differentiation of
soils receiving no fertilizers, mineral fertilizers or cattle
manure (Wolsing and Prieme, 2004). Significant
differences in phospholipid fatty acid (PLFA) profiles
of bacteria and fungi were found in soils from organic or
conventional farming of the sustainable agricultural
farming system (SAFS) project (Bossio et al., 1998;
Lundquist et al., 1999). Peacock et al. (2001) detected
significant differences in bacterial PLFA profiles among
FYM amended, minerally fertilized and control soils.
Carpenter-Boggs et al. (2000) have compared unferti-
lized, minerally fertilized, and compost-amended soils
based on bacterial fatty acid methyl ester (FAME)
profiles, which revealed strong differences between
compost-amendment and mineral or no fertilizer
application, but could not distinguish between minerally
and unfertilized soils as well as between bio-organically
and bio-dynamically treated soils. Over all, these reports
and the data presented in the present study indicate that
gelelectrophoreses
strongdifferences
F. Widmer et al./Applied Soil Ecology 33 (2006) 294–307304

Page 12

input of fertilizer and FYM play an important role by
representing a general enrichment in organic substrates,
and therefore promoting growth of microbial commu-
nities (De Fede et al., 2001; Grayston et al., 2001).
Crop effects of winter wheat and grass-clover
allowed to explain 8.6% of CLSU data variability,
and were supported by 39% of the CLSU values that
significantly (p < 0.05) changed due to the two crops
without interaction with farming systems (Table 4).
Only 7.4% of total variability in the T-RFLP data set
was explained by effects related to the two crops.
Twenty two percent of the bacterial T-RFs were
significantly (p < 0.05) changing in relation to crop
effects without showing interaction with the farming
systems (Table 5). Among the FYM treated soils, crop
effects significantly (p < 0.01) dominated changes in
CLSU (Fig. 2 and Table 3) and T-RFLP (Fig. 3) profiles
revealing consistency between the two analytical
methods and validity of the results obtained. These
findings were supported by results from others showing
a significant influence of plant species on soil microbial
community structures determined by CLSU (Grayston
et al., 1998; Larkin, 2003) and soil DNA analyses
(Buckley and Schmidt, 2001; Smalla et al., 2001;
Marschner et al., 2004). Several studies showed, that
soil bacterial community structures depend on crop or
other plant species (Grayston et al., 1998; Bardgett
et al., 1999; Smalla et al., 2001; Larkin, 2003). Kuske
et al. (2002), for example, were able to clearly
distinguish three different grass species based on
rhizosphere bacterial community structures by using
a T-RFLP approach on SSU rDNA. In our study, plant
effects occurred on a lower level as compared to
relatively strong effects induced by fertilizers. Kennedy
etal.(2004)reportedthatfertilizeramendments,suchas
combination of lime and nitrogen influenced soil
bacterial T-RFLP profiles significantly stronger than
different grassland species. Effects of different plant
species on bacterial community
agriculturally improved and unimproved grassland soils
based on PLFA analyses clearly showed that effects of
plant species also depended on the management of the
soil (Innes et al., 2004). They found that effects on
PLFAprofilesof all plantspeciesinvestigateddepended
on whether they were growing in improved or
unimproved soils. This supports our findings, that crops
dominantly influenced bacterial community composi-
tion in the FYM treated soils, but not in minerally
fertilized and unfertilized soils. In agricultural practice,
however, it may be difficult to separate plant-derived
effects from those of fertilizers, since each crop has its
specific management requirements, which may lead to
composition in
strong interactions of the two factors. Overall, soil
bacterial community profiling based on CLSU with
Biolog EcoPlates and SSU rRNA gene T-RFLP analysis
yielded largely consistent results.
5. Conclusions
The polyphasic approach to evaluate and compare
factors affecting microbiological soil characteristics in
the DOK long-term field experiment allowed to
differentiate the impact of specific farming systems
and selected crops. Soil biological parameters like
microbial biomass and DNA content were strongly
influenced by the farming systems, whereas winter
wheat and grass-clover, representing two positions in
the crop rotation, did not affect these parameters to a
significant extent. Differences among the FYM-based
farming systems BIODYN, BIOORG and CONFYM
were not significant. CLSU and T-RFLP analyses
revealed that the main impact of FYM application was
followed by intermediate effects of the two crops winter
wheat and grass-clover. Smallest and insignificant
effects on soil bacterial communities were detected
among the three farming systems BIODYN, BIOORG
andCONFYM.Most of the specific substrateutilization
values and T-RF abundances were affected by the
specific treatments in the DOK soils, indicating
extended shifts in the soil bacterial communities. These
results suggest that for successful soil quality manage-
ment, input of FYM and crop rotation are of major
importance.
Acknowledgements
Paul Ma ¨der, Roland Ko ¨lliker and David Dubois are
acknowledged for valuable contributions to this manu-
script. We are grateful to Vit Fejfar for his assistance
with laboratory analyses. Special thanks go to the field
crew of the DOK-experiment and to the farmers and
representatives of the Swiss farming organizations
involved. This study was financially supported by the
Swiss Federal Office for Agriculture and by a grant of
the Swiss National Science Foundation.
References
Amann, R., Ludwig, W., 2000. Ribosomal RNA-targeted nucleic acid
probesforstudiesin microbialecology.FEMSMicrobiol. Rev.24,
555–565.
Bardgett, R.D., Mawdsley, J.L., Edwards, S., Hobbs, P.J., Rodwell,
J.S., Davies, W.J., 1999. Plant species and nitrogen effects on soil
biological properties of temperate upland grasslands. Funct. Ecol.
13, 650–660.
F. Widmer et al./Applied Soil Ecology 33 (2006) 294–307305

Page 13

Besson, J.M., Niggli, U., 1991. DOK-Versuch: vergleichende Lang-
zeit-Untersuchung in den drei Anbausystemen biologisch-dyna-
misch, organisch-biologisch und konventionell. I. Konzeption des
DOK-Versuches: 1 und 2. Fruchtfolgeperiode. Schweizerische
Landwirtschaftliche Forschung 31, 79–109.
Blackwood, C.B., Marsh, T., Kim, S.H., Paul, E.A., 2003. Terminal
restriction fragment length polymorphism data analysis for quan-
titative comparison of microbial communities. Appl. Environ.
Microbiol. 69, 926–932.
Blagodatskaya, E.V., Blagodatskii, S.A., Anderson, T.H., 2003. Quan-
titative isolation of microbial DNA from different types of soils of
natural and agricultural ecosystems. Microbiology 72, 744–749.
Borcard,D.,Legendre,P.,Drapeau,P.,1992.Partiallingoutthe spatial
component of ecological variation. Ecology 73, 1045–1055.
Bossio, D.A., Scow, K.M., Gunapala, N., Graham, K.J., 1998. Deter-
minants of soil microbial communities: effects of agricultural
management, season, and soil type on phospholipid fatty acid
profiles. Microb. Ecol. 36, 1–12.
Braker, G., Ayala-Del-Rio, H.L., Devol, A.H., Fesefeldt, A., Tiedje,
J.M., 2001. Community structure of denitrifiers, bacteria, and
archaea along redox gradients in Pacific Northwest marine sedi-
ments by terminal restriction fragment length polymorphism
analysis of amplified nitrite reductase (nirS) and 16S rRNA genes.
Appl. Environ. Microbiol. 67, 1893–1901.
Brodie, E., Edwards, S., Clipson, N., 2003. Soil fungal community
structure in a temperate upland grassland soil. FEMS Microbiol.
Ecol. 45, 105–114.
Buckley, D.H., Schmidt, T.M., 2001. The structure of microbial
communities in soil and the lasting impact of cultivation. Microb.
Ecol. 42, 11–21.
Buckley, D.H., Schmidt, T.M., 2003. Diversity and dynamics of
microbial communities in soils from agro-ecosystems. Environ.
Microbiol. 5, 441–452.
Bundt, M., Widmer, F., Pesaro, M., Zeyer, J., Blaser, P., 2001.
Preferential flow paths: biological ‘hot spots’ in soils. Soil Biol.
Biochem. 33, 729–738.
Bu ¨rgmann, H., Pesaro, M., Widmer, F., Zeyer, J., 2001. A strategy for
optimizing quality and quantity of DNA extracted from soil. J.
Microbiol. Methods 45, 7–20.
Carpenter-Boggs, L., Kennedy, A.C., Reganold, J.P., 2000. Organic
and biodynamic management: effects on soil biology. Soil Sci.
Soc. Am. J. 64, 1651–1659.
De Fede, K.L., Panaccione, D.G., Sexstone, A.J., 2001. Characteriza-
tion of dilution enrichment cultures obtained from size-fractio-
nated soil bacteria by Biolog1community-level physiological
profiles and restriction analysis of 16s rRNA genes. Soil Biol.
Biochem. 33, 1555–1562.
Doran, J.W., Zeiss, M.R., 2000. Soil health and sustainability: mana-
ging the biotic component of soil quality. Appl. Soil Ecol. 15, 3–
11.
Dunbar, J., Ticknor, L.O., Kuske, C.R., 2001. Phylogenetic specificity
and reproducibility and new method for analysis of terminal
restriction fragment profiles of 16S rRNA genes from bacterial
communities. Appl. Environ. Microbiol. 67, 190–197.
Eidg.Volkswirtschaftsdepartement, 1997. Verordnung u ¨ber die biolo-
gische Landwirtschaft und die Kennzeichnung biologisch produ-
zierter Erzeugnisse und Leensmittel (Bio-Verordnung) (http://
www.admin.ch/ch/d/sr/c910_18.html). Systematische Sammlung
des Bundesrechts: 910.18.
Eidg.Volkswirtschaftsdepartement, 1998. Verordnung u ¨ber die Dir-
ektzahlungen an die Landwirtschaft (Direktzahlungsverordnung,
DZV) - Richtlinien fu ¨r den o ¨kologischen Leistungsnachweis
(http://www.admin.ch/ch/d/sr/c910_13.html).
Sammlung des Bundesrechts: 910.13.
Fisher,M.M.,Triplett,E.W.,1999.Automatedapproachforribosomal
intergenic spacer analysis of microbial diversity and ITS applica-
tion to freshwater bacterial communities. Appl. Environ. Micro-
biol. 65, 4630–4636.
Fliessbach, A., Ma ¨der, P., 1997. Carbon source utilization by micro-
bial communities in soils under organic and conventional farming
practice. In: Insam, H., Rangger, A. (Eds.), Microbial Commu-
nities: Functional Versus Structural Approaches. Springer-Verlag,
Berlin, Germany, pp. 109–120.
Fliessbach, A., Ma ¨der, P., 2004. Short- and long-term effects on soil
microorganisms of two potato pesticide spraying sequences with
either glufosinate or dinoseb as defoliants. Biol. Fertil. Soils 40,
268–276.
Garland, J.L., Mills, A.L., 1991. Classification and characterization of
heterotrophic microbial communities on the basis of patterns of
community-level sole-carbon-source utilization. Appl. Environ.
Microbiol. 57, 2351–2359.
Gomez, E., Bisaro, V., Conti, M., 2000. Potential C-source utilization
patterns of bacterial communities as influenced by clearing and
landuseinaverticsoilofArgentina.Appl.SoilEcol.15,273–281.
Grayston, S.J., Griffith, G.S., Mawdsley, J.L., Campbell, C.D., Bard-
gett, R.D., 2001. Accounting for variability in soil microbial
communitiesoftemperateuplandgrasslandecosystems.SoilBiol.
Biochem. 33, 533–551.
Grayston, S.J., Wang, S.Q., Campbell, C.D., Edwards, A.C., 1998.
Selective influence of plant species on microbial diversity in the
rhizosphere. Soil Biol. Biochem. 30, 369–378.
Hartmann, M., Frey, B., Ko ¨lliker, R., Widmer, F., 2005. Semi-auto-
mated genetic analyses of soil microbial communities: compar-
ison of T-RFLP and RISA based on descriptive and discriminative
statistical approaches. J. Microbiol. Methods 63, 349–360.
Heuer,H.,Krsek,M.,Baker,P.,Smalla,K.,Wellington,E.M.H.,1997.
Analysis of actinomycete communities by specific amplification
ofgenesencoding16SrRNAandgel-electrophoretic separationin
denaturing gradients. Appl. Environ. Microbiol. 63, 3233–3241.
Hill,G.T.,Mitkowski,N.A.,Aldrich-Wolfe,L.,Emele,L.R.,Jurkonie,
D.D., Ficke, A., Maldonado-Ramirez, S., Lynch, S.T., Nelson,
E.B.,2000.Methodsforassessingthecompositionanddiversityof
soil microbial communities. Appl. Soil Ecol. 15, 25–36.
Innes, L., Hobbs, P.J., Bardgett, R.D., 2004. The impacts of individual
plant species on rhizosphere microbial communities in soils of
different fertility. Biol. Fertil. Soils 40, 7–13.
Joergensen,R.G.,1996.Thefumigation-extractionmethodtoestimate
soil microbial biomass: calibration of the k(EC) value. Soil Biol.
Biochem. 28, 25–31.
Johnson, D., Booth, R.E., Whiteley, A.S., Bailey, M.J., Read, D.J.,
Grime, J.P., Leake, J.R., 2003. Plant community composition
affects the biomass, activity and diversity of microorganisms in
limestone grassland soil. Eur. J. Soil Sci. 54, 671–677.
Kennedy, A.C., 1999. Bacterial diversity in agroecosystems. Agric.
Ecosyst. Environ. 74, 65–76.
Kennedy, N., Brodie, E., Connolly, J., Clipson, N., 2004. Impact of
lime, nitrogen and plant species on bacterial community structure
in grassland microcosms. Environ. Microbiol. 6, 1070–1080.
Kirk, J.L.,Beaudette,L.A.,Hart,M., Moutoglis, P., Khironomos, J.N.,
Lee, H., Trevors, J.T., 2004. Methods of studying soil microbial
diversity. J. Microbiol. Methods 58, 169–188.
Kuske, C.R., Ticknor, L.O., Miller, M.E., Dunbar, J.M., Davis, J.A.,
Barns, S.M., Belnap, J., 2002. Comparison of soil bacterial
communities in rhizospheres of three plant species and the inter-
Systematische
F. Widmer et al./Applied Soil Ecology 33 (2006) 294–307 306

Page 14

spaces in an arid grassland. Appl. Environ. Microbiol. 68, 1854–
1863.
Larkin, R.P., 2003. Characterization of soil microbial communities
under different potato cropping systems by microbial population
dynamics, substrate utilization, and fatty acid profiles. Soil Biol.
Biochem. 35, 1451–1466.
Liu, W.T., Marsh, T.L., Cheng, H., Forney, L.J., 1997. Characteriza-
tion of microbial diversity by determining terminal restriction
fragment length polymorphisms of genes encoding 16S rRNA.
Appl. Environ. Microbiol. 63, 4516–4522.
Lundquist, E.J., Scow, K.M., Jackson, L.E., Uesugi, S.L., Johnson,
C.R., 1999. Rapid response of soil microbial communities from
conventional, low input, and organic farming systems to a wet/dry
cycle. Soil Biol. Biochem. 31, 1661–1675.
Ma ¨der, P., Fliessbach, A., Dubois, D., Gunst, L., Fried, P., Niggli, U.,
2002. Soil fertility and biodiversity in organic farming. Science
296, 1694–1697.
Marschner, P., Crowley, D., Yang, C.H., 2004. Development of
specific rhizosphere bacterial communities in relation to plant
species, nutrition and soil type. Plant Soil 261, 199–208.
Marschner, P., Kandeler, E., Marschner, B., 2003. Structure and
function of the soil microbial community in a long-term fertilizer
experiment. Soil Biol. Biochem. 35, 453–461.
Marstorp,H.,Guan,X.,Gong,P.,2000.RelationshipbetweendsDNA,
chloroform labile C and ergosterol in soils of different organic
matter contents and pH. Soil Biol. Biochem. 32, 879–882.
Miller, D.N., Bryant, J.E., Madsen, E.L., Ghiorse, W.C., 1999. Eva-
luation and optimization of DNA extraction and purification
procedures for soil and sediment samples. Appl. Environ. Micro-
biol. 65, 4715–4724.
Muyzer, G., 1999. DGGE/TGGE a method for identifying genes from
natural ecosystems. Curr. Opin. Microbiol. 2, 317–322.
O’Donnell, A.G., Seasman, M., Macrae, A., Waite, I., Davies, J.T.,
2001. Plants and fertilisers as drivers of change in microbial
community structure and function in soils. Plant Soil 232, 135–
145.
Peacock, A.D., Mullen, M.D., Ringelberg, D.B., Tyler, D.D., Hedrick,
D.B., Gale, P.M., White, D.C., 2001. Soil microbial community
responses to dairy manure or ammonium nitrate applications. Soil
Biol. Biochem. 33, 1011–1019.
Pochon, J., Tardieux, P., 1962. Techniques d’analyse en microbiologie
du sol. Editions de la Tourelle, St. Mande ´, France, 111 pp.
Powlson, D.S., Johnston, A.E., 1994. Long-term field experiments:
their importance in understanding sustainable land use. In:
Greenland, D.J., Szabolcs, I. (Eds.), Soil Resilience and Sus-
tainable Land Use. CAB International, Wallingford, pp. 367–
394.
Preston-Mafham, J., Boddy, L., Randerson, P.F., 2002. Analysis of
microbial community functional diversity using sole-carbon-
source utilisation profiles—a critique. FEMS Microbiol. Ecol.
42, 1–14.
Roose-Amsaleg,C.L., Garnier-Sillam,E., Harry, M., 2001.Extraction
and purification of microbial DNA from soil and sediment sam-
ples. Appl. Soil Ecol. 18, 47–60.
Schwieger, F., Tebbe, C.C., 1998. A new approach to utilize PCR-
single-strand-conformation polymorphism for 16S rRNA gene-
based microbial community analysis. Appl. Environ. Microbiol.
64, 4870–4876.
Shaffer, J.P., 1995. Multiple hypothesis-testing. Annu. Rev. Psychol.
46, 561–584.
Smalla, K., Wieland, G., Buchner, A., Zock, A., Parzy, J., Kaiser, S.,
Roskot, N., Heuer, H., Berg, G., 2001. Bulk and rhizosphere soil
bacterial communities studied by denaturing gradient gel electro-
phoresis: plant-dependentenrichment
revealed. Appl. Environ. Microbiol. 67, 4742–4751.
terBraak,C.J.F.,Smilauer,P.,2002.CANOCOReferenceManualand
CanoDraw for Windows User’s Guide: Software for Canonical
Community Ordination (Version 4.5) Microcomputer Power,
Ithaca, NY, USA.
Theron, J., Cloete, T.E., 2000. Molecular techniques for determining
microbial diversity and community structure in natural environ-
ments. Crit. Rev. Microbiol. 26, 37–57.
Tiedje, J.M., Asuming-Brempong, S., Nusslein, K., Marsh, T.L.,
Flynn, S.J., 1999. Opening the black box of soil microbial
diversity. Appl. Soil Ecol. 13, 109–122.
Torsvik, V., Goksoyr, J., Daae, F.L., 1990. High diversity in DNA of
soil bacteria. Appl. Environ. Microbiol. 56, 782–787.
Torsvik, V., Ovreas, L., 2002. Microbial diversity and function in soil:
from genes to ecosystems. Curr. Opin. Microbiol. 5, 240–245.
Torsvik, V., Ovreas, L., Thingstad, T.F., 2002. Prokaryotic diversity-
magnitude, dynamics, and controlling factors. Science 296, 1064–
1066.
Vance, E.D., Brookes, P.C., Jenkinson, D.S., 1987. An extraction
method for measuring soil microbial biomass C. Soil Biol. Bio-
chem. 19, 703–707.
Widmer, F., Fliessbach, A., Laczko, E., Schulze-Aurich, J., Zeyer, J.,
2001. Assessing soil biological characteristics: a comparison of
bulk soil community DNA-, PLFA-, and BiologTM-analyses. Soil
Biol. Biochem. 33, 1029–1036.
Wolsing,M., Prieme, A., 2004. Observation of high seasonal variation
in community structure of denitrifying bacteria in arable soil
receiving artificial fertilizer and cattle manure by determining
T-RFLP of nir gene fragments. FEMS Microbiol. Ecol. 48, 261–
271.
and seasonal shifts
F. Widmer et al./Applied Soil Ecology 33 (2006) 294–307307