Blackwell Publishing LtdOxford, UKEMIEnvironmental Microbiology1462-2912Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd.

No claim to original French government works2006812

21622169Original ArticleMicrobial diversity erosion and soil functioningS. Wertz
et al

Environmental Microbiology (2006) 8(12), 2162–2169 doi:10.1111/j.1462-2920.2006.01098.x

Maintenance of soil functioning following erosion of

microbial diversity

Sophie Wertz,1 Valérie Degrange,1 James I. Prosser,2 bacterial species reported in the literature. The reduc-
Franck Poly,1 Claire Commeaux,1 Thomas Freitag,2 tion of the diversity of the functional groups observed
Nadine Guillaumaud1 and Xavier Le Roux1* from genetic fingerprints did not impair the associ-
1
UMR 5557 Ecologie Microbienne (CNRS-Université Lyon ated functioning of these groups, i.e. carbon mineral-
1, USC INRA 1196), bât. G. Mendel, 43 boulevard du 11 ization, denitrification and nitrification. This was
Novembre 1918, 69622 Villeurbanne, France. remarkable, because the amplitude of diversity ero-
2
School of Biological Sciences, University of Aberdeen, sion generated by the dilution approach was huge
Cruickshank Building, St Machar Drive, Aberdeen, (level of bacterial species loss was estimated to be
Scotland, AB24 3UU, UK. around 99.99% for the highest dilution). Our results
demonstrate that the vast diversity of the soil micro-
biota makes soil ecosystem functioning largely insen-
Summary
sitive to biodiversity erosion even for functions
The paradigm that soil microbial communities, being performed by specialized groups.
very diverse, have high functional redundancy levels,
so that erosion of microbial diversity is less important
Introduction
for ecosystem functioning than erosion of plant or
animal diversity, is often taken for granted. However, Unprecedented rates of species extinction (Chapin et al.,
this has only been demonstrated for decomposition/ 2000) resulting from anthropogenic disturbances have
respiration functions, performed by a large proportion prompted extensive research on the impact of biodiversity
of the total microbial community, but not for spe- losses on ecosystem functioning (Loreau et al., 2002). A
cialized microbial groups. Here, we determined the fundamental question is whether ecosystems or functional
impact of a decrease in soil microbial diversity on soil groups with declining numbers of species can maintain
ecosystem processes using a removal approach, in process rates essential for ecosystem sustainability. Dur-
which less abundant species were removed preferen- ing the last decade, several studies have provided evi-
tially. This was achieved by inoculation of sterile soil dence of the functional importance of biodiversity to
microcosms with serial dilutions of a suspension ecosystem processes (e.g. Hooper and Vitousek, 1997;
obtained from the same non-sterile soil and subse- Symstad and Tilman, 2001; Wardle and Zackrisson,
quent incubation, to enable recovery of community 2005). A recent synthesis (Loreau et al., 2002) stresses
size. The sensitivity to diversity erosion was that most biodiversity-ecosystem functioning (B-EF) stud-
evaluated for three microbial functional groups with ies have been carried out on a small number of processes
known contrasting taxonomic diversities (ammonia and systems, in particular primary production and nutrient
oxidizers < denitrifiers < heterotrophs). Diversity ero- retention in grasslands and, to a lesser extent, litter
sion within each functional group was characterized decomposition in soil and primary production in aquatic
using molecular fingerprinting techniques: ribosomal systems (McGrady et al., 1997; Naem and Li, 1997;
intergenic spacer analysis (RISA) for the eubacterial Mikola and Setälä, 1998; Wardle, 2002).
community, denaturing gradient gel electrophoresis Soil microbes possess vast phylogenetic and functional
(DGGE) analysis of nirK genes for denitrifiers, and diversity and are key drivers of energy flow and nutrient
DGGE analysis of 16S rRNA genes for betaproteobac- cycling. Understanding the functional role of microbial
terial ammonia oxidizers. In addition, we simulated diversity is therefore crucial for assessment of (i) the
the impact of the removal approach by dilution on the generic relevance of B-EF relationships (Loreau et al.,
number of soil bacterial species remaining in the 2002) and (ii) the impact of global change on terrestrial
inoculum using values of abundance distribution of ecosystems. Many studies have reported the existence of
correlation or lack of correlation between changes in soil
microbial diversity and functioning in response to different
Received 28 March, 2006; accepted 14 June, 2006. *For
correspondence. E-mail [email protected]; Tel. (+33)
treatments (e.g. Avrahami et al., 2003; Seghers et al.,
4 72 43 13 79; Fax (+33) 4 72 43 12 23. 2003; Webster et al., 2005; Patra et al., 2006). However,

Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd.
No claim to original French government works
 Microbial diversity erosion and soil functioning 2163
in these studies, impacts of changes in microbial diversity 10
and environmental factors on functioning can be con- NS
founded. Studies directly addressing B-EF relationships 8

log (heterotrophs
for soil microorganisms have employed an assembly

g dry soil)
approach on a small number of symbiotic or litter- 6
decomposing fungi (van der Heijden et al., 1998; Jonsson
et al., 2001; Setälä and McLean, 2004). Others (Salonius,

-1
4
1980; Degens, 1998; Griffiths et al., 2000; 2001; 2004)
have used a removal approach. However, an important
2
restriction of the latter studies was that they focused on a
general and widespread function (decomposition), involv-
0
ing a wide range of substrates metabolized by microor-
ganisms distributed throughout the prokaryotic kingdoms, 10
thereby involving a large proportion of the total soil micro-
bial community that likely possesses a high level of func- 8 a
tional redundancy. There is no such study for soil ab ab ab b
ab

log (denitrifiers
ecosystem functions performed by more specialized, less

g dry soil)
6
diverse microbial functional groups. The generic value of
the statement that soil microbial communities have high
functional redundancy levels and that erosion of microbial 4

-1
diversity has little impact on ecosystem functioning
beyond decomposition process has therefore never been 2
critically tested and thus never demonstrated (Wolters
et al., 2000; Bengtsson et al., 2002; Nannipieri et al., 0
2003).
10
The objective of this study was to quantify the impact
of a decrease in soil microbial diversity on general but also
log (ammonia-oxidizers

8
specialized soil ecosystem functions using a removal
approach in which less abundant species were removed a a
b
g dry soil)

first. Our study focused on the impact of biodiversity ero- 6

sion within microbial functional groups with contrasting b
taxonomic diversities (i.e. ammonia oxidizers < denitrifiers 4
c
-1

< all bacterial heterotrophs) on the functioning of these

groups. To address this issue, we established a range of 2
microbial diversities by reducing richness through inocu-
lation of sterile soil microcosms with serial dilutions of a 0
suspension obtained from the same non-sterile soil. -1 -3 -4 -5 -6 -8
10 10 10 10 10 10
Treatment
Results
Fig. 1. Abundance of culturable heterotrophs, denitrifiers and ammo-
Abundances of bacteria within functional groups
nia oxidizers in sterile soil microcosms inoculated with a non-sterile
soil suspension, giving final dilutions of 10−1−10−8 of the non-sterile
The final abundances (after incubation for 19 weeks) of soil, after incubation for 19 weeks. Bars represent standard errors
heterotrophic bacteria of sterile soils microcosms inocu- (n = 4). Values with different letters differ significantly (P < 0.05).
lated with dilutions of non-sterile soil (treatments named
10−1−10−8) were similar in all microcosms (Fig. 1). For
denitrifying and ammonia oxidizing bacteria, similar final
abundances were obtained among treatments 10−1−10−6 tional group decreased progressively from microcosms
and 10−1−10−3, respectively, and were lower in microcosms inoculated with low soil dilutions to those inoculated with
inoculated with higher dilutions (Fig. 1). high dilutions (Fig. 2). In microcosms inoculated with low
dilutions, some differences in molecular fingerprints were
observed between replicates, but fingerprints of commu-
Diversity of each functional group
nities inoculated with high dilutions varied more between
For treatments with similar abundances after the incuba- replicates, even for the most intense bands of the finger-
tion for 19 weeks (Fig. 1), the richness index of each func- prints (Figs 3 and 4).

Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 8, 2162–2169
No claim to original French government works
2164 S. Wertz et al.

30 Simulated effect of dilution on erosion of bacterial diversity

a Simulation results (Fig. 6) show that, assuming that the
a
b abundance distribution of bacterial species reported by
b b Gans and colleagues (2005) were valid for our soil, the
20
c level of species loss in the inoculum were around 21.8%,
86.5%, 97.7% and 99.99% for dilution treatments 10−3, 10−4,
R

10−5 and 10−8 respectively. Simulations also show that a

10 pool of several hundreds of bacterial species remained in
the inoculum for the 10−8 dilution treatment (inset in Fig. 6).

Discussion
0
30 The aim of our study was to characterize the impact of
a a
the erosion of soil microbial diversity on the functioning of

20 b b 1 2 3 4 5 6
b
R

10

0
30

a
20

b
R

10

10 -1 10 -3 10 -4 10 -5 10 -6 10 -8
Treatment
Fig. 2. Values of richness index (R) computed from fingerprints of
(top) eubacteria, (middle) denitrifier and (bottom) ammonia oxidizer
communities for treatments with similar functional group size. For
details, see legend of Fig. 1.

Functioning of each functional group

Carbon mineralization did not differ significantly between

treatments (10−1−10−8). Similarly, denitrification and nitrifi-
cation did not vary significantly between treatments
(10−1−10−6 and 10−1−10−3, respectively) for which the abun-
dance of the functional group was similar (Fig. 5). In addi-
tion, for the three activities, no significant correlations
were obtained between activity level and treatment Fig. 3. Genetic profiles of bacterial community in four replicate soil
microcosms of treatment 10−8 after incubation for 19 weeks. Lanes 1
(0.19 < P < 0.87) or activity level and richness index and 6: 100 bp ladder; lanes 2–5: replicate ribosomal intergenic
(0.21 < P < 0.93) (data not shown). spacer analysis (RISA) profiles.

Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 8, 2162–2169
No claim to original French government works
 Microbial diversity erosion and soil functioning 2165

100 analyses of the relationship between diversity and func-

tion were restricted to those treatments with similar abun-
80 dances. The fact that we obtained, after recolonization,
Dissimilarity (%)

lower abundances of denitrifiers and ammonia oxidizers

in soil inoculated with high dilutions compared with soil
60
inoculated with lower dilutions is likely because these
organisms comprise decreasing proportions of the total
40
community, and because ammonia oxidizers have low
specific growth rates (Prosser, 1989).
20 Assessment of diversity within each functional group
revealed that the experimental procedure we used pro-
0 duced a progressive decrease in richness for each func-
tional group from microcosms inoculated with low soil
100

2.0
80
NS
Dissimilarity (%)

Carbon mineralization
1.5

(µg C h g dry soil)

60

40 1.0

-1 -1
20
0.5

0
0.0
100
1.0

80 NS
0.8
(µg N h g dry soil)
Dissimilarity (%)

Denitrification

60
0.6
-1 -1

40 0.4

20 0.2

0 0.0
-1 -3 -4 -5 -6 -8 2.5
10 10 10 10 10 10
Treatment NS
2.0
(µg N h-1 g-1 dry soil)

Fig. 4. Dissimilarity percentages of the genetic structures within four

replicate soil microcosms as a function of dilution level (for each
Nitrification

1.5
functional group, only treatments with similar group size are taken
into account). Dissimilarity percentages were computed from finger-
prints of (top) the eubacterial (middle) denitrifying and (bottom) 1.0
ammonia oxidizing communities. For details, see legend of Fig. 1.

0.5
a soil ecosystem, focusing on both general and special-
ized soil functions (carbon mineralization, denitrification 0.0
and nitrification). A range of microbial diversities was 10-1 10-3 10-4 10-5 10-6 10-8
established by inoculation of sterile soil microcosms with Treatment
serial dilutions of a suspension obtained from the same
Fig. 5. (Top) Carbon mineralization, (middle) denitrifying enzyme
non-sterile soil and subsequent incubation, to enable activity, and (bottom) nitrifying enzyme activity for treatments with
recovery of community sizes. For each functional group, similar functional group size. For details, see legend of Fig. 1.

Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 8, 2162–2169
No claim to original French government works
2166 S. Wertz et al.
Fig. 6. Simulated impact of a removal
1.0E+07
Dilution Simulated Simulated # approach by suspension/dilution on the number
level erosion (%) remaining species of soil bacterial species remaining in the inoc-
Simulated number of remaining

10-1 0 8.3 x 106 ulum. Mathematical simulation was made using

10-3 21.8 6.5 x 106 the frequency distribution of bacterial species
bacterial species (g soil)

8.0E+06 reported by Gans and colleagues (2005) for a

10-4 86.5 1.1 x 106
managed soil. Extinction of a species was
-1

10-5 97.7 1.9 x 105 assumed to occur when less than one member
10-6 99.6 3.3 x 104 of that species remained in the inoculum at a
6.0E+06 given level of dilution. The inset provides, for
10-8 99.99 8.4 x 10 2
each dilution level, the erosion level induced
(% of species removed) and the simulated num-
ber of species remaining.
4.0E+06

2.0E+06

0.0E+00
-2 -4 -6 -8 -10
0 10 10 10 10 10
Dilution level

dilutions to those inoculated with high dilutions. This is In microcosms inoculated with low dilutions, differences
consistent with the results obtained by Griffiths and in molecular fingerprints were observed between repli-
colleagues (2000; 2001) for the soil eubacterial com- cates. Given the good repeatability of the amplification/
munity using removal approaches. Molecular fingerprint fingerprinting procedure used (i.e. the same profile was
approaches are adequate to evaluate relative changes in obtained when a given sample was run twice), this was
the diversity of microbial communities, but do not allow likely due to the complexity of soil microbial communities
an accurate quantitative analysis of this diversity. Thus, that does not allow achievement of exact replication of
the richness index computed in our study clearly shows community structure. As observed by Griffiths and col-
that the erosion scenario used did induce a progressive leagues (2001) for the soil eubacterial community, varia-
decrease in microbial diversity, but did not allow us to tion between replicates of fingerprints of communities
quantify the amplitude of microbial diversity erosion. inoculated with high dilutions was greater than those from
Methods available to precisely quantify microbial diver- low dilutions, due to stochastic effects associated with
sity such as DNA reassociation are highly time-consum- species removal and initial stages of colonization.
ing and currently prohibitive for application to large Our results show that, despite the progressive decrease
numbers of samples. Similarly, the survey size required in richness of eubacterial community, carbon mineraliza-
for accurate analysis of microbial diversity is impracti- tion was unaffected, which is consistent with results of
cally large when using the cloning/sequencing approach, Griffiths and colleagues (2001). In addition, our results
in particular if less abundant species must be accurately show that decreasing the richness of denitrifying and
taken into account (Hughes, 1986). This is why we esti- ammonia-oxidizing communities did not affect their asso-
mated the amplitude of diversity erosion generated by ciated soil functions (denitrification and nitrification). Thus,
the dilution approach by mathematical simulation, using significant decrease in richness in soil communities did
richness values and abundance frequency distributions not impair soil ecosystem function, highlighting the low
of bacterial species reported for a soil (Gans et al., sensitivity of the functioning of soil microbial communities
2005) for which the value of bacterial community size to diversity erosion even for the most specialized commu-
was consistent with the total number of cultivable bacte- nities. The variability observed in community structure
ria measured on our study soil. Our simulation results between replicates corresponding to high dilutions indi-
show that the amplitude of diversity erosion generated cates that our data cannot be explained by consistent
by the suspension/dilution approach was huge (level of selection for some species.
species loss would be c. 99.99% for dilution treatment The maintenance of functioning of microbial functional
10−8). Because data on abundance distribution of spe- groups does not necessarily suggest that all microbial
cies are available only for the total bacterial community, species would play similar roles and that no functional
we could not evaluate the effect of the suspension/dilu- complementarity exists among species. One hypothesis
tion approach for less diverse communities, i.e. denitrifi- is that less abundant species, that are preferentially lost,
ers and ammonia oxidizers. are functionally less important. However, the high diversity

Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 8, 2162–2169
No claim to original French government works
 Microbial diversity erosion and soil functioning 2167
of soil microorganisms suggests that a more probable Inocula for soil microcosms were prepared by homogeniz-
hypothesis is that the number of species remaining after ing 125 g of soil (equivalent dry mass) in 250 ml of sterile
diversity erosion was sufficiently high to allow mainte- demineralized water by grinding with a pestle and a mortar,
followed by serial 10-fold dilution in demineralized water. Soil
nance of functioning. This is supported by our simulation
microcosms, consisting of 30 g of sterilized soil (equivalent
of the dilution effect. Indeed, we simulated that several dry mass) in flasks, were inoculated with 6 ml from each
hundreds of bacterial species could remain in the inocu- dilution giving inocula equivalent to 10−1−10−8 g of non-sterile
lum for the highest dilution treatment (10−8), which can soil g−1 sterile soil. For each inoculum level, four replicates
provide significant scope for functional diversity. were established, and seven microcosms containing sterile
More generally, our approach involving progressive soil were established as controls. Soil microcosms were incu-
removal of soil bacterial species according to their abun- bated at 20°C and moisture content was maintained at 30%,
equivalent to 70% water-holding capacity (WHC), by addition
dances is a realistic scenario of species loss. Indeed,
of sterile demineralized water. During 19 weeks of incubation,
Gans and colleagues (2005) showed that rare taxa that we surveyed microbial colonization (i.e. recovery of abun-
represented > 99.9% of the total number of bacterial taxa dance of each functional group) in inoculated soil micro-
were purged following exposure to high metal concentra- cosms (Appendix S1). After incubation for 19 weeks, four
tion. In our approach, progressive removal of species replicate microcosms per dilution level were sampled for
implies (i) an initial loss of less abundant species due to assessment of bacterial abundance, community structure
dilution, and (ii) differences in regrowth abilities of and and soil function.
interactions between the remaining species during the
recolonization phase. So, the final outcome of diversity
Enumeration of bacteria within functional groups
erosion for soil ecosystem functioning results from the
loss of organisms belonging to a given functional group Cultivable heterotrophic, denitrifying and ammonia oxidizing
but also from the collateral response of other organisms, bacteria were enumerated by the MPN technique as
described by Patra and colleagues (2005), except that 4 g of
which is itself of great interest for predicting the impact of
soil samples and fivefold serial dilutions were used. The lack
soil biodiversity erosion (Diaz et al., 2003).
of contamination during incubation was verified by enumera-
The three functional groups investigated represent dif- tion of cultivable heterotrophic, denitrifying and ammonia oxi-
ferent degrees of specialism. Carbon mineralization is dizing bacteria in control sterile soils.
carried out by the majority of bacterial heterotrophs;
approximately 0.1–5% of cultured bacterial species in soil
carry out denitrification (Tiedje, 1994); and soil ammonia Measurements of richness index for each functional group
oxidation is restricted to a single monophyletic group of DNA was extracted from 0.5 g of soil using the fast DNA SPIN
betaproteobacterial ammonia oxidizers (Kowalchuk and Kit for soil (BIO 101 Systems; Qbiogene, Carlsbad, CA, USA).
Stephen, 2001). Although additional studies testing the Community structure of eubacterial community was charac-
effect of other scenarios of diversity loss are needed, the terized by rRNA intergenic spacer (Ranjard et al., 2000),
consistency of our results in the three functional groups involving polymerase chain reaction (PCR) amplification of
studied indicates that high levels of functional redundancy the intergenic spacer region between the sequence coding
the large and the small ribosome subunits. Amplification was
are likely to be observed broadly for other functions and
carried out using a thermocycler (T personal, Biometra, Göt-
functional groups, and that biogeochemical cycles will be tingen, Germany) with an initial denaturation step of 5 min at
relatively unaffected by diversity loss if all functional groups 94°C, followed by 25 cycles of 94°C for 1 min, annealing at
are not eliminated and if their abundance can be recovered. 55°C for 1 min and elongation at 72°C for 1 min, with terminal
This is an important characteristic for the sustainability of elongation at 72°C for 5 min. Polymerase chain reaction prod-
terrestrial ecosystems in the context of global change. ucts were loaded on a 5% non-denaturing acrylamide gel
(Euromedex, Mundolsheim, France) that was run for 15 h at
65 V in 1× TBE buffer (Bio-Rad, Ivry sur Seine, France).
Community structure of denitrifying bacteria was charac-
Experimental procedures terized by denaturing gradient gel electrophoresis (DGGE)
Soil microcosms analysis of nirK gene fragments amplified by nested PCR,
using the primers of Liu and colleagues (2003) for the first
The 0–15 cm layer of a permanent grazed pasture soil PCR step and the same primers, modified by addition of a
located at Theix (45° North, 2° East) in France was collected GC clamp to the 5′ end of the reverse primer, in the second
(for soil characteristics, see Le Roux et al., 2003). Fresh soil PCR step. The final reagent concentrations for PCR were
was sieved (2 mm diameter mesh) and homogenized. Soil 1 µM primers, 200 µM of each dNTP, 1.75 U of Taq (Qbio-
was sterilized by 100 kGy gamma irradiation from a 60Co gene, Carlsbad, USA), in 50 µl of 10 mM Tris-HCl, 50 mM
source (Ionisos, Dagneux, France) and stored at 4°C for KCl, 0.1% Triton X-100, 1.5 mM MgCl2, pH 9. Thermocycling
5 weeks before use. Soil sterility was checked by enumera- conditions were: 5 min at 94°C followed by a touchdown
tion of heterotrophic bacteria by the most probable number between 72°C and 67°C (one cycle of 94°C for 30 s, 72°C
(MPN) technique (Alexander, 1982). for 1 min and 72°C for 1 min) and four subsequent cycles in

Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 8, 2162–2169
No claim to original French government works
2168 S. Wertz et al.
which the annealing temperature was decreased by 1°C/ total number of cultivable bacteria measured on our native
cycle to reach 68°C. Touchdown was followed by 25 cycles study soil. Gans and colleagues (2005) estimated the rich-
(30 s at 94°C, 1 min at 67°C and 1 min at 72°C) followed by ness and frequency distribution of bacterial species on their
7 min extension at 72°C. Denaturing gradient gel electro- soil by analysis of DNA reassociation kinetics accounting for
phoresis analysis of PCR products employed the D-Code uneven abundance of bacterial species. We mathematically
Universal Mutation Detection System (Bio-Rad) with a 6% simulated the impact of a dilution approach assuming that
polyacrylamide gel containing a gradient of 35–65% denatur- extinction of a species occurred when less than one member
ant, 100% denaturing solution being defined as 7 M urea and of that species remained in the inoculum at a given level of
40% formamide. Gels were run for 4.5 h at 150 V in 1× TAE dilution. For each dilution level, we computed the erosion
buffer at 60°C. level induced (% of species removed) and the number of
Ammonia oxidizing community structure was characterized species remaining.
by PCR-DGGE analysis of 16S rRNA genes with primary
amplification by CTO189f and CTO654r primers (Kowalchuk
et al., 1997), targeting the majority of betaproteobacterial Acknowledgements
ammonia oxidizers, and secondary amplification using bac-
terial 357f-GC and 518r primers (Muyzer et al., 1993). Ampli- We thank the Alliance program and IFR41 (University Lyon
fication conditions and DGGE analysis are as presented by I) for support. S. Wertz acknowledges funding of a postgrad-
Freitag and Prosser (2003). uate studentship by the French Ministry of Research. We
All gels were stained, photographed and analysed using thank S. Lavorel and K. Killham for helpful comments.
BiocaptMW (Vilber-Lourmat, France) and Gel ComparII soft-
ware (Applied Maths, Kortrijk, Belgium). For each functional
group, richness (R) was estimated as the number of bands. References
For the three bacterial communities, the percentage of dis-
similarity in genetic structures within replicates of each treat- Alexander, M. (1982) Most probable number method for
ment was computed using PRIMER software (PRIMER-E, microbial populations. In Methods of Soil Analysis, Part II:
Plymouth, UK). Chemical and Microbiological Methods. Page, A.L., Miller,
R.H., and Keeney, D.R. (eds) Madison, WI, USA: American
Society of Agronomy, pp. 815–820.
Soil ecosystem function Avrahami, S., Liesack, W., and Conrad, R. (2003) Effects of
temperature and fertilizer on activity and community struc-
Carbon mineralization was determined by incubating 2 g of ture of soil ammonia-oxidizers. Environ Microbiol 5: 691–
soil in closed plasma flasks for 7 h at 28°C. CO2 concentra- 705.
tion was determined in gas samples after 3, 5 and 7 h using Bengtsson, J., Engelhardt, K., Giller, P., Hobbie, S.,
a gas chromatograph (Agilent P200 Microcatharometer, Lawrence, D., Vilà, M., and Wolters, V. (2002) Slippin’ and
USA). slidin’ between the scales: the scaling components of
Denitrifying and nitrifying enzyme activities were deter- biodiversisty-ecosystem functioning relations. In Biodiver-
mined as production rates of N2O (C2H2 was used to inhibit sity and Ecosystem Functioning: Synthesis and Perspec-
N2O-reductase activity) and NO3–, respectively, over short tives. Loreau, M., Naeem, S., and Inchausti, P. (eds)
time periods (8 h at 28°C). The methods used are described Oxford, UK: Oxford University Press, pp. 209–220.
by Patra and colleagues (2005) except that 2 g of soil sam- Chapin,F.S., III, Zavaleta, E.S., Eviner, V.T., Naylor, R.L.,
ples were used. Vitousek, P.M., Reynolds, H.L., et al. (2000) Conse-
quences of changing biodiversity. Nature 405: 234–242.
Degens, B.P. (1998) Decreases in microbial functional diver-
Statistical analyses sity do not result in corresponding changes in decomposi-
Analysis of variance was performed to test if the size, rich- tion under different moisture conditions. Soil Biol Biochem
ness index and activity of each functional group differed 30: 1989–2000.
between dilution treatments. Then, for each variable, Dun- Diaz, S., Symstad, A.J., Chapin, F.S., III, Wardle, D.A., and
can’s test was used to determine if means differed signifi- Huenneke, L.F. (2003) Functional diversity revealed by
cantly. The significance of linear regressions between (i) removal experiments. Trends Ecol Evol 18: 140–146.
activity of each functional group, and (ii) dilution treatments Freitag, T.E., and Prosser, J.I. (2003) Community structure
or richness of the community were also tested. of ammonia-oxidizing bacteria within anoxic marine sedi-
ments. Appl Environ Microbiol 69: 1359–1371.
Gans, J., Murray, W., and Dunbar, J. (2005) Computational
Mathematical simulation of erosion of bacterial diversity improvements reveal great bacterial diversity and high
induced by dilution metal toxicity in soil. Science 309: 1387–1389.
Griffiths, B.S., Ritz, K., Bardgett, R.D., Cook, R., Christensen,
We simulated the impact of the removal approach by suspen- S., Ekelund, F., et al. (2000) Ecosystem response of pas-
sion/dilution on the number of soil bacterial species remain- ture soil communities to fumigation-induced microbial
ing in the inoculum. Mathematical simulation was made using diversity reductions: an examination of the biodiversity-
the total richness and frequency distribution curve reported ecosystem function relationship. Oikos 90: 279–294.
by Gans and colleagues (2005) for a managed soil for which Griffiths, B.S., Ritz, K., Wheatley, R., Kuan, H.L., Boag, B.,
values of bacterial community size were consistent with the Christensen, S., et al. (2001) An examination of the

Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 8, 2162–2169
No claim to original French government works
 Microbial diversity erosion and soil functioning 2169
biodiversity-ecosystem function relationship in arable soil on microbial functional groups involved in soil N dynamics.
microbial communities. Soil Biol Biochem 33: 1713–1722. Ecol Monogr 75: 65–80.
Griffiths, B.S., Kuan, H.L., Ritz, K., Glover, L.A., McCaig, Patra, A.K., Abbadie, L., Clays-Josserand, A., Degrange, V.,
A.E., and Fenwick, C. (2004) The relationship between Grayston, S.J., Guillaumaud, N., et al. (2006) Effects of
microbial community structure and functional stability, management regime and plant species on the enzyme
tested experimentally in an upland pasture soil. Microb activity and genetic structure of N-fixing, denitrifying and
Ecol 47: 104–113. nitrifying bacterial communities in grassland soils. Environ
van der Heijden, M.G.A., Klironomos, J.N., Ursic, M., Mou- Microbiol 8: 1005–1016.
toglis, P., Streitwolf-Engel, R., Boller, T., et al. (1998) Mych- Prosser, J.I. (1989) Autotrophic nitrification in bacteria. Adv
orrhizal fungal diversity determines plant biodiversity, Microbiol Physiol 30: 125–181.
ecosysteem variability and productivity. Nature 396: 69–72. Ranjard, L., Poly, F., Combrisson, J., Richaume, A., Gour-
Hooper, D.U., and Vitousek, P.M. (1997) The effect of plant bière, F., Thioulouse, J. and Nazaret, S. (2000) Heteroge-
composition and diversity on ecosystem processes. neous cell density and genetic structure of bacterial pools
Science 277: 1302–1305. associated with various soil microenvironments as deter-
Hughes, R.G. (1986) Theories and models of species abun- mined by enumeration and DNA fingerprinting approach
dance. Am Nat 128: 879–899. (RISA). Microb Ecol 39: 263–272.
Jonsson, L.M., Nilsson, M.-C., Wardle, D.A., and Zackrisson, Salonius, P.O. (1980) Metabolic capabilities of forest soil
O. (2001) Context dependent effects of ectomychorrhizal microbial populations with reduced species diversity. Soil
species richness on tree seedling productivity. Oikos 93: Biol Biochem 13: 1–10.
353–364. Seghers, D., Verthé, K., Reheul, D., Bulcke, R., Siciliano,
Kowalchuk, G.A., and Stephen, J.R. (2001) Ammonia- S.D., and Top, E.M. (2003) Effect of long-term herbicide
oxidizing bacteria: a model for molecular microbial ecology. applications on the bacterial community structure and func-
Annu Rev Microbiol 55: 485–529. tion in an agricultural soil. FEMS Microbiol Ecol 46: 139–146.
Kowalchuk, G.A., Stephen, J.R., De Boer, W., Prosser, J.I., Setälä, H., and McLean, M.A. (2004) Decomposition rate of
Embley, T.M., and Woldendorp, J.W. (1997) Analysis of organic substrates in relation to the species diversity of soil
ammonia-oxidizing bacteria of the β subdivision of the saprophytic fungi. Oecologia 139: 98–107.
class Proteobacteria in coastal sand dunes by denaturing Symstad, A.J., and Tilman, D. (2001) Diversity loss, recruit-
gradient gel electrophoresis and sequencing of PCR- ment limitation, and ecosystem functioning: lessons
amplified 16S ribosomal DNA fragments. Appl Environ learned from a removal experiment. Oikos 92: 424–435.
Microbiol 63: 1489–1497. Tiedje, J.M. (1994) Denitrifiers. In Methods of Soil Analysis.
Le Roux, X., Bardy, M., Loiseau, P., and Louault, F. (2003) Part 2. SSSA Book Series 5. Weaver, R.W., Angle, J.S.,
Stimulation of soil nitrification and denitrification by grazing and Bottomley, P.S. (eds). Madison, WI: SSSA, pp. 245–
in grasslands: do changes in plant species composition 267.
matter? Oecologia 137: 417–425. Wardle, D.A. (2002) Communities and Ecosystems: Linking
Liu, X., Tiquia, S.M., Holguin, G., Wu, L., Nold, S.C., Devol, the Aboveground and Belowground Components. Prince-
A.H., et al. (2003) Molecular diversity of denitrifying genes ton, NJ, USA: Princeton University Press.
in continental margin sediments within the oxygen-deficient Wardle, D.A., and Zackrisson, O. (2005) Effects of species
zone off the Pacific Coast of Mexico. Appl Environ Micro- and functional group loss on island ecosystem properties.
biol 69: 3549–3560. Nature 435: 806–810.
Loreau, M., Naeem, S., and Inchausti, P. (2002) Biodiversity Webster, G., Embley, M.T., Freitag, T.E., Smith, Z., and
and Ecosystem Functioning: Synthesis and Perspectives. Prosser, J.I. (2005) Links between ammonia oxidizer spe-
Oxford, UK: Oxford University Press. cies composition, functional diversity and nitrification kinet-
McGrady-Steed, J., Harris, P.M., and Morin, P.J. (1997) ics in grassland soils. Environ Microbiol 5: 676–684.
Biodiversity regulates ecosystem predictability. Nature Wolters, V., Silver, W.L., Bignell, D.E., Coleman, D.C.,
390: 162–165. Lavelle, P., van der Putten, W.H., et al. (2000) Effects of
Mikola, J., and Setälä, H. (1998) Relating species diversity global changes on above- and belowground biodiversity in
to ecosystem functioning: mechanistic backgrounds and terrestrial ecosystems: implications for ecosystem func-
experimental approach with a decomposer food web. tioning. Bioscience 50: 1089–1098.
Oikos 83: 180–194.
Muyzer, G., De Waal, E.C., and Uitterlinden, A.G. (1993) Supplementary material
Profiling of complex microbial populations by denaturing
gradient gel electrophoresis analysis of polymerase chain The following supplementary material is available for this
reaction-amplified genes coding for 16S rRNA. Appl Envi- article online:
ron Microbiol 59: 695–700. Appendix S1. Abundances of culturable heterotrophs dur-
Naem, S., and Li, S. (1997) Biodiversity enhances ecosystem ing incubation time for 19 weeks (135 days) in sterile soil
reliability. Nature 390: 507–509. microcosms inoculated with non-sterile soil suspensions. For
Nannipieri, P., Asher, J., Ceccherini, M.T., Landi, L., Pietra- legibility, kinetics for treatments 10−1, 10−3 and 10−6 only are
mellara, G., Renella, G. (2003) Microbial diversity and soil presented. The abundance of culturable heterotrophs in the
functions. Eur J Soil Sci 54: 655–670. native, non-sterile soil is indicated for comparison.
Patra, A.K., Abbadie, L., Clays-Josserand, A., Degrange, V., This material is available as part of the online article from
Grayston, S.J., Loiseau, P., et al. (2005) Effects of grazing http://www.blackwell-synergy.com

Journal compilation © 2006 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 8, 2162–2169
No claim to original French government works