ECOSYSTEM CARBON CYCLING: RELATIONSHIPS TO SOIL MICROBIAL
COMMUNITY STRUCTURE
by
MICHAEL SCOTT STRICKLAND
(Under the Direction of Mark A. Bradford)
ABSTRACT
Many processes related to carbon-cycling in terrestrial ecosystems are carried out by soil
microbes. Our understanding of the relationship between carbon cycling and microbial
community structure is, at best, rudimentary. The role of microbial community structure
in carbon-cycling processes has, for the most part, been treated as either functionally
redundant or driven by binary distinctions (e.g. fungal:bacterial ratios). It was the purpose
of this work to test these generalizations and advance our understanding how soil
microbes regulate ecosystem carbon dynamics. This purpose was accomplished using
field-based observation and controlled lab-based experimentation. In the first of the field-
based studies, I determine the mineralization dynamics of glucose and its relationship to
land-use, microbial community and edaphic characteristics across a southeastern U.S.
landscape. I find that the size, activity, and fungal:bacterial dominance of the microbial
community is unrelated to glucose mineralization but that land-use and the underlying
edaphic variable, soil phosphorus, explains spatial variation in this process. In the second
field-based study, I examine how an invasive grass affects belowground carbon-cycling
in a southeastern forest. Again, I find no role of fungal:bacterial dominance but observe
that the presence of the invasive grass accelerates carbon-cycling and depletes soil carbon
stocks, likely via ‘priming’ of the microbial community’s activity. Using lab-based
experimentation, I explore the role of microbial community structure separate to the
influence of confounding variables. I find that litter mineralization rates are dependent on
the microbial community and that these communities exhibit ‘home-field advantage’.
These results refute theories related to functional redundancy. In a follow-up experiment,
I find that the perception of litter quality, of two chemically-distinct litters, is dependent
on the microbial community. Specifically, communities sourced from herbaceous habitats
perceive more chemically-simple litter to be of higher quality than more chemically-
complex litter, whereas communities from forest habitats mineralize both litters similarly.
Together, my findings show that although microbial community structure may appear
unrelated to carbon dynamics in situ, controlled-experimentation reveals that this
structure may play an important role in determining rates of ecosystem processes. Future
work needs to resolve why there is an apparent disconnect between results from
observational and controlled-experimental studies.
INDEX WORDS: carbon-cycling, microbial communities, bacteria, fungi, land-use,
invasive species, glucose, stable isotopes, priming effect,
preferential substrate utilization, leaf litter
ECOSYSTEM CARBON CYCLING: RELATIONSHIPS TO SOIL MICROBIAL
COMMUNITY STRUCTURE
by
MICHAEL SCOTT STRICKLAND
B.A., William Jewell College, 2005
A Dissertation Submitted to the Graduate Faculty of The University of Georgia in Partial
Fulfillment of the Requirements for the Degree
DOCTOR OF PHILOSOPHY
ATHENS, GEORGIA
2009
© 2009
Michael Scott Strickland
All Rights Reserved
ECOSYSTEM CARBON CYCLING: RELATIONSHIPS TO SOIL MICROBIAL
COMMUNITY STRUCTURE
by
MICHAEL SCOTT STRICKLAND
Major Professor: Mark A. Bradford
Committee: Mac A. Callaham
Lisa A. Donovan
Noah Fierer
Jacqueline E. Mohan
Electronic Version Approved:
Maureen Grasso
Dean of the Graduate School
The University of Georgia
August 2009
iv
DEDICATION
I dedicate this dissertation to Delmer and Minnie Strickland, the best parents in
the world. I could not have accomplished this without your love and support.
v
ACKNOWLEDGEMENTS
I would like to thank my family, friends, and colleagues who have helped me
make it through this process in more ways than you can ever imagine. Catherine, thank
you for your love and sticking by me during this, I know I can be grumpy at times.
You’re my best friend and I look forward to spending the rest of my life with you. Mom,
Dad, John, Donald, Tisha, Lisa, Sophia, Shawn thanks for being the most loving and
supportive family I could ever hope for. Winston thanks for being a good dog. Paul
thanks for introducing me to ecology and the soil. Thanks to Becky, Bruce, Kristina,
Keith, Carol, Chris, Laura, Jesse, Derek, Chip, Julie, Ching-Yu, Carlye, Carrie, Mark G.,
and the rest of the EcoGhetto and friends for making my time in Georgia a lot of fun.
Thanks to Johannes, Becky, Bruce, Yolima, Krista, and all my other collaborators you
have made me think about the soil in many different ways. Bradford Lab, Chris, Ashley,
Ken, Augustina, Jaya, Tara, Ernie, Chad, Karen, Calley, Brian, and Jimmy thanks for
your help, friendship, and just making the lab a fun place to be. Tom and the ACL thanks
for running all my samples. Special thanks to my committee, Mac, Lisa, Noah, and Jackie
for your guidance. Mark, special thanks to you for putting up with me, pushing me to
achieve, and being a damn good advisor. Thanks to the staff, students, and faculty of the
Odum School. Finally, thanks to everyone else that I may have unintentionally left out.
This is really something that would be impossible to accomplish without the love,
friendship, and support of many people. This dissertation was supported by a grant from
vi
the Andrew W. Mellon Foundation and a Doctoral Dissertation Improvement Grant from
the National Science Foundation.
vii
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS .................................................................................................v
CHAPTER
1 INTRODUCTION AND LITERATURE REVIEW .........................................1
2 RATES OF IN SITU CARBON MINERALIZATION IN RELATION TO
LAND-USE, MICROBIAL COMMUNITY AND EDAPHIC
CHARACTERISTICS ................................................................................52
3 GRASS INVASION OF A HARDWOOD FOREST IS ASSOCIATED WITH
DECLINES IN BELOWGROUND CARBON POOLS .............................99
4 TESTING THE FUNCTIONAL SIGNIFICANCE OF MICROBIAL
COMMUNITY COMPOSITION..............................................................147
5 LITTER QUALITY IS IN THE EYE OF THE BEHOLDER: INITIAL
DECOMPOSITION RATES AS A FUNCTION OF INOCULUM
CHARACTERISTICS ...............................................................................184
6 CONCLUSIONS............................................................................................220
APPENDICES
A REFERENCES USED TO CONSTRUCT FIGURE 1.1 ..............................231
B REFERENCES USED TO CONSTRUCT FIGURE 1.2 ..............................244
C CANDIDATE MODEL SET .........................................................................247
viii
D FIGURE SHOWING SOIL RESPIRATION DURING THE PULSE-CHASE
EXPERIMENT ..........................................................................................250
E PARAMETER ESTIMATES ........................................................................251
F TABLE SHOWING THE INITIAL %C, %N, AND THE C:N RATIOS OF
THE LITTERS AND THE SOIL INOCULA ...........................................253
G BRAY-CURTIS SIMILARITY IN MICROBIAL COMMUNITY
FUNCTION ...............................................................................................254
H EDAPHIC CHARACTERISTICS FOR THE CALHOUN PLOTS .............258
1
CHAPTER 1
INTRODUCTION AND LITERATURE REVIEW
Fungal-to-bacterial dominance in soil – A review1
Soil microbial communities are an integral component of many ecosystem level
processes from N-fixation and methane oxidation to the mineralization of both simple
and complex organic compounds (Bedard and Knowles 1989, Zak et al. 2006, Jackson et
al. 2007). Because of this, the role that these communities play at the ecosystem level has
become a major line of study in both soil and microbial ecology (Jackson et al. 2007).
Although a seemingly straightforward proposition, actually gaining an understanding of
these communities regarding their relationship to environmental factors and ecosystem
function, as well as developing the methods to accurately assess them, has often proven
difficult (Barns et al. 1999). One reason for this is likely the fact that these soil
communities are some of the most diverse on Earth, with in excess of 4 × 104
species
found in a single soil sample (Torsvik et al. 2002, Fierer et al. 2007b). A second
contributor to these difficulties is the opaque nature of the soil environment itself, which
makes the direct observation of soil communities nearly, if not totally, impossible.
Given both of these difficulties and likely others, the predominant approach to
understanding soil microbial communities has often been to simplify the community by
dividing it into two categories of ecologically meaningful groups (Koch 2001). An
example of this is the idea of copiotrophs versus oligotrophs where copiotrophs are
1 Strickland, M.S., J. Rousk. To be submitted to Soil Biology & Biochemistry
2
organisms which thrive under high resource conditions and oligotrophs thrive under low
resource conditions (Poindexter 1981, Koch 2001, Fierer et al. 2007a). Other examples of
these sorts of distinctions include Winogradsky‘s idea of autochthonous versus
zymogenous microbial biomass (Winogradsky 1924) and r-versus K-selected organisms
(Fierer et al. 2007a). Although these sorts of distinctions have aided greatly in our
theoretical understanding of soil microbial communities, they all suffer from the fact that
they are largely unmeasurable in the environment. That is, assessing whether or not a
community is largely composed of copiotrophs or oligotrophs, for example, requires
researchers to both know the abundance of specific taxonomic groups in a given soil as
well as their life-history characteristics (Jackson et al. 2007, Green et al. 2008). Such an
understanding of specific organisms is currently lacking, making measurements based on
these community distinctions impossible (Jackson et al. 2007, Green et al. 2008).
One measure of the soil microbial community which is likely to be relevant to
both taxonomy and life history characteristics is the dominance of fungi to bacteria
(Wardle et al. 2004, van der Heijden et al. 2008). Although the methods used to
determine fungal:bacterial dominance should not necessarily be considered easy, fungi
and bacteria in general differ with regards to key traits (e.g. PLFA markers) which allow
researchers to distinguish between them (Anderson and Domsch 1973, Frostegård and
Bååth 1996). Furthermore, these two groups are expected to differ with regards to their
life history characteristics which in turn are likely to influence both their response to
environmental change as well as their impact on ecosystem processes (Hendrix et al.
1986, Holland and Coleman 1987, Yeates et al. 1997, Bardgett and McAlister 1999).
That is the fungal:bacterial dominance of a soil community is both taxonomically and
3
likely ecologically relevant (van der Heijden et al. 2008). For this reason as well as the
difficulties associated with making other categorical distinctions of the microbial
community, assessing soil communities based on their fungal:bacterial dominance has
been widely used (Bardgett et al. 1996, Yeates et al. 1997, Bardgett and McAlister 1999,
Rousk et al. 2009, Rousk and Nadkarni 2009). This simple but elegant approach has
provided insight into the opaque soil system and led to concepts which are widely used
and discussed in today‘s soil ecology literature (Figure 1).
The purpose of this review will be to examine many of these concepts. First we
will briefly review and discuss the prominent methods employed to assess
fungal:bacterial dominance (Section I). In Section II, we will examine some of the major
environmental factors which are likely to lead to differences in the fungal:bacterial
dominance of soil microbial communities. The third section will discuss the relationship
between fungal:bacterial dominance and two major ecosystem processes: carbon (C)
sequestration and litter decomposition. Finally, we will conclude by examining our
current understanding of fungal:bacterial dominance and highlight areas where greater
knowledge may lead to advances in this concept. Ultimately, we hope that this review
will provide a general overview on concepts related to the fungal:bacterial dominance of
soil microbial communities.
Section I – Techniques to measure fungal:bacterial dominance
Over the years a wide array of techniques have been developed which allow for the
assessment of soil microbial communities based on their fungal:bacterial dominance
(Joergensen and Wichern 2008). Although many of these measures utilize different
4
approaches, they are all driven by the use of key distinctions which can be made between
fungi and bacteria. For example, Frostegård and Bååth (1996) used PLFA‘s which were
distinct to bacteria and fungi to estimate the biomass of each group in soil. Alternatively,
Anderson and Domsch (1973) utilized selective inhibition with antibiotics which inhibit
either fungi or bacteria to similarly estimate their dominance. Both of these techniques, as
well as direct observation (i.e. microscopy) and the use of cell wall derived indicators of
fungi and bacteria (e.g. chitin and muramic acid), were recently reviewed by Joergensen
and Wichern (2008). This review found that across major land-use groups (i.e. cultivated,
grassland, forest, and litter layers) these methods generally gave comparable results
(Joergensen and Wichern 2008). The fact that each of these methods is ultimately an
estimate of the biomass of the fungi and bacteria may have led to these similarities.
Measures of whole microbial biomass using different methods are often found to
correlate across large spatial scales but are not necessarily 1:1 at smaller spatial scales
(Wardle and Ghani 1995). This discrepancy as resolution increases is largely attributed to
the possibility that different biomass measures actually gauge slightly different
components/characteristics of the microbial community (Wardle and Ghani 1995).
Whether this holds true for estimates of fungal:bacterial dominance determined via
different methods remains unknown. Furthermore, it is often difficult to ascertain the
status/activity of the microbial biomass in soil (Rousk and Bååth 2007b). Meaning that
methods such as these make the assumption that the biomass of a given group is a direct
reflection of that groups contribution to a certain ecosystem process or its response to
environmental change (Wright et al. 1995, Bååth 1998). Such assumptions have been
frequently questioned (Wright et al. 1995, Bååth 1998, Rousk and Bååth 2007b).
5
A recent alternative measure for assessing fungal:bacterial dominance, is the
use of DNA-based approaches. One of the more widely used of these approaches is
quantitative polymerase chain reaction (qPCR). This approach uses the accumulation of a
florescent reporter molecule during the PCR reaction coupled with primers specific to
either bacteria or fungi (which typically target the 16s or 18s rRNA gene segments,
respectively) to determine the relative abundance of these groups (Fierer et al. 2005).
qPCR represents both a rapid and quantifiable approach to assess fungal:bacterial
dominance in soils (Fierer et al. 2005). However, like with the above biomass measures
there are caveats associated with this technique. Foremost, is that fungal:bacterial
dominance as determined by qPCR may not be indicative of the true abundances of these
groups in soil especially on a per biomass basis. Reasons for this include the fact that
fungal cells may include many or no nuclei, leading to an over or underestimation of
fungal abundance, DNA extraction efficiencies may differ between these two groups, and
the amplification of genes may not be consistent across all fungal and bacterial taxa
(Martin-Laurent et al. 2001). Finally, like with biomass measures, assessments of
fungal:bacterial dominance based on DNA may not necessarily be related to a given
ecosystem process or response to environmental change. This is because DNA is present
in both active and inactive cells (Nocker and Camper 2009). Fungal:bacterial dominance
determined via RNA based approaches may ultimately prove more informative when
assessing relationships which are likely to be driven by active organisms (Poulsen et al.
1993, Aviv et al. 1996, Vestergard et al. 2008, Nocker and Camper 2009).
Another method which is distinct from biomass and DNA based approaches, is
the determination of bacterial and fungal growth (Bååth 1992, Rousk and Bååth 2007b).
6
This method uses leucine/thymidine incorporation to measure bacterial growth and
acetate-in-ergosterol incorporation to measure fungal growth (Bååth 1992, Rousk and
Bååth 2007b). Both techniques estimate the biomass production of bacteria or fungi and
may provide a direct estimate of the relationship between these groups and a given
ecosystem process or their response to environmental change (Rousk and Bååth 2007b).
In fact these techniques have been successfully used to determine relative changes in
fungal and bacterial contribution to decomposition (Rousk and Bååth 2007a). However, a
lack of adequate conversion factors have made quantification of bacterial and fungal
growth in units C difficult, although developments suggest progress for both bacteria
(Bååth 1994) and fungi (Rousk and Bååth, 2007b). Furthermore the accuracy of these
assessments of fungal:bacterial dominance are contingent upon certain requirements
being met such as balanced growth or that all bacterial or fungal taxa can take up their
respective indicator molecule through the cell membrane (Winding et al. 2005).
In summary, all the techniques used to determine fungal:bacterial dominance
in soil are by necessity indirect and are thus susceptible to various confounding factors
that should be carefully considered when relating them to either ecosystem processes or a
community‘s response to environmental change. It is also likely that these relationships
will often be dependent on the method used to assess fungal:bacterial dominance. The
possibility that different methodologies lead to different results or at least
stronger/weaker relationships should be addressed in the future. This will likely require
collaborations between researchers whose expertise lay in assessing fungal:bacterial
dominance via differing methods. Currently the wide array of methods used to assess
fungal:bacterial dominance and the fact that researchers often relate this ratio to a wide
7
array of ecosystem processes and environmental factors makes a formal meta-analysis
nearly impossible. However, generalizations related to the environmental factors likely to
affect fungal:bacterial dominance and the impact fungal:bacterial dominance has on
ecosystem processes can still be made.
Section II – Environmental factors affecting fungal:bacterial dominance
Much of the rationale for using fungal:bacterial dominance as an assessment of the soil
microbial community rests on the proposition that these two groups of organisms respond
dissimilarly to an array of environmental factors (van der Heijden et al. 2008). Such
environmental factors include physical disturbance of the soil (e.g. tilling), nutrient
availability, soil moisture, and soil pH. As these factors and others differ in direction
and/or intensity then the general expectation is that fungi and bacteria will often exhibit
contrasting responses (Hendrix et al. 1986, Holland and Coleman 1987, van der Heijden
et al. 2008). The general consensus has been that highly disturbed and/or nutrient rich
soils are dominated by bacteria whereas fungi dominate less disturbed and/or nutrient
poor soils (Bardgett and McAlister 1999, van der Heijden et al. 2008). For the remainder
of this section we will explore some of the key environmental factors likely to lead to
differences in the fungal:bacterial dominance of soil communities.
Physical disturbance and tilling effects
Much of the interest in the fungal:bacterial dominance of soils has been related to
agricultural practices (Hendrix et al. 1986, Holland and Coleman 1987, Drijber et al.
2000, Bailey et al. 2002, Helgason et al. 2009). Hendrix et al. (1986) proposed that no-till
8
agricultural practices would result in a fungal dominated system as opposed to the
bacterial dominated one expected under conventional tillage practices. This expected
outcome is largely based on key differences in the growth forms of fungi and bacteria
(Hendrix et al. 1986). Most fungi exhibit a hyphal growth form while most soil bacteria
are present as individual cells (Hendrix et al. 1986). The hyphal growth form of fungi
likely conveys several advantages to these organisms. Chief of which is the translocation
of nutrients and resources from microsites where these are abundant to sites where they
are limiting via hyphal networks (Hendrix et al. 1986, Frey et al. 2003). This
phenomenon has been commonly referred to as the hyphal bridge (Hendrix et al. 1986,
Frey et al. 2003). This fungal advantage though may be unfavourable under conventional
agricultural practices where the hyphae are continuously severed or disrupted due to
tillage (Hendrix et al. 1986, Helgason et al. 2009).
In addition to any direct damage caused by tillage, any advantage fungi gain
from hyphal bridges would likely be negated (Hendrix et al. 1986). Bacteria on the other
hand would be relatively unaffected by tillage due to their presence as individual cells
(Hendrix et al. 1986). Furthermore, tillage homogenizes the soil, evenly distributing
nutrients and resources. Just as surely as severing hyphae, this too negates many of the
advantages hyphal bridges convey fungi (Bardgett et al. 1996, Bardgett et al. 1999,
Bardgett and McAlister 1999). Add to this the generally held idea that bacteria are more
efficient colonizers of the organic resources (e.g. newly available SOM and dead
microorganisms) made available post-till than are fungi and it seems safe to conclude that
fungal:bacterial dominance will decrease under conventional tillage practices (Holland
and Coleman 1987). Similar arguments have been made to describe the negative impacts
9
earthworms have on fungal dominance as well as other factors which physically disturb
the soil (Butenschoen et al. 2007). Empirical evidence in support of such impacts on
fungal:bacterial dominance are far from definitive however.
Of five studies which explicitly compared the fungal:bacterial dominance of
communities exposed to conventional tillage practices to those exposed to no-till
practices, none found consistent evidence of tillage effects (Frey et al. 1999, Bailey et al.
2002, Pankhurst et al. 2002, Spedding et al. 2004, Helgason et al. 2009). Of these studies,
three found that changes to no-till or a reduction in tillage did lead to an overall increase
in microbial biomass but this increase was proportional for both bacteria and fungi
(Pankhurst et al. 2002, Spedding et al. 2004, Helgason et al. 2009). Bailey et al. (2002)
found that fungal dominance determined via selective inhibition increased under no-till
practices but the opposite was true when fungal dominance was determined using PLFA.
Although the PLFA measurements in this study were not replicated, it does suggest that
different methodologies may lead to dissimilar results (see Section I). Fungal:bacterial
dominance was determined via PLFA by Frey et al. (1999) and they found that fungal
dominance did increase under no-till practices. However, closer examination of these
results showed that tillage covaried with soil moisture and once this was accounted for,
tillage had no impact on fungal:bacterial dominance (Frey et al. 1999). Although studies
such as these raise doubt concerning the impact that physical disturbance of soil will have
on the fungal:bacterial dominance of the microbial community, they also demonstrate
that other factors may well influence the relative dominance of these two groups.
10
Direct and indirect effects of nutrients
Another characteristic likely to influence the fungal:bacterial dominance of soil microbial
communities is nutrient availability (Bardgett and McAlister 1999, Suzuki et al. 2009).
This like physical disturbance has been largely related to agricultural based management
decisions such as fertilizer inputs (Bardgett et al. 1996, Bardgett et al. 1999, Bardgett and
McAlister 1999). In many ways the hypothesized impacts fertilizer inputs are likely to
have on fungal:bacterial dominance are similar to those hypothesized for tillage. For
example, fertilizer applications in effect homogenize soil nutrient availability and likely
negate any advantages associated with the fungal growth form.
Stoichiometric differences between bacteria and fungi is another possible
reason why fungal:bacterial dominance is expected to decrease as nutrient inputs into the
soil increase (van der Heijden et al. 2008). Generally the carbon:nitrogen (C:N) ratio of
bacterial biomass is expected to be ~ 3-6 while the C:N ratio of fungal biomass is
expected to be ~ 5-15 (McGill et al. 1981). Because it can generally be assumed that
fungal biomass has a greater C:N ratio than bacterial biomass then nutrient applications
are likely to favour bacteria over fungi (Gusewell and Gessner 2009). This rationale is
not solely limited to explaining fertilizer inputs. It has also been used to explain the
effects of litter quality on fungal:bacterial dominance (see Section III for further details
related to this). However, whether or not the biomass stoichiometry of bacteria is distinct
enough from that of fungi to rationalize such an outcome is relatively unexplored.
Much of what is known about bacterial biomass C:N can be attributed to
cultured organisms and more often than not these organisms were grown on relatively
nutrient rich media (McGill et al. 1981). Bacteria under more natural settings are less
11
likely to encounter such nutrient rich environments and their actual C:N ratio may be
much higher than previously assumed. In fact a compilation of studies examining
communities of aquatic bacteria and cultures grown on nutrient poor media have found
that bacteria range in C:N from ~ 3-12 (Fig. 2). Furthermore, assessments of the C:N
ratio of fungi, based largely on sporocarps, show a range of ~3-51 for mycorrhizal fungi
and ~ 4-60 for saprotrophic fungi (Fig. 2). This in the least suggests the possibility that a
large proportion of fungi and bacteria, in general, overlap with regards to biomass C:N
(Fig. 2). In soils, less is known concerning this possible overlap in the fungal and
bacterial C:N ratio largely because of the difficulty in both measuring pure bacterial
biomass in situ and culturing specific soil bacteria (Madsen 2005, Kreuzer-Martin 2007).
Cleveland and Liptzin (2007) have provided some circumstantial evidence that
the C:N ratio of soil bacteria and fungi may not be as distinct a characteristic as
previously thought. They found no difference in microbial biomass C:N ratios across a
range of ecosystem types where fungal:bacterial dominance would be expected to differ
(Cleveland and Liptzin 2007). If bacteria and fungi are distinct with regards to their
biomass C:N ratio then the expectation would have been that total microbial biomass C:N
in fungal-dominated systems would have been greater than in bacteria-dominated
systems. This was not the case and in fact forests which are expected to have greater
fungal dominance than grasslands actually had a lower microbial biomass C:N ratio.
These results are suggestive of either a significant overlap in the C:N ratio of bacteria and
fungi or that the most abundant organisms in either group are stoichiometrically similar.
Regardless of this lack of stoichiometric distinction, several studies have
found that N fertilization impacts fungal:bacterial dominance. A prominent example is
12
Bardgett and McAlister (1999), who observed that pastures which received little or no
inputs of fertilizer N typically had a greater fungal:bacterial dominance than did sites
which received large inputs of fertilizer N. An experimental component of Bardgett and
McAlister‘s (1999) study, however, found that fertilizer cessation did not lead to a
significant increase in fungal:bacterial dominance after 6 years. They suggested that this
may have been due to high residual fertility.
Similar studies have also reported conflicting results for both pasture and
forest sites (de Vries et al. 2006, Demoling et al. 2008). de Vries et al. (2006) found using
a 2 year field-based experimental study that low fertilization rates led to an increase in
fungal:bacterial dominance. On the other hand, a follow up field observation found that
fungal:bacterial dominance was not related to changes in management intensity (de Vries
et al. 2007). This was because both bacterial and fungal biomass responded similarly to
changes in management intensity (de Vries et al. 2007). Demoling et al. (2008) found that
across an N deposition gradient in coniferous forests that fungal:bacterial dominance
(estimated using PLFA) increased as deposition decreased. Interestingly however,
Demoling et al. (2008) did not find a concomitant change in fungal growth (estimated
using acetate-in-ergosterol incorporation). The likely explanation for this was attributed
to a shift toward more mycorrhizal fungi as N-deposition decreased since the fungal
growth estimate (i.e. acetate-in-ergosterol incorporation) is likely only relevant to
saprotrophs. This increase in the biomass of mycorrhizal fungi raises an interesting and
less direct means that changes in nutrient levels are likely to affect fungal:bacterial
dominance.
13
The nutrient acquisition strategies of plants often change concomitantly with
changes in the concentration of soil nutrients (Lambers et al. 2008). In a review by
Lambers et al. (2008), plants in young soils which are typically rich in P but poor in N
were associated with relatively low levels of mycorrhizae and the same was true for
ancient soils, poor in both nutrients. On the other hand, the association between plants
and mycorrhizae are typically highest in soils where the N:P ratio is greatest (Lambers et
al. 2008). Soils typical of this sort of nutrient status are by large found in temperate zones
around the world where many studies investigating fungal:bacterial dominance have
likely been conducted. Additionally, changes in the status of soil nutrients have been
shown to dramatically impact mycorrhizal fungi with increased nutrient additions often
leading to their decline (Oehl et al. 2004, Cruz et al. 2009, but see Birkhofer et al. 2008).
These findings would suggest that plant nutrient status is likely to be a major factor
impacting the fungal:bacterial dominance of the microbial community (van der Heijden
et al. 2008). If researchers are interested in the effect that nutrient status has on
fungal:bacterial dominance then close consideration of mycorrhizal fungi may not be
necessary. On the other hand, if researchers are interested in relating ecosystem processes
to fungal:bacterial dominance then the distinction between mycorrhizal and saprotrophic
fungi is likely to prove important. Especially given the likely distinct ecological strategies
exhibited by these two groups.
Other effects
Other factors which have been proposed to impact fungal:bacterial dominance include
soil pH, temperature, and moisture (Holland and Coleman 1987, Rousk et al. 2009). Of
14
these soil pH, considered a sort of master variable, is likely to have significant impacts on
the fungal:bacterial dominance of microbial communities (Fierer and Jackson 2006).
Recent work has demonstrated that one of the major controls in soil on both the diversity
and composition of bacterial communities is pH (Fierer and Jackson 2006, Lauber et al.
In Review). This work has shown that bacterial diversity is related unimodialy to soil pH
with a peak in diversity found around pH 6-7 (Fierer and Jackson 2006, Lauber et al. In
Review). Soil pH also appears to be related to fungal dominance but this possible
relationship has been less intensively studied (Joergensen and Wichern 2008, Rousk et al.
2009). In general the expectation has been that fungi are likely more acid tolerant than are
bacteria which in turn leads to increased fungal dominance in acidic soils (Hogberg et al.
2007, Joergensen and Wichern 2008, Rousk et al. 2009).
This though does not appear to be a conclusive pattern with alterations in pH
sometimes resulting in increased, decreased, or unchanged levels of fungal dominance
(Marstorp et al. 2000, Bååth and Anderson 2003, Hogberg et al. 2007, Rousk et al. 2009).
One possible explanation for these seemingly inconsistent patterns may be that like soil
bacteria specific fungal taxa are impacted to a greater or lesser degree by changes in pH.
For example, Rousk et al. (2009) suggested that such observed variation may in part be
driven by an increase in ectomycorrhizal fungi in acidic soils due ultimately to a shift in
plant species composition. In this same work, they found that, like bacterial diversity,
fungal dominance followed a similar unimodal pattern (Rousk et al. 2009). They also
suggest that this pattern was largely driven by saprotrophic fungi because no
ectomycorrhizae were present at their study site (Rousk et al. 2009). Another possibility
is that fungal-dominance or the composition of the fungal community is simply not
15
impacted to the same degree by pH as the bacterial community is. Work conducted across
a land-use gradient in the Southeastern U.S. found the relative abundance of fungal taxa
were more strongly related to soil P and C:N ratios than any other edaphic characteristic
(Lauber et al. 2008). If this pattern holds true across a range of systems then it might
imply that the fungal component of the soil microbial community is likely to be impacted
to a lesser degree than is the bacterial community by changes in soil pH.
Our understanding of the impact that moisture and temperature have on soil
microbial communities has often been related to their role as mediators of ecosystem
processes such as decomposition (Aerts 1997). However, given the likelihood that
temperature and moisture regimes are apt to change both in degree and frequency across
broad geographical regions in the face of global climate change, these factors and their
effect on fungal:bacterial dominance have garnered increased attention (Allison and
Treseder 2008). In general as with most expected effects of environmental perturbations,
it has been proposed that fungi will exhibit less of a response to changing temperature
and moisture than will bacteria (Holland and Coleman 1987). This is in many ways
related to the fact that fungi have thick cells walls composed of chitin which are apt to be
more resistant to desiccation caused by warm, dry conditions or to the stress caused by
rapid changes in these conditions (Holland and Coleman 1987).
Frey et al. (1999) found, however, that fungal biomass was positively related
to soil moisture in cultivated settings while bacterial biomass remained relatively
constant. Yeilding a fungal:bacterial dominance that was positively related to soil
moisture (Frey et al. 1999). Similarly, Gordon et al. (2008) found that drying and
subsequent rewetting events led to a decrease in fungal:bacterial dominance with no
16
change in bacterial biomass but a significant change in fungal biomass. Both studies
show, in contrast to earlier expectations, that fungi are likely to be affected to a greater
extent by changes in soil moisture than are bacteria. Frey et al. (1999) proposed that
because bacteria are generally expected to inhabit soil pore spaces whereas fungal hyphae
are generally found on the exterior of soil aggregates then bacteria are in effect physically
protected from desiccation to some extent. Extending on this idea, bacteria would also be
less affected by mild drying/rewetting events. This is not to say that bacteria are in some
way physiologically more resistant/resilient to moisture effects but the habitat likely
occupied by these organisms may well infer greater resistance/resilience (Six et al. 2006).
If this is true then the protection from moisture stress inferred on the bacterial component
of the microbial community is likely to be influenced greatly by soil texture (Six et al.
2006).
Temperature effects like moisture do not appear to follow the general
consensus. In fact, the response of both bacteria and fungi to changing temperature
appears to largely be in the same direction resulting in little to no change in
fungal:bacterial dominance (Allison and Treseder 2008, Bárcenas-Moreno et al. 2009).
For example, Allison and Treseder (2008) found that the relative abundance (determined
via qPCR) of both bacteria and fungi declined by ~ 50% under experimental soil
warming. This would have resulted in no change in fungal:bacterial dominance although
specific fungal species were affected by warming (Allison and Treseder 2008). Another
study, found that the actual growth rates of fungi and bacteria (determined via
leucine/thymidine and acetate-in-ergosterol incorporation) responded similarly to
changing temperatures (Bárcenas-Moreno et al. 2009). It would appear that, in the
17
absence of concomitant moisture effects related to temperature and barring any
differences in acclimatization between these groups, there is no clear evidence that
temperature itself influences fungal:bacterial dominance.
Section III – Functional implications of fungal:bacterial dominance
As detailed in Section II, above, there is the expectation in soil ecology that fungi and
bacteria differ with regards to several key traits (van der Heijden et al. 2008). These
differences on one hand are expected to relate to how each group is likely to respond to
environmental change and can largely be summed up, albeit highly simplified, as a
relationship where what is favorable to bacteria is unfavorable to fungi or vice versa.
Extending these ideas has led to the implication of fungal:bacterial dominance as a
determinant of ecosystem processes (Hendrix et al. 1986, Bardgett and McAlister 1999,
van der Heijden et al. 2008). In many ways this is a logical extension of the expected
differences between these two groups and is comparable, for example, to ideas relating
plant community composition to primary production (Tilman et al. 2001). This has led to
the suggestion that changes in fungal:bacterial dominance are likely to be related to such
ecosystem processes as decomposition, nutrient cycling, C-sequestration potential, and
ecosystem self-regulation (Hendrix et al. 1986, Bardgett and McAlister 1999, Bailey et
al. 2002, van der Heijden et al. 2008).
It is the purpose of this section to explore the rationale for these linkages
between fungal:bacterial dominance and ecosystem processes and to discuss the evidence
in favor and against such linkages. Clearly, the rationale for many of the linkages
between fungal:bacterial dominance and ecosystem processes are closely related to the
18
environmental response of each group, as described in Section II above, and therefore
some redundancy between these sections is necessary. We have clearly cross referenced
such occurrences. Furthermore, given the wide-array of ecosystem process which
fungal:bacterial dominance may relate to, We have limited ourselves to the exploration of
two processes: C-sequestration and decomposition. Both of these should serve as
sufficient case studies of the rationale linking fungal:bacterial dominance to ecosystem
processes and also provide a broad overview of the other processes expected to be
affected.
C-sequestration and fungal-to-bacterial dominance
The C-sequestration potential of ecosystems has gained increased interest due to current
concerns pertaining to climate change (Bailey et al. 2002). The fungal:bacterial
dominance of a particular site has often been associated with that site‘s C-sequestration
potential with a greater potential associated with fungal dominance and a lesser one
associated with bacterial dominance (Jastrow et al. 2007). Much of the evidence to date
which relates fungal:bacterial dominance to C-sequestration has largely been gained from
comparisons of intensively managed and less intensively managed systems. For example,
Bailey et al. (2002) compared five land-use pairs which differed in management intensity
and found that total soil C and fungal activity were both higher for four of the five pairs
when management intensity was lower. Guggenberger et al. (1999) found the same for
three of six sites but Busse et al. (2009) found almost the opposite with fungal-dominated
sites having lower total soil C than bacterial dominated ones. The hypothesis that greater
fungal dominance is synonymous with a greater C-sequestration potential has become a
19
widely held idea in soil ecology and like many ideas linking fungal:bacterial dominance
to ecosystem processes is based on differences in specific traits between fungi and
bacteria (Guggenberger et al. 1999, Bailey et al. 2002, Six et al. 2006, van der Heijden et
al. 2008). Biomass stoichiometry, C-use efficiency (CUE), and the decomposability of
their respective necromass are all examples of fungal and bacterial traits which are
expected to differ and in turn impact C-sequestration.
It has been suggested that on average bacterial and fungal biomass will differ
with regards to their stoichiometry (see Section I). Closely related to these expected
differences in biomass stoichiometry and another trait expected to relate to greater C-
sequestration in fungal-dominated systems is CUE also known as growth yield efficiency
(e.g. Six et al 2006). It is expected that fungi on average produce more biomass-C per
unit of C metabolized than do bacteria which again leads to a greater proportion of C
stored in a fungal-dominated systems (high fungal:bacterial dominance) when compared
to bacterial-dominated systems (low fungal:bacterial dominance).
Six et al. (2006) conducted an extensive review on the CUE of bacteria and
fungi and found a significant degree of overlap between these two groups. More direct
experimental tests conducted by Thiet et al. (2006) found no difference in CUE between
sites differing in fungal:bacterial dominance and concluded that many of the differences
in CUE associated with changes in fungal:bacterial dominance were purely due to
methodological issues. In this experiment, though, Thiet et al. (2006) only determined the
amount of C mineralized and the amount remaining in the soil. They did not determine
the actual accumulation of C into the biomass of the respective groups. Furthermore,
whether or not bacterial and fungal CUE is similar under different environmental stresses
20
in soils is currently unknown. It might be expected that fungi or bacteria are more or less
resistant to certain types of stress such as drying/rewetting events (see Section II) and that
these stress events will lead to subsequent impacts on CUE. Another factor to consider is
how substrate C:N ratios influence CUE. It has been shown that as substrate C:N ratios
increase, a decrease in decomposer CUE is noted (Manzoni et al. 2008). Whether fungi in
comparison to bacteria reduce their CUE less as substrate C:N ratios increase is unknown
and like many other factors related to fungal and bacterial CUE warrants further
investigation.
The actual amount of time that C is stored in living fungal or bacterial biomass
is another of the proposed links between fungal:bacterial dominance and C-sequestration
(Six et al. 2006). Although there have been relatively few reports of biomass turnover for
either group, Bååth (1994) estimated that bacterial biomass turnover ranged from days to
weeks, averaging about one week. A similar estimate of turnover times for fungi showed
that turnover was on the order of months (i.e. ~ 130-150 days [Rousk and Bååth 2007b]).
This would suggest that C is stored longer in living fungal biomass than it is in living
bacterial biomass and ultimately may lead to an increased residence time for C associated
with the fungal component of the microbial community.
Like living biomass the decomposition of microbial necromass is another
anticipated factor linking fungal:bacterial dominance to C-sequestration (Guggenberger
et al. 1999). This is based on the expectation that fungal biomass is more chemically
recalcitrant than is bacterial biomass. The rationale for the greater chemical recalcitrance
of fungal biomass is due to cell-wall components which include chitin and membrane
components such as the potentially recalcitrant ergosterol (Mille-Lindblom et al. 2004,
21
Zhao et al. 2005). Also, fungal biomass, as discussed above, is expected to have a wider
C:N ratio than is bacterial biomass (McGill et al. 1981). Thus if a greater proportion of a
systems dead microbial biomass is composed of fungi then the decomposition of that
biomass may progress more slowly than if a greater proportion of that biomass were
derived from bacteria (Guggenberger et al. 1999). Concurrently, more C would remain
locked up over the same period of time in fungal rather than bacterial necromass leading
to increased C-sequestration in fungal dominated systems (Guggenberger et al. 1999).
However, there have been relatively few studies which have explicitly looked
at the decomposition rates of bacterial versus fungal biomass but in at least one instance
the decomposition of fungal biomass was equivalent to that of bacterial biomass (Li and
Brune 2005a). Although fungal and bacterial biomass in this study was only sourced from
two cultured organisms and may not represent the full spectrum of possibilities, it none
the less demonstrates that the necromass of certain species of bacteria and fungi may be
equally decomposable (Li and Brune 2005a). On the other hand, Nakas and Klien (1979)
found that the cell wall components of two bacteria species, in general, decomposed at
faster rates than did three fungal species. One possible explanation for this discrepancy
between studies may be that Li and Brune (2005) utilized a gram positive bacterium
whereas Nakas and Klien (1979) utilized a gram negative. Gram positive bacteria have a
thick cell wall composed of peptidoglycan whereas the cell wall of gram negative
bacteria is composed of a much thinner layer of peptidoglycan and an outer
lipopolysaccharide membrane. Since the cell walls of gram positive bacteria are
composed primarily of peptidoglycan then the decomposition of this groups cell wall
constituents may proceed at a slower rate than that of gram negative bacteria and may
22
even be slower than the chitinous cell walls of fungi. In fact a comparison, based on two
studies, of the mineralization of chitin and peptidoglycan shows that chitin is mineralized
at an equivalent or ~30% greater rate than peptidoglycan (Li and Brune 2005b, 2005a).
Although it can be argued that there is overlap in many of the characteristics
of bacteria and fungi which are expected to impact C-sequestration, perhaps the most
plausible mechanism leading to enhanced C-sequestration in a fungal-dominated system
can be attributed to the fungal growth form. The majority of fungi exhibit a hyphal
growth form which allows these organisms to grow into new habitats in order to access
new substrates (Hendrix et al. 1986, Guggenberger et al. 1999). In comparison, bacteria,
with the exception of the actinomycetes, are typically relegated to growth on the surface
layers of substrates (Guggenberger et al. 1999). This difference between bacteria and
fungi may lead to greater C-sequestration when a system is fungal-dominated for at least
two reasons. First, if fungal hyphae grow into a habitat such as a soil aggregate and those
hyphae die then they are largely protected from decomposition (Guggenberger et al.
1999, Six et al. 2006). Second, fungal activity may lead to increased aggregation via both
mechanical processes (e.g. entanglement of soil particles by hyphae; Guggenberger et al.
1999, Six et al. 2006) and chemical processes (e.g. the exudation of glomalin
[Guggenberger et al. 1999, Pikul et al. 2009]). An increase in soil aggregation in turn
may lead to incorporation of SOC into the aggregate which is then biologically
unavailable and sequestered (Guggenberger et al. 1999, Wilson et al. 2009). As fungal-
dominance increases then the likelihood of either of these two occurrences may
subsequently increase. Yet, many other physical and chemical properties of the soil are
also likely to influence soil aggregation and fungal-dominance may only be one of many
23
at play (Wilson et al. 2009). The importance of fungal-dominance as it pertains to
aggregate formation and possibly C-sequestration is an area where research is needed.
Litter decomposition and fungal-to-bacterial dominance
The decomposition of foliar leaf litter is perhaps one of the most widely studied processes
in terrestrial ecosystems and is expected to be influenced by the microbial community.
The role that fungi play in this process is expected to be distinct from the role played by
bacteria such that the outcome of this process will be dependent on the fungal:bacterial
dominance of the community (de Boer et al. 2005, van der Wal et al. 2006, Meidute et al.
2008). As litter recalcitrance increases then the role of fungi in decomposition is also
expected to increase (van der Heijden et al. 2008). To some degree this is based on the
C:N ratio of the leaf litter in question with a greater initial C:N expected to favor fungi
and a smaller initial C:N expected to favor bacteria (Gusewell and Gessner 2009).
Several mechanisms have been proposed to justify this expectation including
stoichiometric differences between bacteria and fungi, the hyphal growth form of fungi
which allows them to grow into leaf litter bypassing the more recalcitrant outer layers of
the leaf, and that hyphal bridges allow fungi to alleviate nutrient limitations associated
with litter decomposition by mining limiting nutrients from other sources (Hendrix et al.
1986, Holland and Coleman 1987). Also built into the expectation that recalcitrant litter
will favor fungal growth is the lignin degrading ability of fungi (de Boer et al. 2005).
There is a general consensus in the literature that fungi are capable of degrading lignin
whereas bacteria are not (de Boer et al. 2005, van der Heijden et al. 2008). This allows
fungi to exploit these structural components of litter leading to greater decomposition
24
rates (de Boer et al. 2005, van der Heijden et al. 2008). In this section we will explore
some of these mechanisms and, as above, assess the validity of relating fungal:bacterial
dominance to leaf litter decomposition as well as the role both bacteria and fungi play in
this process.
The argument that litter with a greater C:N ratio is likely to be primarily
decomposed by fungi is in part based on the expectation that fungi have a greater biomass
C:N ratio than bacteria (McGill et al. 1981). However, as mentioned above (see Section
II) there may be significant overlap between these two groups and further examination
with specific emphasis paid to those organisms inhabiting litter is necessary. In lieu of
such investigations, Güsewell and Gessner (2009) have provided research which suggests
that the role of fungi and bacteria in the decomposition of leaf litter may not be as clear
cut as once expected. In their investigation they found that bacteria dominated leaf litter
with a greater C:N ratio when phosphorous (P) was not limiting and this was attributed to
heterotrophic N-fixation by the litter inhabiting bacteria. Under these conditions bacteria
dominated when litter had a greater C:N ratio and ample P but fungi dominated when the
C:N ratio was lower and P was limiting (Gusewell and Gessner 2009). Comparing
treatments where there was a greater dominance of bacteria to those with a greater
dominance of fungi showed similar overall decomposition rates and in at least one
instance the bacterial-dominated treatment had a higher rate (Gusewell and Gessner
2009). Other studies have demonstrated that the role of bacteria and fungi in the
decomposition of recalcitrant litter resources may be in part dependent on the specific
organisms involved in the process (Hunt et al. 1988, Gholz et al. 2000, Ayres et al. 2009,
Strickland et al. 2009a, Strickland et al. 2009b). In one of these studies the decomposition
25
of a recalcitrant litter resource was correlated both positively and negatively to both
bacterial and fungal taxa (Strickland et al. 2009b). Although correlative, such a result
does suggest the possibility that bacteria and fungi may overlap with regards to their
degradative potentials but other explanations are also plausible such as antagonistic
(Mille-Lindblom and Tranvik 2003; Mille-Lindblom et al. 2006) or synergistic
(Bengtsson 1992) interactions between differing groups (Strickland et al. 2009b).
Both of these studies however were short-term (between 5 and 7 weeks) when
compared to most other studies and may not have captured the successional dynamics
associated with litter decomposition. For example, Poll et al. (2008) found in a
microcosm study that initial colonization of litter was dominated by bacteria and latter
stages where dominated by fungi. This was presumably driven by bacteria mineralizing
the soluble, easily-available compounds leaving the more recalcitrant compounds behind
(Poll et al. 2008). Although this may be true in some cases, it is not true for all litter types
or for all microbial communities. Strickland et al. (2009a) found that in some cases
bacterial dominance actually increased during decomposition and this was dependent on
both the microbial community in question and the litter being degraded.
Another key argument made in favor of fungi being the primary decomposers of
recalcitrant litter resources is based on their lignin degrading abilities (Tuomela et al.
2000, de Boer et al. 2005). It has been known for some time that fungi possess the
necessary enzymes capable of degrading lignin while few bacteria have been documented
which are capable of performing this same role (de Boer et al. 2005). Work has been
conducted which demonstrates that in the presence of lignin degrading fungi, bacterial
community composition changes. These bacteria are often ascribed to supporting roles in
26
lignin decomposition whereby they scavenge simple C-compounds released by fungal
enzymes (Tornberg et al. 2003, Folman et al. 2008). However, recent work conducted by
Vargas-García et al. (2007) in the field of biodegradation questions this. Vargas-García et
al. (2007) isolated six strains of bacteria and seven strains of fungi from compost heaps
and then measured their lignin degrading abilities. They found that four of the seven
fungal strains led to a significant reduction in lignin. On the other hand, all six strains of
bacteria led to a significant reduction in lignin. Furthermore, they found that one strain of
bacteria (Bacillus licheniformis) demonstrated the greatest lignin degrading capability.
Other results from this field have shown similar capabilities for bacteria but whether
these outcomes will hold under field settings is as yet unknown (Perestelo et al. 1996).
Perhaps, like with C-sequestration, one of the most direct lines of evidence in
support of greater decomposition of recalcitrant litter by fungi can be made by examining
the fungal growth form. On one hand, fungi are expected to grow into litter material
bypassing the recalcitrant outer-layers (Hendrix et al. 1986). This is likely true and is an
ability that most bacteria lack but the actual importance of this mechanism to
decomposition is largely unknown and unexplored. It also disregards the comminuting
effects of other litter inhabiting organisms, such as collembola, mites, and earthworms,
which may increase the surface area of litter available for bacterial colonization (Wardle
2002). Thus if soil fauna are in low abundance then the role of fungi in litter
decomposition may be more important but if soil fauna are in greater abundance then the
role of bacteria may be as important if not more so than that of fungi (Wardle 2002).
Another advantage that the growth form of fungi might incur is related to the idea
of hyphal bridges/networks which allow for the translocation and integration of C and
27
nutrients across a wide spatial area (Briones and Ineson 1996, Frey et al. 2000, Frey et al.
2003). Because fungal hyphae can cover a great distance this allows fungi to subsidize
the decomposition of a nutrient poor litter resource with nutrients translocated from a
nutrient rich litter resource (Briones and Ineson 1996, Tiunov 2009). In fact it is this
mechanism which has often been used to explain increased rates of decomposition
associated with recalcitrant litters in litter mixtures (Tiunov 2009). This advantage of the
fungal growth form though may also be negatively impacted by comminuting effects
(Butenschoen et al. 2007). Tiunov (2009) found that in litter mixtures as particle size of
the litters decrease then positive effects on decomposition rates were negated. The
explanation proposed for this result was that inhibitory compounds, such as phenolics,
could more easily diffuse throughout the litter layer thus negatively impacting microbial
decomposers (Tiunov 2009). It is also likely that communation impacts fungi in many of
the same ways that tilling does (see Section II).
To conclude this section, we hope to have touched on how fungal:bacterial
dominance is expected to relate to at least two ecosystem processes, C-sequestration and
decomposition, and these two processes where meant to serve as case studies of this
relationship. Although it should be noted that many of the arguments used in relating
fungal:bacterial dominance to either C-sequestration or decomposition are also used to
relate it to other ecosystem processes. We also hope that this section has pointed out
some of the research needed to evaluate the linkage between fungal:bacterial dominance
and ecosystem processes.
28
Section IV – Conclusions
Fungal:bacterial dominance has provided soil ecologists with a metric for assessing both
the environmental impacts on and the functional implications of soil microbial
communities. It has become both widely utilized and has provided many theoretical
advances. It was the goal of this review to assess the implications of fungal:bacterial
dominance as it relates to our current understanding of the soil environment in the hope
of determining its efficacy. We believe that we have highlighted key areas where this
assessment of the microbial community meets the general expectations, areas where it
diverges from these expectations, and areas were more information is warranted.
For example, fungal biomass typically increases as management intensity
decreases (i.e. tillage and fertilization). This result has been by and large the expectation
for fungi under such changes (Hendrix et al. 1986, Bardgett and McAlister 1999). On the
other hand, bacterial biomass also tended to increase as management intensity decreased
typically leading to little if any change in fungal:bacterial dominance (Hendrix et al.
1986, Bardgett and McAlister 1999). This is generally underemphasized and suggests
that management intensity related to both tillage and fertilizer inputs may result in
parallel effects on both fungi and bacteria. Such effects on these two groups may also be
expected due to temperature change. It might also be expected that bacteria and fungi
significantly overlap with regards to the key traits expected to differentiate them such as
biomass stoichiometry, CUE, and even their ability to decompose organic substrates. This
is not a novel idea however. The preeminent microbiologist, Jackson W. Foster stated, ―It
appears incongruous to delineate metabolism of bacteria from that of fungi, since the two
29
overlap in so many respects and since each group displays, more than any other groups of
microorganisms, a kaleidoscopic array of metabolic diversity.‖
This may in many ways hold true especially given the large degree of taxonomic
diversity found in both groups. In fact great diversity has often been a major pillar of
arguments related to the functional redundancy of the whole microbial community and
might also be applicable to the environmental response of fungi and bacteria (Andren and
Balandreau 1999, Behan-Pelletier and Newton 1999, Wall and Virginia 1999). Although
functional redundancy of whole soil communities has recently been questioned (Reed and
Martiny 2007, Strickland et al. 2009a), this does not negate the possibility that fungi and
bacteria on a global scale may have significant trait overlap. This might be one reason
why changes in fungal:bacterial dominance are variable under similar conditions.
As more studies are conducted across regions which explore the impacts on
fungal:bacterial dominance, then formal meta-analyses may be conducted which will
ultimately provide broad generalizations related to the impact that environmental factors
have on this measure of microbial community structure. It will also be necessary to
formally assess many of the expected differences between fungi and bacteria under field
conditions. For example, assessments of bacterial and fungal C:N ratios (especially
bacteria) would prove enlightening and demonstrate whether or not there is significant
overlap between these groups. In situ measures of fungal and bacterial biomass C:N
ratios, until recently, have proven difficult if not impossible to attain. However, advances
in the use of nano-scale secondary ion mass spectrometry (NanoSIMS) may allow
researchers to explore these soil communities in ever greater detail (Herrmann et al.
2007). Using this method researcher will be able to determine the elemental composition
30
of individual cells and paired with stable isotopes, determine the biomass incorporation
of specific elements (Herrmann et al. 2007).
The functional implications of fungal:bacterial dominance will likely also need
further assessment. we did note that one of the key attributes of fungi likely to influence
ecosystem processes is the fungal growth form. In many instances, when fungal biomass
increased then soil C also tended to increase, although it is worth noting that this was not
always associated with a decline in bacterial biomass. However, given this fact and
evidence that fungi and bacteria are not clearly distinguishable via other characteristics
(e.g. biomass C:N, CUE) then the hyphal growth form of fungi may be the trait most
likely to influence such ecosystem processes. This highlights the need to fully assess the
likely mechanism by which a change in fungal:bacterial dominance might influence a
given ecosystem process and whether or not that mechanism is clearly distinct for
bacteria and fungi. Then measurements clearly related to the assessment of the given
mechanism should be employed to determine if fungal:bacterial dominance is truly at
play. For example, if C-sequestration is the ecosystem process in question and it is
decided that a change in the amount of fungal hyphae are likely to influence this process
then measurements of fungal biomass or hyphae would likely be the most applicable.
Finally, we feel that it is important to realize that the implementation of
fungal:bacterial dominance likely arose because it provided researchers with a
taxonomically and ecologically tractable measure of the microbial community. Although
this has provided researchers with a more complete view of the opaque soil environment,
more recent methodological advances related to the soil community may provide even
further insight. For example, via the use of DNA-based approaches researchers are now
31
capable of having an incredibly detailed taxonomic view of the soil microbial
community. Such detail has allowed researchers to relate specific taxonomic groups of
fungi and bacteria to environmental change as well as ecosystem processes (Lauber et al.
2008, Strickland et al. 2009b, Lauber et al. In Review). Also coupling these relationships
to advances in culturing techniques and growth based measures may ultimately allow
researchers to correlate specific fungal and bacterial species to a range of ecological
classifications (Madsen 2005). In the end fungal:bacterial dominance has provided many
insights related to an array of environmental factors and ecosystem processes and has, in
the least, provided a stepping stone which allows for the assessment of soil microbial
communities. It seems that the more manageable complexity of the microbial community
achieved when dividing it into fungi and bacteria historically has enabled a quantitative
and more mechanistic understanding of soil ecology. We predict that this approach has
not yet been exhausted, and that insight may still be gained from it. Yet, future research
will prove if this assessment of the microbial community will prove more reliable and/or
informative in the face of new techniques also aimed at assessing these same soil
communities.
Introduction to the dissertation
The above review was meant to highlight one of the predominant ways in which soil
microbial communities are assessed, that is via its fungal:bacterial dominance. It was this
assessment, at least initially, which was the principal driver of my dissertation research.
In fact it was the relationship between the fungal:bacterial dominance of the microbial
community and ecosystem C-cycling which formed the basis for the two field studies I
32
conducted. For these two studies, I wanted to examine the relationship between
fungal:bacterial dominance first under differing land-use regimes and second under an
invasive plant species. My expectation was that the fungal:bacterial dominance of the
microbial community would differ in the face of these changes and would subsequently
influence ecosystem processes related to C-cycling. Although the first of these
expectations did hold some truth often it was other characteristics of the microbial
community which likely influenced C-processes.
For instance in Chapter 2, I set out to examine how changing land-use/land-
management regimes influence the soil microbial community as well as the
mineralization of the low molecular weight organic carbon (LMWOC) compound,
glucose. I found that fungal:bacterial dominance was not influenced by land-use/land-
management regimes. Not surprisingly then was that fungal:bacterial dominance was
unrelated to differences in the mineralization of glucose and other factors such as land-
use and extractable soil P were more strongly related to glucose mineralization. Given
that the microbial community acts as a direct control on the mineralization of glucose,
this was one of the first signs that measures of fungal:bacterial dominance may not
capture all of the intricacies of the soil microbial community which in turn influence
ecosystem C-cycling.
Further evidence of this was found when I explored the impact that the invasive
grass, Microstegium vimineum, had on microbial communities and the C-processes
regulated by them (see Chapter 3). Again I expected that the microbial community would
be dominanted by bacteria in the presence of M. vimineum with the opposite occurring in
the surrounding forest matrix. However, no difference in fungal:bacterial dominance was
33
detected but other characteristics related to the microbial community were noted with
distinct effects on C-processes. Results from this study in addition to those described in
Chapter 2 suggested that other factors related to the microbial community were likely at
play and so I decided to investigate these possibilities using controlled lab-based
experimentation.
In the first of these experimental studies (Chapter 4), I explored the functional
redundancy of soil microbial communities. First, I found evidence which suggested that
microbial communities were not functionally redundant with regards to the
mineralization of leaf litter. Additionally, there was evidence which suggested that
communities which shared a common history with a given litter resource often
mineralized a greater proportion of that litter (i.e. ‗home-field advantage‘ [Gholz et al.
2000]). I also noted that across the course of decomposition microbial communities were
taxonomically distinct and this was related to both the bacterial and fungal components of
the microbial community.
With regards to both home-field advantage and taxonomically distinct microbial
communities, I further explored the role of these communities in a follow-up experiment
examining the decomposition of novel litter resources (Chapter 5). In this work, I found
that not all microbial communities perceived litter quality equally. I found that the
perception of litter quality was in part related to the environment that the microbial
community was sourced from as well as the actual composition of that community.
Together, each of these studies has expanded on our understanding of the relationship
between the structure of soil microbial communities and ecosystem C-cycling.
34
References
Aerts, R. 1997. Climate, leaf litter chemistry and leaf litter decomposition in terrestrial
ecosystems: A triangular relationship. Oikos 79:439-449.
Allison, S. D., and K. K. Treseder. 2008. Warming and drying suppress microbial activity
and carbon cycling in boreal forest soils. Global Change Biology 14:2898-2909.
Anderson, J. P., and K. H. Domsch. 1973. Quantification of bacterial and fungal
contributions to soil respiration. Archiv Fur Mikrobiologie 93:113-127.
Andren, O., and J. Balandreau. 1999. Biodiversity and soil functioning - from black box
to can of worms? Applied Soil Ecology 13:105-108.
Aviv, M., H. Giladi, A. B. Oppenheim, and G. Glaser. 1996. Analysis of the shut-off of
ribosomal RNA promoters in Escherichia coli upon entering the stationary phase
of growth. Fems Microbiology Letters 140:71-76.
Ayres, E., H. Steltzer, B. L. Simmons, R. T. Simpson, J. M. Steinweg, M. D. Wallenstein,
N. Mellor, W. J. Parton, J. C. Moore, and D. H. Wall. 2009. Home-field
advantage accelerates leaf litter decomposition in forests. Soil Biology &
Biochemistry 41:606-610.
Bååth, E. 1992. Thymidine incorporation into macromolecules of bacteria extracted from
soil by homogenization centrifugation. Soil Biology & Biochemistry 24:1157-
1165.
Bååth, E. 1994. Measurement of protein-synthesis by soil bacterial assemblages with the
leucine incorporation technique. Biology and Fertility of Soils 17:147-153.
35
Bååth, E. 1998. Growth rates of bacterial communities in soils at varying pH: A
comparison of the thymidine and leucine incorporation techniques. Microbial
Ecology 36:316-327.
Bååth, E., and T. H. Anderson. 2003. Comparison of soil fungal/bacterial ratios in a pH
gradient using physiological and PLFA-based techniques. Soil Biology &
Biochemistry 35:955-963.
Bailey, V. L., J. L. Smith, and H. Bolton. 2002. Fungal-to-bacterial ratios in soils
investigated for enhanced C sequestration. Soil Biology & Biochemistry 34:997-
1007.
Bárcenas-Moreno, G., M. Gomez-Brandon, J. Rousk, and E. Bååth. 2009. Adaptation of
soil microbial communities to temperature: comparisons of fungi and bacteria in a
laboratory experiment. Global Change Biology doi: 10.1111/j.1365-
2486.2009.01882.x.
Bardgett, R. D., P. J. Hobbs, and A. Frostegård. 1996. Changes in soil fungal:bacterial
biomass ratios following reductions in the intensity of management of an upland
grassland. Biology and Fertility of Soils 22:261-264.
Bardgett, R. D., R. D. Lovell, P. J. Hobbs, and S. C. Jarvis. 1999. Seasonal changes in
soil microbial communities along a fertility gradient of temperate grasslands. Soil
Biology & Biochemistry 31:1021-1030.
Bardgett, R. D., and E. McAlister. 1999. The measurement of soil fungal : bacterial
biomass ratios as an indicator of ecosystem self-regulation in temperate meadow
grasslands. Biology and Fertility of Soils 29:282-290.
36
Barns, S. M., S. L. Takala, and C. R. Kuske. 1999. Wide distribution and diversity of
members of the bacterial kingdom Acidobacterium in the environment. Applied
and Environmental Microbiology 65:1731-1737.
Bedard, C., and R. Knowles. 1989. Physiology, Biochemistry, and Specific Inhibitors of
CH4, NH4+, and Co Oxidation by Methanotrophs and Nitrifiers. Microbiological
Reviews 53:68-84.
Behan-Pelletier, V., and G. Newton. 1999. Computers in biology - Linking soil
biodiversity and ecosystem function - The taxonomic dilemma. Bioscience
49:149-153.
Bengtsson, G., 1992. Interactions between fungi, bacteria and beech leaves in a stream
microcosm. Oecologia 89:542-549.
Birkhofer, K., T. M. Bezemer, J. Bloem, M. Bonkowski, S. Christensen, D. Dubois, F.
Ekelund, A. Fliessbach, L. Gunst, K. Hedlund, P. Mader, J. Mikola, C. Robin, H.
Setala, F. Tatin-Froux, W. H. Van der Putten, and S. Scheu. 2008. Long-term
organic farming fosters below and aboveground biota: Implications for soil
quality, biological control and productivity. Soil Biology & Biochemistry
40:2297-2308.
Briones, M. J. I., and P. Ineson. 1996. Decomposition of eucalyptus leaves in litter
mixtures. Soil Biology & Biochemistry 28:1381-1388.
Busse, M. D., F. G. Sanchez, A. W. Ratcliff, J. R. Butnor, E. A. Carter, and R. F. Powers.
2009. Soil carbon sequestration and changes in fungal and bacterial biomass
following incorporation of forest residues. Soil Biology & Biochemistry 41:220-
227.
37
Butenschoen, O., C. Poll, R. Langel, E. Kandeler, S. Marhan, and S. Scheu. 2007.
Endogeic earthworms alter carbon translocation by fungi at the soil-litter
interface. Soil Biology & Biochemistry 39:2854-2864.
Cleveland, C. C., and D. Liptzin. 2007. C : N : P stoichiometry in soil: is there a
"Redfield ratio" for the microbial biomass? Biogeochemistry 85:235-252.
Cruz, A. F., C. Hamel, K. Hanson, F. Selles, and R. P. Zentner. 2009. Thirty-seven years
of soil nitrogen and phosphorus fertility management shapes the structure and
function of the soil microbial community in a Brown Chernozem. Plant and Soil
315:173-184.
de Boer, W., L. B. Folman, R. C. Summerbell, and L. Boddy. 2005. Living in a fungal
world: impact of fungi on soil bacterial niche development. Fems Microbiology
Reviews 29:795-811.
de Vries, F. T., J. Bloem, N. van Eekeren, L. Brusaard, and E. Hoffland. 2007. Fungal
biomass in pastures increases with age and reduced N input. Soil Biology &
Biochemistry 39:1620-1630.
de Vries, F. T., E. Hoffland, N. van Eekeren, L. Brussaard, and J. Bloem. 2006.
Fungal/bacterial ratios in grasslands with contrasting nitrogen management. Soil
Biology & Biochemistry 38:2092-2103.
Demoling, F., L. O. Nilsson, and E. Bååth. 2008. Bacterial and fungal response to
nitrogen fertilization in three coniferous forest soils. Soil Biology & Biochemistry
40:370-379.
38
Drijber, R. A., J. W. Doran, A. M. Parkhurst, and D. J. Lyon. 2000. Changes in soil
microbial community structure with tillage under long-term wheat-fallow
management. Soil Biology & Biochemistry 32:1419-1430.
Fierer, N., M. A. Bradford, and R. B. Jackson. 2007a. Toward an ecological classification
of soil bacteria. Ecology 88:1354-1364.
Fierer, N., M. Breitbart, J. Nulton, P. Salamon, C. Lozupone, R. Jones, M. Robeson, R.
A. Edwards, B. Felts, S. Rayhawk, R. Knight, F. Rohwer, and R. B. Jackson.
2007b. Metagenomic and small-subunit rRNA analyses reveal the genetic
diversity of bacteria, archaea, fungi, and viruses in soil. Applied and
Environmental Microbiology 73:7059-7066.
Fierer, N., J. A. Jackson, R. Vilgalys, and R. B. Jackson. 2005. Assessment of soil
microbial community structure by use of taxon-specific quantitative PCR assays.
Applied and Environmental Microbiology 71:4117-4120.
Fierer, N., and R. B. Jackson. 2006. The diversity and biogeography of soil bacterial
communities. Proceedings of the National Academy of Sciences of the United
States of America 103:626-631.
Folman, L. B., P. Gunnewiek, L. Boddy, and W. de Boer. 2008. Impact of white-rot fungi
on numbers and community composition of bacteria colonizing beech wood from
forest soil. Fems Microbiology Ecology 63:181-191.
Frey, S. D., E. T. Elliott, and K. Paustian. 1999. Bacterial and fungal abundance and
biomass in conventional and no-tillage agroecosystems along two climatic
gradients. Soil Biology & Biochemistry 31:573-585.
39
Frey, S. D., E. T. Elliott, K. Paustian, and G. A. Peterson. 2000. Fungal translocation as a
mechanism for soil nitrogen inputs to surface residue decomposition in a no-
tillage agroecosystem. Soil Biology & Biochemistry 32:689-698.
Frey, S. D., J. Six, and E. T. Elliott. 2003. Reciprocal transfer of carbon and nitrogen by
decomposer fungi at the soil-litter interface. Soil Biology & Biochemistry
35:1001-1004.
Frostegård, A., and E. Bååth. 1996. The use of phospholipid fatty acid analysis to
estimate bacterial and fungal biomass in soil. Biology and Fertility of Soils 22:59-
65.
Gholz, H. L., D. A. Wedin, S. M. Smitherman, M. E. Harmon, and W. J. Parton. 2000.
Long-term dynamics of pine and hardwood litter in contrasting environments:
toward a global model of decomposition. Global Change Biology 6:751-765.
Gordon, H., P. M. Haygarth, and R. D. Bardgett. 2008. Drying and rewetting effects on
soil microbial community composition and nutrient leaching. Soil Biology &
Biochemistry 40:302-311.
Green, J. L., B. J. M. Bohannan, and R. J. Whitaker. 2008. Microbial biogeography:
From taxonomy to traits. Science 320:1039-1043.
Guggenberger, G., S. D. Frey, J. Six, K. Paustian, and E. T. Elliott. 1999. Bacterial and
fungal cell-wall residues in conventional and no-tillage agroecosystems. Soil
Science Society of America Journal 63:1188-1198.
Gusewell, S., and M. O. Gessner. 2009. N : P ratios influence litter decomposition and
colonization by fungi and bacteria in microcosms. Functional Ecology 23:211-
219.
40
Helgason, B. L., F. L. Walley, and J. J. Germida. 2009. Fungal and bacterial abundance
in long-term no-till and intensive-till soils of the Northern Great Plains. Soil
Science Society of America Journal 73:120-127.
Hendrix, P. F., R. W. Parmelee, D. A. Crossley, D. C. Coleman, E. P. Odum, and P. M.
Groffman. 1986. Detritus food webs in conventional and no-tillage
agroecosystems. Bioscience 36:374-380.
Herrmann, A. M., K. Ritz, N. Nunan, P. L. Clode, J. Pett-Ridge, M. R. Kilburn, D. V.
Murphy, A. G. O'Donnell, and E. A. Stockdale. 2007. Nano-scale secondary ion
mass spectrometry - A new analytical tool in biogeochemistry and soil ecology: A
review article. Soil Biology & Biochemistry 39:1835-1850.
Hogberg, M. N., P. Hogberg, and D. D. Myrold. 2007. Is microbial community
composition in boreal forest soils determined by pH, C-to-N ratio, the trees, or all
three? Oecologia 150:590-601.
Holland, E. A., and D. C. Coleman. 1987. Litter Placement Effects on Microbial and
Organic-Matter Dynamics in an Agroecosystem. Ecology 68:425-433.
Hunt, H. W., E. R. Ingham, D. C. Coleman, E. T. Elliott, and C. P. P. Reid. 1988.
Nitrogen limitation of production and decomposition in prairie, mountain
meadow, and pine forest. Ecology 69:1009-1016.
Jackson, R. B., N. Fierer, and J. P. Schimel. 2007. New directions in microbial ecology.
Ecology 88:1343-1344.
Jastrow, J. D., J. E. Amonette, and V. L. Bailey. 2007. Mechanisms controlling soil
carbon turnover and their potential application for enhancing carbon
sequestration. Climatic Change 80:5-23.
41
Joergensen, R. G., and F. Wichern. 2008. Quantitative assessment of the fungal
contribution to microbial tissue in soil. Soil Biology & Biochemistry 40:2977-
2991.
Koch, A. L. 2001. Oligotrophs versus copiotrophs. Bioessays 23:657-661.
Kreuzer-Martin, H. W. 2007. Stable isotope probing: Linking functional activity to
specific members of microbial communities. Soil Science Society of America
Journal 71:611-619.
Lambers, H., J. A. Raven, G. R. Shaver, and S. E. Smith. 2008. Plant nutrient-acquisition
strategies change with soil age. Trends in Ecology & Evolution 23:95-103.
Lauber, C., M. Hamady, R. Knight, and N. Fierer. In Review. Soil pH as a predictor of
soil bacterial community structure at the continental scale: a pyrosequencing-
based approach. Applied and Environmental Microbiology.
Lauber, C. L., M. S. Strickland, M. A. Bradford, and N. Fierer. 2008. The influence of
soil properties on the structure of bacterial and fungal communities across land-
use types. Soil Biology & Biochemistry 40:2407-2415.
Li, X. Z., and A. Brune. 2005a. Digestion of microbial biomass, structural
polysaccharides, and protein by the humivorous larva of Pachnoda ephippiata
(Coleoptera : Scarabaeidae). Soil Biology & Biochemistry 37:107-116.
Li, X. Z., and A. Brune. 2005b. Selective digestion of the peptide and polysaccharide
components of synthetic humic acids by the hurnivorous larva of Pachnoda
ephippiata (Coleoptera : Scarabaeidae). Soil Biology & Biochemistry 37:1476-
1483.
42
Madsen, E. L. 2005. Identifying microorganisms responsible for ecologically significant
biogeochemical processes. Nature Reviews Microbiology 3:439-446.
Manzoni, S., R. B. Jackson, J. A. Trofymow, and A. Porporato. 2008. The global
stoichiometry of litter nitrogen mineralization. Science 321:684-686.
Marstorp, H., X. Guan, and P. Gong. 2000. Relationship between dsDNA, chloroform
labile C and ergosterol in soils of different organic matter contents and pH. Soil
Biology & Biochemistry 32:879-882.
Martin-Laurent, F., L. Philippot, S. Hallet, R. Chaussod, J. C. Germon, G. Soulas, and G.
Catroux. 2001. DNA extraction from soils: Old bias for new microbial diversity
analysis methods. Applied and Environmental Microbiology 67:2354-2359.
McGill, W. B., H. W. Hunt, R. G. Woodmansee, and J. O. Reuss. 1981. PHOENIX, a
model of the dynamics of carbon and nitrogen in grassland soils. Pages 49-116 in
F. E. Clark and T. Rosswall, editors. Terrestrial Nitrogen Cycles. Ecological
Bulletins, Stockholm, Sweden.
Meidute, S., F. Demoling, and E. Bååth. 2008. Antagonistic and synergistic effects of
fungal and bacterial growth in soil after adding different carbon and nitrogen
sources. Soil Biology & Biochemistry 40:2334-2343.
Mille-Lindblom, C., Fischer, H., Tranvik, L.J., 2006. Antagonism between bacteria and
fungi: substrate competition and a possible tradeoff between fungal growth and
tolerance toward bacteria. Oikos 113:233-242.
Mille-Lindblom, C., E. von Wachenfeldt, and L. J. Tranvik. 2004. Ergosterol as a
measure of living fungal biomass: persistence in environmental samples after
fungal death. Journal of Microbiological Methods 59:253-262.
43
Mille-Lindblom, C., Tranvik, L.J., 2003. Antagonisms between bacteria and fungi on
decomposing aquatic plant litter. Microbial Ecology 45:173-182.
Nakas, J. P., and D. A. Klein. 1979. Decomposition of microbial cell components in a
semi-arid grassland soil. Applied and Environmental Microbiology 38:454-460.
Nocker, A., and A. K. Camper. 2009. Novel approaches toward preferential detection of
viable cells using nucleic acid amplification techniques. FEMS Microbiology
Letters 291:137-142.
Oehl, F., E. Sieverding, P. Mader, D. Dubois, K. Ineichen, T. Boller, and A. Wiemken.
2004. Impact of long-term conventional and organic farming on the diversity of
arbuscular mycorrhizal fungi. Oecologia 138:574-583.
Pankhurst, C. E., C. A. Kirkby, B. G. Hawke, and B. D. Harch. 2002. Impact of a change
in tillage and crop residue management practice on soil chemical and
microbiological properties in a cereal-producing red duplex soil in NSW,
Australia. Biology and Fertility of Soils 35:189-196.
Perestelo, F., A. Rodriguez, R. Perez, A. Carnicero, G. delaFuente, and M. A. Falcon.
1996. Isolation of a bacterium capable of limited degradation of industrial and
labelled, natural and synthetic lignins. World Journal of Microbiology &
Biotechnology 12:111-112.
Pikul, J. L., G. Chilom, J. Rice, A. Eynard, T. E. Schumacher, K. Nichols, J. M. F.
Johnson, S. Wright, T. Caesar, and M. Ellsbury. 2009. Organic matter and water
stability of field aggregates affected by tillage in South Dakota. Soil Science
Society of America Journal 73:197-206.
44
Poindexter, J. S. 1981. Oligotrophy - Fast and famine existence. Advances in Microbial
Ecology 5:63-89.
Poll, C., S. Marhan, J. Ingwersen, and E. Kandeler. 2008. Dynamics of litter carbon
turnover and microbial abundance in a rye detritusphere. Soil Biology &
Biochemistry 40:1306-1321.
Poulsen, L. K., G. Ballard, and D. A. Stahl. 1993. Use of ribosomal-RNA fluorescence in
situ hybridization for measuring the activity of single cells in young and
established biofilms. Applied and Environmental Microbiology 59:1354-1360.
Reed, H. E., and J. B. H. Martiny. 2007. Testing the functional significance of microbial
composition in natural communities. FEMS Microbiology Ecology 62:161-170.
Rousk, J., and E. Bååth. 2007a. Fungal and bacterial growth in soil with plant materials
of different C/N ratios. Fems Microbiology Ecology 62:258-267.
Rousk, J., and E. Bååth. 2007b. Fungal biomass production and turnover in soil estimated
using the acetate-in-ergosterol technique. Soil Biology & Biochemistry 39:2173-
2177.
Rousk, J., P. C. Brookes, and E. Bååth. 2009. Contrasting soil pH effects on fungal and
bacterial growth suggest functional redundancy in carbon mineralization. Applied
and Environmental Microbiology 75:1589-1596.
Rousk, J., and N. M. Nadkarni. 2009. Growth measurements of saprotrophic fungi and
bacteria reveal differences between canopy and forest floor soils. Soil Biology &
Biochemistry 41:862-865.
Schimel, J., T. C. Balser, and M. Wallenstein. 2007. Microbial stress-response
physiology and its implications for ecosystem function. Ecology 88:1386-1394.
45
Six, J., S. D. Frey, R. K. Thiet, and K. M. Batten. 2006. Bacterial and fungal
contributions to carbon sequestration in agroecosystems. Soil Science Society of
America Journal 70:555-569.
Spedding, T. A., C. Hamel, G. R. Mehuys, and C. A. Madramootoo. 2004. Soil microbial
dynamics in maize-growing soil under different tillage and residue management
systems. Soil Biology & Biochemistry 36:499-512.
Strickland, M. S., C. Lauber, N. Fierer, and M. A. Bradford. 2009a. Testing the
functional significance of microbial community composition. Ecology 90:441-
451.
Strickland, M. S., E. Osburn, C. Lauber, N. Fierer, and M. A. Bradford. 2009b. Litter
quality is in the eye of the beholder: initial decomposition rates as a function of
inoculum characteristics. Functional Ecology 23:627-636.
Suzuki, C., K. Nagaoka, A. Shimada, and M. Takenaka. 2009. Bacterial communities are
more dependent on soil type than fertilizer type, but the reverse is true for fungal
communities. Soil Science and Plant Nutrition 55:80-90.
Thiet, R. K., S. D. Frey, and J. Six. 2006. Do growth yield efficiencies differ between soil
microbial communities differing in fungal: bacterial ratios? Reality check and
methodological issues. Soil Biology & Biochemistry 38:837-844.
Tilman, D., P. B. Reich, J. Knops, D. Wedin, T. Mielke, and C. Lehman. 2001. Diversity
and productivity in a long-term grassland experiment. Science 294:843-845.
Tiunov, A. V. 2009. Particle size alters litter diversity effects on decomposition. Soil
Biology & Biochemistry 41:176-178.
46
Tornberg, K., E. Bååth, and S. Olsson. 2003. Fungal growth and effects of different wood
decomposing fungi on the indigenous bacterial community of polluted and
unpolluted soils. Biology and Fertility of Soils 37:190-197.
Torsvik, V., L. Ovreas, and T. F. Thingstad. 2002. Prokaryotic diversity - Magnitude,
dynamics, and controlling factors. Science 296:1064-1066.
Tuomela, M., M. Vikman, A. Hatakka, and M. Itavaara. 2000. Biodegradation of lignin
in a compost environment: a review. Bioresource Technology 72:169-183.
van der Heijden, M. G. A., R. D. Bardgett, and N. M. van Straalen. 2008. The unseen
majority: soil microbes as drivers of plant diversity and productivity in terrestrial
ecosystems. Ecology Letters 11:296-310.
van der Wal, A., J. A. van Veen, A. S. Pijl, R. C. Summerbell, and W. de Boer. 2006.
Constraints on development of fungal biomass and decomposition processes
during restoration of arable sandy soils. Soil Biology & Biochemistry 38:2890-
2902.
Várgas-Garcia, M. C., F. Suarez-Estrella, M. J. Lopez, and J. Moreno. 2007. In vitro
studies on lignocellulose degradation by microbial strains isolated from
composting processes. International Biodeterioration & Biodegradation 59:322-
328.
Vestergard, M., F. Henry, J. I. Rangel-Castro, A. Michelsen, J. I. Prosser, and S.
Christensen. 2008. Rhizosphere bacterial community composition responds to
arbuscular mycorrhiza, but not to reductions in microbial activity induced by
foliar cutting. Fems Microbiology Ecology 64:78-89.
47
Wall, D. H., and R. A. Virginia. 1999. Controls on soil biodiversity: insights from
extreme environments. Applied Soil Ecology 13:137-150.
Wardle, D. A. 2002. Communities and Ecosystems: Linking the Aboveground and
Belowground Components. Princeton University Press, Princeton, NJ, USA.
Wardle, D. A., R. D. Bardgett, J. N. Klironomos, H. Setala, W. H. van der Putten, and D.
H. Wall. 2004. Ecological linkages between aboveground and belowground biota.
Science 304:1629-1633.
Wardle, D. A., and A. Ghani. 1995. Why is the strength of relationships between pairs of
methods for estimating soil microbial biomass often so variable. Soil Biology &
Biochemistry 27:821-828.
Wilson, G. W. T., C. W. Rice, M. C. Rillig, A. Springer, and D. C. Hartnett. 2009. Soil
aggregation and carbon sequestration are tightly correlated with the abundance of
arbuscular mycorrhizal fungi: results from long-term field experiments. Ecology
Letters 12:452-461.
Winding, A., K. Hund-Rinke, and M. Rutgers. 2005. The use of microorganisms in
ecological soil classification and assessment concepts. Ecotoxicology and
Environmental Safety 62:230-248.
Winogradsky, S. 1924. Sur la microflora autochtone de la terre arable. Comptes Rendus
Hebdomadaire des Séances de l'Academie des Sciences, Paris 178:1236-1239.
Wright, D. A., K. Killham, L. A. Glover, and J. I. Prosser. 1995. Role of pore-size
location in determining bacterial-activity during predation by protozoa in soil.
Applied and Environmental Microbiology 61:3537-3543.
48
Yeates, G. W., R. D. Bardgett, R. Cook, P. J. Hobbs, P. J. Bowling, and J. F. Potter.
1997. Faunal and microbial diversity in three Welsh grassland soils under
conventional and organic management regimes. Journal of Applied Ecology
34:453-470.
Zak, D. R., C. B. Blackwood, and M. P. Waldrop. 2006. A molecular dawn for
biogeochemistry. Trends in Ecology & Evolution 21:288-295.
Zhao, X. R., Q. Lin, and P. C. Brookes. 2005. Does soil ergosterol concentration provide
a reliable estimate of soil fungal biomass? Soil Biology & Biochemistry 37:311-
317.
49
Figure Legends
Figure 1.1 The prevalence of fungal:bacterial dominance in the literature today showing
(A) the number of publications found using the search terms ―fungal:bacterial‖ and ―soil‖
and (B) the number of citations referring to those same publications from 1991 to 2008.
Both works specifically related to fungal:bacterial dominance and the citations of these
works have increased during this period of time. The number of publications and citations
were identified using Web of Science. See Appendix A for the list of articles used to
construct this figure.
Figure 1.2 The distribution of the C:N ratios of bacteria, saprotrophic fungi, and
mycorrhizal fungi. Shown are box plots and distributions, notably fewer measurements of
the C:N ratio of bacteria have been made. Results were compiled from the literature and
included 1,099 observations from 18 studies. Most of the data on fungi was obtained in
the field from sporocarps whereas most of the data for bacteria was obtained from
cultures or aquatic samples. Closed circles = Mycorrhizal fungi; Open circles =
Saprotrophic fungi; Closed squares = Bacteria. See Appendix B for the list of articles
used to construct this figure.
50
Figure 1.1
(A)P
ubli
cati
ons
0
2
4
6
8
10
12
14
16
(B)
Year
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010
Cit
atio
ns
0
100
200
300
400
500
51
Figure 1.2
C:N ratio
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
Coun
t
0
20
40
60
80
100
120
140
160
180
Mycorrhizal fungiSaprotrophic fungiBacteria
52
CHAPTER 2
RATES OF IN SITU CARBON MINERALIZATION IN RELATION TO LAND-USE,
MICROBIAL COMMUNITY AND EDAPHIC CHARACTERISTICS2
2 Strickland, M.S., M.A. Callaham Jr., C.A. Davies, C.L. Lauber, K. Ramirez, D.D. Richter Jr., N. Fierer,
M.A. Bradford. To be submitted to Soil Biology & Biochemistry.
53
Abstract
Plant-derived carbon compounds enter soils in a number of forms; two of the most
abundant being leaf litter and rhizodeposition. Our knowledge concerning the
predominant controls on the cycling of leaf litter far outweighs that for rhizodeposition
even though the constituents of rhizodeposits includes a cocktail of low molecular weight
organic compounds which represent a rapidly cycling source of carbon, readily available
to soil microbes. We determined the mineralization dynamics of a major rhizodeposit,
glucose, and its relationship to land-use, microbial community and edaphic characteristics
across a landscape in the southeastern United States. The landscape consists of cultivated,
pasture, pine plantation, and hardwood forest sites (n=3). Mineralization dynamics are
resolved in both winter and summer using an in situ 13
C-glucose pulse-chase approach.
Mineralization rates of the labeled glucose declined exponentially across the 72 h
measurement periods. This pattern and absolute mineralization rates are consistent across
seasons. An information-theoretic approach revealed that land-use is a moderately strong
predictor of cumulative glucose mineralization. Measures assessing the size, activity,
and/or composition of the microbial community were poor predictors of glucose
mineralization. The strongest predictor of glucose mineralization was soil-extractable
phosphorus. It was positively related to glucose mineralization across seasons and
explained 60% and 48% of variation in cumulative glucose mineralization in the summer
and winter, respectively. We discuss potential mechanisms underlying the relationship
between soil phosphorus and glucose mineralization. Our results suggest that specific soil
characteristics often related to land-use and/or land-management decisions may be strong
predictors of glucose mineralization rates across a landscape. We emphasize the need for
54
future research into the role of soil phosphorus availability and land-use history in
determining soil organic carbon dynamics.
Keywords: Soil microbial communities, root exudates, low molecular weight
compounds, fungal-to-bacterial ratios, land-use, rhizosphere, carbon cycling,
decomposition
Introduction
Much of our understanding regarding the cycling of plant-derived carbon (C)-compounds
has been developed from studies that examine the decomposition of leaf litter (Melillo et
al. 1982, Couteaux et al. 1995, Aerts 1997, Gholz et al. 2000, Harmon et al. 2009).
However, leaf litter is not the only plant-derived input of C to terrestrial systems. Of
growing interest is the role that rhizodeposits play in ecosystem level C-processes (van
Hees et al. 2005, Pollierer et al. 2007, Hogberg et al. 2008, Phillips et al. 2008).
Rhizodeposits are in part composed of low molecular weight organic C (LMWOC)
compounds, such as sugars (mono- and disaccharides), amino acids, and organic acids
(Rovira 1969, Grayston et al. 1997, Dakora and Phillips 2002). Although the actual input
of LMWOC compounds into a system have been estimated to account for less than 30%
of total soil respiration per year (van Hees et al. 2005), the overall impact of these
compounds on C-dynamics may be large (van Hees et al. 2005, Hogberg et al. 2008,
Phillips et al. 2008). In part, this is because LMWOC compounds represent a highly
labile, rapidly-cycling, form of C which is immediately available to the soil microbial
community for growth, maintenance and respiration (van Hees et al. 2005, Hogberg et al.
2008). Indeed, the turnover of LMWOC compounds may explain significant variation in
55
soil respiration rates (Gu et al. 2004, Bengtson and Bengtsson 2007). Furthermore, the
degree to which rhizodeposition interacts with the microbial community is likely to
influence ecosystem C-sequestration, with increases or decreases in soil organic C pools
expected under differing input rates of LMWOC compounds (Bradford et al. 2008a,
Bradford et al. 2008b, Blagodatskaya et al. 2009). Less well understood is how variation
in land-use, microbial community and edaphic characteristics will impact the
mineralization rates of such rhizodeposits.
Overall there is the expectation that the size, activity, and composition of the soil
microbial community will influence the mineralization rates of LMWOC compounds
(van Hees et al. 2005, Fierer et al. 2007, Allison and Martiny 2008, Green et al. 2008, but
see Kemmitt et al. 2008). An increase in the size and/or activity of the microbial
community is expected to increase mineralization (van Hees et al. 2002, van Hees et al.
2005, Six et al. 2006). The influence of composition, on the other hand, is less clear.
Compositional changes in the soil microbial community, associated with broad
distinctions in life history strategies, are likely to have the greatest impact on
mineralization dynamics. For example, using one of the traditional compositional
distinctions, microbial communities largely composed of zymogenous (i.e. fresh organic
matter decomposers) as opposed to autochthonous (i.e. soil organic matter decomposers)
populations of microbes may have higher rates of LMWOC mineralization (Winogradsky
1924, Fontaine and Barot 2005). A phylogenetic distinction might posit that communities
dominated by bacteria will have greater mineralization rates than those dominated by
fungi (Bardgett and McAlister 1999, Six et al. 2006).
56
It is also feasible that microbial communities which have previously been exposed
to a specific LMWOC compound will subsequently influence the future mineralization of
that compound (Saggar et al. 2000). This expectation is similar to the hypothesis of
‗home-field advantage‘, which was developed to explain differences in litter
decomposition rates not accounted for by climate and/or litter quality (Hunt et al. 1988,
Gholz et al. 2000, Ayres et al. 2009, Strickland et al. 2009a). The hypothesis suggests
that communities pre-exposed to a given leaf litter are likely to mineralize that litter more
rapidly than communities for which that litter is novel (Gholz et al. 2000, Ayres et al.
2009). Although the underlying mechanisms for home-field advantage may differ for
mineralization rates of litter and root exudates, the scenario suggests that communities
which derive the bulk of their C from rhizodeposits and/or LMWOC compounds may
have greater mineralization rates than communities deriving their C from another source
(Saggar et al. 2000, Stevenson et al. 2004). Based on this possibility, the mineralization
rates of LMWOC compounds may be dependent on the plant community and factors such
as land-use/land-cover, season, and nutrient availability (Grayston et al. 1997,
Warembourg et al. 2003, van Hees et al. 2005). For example, where the dominant plant
cover is grass (e.g. pastures), inputs of LMWOC compounds may be a more important C
source than in forests, explaining higher mineralization rates of these compounds in
herbaceous communities (Stevenson et al. 2004). Similarly, these inputs may vary
dependent on both season and the nutrient demands of the plant itself (Nguyen 2003, Bais
et al. 2006). Finally, factors typically related to soil C dynamics, such as temperature,
moisture, pH, soil texture and nutrient availability, may all influence mineralization rates
of LMWOC compounds (Saggar et al. 2000, Fierer and Jackson 2006, Yuste et al. 2007,
57
Bradford et al. 2008a, Bradford et al. 2008b). Specifically, each of these factors has been
related to effects on the activity and/or composition of the soil microbial community but
their effect on the mineralization rates of LMWOC compounds is lacking (van Hees et al.
2005).
One of the key constituents of LMWOC compounds is glucose (Dakora and
Phillips 2002, Nguyen 2003, van Hees et al. 2005). Glucose is likely to be found in the
rhizodeposits of nearly all plant species under an array of environmental conditions and is
also found in the leachates of forest floor litter (Dakora and Phillips 2002, Nguyen 2003,
van Hees et al. 2005). In light of the significance of glucose as a rhizodeposit and
leachate-constituent, microbial mineralization of glucose has been widely studied. As
early as the 1980s researchers have been adding glucose to soil and tracking the resulting
mineralization dynamics. Notably, Voroney et al. (1989) found that the initial
mineralization of glucose was rapid but that a large proportion (~ 25%) of it was later
stabilized in the system, presumably as microbially-derived products. Using
concentrations more typical of those found in situ, mineralization of glucose has been
shown to follow Michaelis-Menten type kinetics (van Hees et al. 2005, van Hees et al.
2008). Soil microbial communities respond rapidly to glucose additions, even when these
additions are in trace amounts, suggesting that at least certain components of the
microbial community may be adapted to it and/or LMWOC compounds in general (De
Nobili et al. 2001, Jones and Murphy 2007, Hanson et al. 2008, Hoyle et al. 2008). Lab-
based studies of this phenomenon have demonstrated that this adaptation may be
contingent on the source environment of the microbial community, with microbes from
grasslands responding more rapidly than those from forests (Jones and Murphy 2007).
58
Such studies have added greatly to our understanding of the microbial response to
glucose but few studies have assessed these dynamics in intact soil systems under field
conditions. The exception is Boddy et al. (2007), who found that glucose mineralization
was more rapid in the field than in the lab; this was attributed to an intact microbial
community in situ associated with the rhizosphere. Less well understood is whether the
mineralization of glucose (or other LMWOC compounds) can be generalized across a
landscape in much the same way that leaf litter decomposition has been (Barlow et al.
2007, Madritch and Cardinale 2007). That is, are there characteristics across space and
time that likely explain variation in the mineralization rates of glucose in situ?
The objective of our study was to determine the landscape-level controls on
glucose mineralization rates. We amended soils with small amounts of 13
C-labeled
glucose and tracked its resultant mineralization, in both the summer and winter, across a
representative rural landscape of the southeastern United States (see Richter et al. 1999,
Richter and Markewitz 2001). This landscape consists of upland soils with four distinct
land-use histories, currently with cultivated crops, pastures, old-field pine stands, and
never cultivated hardwood forests. We expected that the size, activity, and composition
of microbial communities, all of which could be considered potential determinants of
glucose mineralization rates, would vary markedly. Our primary expectation was that
either one or all three of these microbial community characteristics, and/or land-use,
would be a major determinant of in situ glucose mineralization rates. More specifically,
we expected, due to likely differences in microbial community characteristics and
rhizospheric inputs, that those sites currently under herbaceous plant cover (i.e. cultivated
and pasture) would exhibit greater mineralization rates than would those under forest
59
cover (i.e. pine and hardwood). To investigate these expectations we used an
information-theoretic approach (Burnham and Anderson 1998) which permitted us to
both determine the probable best model in a suite of candidate models, in addition to the
most important model parameter, for explaining variation in glucose mineralization in
situ. We assessed a secondary set of models that, in addition to microbial community
characteristics and land-use, included a suite of soil characteristics which have been
speculated to influence glucose mineralization (e.g. texture). To justify the use of an
information theoretic approach we include potential explanatory variables for which there
is an established mechanistic rationale for why they might explain variation in the
response variable (see Burnham and Anderson 1998). These rationales are provided in
the Methods. Our overall objective was to compare multiple, ecologically relevant
models to provide a greater mechanistic understanding of mineralization dynamics of
LMWOC compounds across a landscape.
Methods
Site descriptions
Sites used in this study were located in, or adjacent to, the Calhoun Experimental Forest
(CEF), which is managed by the USDA Forest Service and located in the Southern
Piedmont physiographic region (approximately 34.5°N, 82°W) of northwestern South
Carolina, USA (Gaudinski et al. 2001, Callaham et al. 2006). The acidic Ultisols soils at
these sites are classified as fine, kaolinitic, thermic Typic Kanhapludults of the Appling,
Cecil, Hiwassee, and Madison series (Richter et al. 2006, Lauber et al. 2008, Grandy et
al. 2009). All sites are on uplands and on interfluves from similar bedrock, thus
60
attempting to control native soil characteristics, geomorphology, and geology. Sites were
within 30 km of each other and each represented one of four land-use practices common
to the region (Lauber et al. 2008, Grandy et al. 2009): annual row-crop agriculture, cattle
pasture, old-field pine forest, and oak-hickory forest (n=3 in all cases).
Management regimes for both the cultivated and pasture sites have been in place
for at least the past 40 years. The cultivated sites are managed by the South Carolina
Department of Natural Resources as wildlife fields with annual crop rotation between
corn, millet, wheat, sorghum, sunflowers, and fallow using conventional tillage practices.
During the winter sampling (start date: 30th
November 2006) two of the three cultivated
sites were fallow while the other (site name: Cultivated 3) was planted in winter wheat. In
the summer sampling (start date: 12 June 2007), cultivated sites were planted with corn
(Cultivated 1), sunflowers (Cultivated 2), and wheat/corn (Cultivated 3). Pasture sites are
dominated by rye and Bermuda grass and are, more or less, continuously cattle-grazed.
Both cultivated and pasture ecosystems are fertilized and limed. Two of the pine
plantations (i.e. Pine 1 and 3) are ~50 years old and consist of an overstory of loblolly
pine with an understory of oak and hickory species. The other pine plantation (i.e. Pine 2)
is a 10-year old loblolly pine monoculture. These pine forests are all growing on old
fields previously cultivated for cotton and have therefore been fertilized and limed but
not during the growth of pine. The hardwood sites are mature oak-hickory stands, ~75
years of age. One hardwood plot (i.e. Hardwood 2) is grazed by cattle and has little
understory vegetation. Average bulk density (for depth 0-7.5 cm) ± 1 S.E. was 1.23±0.15,
1.66±0.10, 1.32±0.09, and 1.16±0.17 g cm-3
for cultivated, pasture, pine, and hardwood
land-uses, respectively.
61
This replicated land-use design allowed us to explore how variation in both the
contemporary and historical management of these sites drove differences, which might be
expected to influence glucose mineralization, in soil properties and microbial
communities (Richter et al. 1999, Richter and Markewitz 2001). Differences in dominant
plant cover per land-use could also be a significant determinant of glucose mineralization
(e.g. Stevenson et al. 2004). It was our goal to select sites representative of the land-use
practices and land-use legacies of this region, enabling a practical insight into the
relationship between in situ glucose mineralization and soil chemical, physical and
microbial community characteristics.
13C-glucose pulse-chase
To determine glucose mineralization rates across sites and seasons, we conducted a 13
C-
glucose pulse-chase experiment in both the winter and summer. The pulse-chase was
conducted by making additions of 99 atom% 13
C-labeled glucose and tracking its
mineralization as 13
CO2. This was accomplished by placing two PVC collars (15.4 cm
dia., inserted 5 cm into the soil) in each land-use plot. Water was then added to each core
24 h prior to sampling in order to alleviate any water stress between plots. Soil CO2
efflux rates were determined using a closed-chamber approach (e.g. Bradford et al. 2001),
where CO2 concentrations were determined at the start and end of a 45 min capping
period. We conducted a pilot study to determine the appropriate capping time and found
that headspace CO2 concentrations increase linearly from 0-45 minutes and flux rate
estimates only begin to decrease after 60 minutes. Although this closed-chamber
approach is likely to have caveats associated with it (i.e. underestimated CO2 efflux
62
rates), it is nonetheless a widely used approach across an array of ecosystem types (Nay
et al. 1994, Franzluebbers et al. 2002). Similar capping times and protocols as described
here have been employed by other researchers (Iqbal et al. 2008, Mo et al. 2008).
Headspace samples were taken with 20 mL SGE gas syringes, transported to the
laboratory in 12 mL Exetainers, and then CO2 concentrations were determined using an
infra-red gas analyzer (IRGA; Li-Cor Biosciences, Lincoln, NE, USA, Model LI-7000).
A second sample was analyzed using continuous flow, isotope-ratio mass spectrometry
(IRMS; Thermo, San Jose, CA, USA) to determine the δ13
C value of the CO2 in the
sample. The initial headspace sampling provided the natural abundance values for the
isotope mixing equations (see below). After this initial sampling, 1 L of 2.5 mM 13
C-
labeled glucose solution (99 atom%) was added to the collars and permitted to drain. The
capping procedure was repeated post addition at 2, 5, 24, 48 and 72 h, permitting a
negative exponential ‗decay‘ of 13
C label to be tracked in the soil CO2 efflux, from which
cumulative mineralization rates were estimated. The amount of C added to each collar
was relatively small: <25 µg C g dry wt soil-1
, which is <0.0001% of the total soil carbon
(from 0 to 7.5 cm depth in a 15.4-cm dia. PVC collar).
The contribution of 13
C-labeled glucose to soil respiration was estimated using
isotope mixing equations. The amount of CO2 derived from glucose was calculated as
follows (sensu Ineson et al. 1996): Cglucose derived = Ctotal × (δ13
Cafter - δ13
Cbefore)/( δ13
Cglucose
- δ13
Cbefore), where Ctotal is the total amount of C respired, δ13
Cafter is the δ13
C value of the
respired C after glucose was added, δ13
Cbefore is the δ13
C value of respired C before
glucose was added (i.e. the natural abundance value), and δ13
Cglucose is the value for the
glucose itself. Values were calculated per PVC collar but for the statistical analyses the
63
mean of the pair of values for each site (per season) was used; in effect treating the pair
of values as ‗experimental repeats‘ and not replicates.
Determination of microbial community structure and soil characteristics
During both the summer and winter pulse-chase, we collected ten individual A horizon
soil cores (8 cm dia., 0 – 7.5 cm depth) from each site using a stratified random approach.
Soil cores were taken from areas adjacent (i.e. within 10 m) to the pulse-chase plots in
each land-use. Soils were sieved (4 mm), homogenized, and stored at +5ºC until
analyzed. Analyses included those used to assess the soil microbial community and key
soil characteristics expected to influence the mineralization of glucose. Mean values per
land-use for both sets of characteristics are presented in Table 2.1 and the rationale for
each is given below.
Although there is debate surrounding the applicability of different methods aimed
at determining the size, activity, and composition of microbial communities, we decided
to use three of the most common (Wardle and Ghani 1995). These included a modified
chloroform fumigation-extraction (CFE) method to assess microbial biomass, substrate-
induced respiration (SIR) as an estimate of activity, and determination of fungal-to-
bacterial dominance via qPCR to estimate composition. All measures of microbial
community size, activity, and composition were determined for both the winter and
summer pulse-chase.
The modified CFE method as described in Fierer and Schimel (2002) and Fierer
and Schimel (2003) allowed us to estimate microbial biomass C. Briefly this was
estimated as the flush of DOC following fumigation with ethanol-free chloroform. Raw
64
values are reported for microbial biomass C; no correction factors were used. The SIR
method we used follows Fierer & Schimel (2003) whereby soil slurries are incubated,
after a 1 h pre-incubation with excess substrate (i.e. autolyzed yeast extract), for 4 h at
20ºC. After the 4-h incubation, SIR is determined via infrared gas analysis of headspace
CO2 concentrations using a static incubation technique (Fierer et al. 2005a). Note that
although both the CFE and SIR method both report measures of microbial biomass, these
measures are not necessarily equivalent especially at finer spatial scales (i.e. <100 km)
(Wardle and Ghani 1995). Where CFE is expected to estimate total biomass, SIR has
been suggested to be a measure of active microbial biomass (Wardle and Ghani 1995).
The composition of the microbial community was assessed as the relative
abundance of fungi-to-bacteria. The relative fungal-to-bacterial dominance of the
microbial community is often taken as an indicator of belowground C processes with
fungal-dominated systems being associated with greater C stores and/or slower C-cycling
(Bardgett and McAlister 1999, Six et al. 2006). We determined these ratios using the
quantitative PCR (qPCR) method described by Fierer et al. (2005b) and Lauber et al.
(2008). Briefly, DNA was isolated from soil (kept at -80ºC until use) using the MoBIO
Power Soil DNA Extraction kit (MoBio Laboratories, Carlsbad, CA) with modifications
described in Lauber et al. (2008). Standard curves were constructed to estimate bacterial
and fungal small-subunit rRNA gene abundances using E. coli 16S rRNA gene, or the
Saccharomyces cerevisiae 18S rRNA gene, according to Lauber et al (2008). Ratios of
fungal to bacterial gene copy numbers were generated by using a regression equation for
each assay relating the cycle threshold (Ct) value to the known number of copies in the
65
standards. All qPCR reactions were run in quadruplicate. Additional details, including the
specific qPCR reaction conditions, are provided in Lauber et al. (2008).
In addition to measures aimed at determining the size, activity, and composition
of the soil microbial community we also identified and determined several soil
characteristics which are likely to influence the mineralization of glucose both directly
and indirectly. These included in the same surficial 7.5-cm deep samples, soil pH, soil
temperature and moisture, soil texture, soil organic material C:N ratio, and extractable P.
Soil pH, moisture, and temperature were determined for both the winter and summer
pulse-chase. Soil texture, C:N, and extractable P were determined on 7 September 2006
as these measures are not expected to vary greatly across seasons (Richter et al. 2006).
Soil pH (1:1, soil:H2O by volume) was measured with a bench-top pH meter . The
overall rationale for including soil pH as a possible indicator of in situ glucose
mineralization is partially related to its role as an indicator of microbial community
composition and diversity (Fierer and Jackson 2006, Lauber et al. 2008, Jones et al.
2009). At broad spatial scales, as well at our study sites, soil pH is related to the relative
abundance of specific microbial taxa, bacterial diversity, and fungal-to-bacterial
dominance (Fierer and Jackson 2006, Lauber et al. 2008, Jones et al. 2009). Additionally,
soil pH may directly impose stress on the microbial community affecting its activity
and/or it may serve as an integrator of many soil characteristics likely to impact both the
soil microbial community and as a result glucose mineralization (Fierer and Jackson
2006).
Soil temperature (determined at 7.5 cm depth) and moisture (determined across 0-
10 cm depth) were determined at each pulse-chase measurement using hand-held
66
moisture and temperature probes. The average of both was estimated for the entire 72 h
period of both the summer and winter pulse-chase. The rationale for determining
temperature and moisture is based on the well known interaction between these factors
and the decomposition of leaf litter (Couteaux et al. 1995, Yuste et al. 2007). This is
likely due to the influence that these two factors have on the activity of the soil microbial
community and it is likely that the interaction between these factors will also influence
the mineralization of glucose.
Soil texture was estimated using a simplified version of the hydrometer method
as described by Gee and Orr (2002). Texture has been shown to relate to the composition
of microbial communities, including across our study landscape (Lauber et al. 2008).
Generally, finer textured soils may protect microbial biomass from both predation and
environmental stress, such as desiccation, and may influence the activity of the microbial
community (Six et al. 2006). This in turn typically leads to a lower turnover of the
microbial biomass, lower activity, possible shifts in community composition, and in turn
likely differences in glucose mineralization (Six et al. 2006).
Soil C:N was determined using an NA1500 CHN Analyser (Carlo Erba
Strumentazione, Milan, Italy). Extractable P was measured on an Alpkem auto-analyzer
(OI Analytical, College Station, TX) using Murphy-Riley chemistry after extraction with
Mehlich I double-acid (H2SO4 –HCl) using a 1:4 mass:volume ratio (Kuo, 1996). Both
soil C:N and extractable P are suggestive of soil nutrient status and have been shown to
influence the mineralization of an array of C-compounds (Bradford et al. 2008a, Bradford
et al. 2008b, Gusewell and Gessner 2009, Strickland et al. 2009b). Both of these soil
characteristics have also been found to relate to the composition of soil microbial
67
communities, specifically fungal communities, at the CEF as well as other ecosystems
(Lauber et al. 2008, Clare et al. 2009, Gusewell and Gessner 2009, Hrynkiewicz et al.
2009). Due to the fact that both soil C:N and P are tied to nutrient availability as well as
microbial community characteristics then they may also influence glucose mineralization.
Statistical analyses
Linear mixed-effects models (Pinheiro and Bates 2000) were used to examine the
relationships between glucose mineralization and soil microbial community
characteristics (i.e. size, activity, and composition), land-use, and select soil
characteristics. For this approach microbial community characteristics, land-use, soil
characteristics, and season (when applicable) were treated as fixed effects. Site identity
was included as a random effect to account for repeated sampling (i.e. season).
Cumulative glucose mineralization was loge-transformed to conform to assumptions of
homoscedasticity (verified using model checking). Parameter estimates were calculated
using restricted maximum likelihood. However, models compared using the information
theoretic approach were constructed using maximum likelihood estimates as the fixed
effects of these models varied (MacNeil et al. 2009). Further assessment of model fit was
done using linear regression within a given season (i.e. the same linear model was fit to
either data gathered in the winter or summer). Finally, analyses were also conducted
which examined the effect of land-use on cumulative glucose mineralization and
mineralization dynamics. We analyzed cumulative glucose mineralization using a linear
mixed effects model where land-use and season were treated as fixed effects and site
identity was treated as a random effect to account for repeated sampling across seasons;
68
land-use and season were allowed to interact in this model. Land-use was further
examined within each season using ANOVA with land-use as a discrete variable and the
Tukey method was used to assess differences between means. We analyzed the
mineralization dynamics across the 72 h period using a linear mixed-effects model where
land-use, season, and time since glucose addition were treated as fixed effects and site
identity was treated as a random effect to account for repeated sampling across the 72 h.
Time was treated as a continuous variable and all three variables were allowed to interact.
Mineralization dynamics were also analyzed within a given season using a similar mixed-
effects model except in this case there was no fixed effect of season. All analyses were
conducted using the freeware statistical package R (http://cran.r-project.org/).
Model selection and comparison was conducted using an information-theoretic
approach (Burnham and Anderson 1998). We initially compared models which related
soil microbial community characteristics and land-use to cumulative glucose
mineralization. This resulted in a set of 30 candidate models plus an additional intercept
only model. This model set included a full factorial combination of soil microbial
community characteristics with and without both land-use and season (see Appendix C
for the full model list). Land-use was also included in this candidate set as a single
parameter and in combination with season. Model parameters were only included as
additive terms in order to both keep the number of candidate models manageable and
because there was no a priori rationale for the inclusion of interactions. Akaike‘s
information criterion for small sample size (AICc) was used as the model selection
criterion (Burnham and Anderson 1998). Model comparisons are based on the difference
in AICc (Δi) with models <2 AICc units apart considered nearly indistinguishable
69
(Burnham and Anderson 1998). Models >10 AICc units from the model with the
minimum AICc value have no support (Burnham and Anderson 1998). We also
calculated Akaike weights (ωi) which allowed us to quantify the support for one model
relative to another, calculate the 95% confidence set of models, and determine the
importance of a given model variable (Burnham and Anderson 1998). The last of these
was calculated as the Σ ωi for all models containing the variable in question (Burnham
and Anderson 1998).
In addition to the initial 30 candidate models we also looked at the role of specific
soil characteristics. The overall rationale for this was to determine, if any, which specific
soil characteristics regardless of their relationship to land-use likely drove differences in
cumulative glucose mineralization. In order to assess this possibility we incorporated five
soil characteristics likely to impact the soil microbial community: pH, the interaction
between soil temperature and moisture, silt + clay content, soil organic material C:N
ratio, and extractable P. The rationale for inclusion of each is given above (see
Determination of microbial community structure and soil characteristics). Each of these
was included as an additive variable in models containing microbial community
characteristics as well as a single factor and as an additive factor with season (see
Appendix C for the full model list). This resulted in a total of 110 candidate models (plus
an intercept only model) which included the original 30 candidate models. Model
selection for this set of candidate models was conducted in the same manner as described
above.
70
Results
Temporal dynamics and cumulative mineralization of glucose
Overall the addition of glucose caused relatively little if any change in total soil
respiration within a land-use, regardless of season (Appendix D). The amount of glucose
respired as 13
CO2 declined across the 72 h measurement period with the highest
mineralization values recorded ~2 h after the addition of glucose and the lowest values
occurring at ~48 or ~72 h (Fig. 2.1). This change in glucose mineralization rates across
time was highly significant (F1, 96 = 278.93; P<0.0001) and an interaction between these
temporal dynamics and season was detected (F1, 96 = 5.27; P<0.05). This interaction was
likely explained by the similar mineralization rates regardless of season at the 2 and 5 h
measurement periods but greater mineralization rates in the winter at the 24, 48, and 72 h
measurement periods. It is worth noting that glucose mineralization rates tended to be
similar across seasons in spite of the fact that total soil respiration was greater in the
summer when compared to the winter (Fig. 2.1, Appendix D). Land-use was also a
significant factor impacting glucose mineralization rates (F3, 8 = 9.89; P<0.01) and this
was consistent for both the winter (F3, 8 = 10.08; P<0.01) and summer (F3, 8 = 9.57;
P<0.01) (Fig. 2.1a,b). In general we noted that pasture and cultivated sites typically had
higher glucose mineralization rates across 72 h than did pine and hardwood sites,
although this pattern was clearer in the summer than in the winter (Fig. 2.1).
Overall, land-use had a significant influence on cumulative glucose mineralization
(F3, 8 = 10.34; P<0.01). Neither season effects, (F1, 8 = 0.45; P = 0.52), nor an interaction
between season and land-use (F3, 8 = 1.36; P = 0.32), were observed for the cumulative
amount of glucose mineralized. Generally, cumulative mineralization was higher in
71
pasture sites while both the pine and hardwood sites tended to be lower (Fig. 2.1). The
cultivated sites were not as easily generalized given that cumulative mineralization
tended to be lower in the winter when compared to the summer. Additionally, during the
summer there was a large degree of variation associated with this land-use, likely due to
differences in crop cover between the three sites (see Methods and Discussion). The high
variation in cultivated sites in the summer may explain why the land-use effect on
cumulative glucose mineralization (Fig. 2.1c, d) was not consistent across seasons. That
is, in the winter a clear land-use effect was noted (F3, 8 = 18.22; P <0.001) but this was
not true for the summer (F3, 8 = 2.64; P = 0.12), where larger error variances were
observed (compare Fig. 2.1c, d). The large variation in cumulative glucose mineralized
across cultivated sites may also explain why the temporal dynamic data (Fig. 2.1b)
revealed a statistically significant land-use effect but the cumulative data (Fig. 2.1d) did
not.
Relatively little of the added glucose was recovered as CO2 and this was true
regardless of season or land-use. The amount of C recovered as CO2 ranged from a low
of 0.57% (Hardwood 3 in the winter) to a high of 4.29% (Cultivated 3 in the summer)
and across all land-uses and both seasons averaged 1.93%. This suggests that
approximately 95-99% of the added glucose-C remained within the soil system. Of
course, given that we did not begin tracking the mineralization of glucose until 2 h after
its addition, and the fact that some of the respired glucose may not have been captured by
our chamber method; the actual amount of C remaining may have been lower. However,
low recovery rates of glucose-C as CO2 are typical when glucose solutions are added to
soils in low concentrations (Boddy et al. 2007).
72
Glucose mineralization and its relationship to microbial community structure and land-
use
We initially set out to determine the importance of land-use and the size, activity, and
composition of the microbial community as they relate to cumulative glucose
mineralization in situ. In order to do this we utilized an information-theoretic approach to
assess a series of 30 candidate models (Burnham and Anderson 1998). Given that land-
use was significantly related to cumulative glucose mineralized (Fig. 2.1), it is not
surprising that the best model (i.e. lowest AICc) included land-use as the sole explanatory
variable (Table 2.2). The probability that this model was the actual best model in this set
was 53% suggesting moderate support for this outcome. However, a model which only
included an intercept term was within 2 AICc units of the top model suggesting that it
was also a plausible top model candidate (Table 2.2). Furthermore, the evidence ratio (i.e.
ω1/ ω2 = 1.89) when comparing these two models is ~2 and we must then assume a high
level of model selection uncertainty when considering these two models (Burnham and
Anderson 1998). Additionally, all of the remaining models, including those that
accounted for the size, activity, or composition of the microbial community, were > 4
AICc units from the top model indicating that there was overall considerably less support
that any of these models was the ‗actual‘ best model (Burnham and Anderson 1998). In
fact when calculating the importance of each model parameter, we found that the size,
activity, and composition of the microbial community were all relatively unimportant in
the prediction of glucose mineralization (ωi<0.10 for all 3; Fig. 2.2a) while land-use was
moderately important (ωi = 0.55; Fig. 2.2a). Parameter estimates for models within the
95% confidence limit are reported in Appendix E.
73
Glucose mineralization and its relationship to microbial community structure, land-use,
and soil characteristics
Given that only marginal support for any one model or model parameter was found in the
initial candidate set, we decided to explore the relationship between several key soil
characteristics at our sites and the cumulative mineralization of glucose. We found that
the best model (i.e. lowest AICc) in this candidate set included soil P concentrations as
the sole explanatory variable (Table 2.3). The probability that this model was the actual
best model in this set was 73% suggesting moderately strong support for this outcome. In
fact no other model was within 2 AICc units of this model and in comparison to the
second ranked model in this set, the evidence ratio was 10.11. This suggests that given
this set of candidate models that it is fairly certain that this model is in fact the top model.
Further supporting this is the fact that no other models were <4 AICc units from this top
model which indicates that there was overall considerably less support for any of the
other 109 models (Burnham and Anderson 1998). Parameter estimates for all models
within the 95% confidence limit are reported in Appendix E.
We also found that soil extractable P was the most important model parameter
and was significantly related to land-use (F3, 8 = 9.53; P<0.01). This is not surprising
given that soil P was both the sole parameter in the top model and was also found in all of
the other top models (i.e. all models within 10 AICc units of the top model). Soil P also
had an extremely high weight of evidence (ωi = 0.99; Fig. 2.2b), again demonstrating its
importance as an indicator of cumulative glucose mineralization. No other parameter was
found to be as important. In fact all other model parameters were > 10 times less likely to
be as important as soil P (Fig. 2.2b). Soil P and cumulative glucose mineralized, after
74
accounting for variation among plots, were positively related (β = 0.48 ± 0.09; F1,10 =
30.04; P<0.001). Regression analysis found that this was generally consistent across
seasons (Fig. 2.3) with soil P explaining 60 and 48% of the variation in cumulative
glucose mineralized in the winter and summer, respectively (Fig. 2.3).
Discussion
In this study we amended four land-uses representative of a rural upland landscape with
13C –labeled glucose and tracked its mineralization across the course of 72 h in both the
summer and winter in order to determine what factors, if any, were related to the
mineralization of this common LMWOC compound in soils. We expected that
cumulative glucose mineralization would be related to land-use and/or the size, activity,
or composition of the soil microbial community. We also evaluated a second set of
putative explanatory variables related to specific soil characteristics. We expected that
these soil characteristics either alone or in combination with microbial community
characteristics would affect cumulative glucose mineralization. Both of these sets of
variables were examined using an information-theoretic approach that allowed us to
determine the probable best model as well as the most important individual parameter for
explaining glucose mineralization. The approach, when applied to evaluate and compare
potential explanatory variables for which there is an established mechanistic rationale
detailing why they might cause variation in a response variable, generates ecologically-
relevant, regression models (see Burnham and Anderson 1998). Our overarching
objective was to advance the understanding of factors that influence the in situ
mineralization of LMWOC compounds across landscapes.
75
Across the course of 72 h we found that glucose mineralization rates decreased
exponentially and this pattern was consistent across land-use and season (Fig. 2.1a, b).
Such results are similar to the findings of an array of lab-based studies (Bremer and
Vankessel 1990, Bremer and Kuikman 1994, Hoyle et al. 2008, van Hees et al. 2008), as
well as the findings of field-based studies where relatively high concentrations of glucose
have been applied (Voroney et al. 1989). The mineralization dynamics observed in this
study where similar to those observed by Boddy et al. (2007), the only other in situ study
which simulated glucose additions at amounts and concentrations comparable to those
expected under root exudation. This may suggest that across a wide array of settings the
mineralization dynamics of glucose and, perhaps other LMWOC compounds, follow a
predictable pattern.
Of the glucose added, we found that as much as 4% was recovered as CO2
meaning that ~96% of glucose-derived C remained in the soil. This is not an uncommon
occurrence, with comparable studies such as Voroney et al. (1989) finding that ~25% of a
large glucose addition remained after 7 years and Boddy et al. (2007) finding that ~70-
80% of a small glucose addition remained after 48 h. In light of such findings, it has been
suggested that relatively large amounts of added/exuded glucose are cycled through the
microbial biomass and may to some degree be stabilized as microbial byproducts in the
soil organic matter (Voroney et al. 1989). However, such processes may be slow and
occur over decadal time scales (Richter et al. 1999). The stabilization (or not) of
LMWOC in SOC sinks is an area requiring considerably more research attention, given
the importance of the inputs of these compounds to soils.
76
We found that land-use affected the cumulative amount of glucose mineralized
(Fig. 2.1c, d). Like temporal mineralization dynamics (Fig. 2.1a, b), we found that
pasture sites typically had greater cumulative glucose mineralization than did pine or
hardwood sites. These results concur with those of Stevenson et al. (2004), who showed,
using lab-based catabolic response profiling, that grassland soils are typically associated
with greater glucose mineralization than forest soils. Cumulative glucose mineralized in
the cultivated sites was on average lower in the winter than in the summer and this was
likely due to the presence of established crop cover in the summer. Specifically,
cumulative mineralization in the cultivated sites in summer was often comparable to
cumulative values for pastures. In the winter two of the cultivated sites were fallow and
the third (Cultivated 3) had freshly sprouted winter wheat. Interestingly, variation in
mineralization rates across cultivated sites in the summer was high (Fig. 2.1d) and this
variation may be due to differences in crop species cover. However, the role of plant-
cover species identity in regulating mineralization dynamics of LMWOC compounds is
relatively unexplored (see Paterson et al. 2007).
Not surprising given that land-use was significantly related to cumulative glucose
mineralized, the best explanatory model in the primary set of models analyzed using the
information theoretic approach contained land-use as the sole parameter (Table 2.2).
Land-use was also found to be the most important parameter for the prediction of
cumulative glucose mineralization (Fig. 2.2a). Surprisingly, all models containing
parameters related to the size, activity, or composition of the microbial community were
likely poor predictors of cumulative glucose mineralization and overall they were not
important parameters (Table 2.2, Fig. 2.2a). A possible explanation for the overall lack of
77
importance for all of these model variables may be the high functional redundancy in the
microbial community with regards the mineralization of glucose (Jones and Murphy
2007, Hanson et al. 2008). However, we only used one method each to determine the
size, activity, and composition of the microbial communities. Although these methods are
commonly and widely employed, they are coarse and may not be indicative of the actual
characteristics of the microbial community involved in the mineralization of glucose.
Future research will need to assess the role of methodology as it pertains to measures of
the microbial community that might explain glucose mineralization dynamics in situ.
Land-use and land-management regimes are likely to transform soil systems in
many ways and this is apparent at the CEF (Richter and Markewitz 2001). Research at
these sites has demonstrated that a suite of edaphic properties (e.g. soil P, texture) are
affected by both the contemporary land-use as well as the legacy of land-management
decisions (Richter and Markewitz 2001, Richter et al. 2006, Lauber et al. 2008, Grandy et
al. 2009). For example, Richter et al. (2006) demonstrates across the same landscape that
soil P is strongly influenced by management history such as fertilization, even for
decades after it was last practiced. Notably, we found that soil P was a parameter of
major importance in explaining cumulative glucose mineralization; it was over 10 times
more important than any other parameter considered in the entire model set (Fig. 2.2b).
Soil P was positively related to cumulative glucose mineralized and this was consistent
across both the winter and summer (Fig. 2.3). These results may suggest that both past
and present land-use/management decisions ultimately influence ecosystem C dynamics
through influences on underlying variables including soil P. Saggar et al. (2000), using a
lab-based study of pasture soils with different fertilizer input histories, observed that
78
glucose mineralization was positively related to soil P concentrations. They suggested
that this occurs because the microbial biomass becomes less efficient at utilizing C-
compounds as soil P becomes limiting. Similarly, Cleveland and Townsend (2006) found
that when labile C (i.e. leached from litter) was plentiful then P-availability was the
primary control on its mineralization under both field and lab conditions. There is also
the likelihood that increasing P-availability increases overall plant biomass, including
root biomass (Saggar et al. 1997, Saggar et al. 2000), leading to an increase in the size of
the microbial community geared toward the mineralization of LMWOC compounds like
glucose. A recent study conducted with soils from our CEF sites demonstrated that the
decomposition of recalcitrant leaf litter material decreased as soil P increased (Strickland
et al. 2009b). Together these studies, and our new results, suggest that soil P may be an
indicator of shifts in the microbial community from (under higher P) organisms which
specialize in utilizing LMWOC compounds to those (under lower P) which utilize more
complex substrates such as leaf litter. This may be similar to distinctions relating certain
components of the microbial community to r- and K-strategists (Fierer et al. 2007,
Blagodatskaya et al. 2009). Alternatively, soil P may simply be co-correlated with
characteristics of the soil microbial community that determine C-dynamics. For example,
it has been shown that soil P is related to the relative abundance of different microbial
taxa (Lauber et al. 2008). Increased resolution beyond the level of fungal-to-bacterial
dominance may provide more relevant indices of community composition when
attempting to predict glucose mineralization (e.g. Hanson et al. 2008).
Soil P may also be related to the ATP supply rate of the microbial community
(Bradford et al. 2008b, Hoyle et al. 2008). Under this expectation soil P is associated with
79
an increased ATP store within the microbial community which may subsequently lead to
greater glucose mineralization. On the other hand, a greater ATP supply rate may favor
soil microbes that maintain a high intracellular concentration of ATP (De Nobili et al.
2001, Hoyle et al. 2008). These types of microorganisms have been described as
exhibiting a ‗metabolically alert‘ life history strategy (De Nobili et al. 2001). Such
organisms might be expected to dominate in systems where LMWOC compounds are the
predominant form of C inputs. If either of these possibilities hold true then assessments
of ATP may be a better indicator of microbial activity with regards to glucose
mineralization than the SIR method employed in this study (but see Landi et al. 2006).
In conclusion, it was the goal of this study to determine the in situ mineralization
of the common rhizodeposit and leachate-constituent, glucose, and its relationship to
land-use, soil microbial community and edaphic characteristics across a landscape.
Surprisingly no measure of the microbial community was related to glucose
mineralization. However, soil P and land-use were predictors of glucose mineralization.
Given that land-use and soil P are intimately related at these sites, there is the indication
that land management decisions which impact soil P may in turn lead to altered
mineralization rates of LMWOC compounds (Richter and Markewitz 2001, Richter et al.
2006). This is not likely a trivial matter given that a large proportion of the Earth‘s soils
are similar to the low P Ultisols at our sites and that many of these same soils are apt to
be or are already impacted by management decisions which increase soil P (Richter and
Markewitz 2001). Such management decisions may lead to increased mineralization of
LMWOC compounds, altering ecosystem C-cycling and C source-sink dynamics. Future
studies will need to address these possibilities and decipher the underlying mechanisms
80
that explain the relationship between soil P and C dynamics. Such studies will further our
understanding of the factors which control the cycling of belowground plant-C inputs and
may ultimately lead to an understanding of LMWOC compound dynamics comparable to
our current understanding of leaf litter dynamics.
Acknowledgements
We gratefully acknowledge funding from the Andrew W. Mellon Foundation to N.F. and
M.A.B., and to M.A.B from the Office of Science (BER), U.S. Department of Energy.
We thank Tom Maddox in the Analytical Chemistry Laboratory of the Odum School of
Ecology, Univ. of Georgia, for element and isotope analyses; Jimmy Blackmon for field
assistance; and Mr. Jack Burnett for access to his property.
References
Aerts, R. 1997. Climate, leaf litter chemistry and leaf litter decomposition in terrestrial
ecosystems: A triangular relationship. Oikos 79:439-449.
Allison, S. D., and J. B. H. Martiny. 2008. Resistance, resilience, and redundancy in
microbial communities. Proceedings of the National Academy of Sciences of the
United States of America 105:11512-11519.
Ayres, E., H. Steltzer, B. L. Simmons, R. T. Simpson, J. M. Steinweg, M. D. Wallenstein,
N. Mellor, W. J. Parton, J. C. Moore, and D. H. Wall. 2009. Home-field
advantage accelerates leaf litter decomposition in forests. Soil Biology &
Biochemistry 41:606-610.
81
Bais, H. P., T. L. Weir, L. G. Perry, S. Gilroy, and J. M. Vivanco. 2006. The role of root
exudates in rhizosphere interations with plants and other organisms. Annual
Review of Plant Biology 57:233-266.
Bardgett, R. D., and E. McAlister. 1999. The measurement of soil fungal : bacterial
biomass ratios as an indicator of ecosystem self-regulation in temperate meadow
grasslands. Biology and Fertility of Soils 29:282-290.
Barlow, J., T. A. Gardner, L. V. Ferreira, and C. A. Peres. 2007. Litter fall and
decomposition in primary, secondary and plantation forests in the Brazilian
Amazon. Forest Ecology and Management 247:91-97.
Bengtson, P., and G. Bengtsson. 2007. Rapid turnover of DOC in temperate forests
accounts for increased CO2 production at elevated temperatures. Ecology Letters
10:783-790.
Blagodatskaya, E. V., S. A. Blagodatsky, T. H. Anderson, and Y. Kuzyakov. 2009.
Contrasting effects of glucose, living roots and maize straw on microbial growth
kinetics and substrate availability in soil. European Journal of Soil Science
60:186-197.
Boddy, E., P. W. Hill, J. Farrar, and D. L. Jones. 2007. Fast turnover of low molecular
weight components of the dissolved organic carbon pool of temperate grassland
field soils. Soil Biology & Biochemistry 39:827-835.
Bradford, M. A., N. Fierer, R. B. Jackson, T. R. Maddox, and J. F. Reynolds. 2008a.
Nonlinear root-derived carbon sequestration across a gradient of nitrogen and
phosphorous deposition in experimental mesocosms. Global Change Biology
14:1113-1124.
82
Bradford, M. A., N. Fierer, and J. F. Reynolds. 2008b. Soil carbon stocks in experimental
mesocosms are dependent on the rate of labile carbon, nitrogen and phosphorus
inputs to soils. Functional Ecology 22:964-974.
Bradford, M. A., P. A. Wookey, P. Ineson, and H. M. Lappin-Scott. 2001. Controlling
factors and effects of chronic nitrogen and sulphur deposition on methane
oxidation in a temperate forest soil. Soil Biology & Biochemistry 33:93-102.
Bremer, E., and P. Kuikman. 1994. Microbial utilization of C14
glucose in soil is affected
by the amount and timing of glucose additions. Soil Biology & Biochemistry
26:511-517.
Bremer, E., and C. Vankessel. 1990. Extractability of microbial C14
and N15
following
addition of variable rates of labeled glucose and (NH4)2SO4 to soil. Soil Biology
& Biochemistry 22:707-713.
Burnham, K. P., and D. R. Anderson. 1998. Model Selection and Multimodel Inference:
A Practical Information-Theoretic Approach. Springer, New York, NY, USA.
Callaham, M. A., D. D. Richter, D. C. Coleman, and M. Hofmockel. 2006. Long-term
land-use effects on soil invertebrate communities in Southern Piedmont soils,
USA. European Journal of Soil Biology 42:S150-S156.
Clare, J. T., D. Johnson, and R. R. E. Artz. 2009. Litter type, but not plant cover,
regulates initial litter decomposition and fungal community structure in a
recolonising cutover peatland. Soil Biology & Biochemistry 41:651-655.
Cleveland, C. C., and A. R. Townsend. 2006. Nutrient additions to a tropical rain forest
drive substantial soil carbon dioxide losses to the atmosphere. Proceedings of the
National Academy of Sciences of the United States of America 103:10316-10321.
83
Couteaux, M. M., P. Bottner, and B. Berg. 1995. Litter decomposition, climate and litter
quality. Trends in Ecology & Evolution 10:63-66.
Dakora, F. D., and D. A. Phillips. 2002. Root exudates as mediators of mineral
acquisition in low-nutrient environments. Plant and Soil 245:35-47.
De Nobili, M., M. Contin, C. Mondini, and P. C. Brookes. 2001. Soil microbial biomass
is triggered into activity by trace amounts of substrate. Soil Biology &
Biochemistry 33:1163-1170.
Fierer, N., M. A. Bradford, and R. B. Jackson. 2007. Toward an ecological classification
of soil bacteria. Ecology 88:1354-1364.
Fierer, N., J. M. Craine, K. McLauchlan, and J. P. Schimel. 2005a. Litter quality and the
temperature sensitivity of decomposition. Ecology 86:320-326.
Fierer, N., J. A. Jackson, R. Vilgalys, and R. B. Jackson. 2005b. Assessment of soil
microbial community structure by use of taxon-specific quantitative PCR assays.
Applied and Environmental Microbiology 71:4117-4120.
Fierer, N., and R. B. Jackson. 2006. The diversity and biogeography of soil bacterial
communities. Proceedings of the National Academy of Sciences of the United
States of America 103:626-631.
Fierer, N., and J. P. Schimel. 2002. Effects of drying-rewetting frequency on soil carbon
and nitrogen transformations. Soil Biology & Biochemistry 34:777-787.
Fierer, N., and J. P. Schimel. 2003. A proposed mechanism for the pulse in carbon
dioxide production commonly observed following the rapid rewetting of a dry
soil. Soil Science Society of America Journal 67:798-805.
84
Fontaine, S., and S. Barot. 2005. Size and functional diversity of microbe populations
control plant persistence and long-term soil carbon accumulation. Ecology Letters
8:1075-1087.
Franzluebbers, K., A. J. Franzluebbers, and M. D. Jawson. 2002. Environmental controls
on soil and whole-ecosystem respiration from a tallgrass prairie. Soil Science
Society of America Journal 66:254-262.
Gaudinski, J. B., S. E. Trumbore, E. A. Davidson, A. C. Cook, D. Markewitz, and D. D.
Richter. 2001. The age of fine-root carbon in three forests of the eastern United
States measured by radiocarbon. Oecologia 129:420-429.
Gee, G. W., and D. Or. 2002. Particle-Size Analysis. Pages 255-293 in J. H. Dane and G.
C. Topp, editors. Methods of Soil Analysis, Part 4: Physical Methods. Soil
Science Society of America, Inc., Madison, Wisconsin, USA.
Gholz, H. L., D. A. Wedin, S. M. Smitherman, M. E. Harmon, and W. J. Parton. 2000.
Long-term dynamics of pine and hardwood litter in contrasting environments:
toward a global model of decomposition. Global Change Biology 6:751-765.
Grandy, A. S., M. S. Strickland, C. Lauber, M. A. Bradford, and N. Fierer. 2009. The
influence of microbial communities, management, and soil texture on soil organic
chemistry. Geoderma 150: 278-286.
Grayston, S. J., D. Vaughan, and D. Jones. 1997. Rhizosphere carbon flow in trees, in
comparison with annual plants: The importance of root exudation and its impact
on microbial activity and nutrient availability. Applied Soil Ecology 5:29-56.
Green, J. L., B. J. M. Bohannan, and R. J. Whitaker. 2008. Microbial biogeography:
From taxonomy to traits. Science 320:1039-1043.
85
Gu, L. H., W. M. Post, and A. W. King. 2004. Fast labile carbon turnover obscures
sensitivity of heterotrophic respiration from soil to temperature: A model analysis.
Global Biogeochemical Cycles 18.
Gusewell, S., and M. O. Gessner. 2009. N : P ratios influence litter decomposition and
colonization by fungi and bacteria in microcosms. Functional Ecology 23:211-
219.
Hanson, C. A., S. D. Allison, M. A. Bradford, M. D. Wallenstein, and K. K. Treseder.
2008. Fungal taxa target different carbon sources in forest soil. Ecosystems
11:1157-1167.
Harmon, M. E., W. L. Silver, B. Fasth, H. Chen, I. C. Burke, W. J. Parton, S. C. Hart, and
W. S. Currie. 2009. Long-term patterns of mass loss during the decomposition of
leaf and fine root litter: an intersite comparison. Global Change Biology 15:1320-
1338.
Hogberg, P., M. N. Hogberg, S. G. Gottlicher, N. R. Betson, S. G. Keel, D. B. Metcalfe,
C. Campbell, A. Schindlbacher, V. Hurry, T. Lundmark, S. Linder, and T.
Nasholm. 2008. High temporal resolution tracing of photosynthate carbon from
the tree canopy to forest soil microorganisms. New Phytologist 177:220-228.
Hoyle, F. C., D. V. Murphy, and P. C. Brookes. 2008. Microbial response to the addition
of glucose in low-fertility soils. Biology and Fertility of Soils 44:571-579.
Hrynkiewicz, K., C. Baum, and P. Leinweber. 2009. Mycorrhizal community structure,
microbial biomass P and phosphatase activities under Salix polaris as influenced
by nutrient availability. European Journal of Soil Biology 45:168-175.
86
Hunt, H. W., E. R. Ingham, D. C. Coleman, E. T. Elliott, and C. P. P. Reid. 1988.
Nitrogen limitation of production and decomposition in prairie, mountain
meadow, and pine forest. Ecology 69:1009-1016.
Ineson, P., M. F. Cotrufo, R. Bol, D. D. Harkness, and H. Blum. 1996. Quantification of
soil carbon inputs under elevated CO2:C3 plants in a C
4 soil. Plant and Soil
187:345-350.
Iqbal, J., R. G. Hu, L. J. Du, L. Lan, L. Shan, C. Tao, and L. L. Ruan. 2008. Differences
in soil CO2 flux between different land use types in mid-subtropical China. Soil
Biology & Biochemistry 40:2324-2333.
Jones, D. L., and D. V. Murphy. 2007. Microbial response time to sugar and amino acid
additions to soil. Soil Biology & Biochemistry 39:2178-2182.
Jones, R. T., M. S. Robeson, C. L. Lauber, M. Hamady, R. Knight, and N. Fierer. 2009.
A comprehensive survey of soil acidobacterial diversity using pyrosequencing
and clone library analyses. ISME Journal 3:442-453.
Kemmitt, S. J., C. V. Lanyon, I. S. Waite, Q. Wen, T. M. Addiscott, N. R. A. Bird, A. G.
O'Donnell, and P. C. Brookes. 2008. Mineralization of native soil organic matter
is not regulated by the size, activity or composition of the soil microbial biomass -
a new perspective. Soil Biology & Biochemistry 40:61-73.
Kuo, S. 1996. Phosphorus. Pages 869-919 in D. L. Sparks, editor. Methods of soil
analysis, Part 3: Chemical Methods. Soil Science Society of America, Madison,
WI, USA.
Landi, L., F. Valori, J. Ascher, G. Renella, L. Falchini, and P. Nannipieri. 2006. Root
exudate effects on the bacterial communities, CO2 evolution, nitrogen
87
transformations and ATP content of rhizosphere and bulk soils. Soil Biology &
Biochemistry 38:509-516.
Lauber, C. L., M. S. Strickland, M. A. Bradford, and N. Fierer. 2008. The influence of
soil properties on the structure of bacterial and fungal communities across land-
use types. Soil Biology & Biochemistry 40:2407-2415.
MacNeil, M. A., N. A. J. Graham, N. V. C. Polunin, M. Kulbicki, R. Galzin, M.
Harmelin-Vivien, and S. P. Rushton. 2009. Hierarchical drivers of reef-fish
metacommunity structure. Ecology 90:252-264.
Madritch, M. D., and B. J. Cardinale. 2007. Impacts of tree species diversity on litter
decomposition in northern temperate forests of Wisconsin, USA: a multi-site
experiment along a latitudinal gradient. Plant and Soil 292:147-159.
Melillo, J. M., J. D. Aber, and J. F. Muratore. 1982. Nitrogen and lignin control of
hardwood leaf litter decomposition dynamics. Ecology 63:621-626.
Mo, J., W. Zhang, W. Zhu, P. Gundersen, Y. Fang, D. Li, and H. Wang. 2008. Nitrogen
addition reduces soil respiration in a mature tropical forest in southern China.
Global Change Biology 14:403-412.
Nay, S. M., K. G. Mattson, and B. T. Bormann. 1994. Biases of chamber methods for
measuring soil CO2 efflux demonstrated with a laboratory apparatus. Ecology
75:2460-2463.
Nguyen, C. 2003. Rhizodeposition of organic C by plants: mechanisms and controls.
Agronomie 23:375-396.
88
Paterson, E., T. Gebbing, C. Abel, A. Sim, and G. Telfer. 2007. Rhizodeposition shapes
rhizosphere microbial community structure in organic soil. New Phytologist
173:600-610.
Phillips, R. P., Y. Erlitz, R. Bier, and E. S. Bernhardt. 2008. New approach for capturing
soluble root exudates in forest soils. Functional Ecology 22:990-999.
Pinheiro, J. C., and D. M. Bates. 2000. Mixed-effects models in S and S-PLUS. Springer
Verlag, New York, New York, USA.
Pollierer, M. M., R. Langel, C. Korner, M. Maraun, and S. Scheu. 2007. The
underestimated importance of belowground carbon input for forest soil animal
food webs. Ecology Letters 10:729-736.
Richter, D. D., D. Markewitz, S. E. Trumbore, and C. G. Wells. 1999. Rapid
accumulation and turnover of soil carbon in a re-establishing forest. Nature
400:56-58.
Richter, D. D., H. L. Allen, J. W. Li, D. Markewitz, and J. Raikes. 2006. Bioavailability
of slowly cycling soil phosphorus: major restructuring of soil P fractions over four
decades in an aggrading forest. Oecologia 150:259-271.
Richter, D. D., and D. Markewitz. 2001. Understanding Soil Change: Soil sustainability
over millenia, centuries, and decades. Cambridge University Press, New York,
New York, USA.
Rovira, A. D. 1969. Plant Root Exudates. Botanical Review 35:35-&.
Saggar, S., C. Hedley, and A. D. Mackay. 1997. Partitioning and translocation of
photosynthetically fixed C14
in grazed hill pastures. Biology and Fertility of Soils
25:152-158.
89
Saggar, S., C. B. Hedley, K. M. Giddens, and G. J. Salt. 2000. Influence of soil
phosphorus status and nitrogen addition on carbon mineralization from C14
-
labelled glucose in pasture soils. Biology and Fertility of Soils 32:209-216.
Six, J., S. D. Frey, R. K. Thiet, and K. M. Batten. 2006. Bacterial and fungal
contributions to carbon sequestration in agroecosystems. Soil Science Society of
America Journal 70:555-569.
Stevenson, B. A., G. P. Sparling, L. A. Schipper, B. P. Degens, and L. C. Duncan. 2004.
Pasture and forest soil microbial communities show distinct patterns in their
catabolic respiration responses at a landscape scale. Soil Biology & Biochemistry
36:49-55.
Strickland, M. S., C. Lauber, N. Fierer, and M. A. Bradford. 2009a. Testing the
functional significance of microbial community composition. Ecology 90:441-
451.
Strickland, M. S., E. Osburn, C. Lauber, N. Fierer, and M. A. Bradford. 2009b. Litter
quality is in the eye of the beholder: initial decomposition rates as a function of
inoculum characteristics. Functional Ecology. 23:627-636.
van Hees, P. A. W., E. Johansson, and D. L. Jones. 2008. Dynamics of simple carbon
compounds in two forest soils as revealed by soil solution concentrations and
biodegradation kinetics. Plant and Soil 310:11-23.
van Hees, P. A. W., D. L. Jones, R. Finlay, D. L. Godbold, and U. S. Lundstomd. 2005.
The carbon we do not see - the impact of low molecular weight compounds on
carbon dynamics and respiration in forest soils: a review. Soil Biology &
Biochemistry 37:1-13.
90
van Hees, P. A. W., D. L. Jones, and D. L. Godbold. 2002. Biodegradation of low
molecular weight organic acids in coniferous forest podzolic soils. Soil Biology &
Biochemistry 34:1261-1272.
Voroney, R. P., E. A. Paul, and D. W. Anderson. 1989. Decomposition of Wheat Straw
and Stabilization of Microbial Products. Canadian Journal of Soil Science 69:63-
77.
Wardle, D. A., and A. Ghani. 1995. Why is the strength of relationships between pairs of
methods for estimating soil microbial biomass often so variable. Soil Biology &
Biochemistry 27:821-828.
Warembourg, F. R., C. Roumet, and F. Lafont. 2003. Differences in rhizosphere carbon-
partitioning among plant species of different families. Plant and Soil 256:347-357.
Winogradsky, S. 1924. Sur la microflora autochthone de la terrearable. Hebd Acad Sci,
Paris 178:1236-1239.
Yuste, J. C., D. D. Baldocchi, A. Gershenson, A. Goldstein, L. Misson, and S. Wong.
2007. Microbial soil respiration and its dependency on carbon inputs, soil
temperature and moisture. Global Change Biology 13:2018-2035.
91
Table 2.1 Mean values for soil and microbial community characteristics per land-use in the winter and summer. Values in parentheses
are standard errors across the three sites in each land-use. Silt + clay content, soil C:N, and soil P were determined in September 2006
only as these measures are not expected to vary greatly across seasons.
Land-use
Microbial
biomass
(g C m-2
)
SIR
(g CO2-C
m-2
h-1
)
F:B pH
Soil
Temperature
(ºC)
Soil
Moisture
(%)
Silt + Clay
(%) Soil C:N
Ext P
(g m-2
)
Cultivated
Winter 5.17
(2.33)
0.08
(0.008)
0.022
(0.013)
5.39
(0.14)
14.2
(0.23)
28.94
(2.72)
38.8
(2.87)
14.68
(1.19)
1.62
(0.22)
Summer 5.26
(1.95)
0.13
(0.011)
0.002
(0.000)
4.51
(0.22)
25.4
(0.52)
19.50
(0.52) -- -- --
Pasture
Winter 4.77
(1.40)
0.11
(0.007)
0.047
(0.027)
5.62
(0.18)
14.2
(0.04)
27.67
(0.04)
21.0
(4.16)
14.11
(0.38)
1.94
(0.30)
Summer 8.81
(3.36)
0.22
(0.024)
0.004
(0.001)
5.15
(0.09)
24.3
(0.08)
15.29
(0.08) -- -- --
Pine
Winter 2.86
(0.58)
0.09
(0.003)
0.008
(0.001)
5.37
(0.33)
14.9
(0.21)
14.12
(0.21)
17.7
(3.13)
23.90
(4.95)
0.55
(0.39)
Summer 6.38
(1.92)
0.15
(0.036)
0.008
(0.003)
4.63
(0.36)
19.9
(0.31)
13.46
(0.31) -- -- --
Hardwood
Winter 5.87
(2.28)
0.12
(0.021)
0.239
(0.159)
6.09
(0.50)
14.2
(0.20)
20.61
(0.20)
30.1
(3.18)
21.78
(2.96)
0.17
(0.11)
Summer 8.83
(6.36)
0.16
(0.028)
0.006
(0.002)
5.15
(0.30)
20.7
(0.42)
17.04
(0.42) -- -- --
92
Table 2.2 The 10 best models (i.e. ΔAICc<10 ) of the initial 30 candidate models considered which explain cumulative 13
C-glucose
mineralized across 72 h (13
CO2).
Model K log(L) AICc Δi AICc ωi Evidence
ratio
Rank
13CO2 = Land*
6 -7.08 40.15 0.00 0.53 1.00 1
13CO2 = 1* 3 -16.38 41.42 1.27 0.28 1.89 2
13CO2 = MicC* 4 -15.73 44.47 4.32 0.06 8.65 3
13CO2 = F:B* 4 -16.25 45.51 5.36 0.04 14.57 4
13CO2 = SIR* 4 -16.36 45.72 5.57 0.03 16.18 5
13CO2 = MicC + SIR 5 -14.67 47.90 7.75 0.01 48.24 6
13CO2 = MicC + Seas. 5 -14.82 48.21 8.06 0.01 56.16 7
13CO2 = Land + MicC 7 -6.29 48.98 8.83 0.01 82.73 8
13CO2 = MicC + F:B 5 -15.39 49.36 9.21 0.01 99.92 9
13CO2 = Land + Seas. 7 -6.82 50.04 9.89 0.00 140.35 10
The table shows the number of parameters (K), maximized log-likelihood (log(L)), AICc, AICc differences (Δi AICc), Akaike weights
(ωi), evidence ratio (ω1/ ωn), and the model rank. In bold are the best top models in this candidate set (ΔAICc<2). Models within the
95% confidence interval are denoted with an * and parameter estimates for these models are given in Appendix E. All models had plot
as a random effect in order to account for repeated sampling at the plot level.
SIR = substrate induced respiration, a measure of microbial activity; MicC = microbial biomass C as determined via CFE, a measure
of the size of the microbial biomass; F:B = fungal-to-bacterial dominance as determined via qPCR, a measure of community
composition; Seas. = Season, either winter or summer when the pulse-chase experiment was conducted; Land = land-use.
93
Table 2.3 The 6 best models (i.e. ΔAICc<10) of the initial 104 candidate models considered which explain cumulative 13
C-glucose
mineralized across 72 h (13
CO2).
Model K log(L) AICc Δi AICc ωi Evidence
ratio
Rank
13CO2 = P*
4 -8.10 29.20 0.00 0.73 1.00 1
13CO2 = P + MicC* 5 -7.63 33.83 4.63 0.07 10.11 2
13CO2 = P + SIR* 5 -7.84 34.26 5.06 0.06 12.53 3
13CO2 = P + Seas.* 5 -7.87 34.30 5.10 0.06 12.82 4
13CO2 = P + F:B 5 -8.06 34.68 5.48 0.05 15.50 5
13CO2 = P + MicC + SIR 6 -5.52 37.04 7.84 0.01 50.31 6
The table shows the number of parameters (K), maximized log-likelihood (log(L)), AICc, AICc differences (Δi AICc), Akaike weights
(ωi), evidence ratio (ω1/ ωn), and the model rank. In bold are the best top models in this candidate set (ΔAICc<2). Models within the
95% confidence interval are denoted with an * and parameter estimates for these models are given in Appendix E. All models had plot
as a random effect in order to account for repeated sampling at the plot level.
SIR, MicC, F:B, Land, and Seas. are the same as in Table 2.2. P = extractable phosphorous.
94
Figure 2.1 Results of the 13
C-glucose pulse-chase conducted during the winter
(December 2006) and summer (June 2007) associated with land-use. Panels a and b show
glucose mineralization dynamics across 72 h for each land-use. Land-use differences
were detected for these dynamics across seasons (P<0.01) and during both the winter
(P<0.01) and summer (P<0.01). Panels c and d show the cumulative amount of glucose
mineralized during the entire 72-h period. Although there was a significant land-use
effect across seasons on cumulative glucose mineralization (P<0.01) this was not
consistent within seasons. That is, a statistically significant land-use effect was observed
in the winter (P<0.001) but not in the summer (P=0.12) for cumulative amounts. Values
are means ± 1S.E. per land-use (n=3) and letters in panel c indicate significant differences
between land-uses based on pair-wise comparisons. Note that mineralization dynamics
and cumulative values were roughly equivalent for a given land-use regardless of season.
Figure 2.2 The relative importance of variables from the primary (a) and secondary (b)
set of candidate models which explain cumulative glucose mineralized. Relative
importance was determined via the sum of Akaike weights (∑ωi) for the models in which
a given variable was present. The closer to 1 a variable is the more important it is in this
given set. Numbers to the right of each bar indicate the number of models that a given
variable was in (e.g. land-use was in 16 of the candidate models).
Figure 2.3 The relationship between cumulative glucose mineralized and soil P for the
winter (a) and summer (b). Notably this relationship was very similar in both the winter
(y = 0.51x + 2.73) and summer (y = 0.45x + 2.90). Cumulative glucose mineralized was
loge-transformed in the regression analysis but untransformed values are shown here.
Each symbol represents a specific plot with Cu1-3, Pa1-3, Pi1-3, and Ha1-3 representing
95
the individual sites (n=3) in the cultivated, pasture, pine, and hardwood land-uses,
respectively.
96
Figure 2.1
(a) Winter (December) Land-use effect: P <0.01
Res
pir
ed 1
3C
O2-C
(m
g m
-2 h
-1)
0
1
2
3
4
Cultivated
Pasture
Pine
Hardwood
(b) Summer (June) Land-use effect: P <0.01
Time since glucose added (h)
0 20 40 60 80
0
1
2
3
4
(c) Winter (December) Land-use effect: P <0.001
Cu
mu
lati
ve
glu
cose
min
eral
ized
(m
g 1
3C
O2-C
m-2
)
0
10
20
30
40
50
60
70
a
c
ab b
(d) Summer (June) Land-use effect: P =0.12
Land-use
Cultivated Pasture Pine Hardwood
0
10
20
30
40
50
60
70
97
Figure 2.2
i
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Model
par
amet
er
P
MicC
SIR
Seas.
F:B
M*T
C:N
Land
pH
S+C 16
16
16
16
16
16
56
56
56
55
Land
MicC
SIR
F:B
Seas. 16
16
16
16
16
(a) Initial candidate model set
(b) Secondary candidate model set
98
Figure 2.3
Winter (December)
0
10
20
30
40
50
60
70
80
Cu1
Cu2
Cu3
Pa1
Pa2Pa3
Pi1
Pi2
Pi3
Ha1
Ha2
Ha3
Summer (June)
Extractable P(g m-2
)
0.0 0.5 1.0 1.5 2.0 2.5
Cu
mu
lati
ve
glu
cose
min
eral
ized
(m
g 1
3C
O2-C
m-2
)
10
20
30
40
50
60
70
80
Cu1
Cu2
Cu3
Pa1 Pa2
Pa3
Pi1
Pi2
Pi3
Ha1
Ha2
Ha3
r2 = 0.60
P<0.01
r2 = 0.48
P<0.01
99
CHAPTER 3
GRASS INVASION OF A HARDWOOD FOREST IS ASSOCIATED WITH
DECLINES IN BELOWGROUND CARBON POOLS 3
3 Strickland, M.S., J.L. DeVore, J.C. Maerz, M.A. Bradford. Submitted to Global Change Biology,
4/23/2009.
100
Abstract
Invasive plant species affect a wide range of ecosystem processes but their impact on
belowground carbon pools is still relatively unexplored. This is particularly true for grass
invasions of forested ecosystems. Such invasions may alter both the quantity and quality
of inputs to forest floors. Dependent on both, two dominant theories, ‗priming‘ and
‗preferential substrate utilization‘, suggest these changes may decrease, increase, or leave
unchanged native plant-derived soil carbon. A decrease is expected under ‗priming‘
theory due to increased activity of the soil microbial community degrading soil carbon.
Under the theory of ‗preferential substrate utilization‘, either an increase or no change is
expected because the invasive plant‘s inputs are used by the microbial community instead
of soil carbon. Here we examine how the widespread invasive Microstegium vimineum
affects belowground carbon-cycling in a southeastern U.S. forest. Following predictions
of priming theory, the presence of M. vimineum is associated with decreases in native-
derived, soil carbon pools. For example, in September 2006 M. vimineum is associated
with 24, 34, 36 and 72% declines in total organic, particulate organic matter,
mineralizable, and microbial biomass carbon, respectively. Soil carbon formed from M.
vimineum inputs does not fully compensate for these decreases, meaning that the sum of
native- and invasive-derived carbon pools are smaller than native-derived carbon pools in
uninvaded plots. Supporting our inferences that carbon-cycling is accelerated under
invasion, the microbial community is more active per unit biomass: added 13
C-glucose is
respired more rapidly in invaded plots. Our work suggests that this invader may
accelerate carbon-cycling in forest soils and deplete carbon stocks. The paucity of studies
investigating impacts of grass invasion on soil carbon-cycling in forests highlights the
101
need to further study M. vimineum and other invasive grasses to assess their impacts on
carbon sink strength and forest fertility.
Key-words: Microstegium vimineum, Japanese stiltgrass, Nepalese browntop, stable
isotopes, carbon sequestration, carbon sink, priming effects, preferential substrate
utilization, annual grass, exotic species
Introduction
Invasive plant species are capable of altering a range of ecosystem processes and
properties from nutrient cycling to species interactions (Mack et al. 2000; Liao et al.
2008). In some cases, these impacts are the result of specific traits that are novel to the
recipient community. For example, nitrogen (N)-fixing invasive species can dramatically
impact decomposition rates and soil fertility in ecosystems lacking native N-fixers
(Vitousek et al. 1987; Vitousek & Walker 1989; Allison & Vitousek 2004). Similarly, the
introduction of invasive grass species can increase rates of fire disturbance and fire
intensity, altering rates of carbon (C)-cycling (Mack et al. 2001; Mack & D'Antonio
2003; Bradley et al. 2006). We know far less about the more general effects of invaders,
particularly the many species that do not exhibit particular traits such as N-fixation.
Consequently, it is difficult to predict whether many invasive plant species will
significantly alter ecosystem processes and/or the communities they invade.
Blumenthal (2006) noted that many invasive plant species are ―high-resource‖
species. That is, these species show lower investment in defensive compounds and tend
to be of higher chemical quality (e.g. lower C:N) than native plant species. Others have
made similar observations (Pattison et al. 1998; Baruch & Goldstein 1999; D'Antonio &
102
Corbin 2003). Evidence also suggests that the presence of invasive plant species is, more
often than not, associated with increased detrital inputs to native ecosystems (Ehrenfeld
2003; Liao et al. 2008). While such changes may not be true for all invasive plants, this
highlights the likelihood that many invaders may alter ecosystem processes simply via
changes in the quantity and/or quality of detrital inputs within a system (D'Antonio &
Corbin 2003). This would be synonymous to those studies which demonstrate that
changes in the composition of native detrital resources within an ecosystem affects that
system‘s decomposition rates (Ball et al. 2008; Harguindeguy et al. 2008). Specifically,
the presence of an invasive plant species may alter the composition of detrital inputs
entering a system via an increase in both the quality and quantity of inputs. The impacts
of these altered inputs on belowground processes and properties, specifically with regard
to soil C, is an area of debate in soil theory (Bradford et al. 2008c) and such impacts as
they relate to invasive plant species are little studied (Hook et al. 2004; Valery et al.
2004; Batten et al. 2005; Litton et al. 2006, 2008).
Evidence on one hand suggests that increases in the degradability of C inputs, as
perceived by the microbial community (Strickland et al. 2009), will have a negligible and
in some cases a positive impact on belowground C pools (Dalenberg & Jager 1981; Wu
et al. 1993; Fontaine & Barot 2005). This is referred to as the ‗preferential substrate
utilization‘ theory and is based on the premise that high quality C sources, such as leaf
litter and rhizodeposits, will be utilized before lower quality soil C sources (Fontaine et
al. 2004a; Fontaine & Barot 2005). Although the mechanism leading to preferential
substrate utilization is debatable, it has been largely assumed that if novel inputs are of
both a high enough quality and quantity in comparison to background inputs then a shift
103
in the microbial community toward organisms which specialize on less recalcitrant
compounds will occur (Blagodatskaya & Kuzyakov 2008; Bradford et al. 2008c). Under
such an expectation the component of the microbial community responsible for the
degradation of more recalcitrant soil C sources may go unchanged leading to little change
in soil C pools (Fontaine et al. 2003). Under more extreme cases, the increase in
organisms specializing on less recalcitrant compounds may negatively impact (e.g.
through competition) those specializing on more recalcitrant resources (e.g. soil C)
leading to a decline in the latter‘s population and an increase in the size of soil C pools
(Fontaine et al. 2003). In contrast, ‗priming effects‘ theory posits that high-quality inputs
give rise to disproportionate impacts on soil C (Fontaine et al. 2003; Fontaine et al.
2004b; Dijkstra & Cheng 2007). Priming is theorized to occur because high quality inputs
stimulate the activity of the microbial community, in particular those groups which
specialize in decomposition of recalcitrant soil C pools. The result is enhanced
degradation of belowground C pools, leading to declines in their size (Fontaine et al.
2004b). It seems reasonable to assume that the impact of invasive plant species on
belowground C pools in native ecosystems will be dependent on the quality and quantity
of invasive inputs relative to native ones, the proportion of labile invasive inputs to
recalcitrant ones, and whether or not the outcomes of preferential substrate utilization or
priming effects theory are realized. Regardless of which outcome occurs, and given that
both theories suggest that a change in soil C-cycling will occur, altered C inputs with
plant invasions will likely alter a site‘s fertility, drought tolerance, C storage potential,
and possibly the aboveground-belowground interactions between organisms (Lal 2004;
Lal et al. 2007).
104
We studied an advancing invasion of Microstegium vimineum (Trin.) A. Camus
(commonly named Japanese stiltgrass or Nepalese browntop) to measure associated
changes in soil C-cycling and C pools. Microstegium vimineum is an annual C4 grass,
which produces relatively little root biomass, and invades the understory of a variety of
forest types across a large portion of the southern and eastern United States (U.S.). The
U.S. Department of Agriculture reports M. vimineum invasions in 25 states including all
states east of the Mississippi River except Vermont, New Hampshire and Maine
(http:/plants.usda.gov/). Research in northeastern U.S. forests demonstrates that M.
vimineum alters soil properties, including the activities of microbial exoenzymes that may
affect soil C-cycling (Kourtev et al. 1998; Ehrenfeld et al. 2001; Kourtev et al. 2002). As
yet no work has examined how M. vimineum invasion impacts belowground C pools. In
fact, little is known generally about the impact of grass invasions on soil C in forest
ecosystems (Kourtev et al. 2003; Mack & D'Antonio 2003; Litton et al. 2008). A recent
meta-analysis of invasive plant impacts on C and N cycling shows only 23% of 94
studies measured some form of belowground C, and of those only two studies looked at
the effect of grass invasion on belowground C pools in forests (Liao et al. 2008). The first
of these studies examined how perennial grass replacement of forest, due to an increase
in fire occurrence, impacted soil C (Mack & D'Antonio 2003). The second, a laboratory
study, compared changes in soil C between invasive understory species (including M.
vimineum) and a single native understory species (Kourtev et al. 2003). Mack &
D'Antonio (2003) found no significant difference in mineralizable soil C between forest
sites where an exotic grass had and had not been removed but did note a greater
percentage of mineralizable soil C in the invaded sites during the dry season. Kourtev et
105
al. (2003) found that soils planted with M. vimineum when compared to soils planted with
a native understory species had a greater percentage of soil organic matter but whether
this difference held for an intact forest invaded by M. vimineum was not determined. A
third and more recent study (Litton et al. 2008) assessed the impacts of invasive grass
removal on soil C. The researchers found decreased rates of soil C flux for forests where
the invasive grass was removed but did not find changes in soil C pools between removal
and invaded plots. Despite the ubiquity of M. vimineum invasions of eastern U.S. forest
understories and recent calls for research investigating understory invaders (Martin et al.
2009), as far as we are aware there has been no assessment of the impacts on soil C pools
of intact, forest ecosystems where uninvaded and grass-invaded areas are compared.
What M. vimineum‘s impact on belowground C pools in an intact forest landscape
will be is unknown, but because the plant has invaded forests over a large portion of the
eastern U.S. understanding its impact is important (Morrison et al. 2007). The objectives
of this study were to compare the size of multiple soil C and N pools across an active M.
vimineum invasion front, measure the contribution of native plant species and M.
vimineum C to soil C pools, and evaluate differences in soil C pools across the invasion
front with changes in soil microbial activity. In our study system C inputs from M.
vimineum can be distinguished from native inputs using stable C isotope ratios because
M. vimineum is the only species to use the C4 photosynthetic pathway. Our overarching
hypothesis was that M. vimineum invasion would alter belowground carbon cycling and
we expected this alteration to be explained by one of two hypotheses (H1 or H2). We
expected that in invaded plots either the predictions of (H1) preferential substrate
utilization theory would be observed through no change or an increase in native plant-
106
derived, soil C pools, or (H2) priming effect theory would be observed by decreases in the
same soil C pools. A significant strength of our study is the investigation of an active
invasion front in an intact forest ecosystem. At the same time this ‗observational‘
approach means we need to consider alternative mechanisms that may contribute to those
responses we document; we explore these through additional measurements and in the
Discussion.
Methods
Site description
Twelve 2 m × 2 m study plots divided into 6 pairs (1 uninvaded and 1 invaded) were
established across the edges of a rapidly progressing M. vimineum invasion front in a
riparian forest within the Whitehall Experimental Forest, Athens, GA, USA (N33º53.27‘
W 83º21.93‘). The site is a former wood lot, managed as mixed hardwood for the past 70
years (Rhett Jackson pers. comm.). There is little evidence that these particular sites were
actually farmed but much if not all of the surrounding land was (Rhett Jackson pers.
comm.). Soils at this site are a sandy loam in the Madison series (Soil Survey Staff 2009).
Bulk density was consistent across the site and was 1.08 g cm-3
. Anecdotal reports
indicate M. vimineum became apparent within the Whitehall Experimental Forest ~15
years ago, but we do not know the exact time of invasion for these study sites but given
their size they are likely < 15 years old. It is likely that M. vimineum went largely
unnoticed until it formed dense lawns and so pinpointing the exact invasion date is
difficult (Martin et al. 2009). Notably, the invaded plot in each pair appeared to have
developed from a discrete, invaded patch and remained discrete while our work was
107
conducted in 2006 and 2007. In the 2008 growing season these patches were joining-up
to form a contiguous cover of M. vimineum. No confounding issues related to invasive
earthworms were noted between invaded and uninvaded sites (see Results). The forest
overstory was composed of Acer rubrum, Quercus nigra, Platanus occidentalis, and
Liquidambar styraciflua. The uninvaded areas of the study site were generally
depauperate in understory plants. The understory plant community was composed of <
5% cover of Vitis rotundifolia, Acer negundo, A. rubrum, Q. nigra, Parthenocissus
quincifolia, Carex sp. Ligustrum sinense, Lonicera japonica, Smilax bona-nox,
Toxicodendron radicans, Euonymous sp., Chasmanthium latifolium, Bignonia capriolata,
and Aster sp. (unpub. data). In the invaded areas, M. vimineum covered >90% of the
understory on average with an average dry green biomass of ~63.04 g m-2
(unpub. data).
M. vimineum began to invade the uninvaded plots towards the end of our study in
2007 and in 2008 had invaded all of these control plots. This suggests that our uninvaded
study plots were hospitable to M. vimineum and the plant had simply not invaded these
areas yet. This is an important strength of our study. We deliberately worked across an
advancing invasion to minimize the likelihood that invaded and uninvaded areas might
exhibit differences in soil C cycling because of an historical factor other than the M.
vimineum invasion. We do recognize, however, that our study design is not a controlled
experiment and our inferences should be interpreted with this in mind. Further, although
our sample plots are spatially independent and our results pronounced, they are limited to
the description of a single forest site. Whether they can be generalized to other similar
sites or to other grass invasions of forest understories is not yet known.
108
Standard sampling regime
Standard sampling consisted of taking three individual A horizon soil cores (8 cm dia., 0
- 10 cm depth) from each plot using a stratified random approach. Soils were sieved (4
mm), homogenized, and stored at +5ºC until analyzed. Standard samples were taken six
times between September 2006 and July 2007 (12 September 2006, 13 November 2006,
19 January 2007, 14 February 2007, 8 May 2007, and 24 July 2007). Microstegium
vimineum was actively growing during the sampling months of September, May, and July
but had senesced by November and did not germinate prior to the January and February
samplings.
Measurements taken during the seasonal sampling regime consisted of
gravimetric soil moisture, pH, soil temperature (taken at 10 cm depth), soil CO2 efflux,
substrate induced respiration (SIR; a measure of microbial biomass), and mineralizable C
(a measure of microbially-available C). Measurements of both moisture and pH were
conducted on two analytical repeats per sample. Gravimetric soil moisture was
determined by drying a soil sub-sample at 105ºC for 24 h and pH (1 : 1, soil : H2O by
volume) was determined using a bench-top pH meter. Soil CO2 efflux, which is a
measure of both heterotrophic and autotrophic respiration, was taken in the field using an
infrared gas analyzer (IRGA; Li-Cor Biosciences, Lincoln, NE, USA, Model LI-8100).
Specifically, given the small plot size, one PVC collar was used per plot (20 cm dia.,
inserted 5 cm into the soil), and one measurement per plot was taken during the late
afternoon. Collars were placed on each sample day and measurements were taken at a
minimum of 1 h after placement. The SIR method we used follows Fierer & Schimel
109
(2003) whereby soil slurries are incubated, after a 1 h pre-incubation with excess
substrate, for 4 h at 20ºC.
Mineralizable C was determined using 60-day C-mineralization assays and is
often used to estimate the pool size of labile C in the soil. This is similar to measurements
of dissolved organic carbon (DOC; see below) but the mineralizable C assays provide a
more direct measure of microbially available C than do measurements of DOC (e.g.
Bradford et al. 2008b). Mineralizable C was determined by maintaining soils at 20C and
65% water-holding capacity (WHC) for 60 d with periodic determinations of respiration
rates using a static incubation technique (Fierer et al. 2005a) and infra-red gas analysis of
headspace CO2 concentrations. Mineralizable C was estimated as the area under the curve
derived by plotting CO2 production against time.
Intensive sampling regime
Measurements for the intensive sampling regime were taken before (12 September 2006)
and after (13 November 2006) M. vimineum had senesced. Due to the elongated growing
season in the southeastern United States, M. vimineum biomass was still green in
September. For the intensive sampling regime, we measured the mineral-associated and
particulate organic matter (POM) C and N pools, the DOC pool, CFE microbial biomass
C and N, and extractable inorganic and organic N. This fractionation approach, whereby
we resolved C pools with different turnover times, provided a powerful approach for
detecting changes in soil C that a single blanket measure of total soil C might have
obscured (see Bradford et al. 2008b; Bradford et al. 2008c).
110
To determine the mineral-associated and POM C and N pools, the fractionation
method described in Bradford et al. (2008c) was used. Briefly, duplicate soil samples (10
g of air-dry soil) from each site were dispersed with NaHMP (30 mL sample-1
) via
shaking for 18 h and then passed through a 53 m sieve. Material < 53 m is considered
mineral-associated and material > 53 m is considered POM. Both mineral and POM
material were dried (105C), ball-milled to a fine powder, and percentage C and N
determined using an NA1500 CHN Analyser (Carlo Erba Strumentazione, Milan, Italy).
Of these two fractions, mineral-associated C pools are expected to have slower turnover
times than are POM C pools (Schlesinger & Lichter 2001). Mineral-associated C pools
are presumed to be primarily microbial-derived C whereas POM pools are presumed to
be primarily plant-derived C (Grandy & Robertson 2007). Given the multi-annual
turnover times of these pools we only measured them during the September sampling; i.e.
further changes in pool size would not have been observed across the time course of our
study.
During both the September and November sampling, DOC and organic and
inorganic N pools were determined on two analytical repeats per sample. Samples were
shaken with 0.5 M K2SO4 for 4 h, filtered, and DOC concentrations determined using a
Total Organic Carbon Analyzer (Shimadzu, Columbia, USA). Dissolved organic N
(DON) and inorganic N pools were quantified on a Lachat autoanalyzer (Milwaukee, WI,
USA). The size of the DOC pool is often determined because it is one estimate of labile
soil C (Bradford et al. 2008c). However, although its turnover time is more rapid than
either POM or mineral-associated C pools, DOC is a mixture of both high and low
111
molecular-weight C compounds and so is likely less labile than the C pool resolved using
our 60-day mineralization assays described above.
Microbial biomass C and N was estimated using a modified chloroform
fumigation-extraction (CFE) method as described in Fierer & Schimel (2002) and Fierer
& Schimel (2003). Microbial biomass C or N was estimated as the flush of DOC or
DON, respectively, following fumigation. Raw values are reported for microbial biomass
C and N; no correction factors are used. By determining microbial biomass C using the
CFE method, we were able to determine the proportion of microbial biomass which was
derived from M. vimineum (see below). Note that the relationship between SIR and the
CFE method for microbial biomass is not necessarily one to one (Wardle & Ghani 1995).
For example, where CFE is expected to estimate total biomass, SIR is suspected to
measure active microbial biomass (Wardle & Ghani 1995). We used the SIR method for
the standard seasonal sampling because it likely provides greater resolution of differences
in soil microbial biomass within a site (Wardle & Ghani 1995). It could not though
provide the information on native versus invasive-derived plant C in the microbial
biomass that the CFE method could.
Finally, during intensive sampling we determined earthworm biomass, litter
chemistry, and the fungal-to-bacterial dominance of the microbial community. M.
vimineum invasion is linked to increased nonnative earthworm biomass in northeastern U.
S. forests (Kourtev et al. 1998; Kourtev et al. 1999; Nuzzo et al. 2009) so we wanted to
account for this potential confounding issue in our study. Earthworm biomass was
determined by hand-extracting earthworms from 20 cm dia., 20 cm deep soil cores.
Hand-extracting is commonly considered to be the most accurate sampling technique for
112
earthworm biomass (Lee 1985; Eisenhauer et al. 2008), which we report here as wet
mass.
Total percentage C, nitrogen (N), cellulose, hemi-cellulose, and lignin were
determined for native forest litter, M. vimineum leaf material, and M. vimineum stem
material (Table 3.1). We also determined the ratio of leaf-to-stem material for M.
vimineum which allowed us to determine whole plant litter chemistry. Total C and N
were determined using an NA1500 CHN Analyzer (Carlo Erba Strumentazione, Milan,
Italy). Fiber concentrations were determined using an Ankom A200 Fiber Analyzer
(Ankom, Macedon, USA). Quantitative inputs of M. vimineum litter and native forest
litter were assessed by taking 0.25 0.25 m quadrats from each plot post-senescence (i.e.
November), drying the sample at 65ºC, hand-sorting the material, and determining its
mass (Table 3.1).
Fungal-to-bacterial dominance of the microbial community is often taken as an
indicator of belowground C processes with fungal-dominated systems being associated
with greater C stores. We determined these ratios using the quantitative PCR (qPCR)
method described by Fierer et al. (2005b) and Lauber et al. (2008). Briefly, DNA was
isolated from soil (kept at -80ºC until use) using the MoBIO Power Soil DNA Extraction
kit (MoBio Laboratories, Carlsbad, CA, USA) with modifications described in Lauber et
al. (2008). Standard curves were constructed to estimate bacterial and fungal small-
subunit rRNA gene abundances. After soil DNA concentrations were normalized, PCR
reactions were carried out which allowed us to generate the ratios of fungal to bacterial
gene copy numbers by using a regression equation for each assay relating the cycle
threshold (Ct) value to the known number of copies in the standards. All qPCR reactions
113
were run in quadruplicate. Additional details, including the specific PCR conditions, are
provided in Lauber et al. (2008).
Determination of M. vimineum derived C
To establish the amount of C derived from M. vimineum, we determined the δ13
C value of
the following C pools: mineralizable C, CFE microbial biomass, DOC, total soil C, POM
C, mineral-associated C, and the litter layer. For mineralizable C, a gas sub-sample taken
on the first day of each 60-day incubation was analyzed. The δ13
C value of the CO2 in the
sample was determined using continuous-flow isotope-ratio mass spectrometry (IRMS;
Thermo, San Jose, CA, USA). For the DOC and CFE biomass a TOC analyzer was
coupled to the IRMS and the remaining solid samples were introduced to the IRMS via
an elemental analyzer. Once δ13
C values were determined, we were able to estimate the
amount of M. vimineum derived C in each of the C pools by using the difference in the
natural abundance, δ13
C values of either the plant or soil C pools. The pre-invasion δ13
C
value for each C pool was determined from the plots where M. vimineum was absent and
due to an absence of any native C4-plant species these pools all exhibited a C3-
photosynthetic value which ranged from -27.72 ± 0.25 ‰ (mean ± 1SE; n =6) for native
POM C to -24.29 ± 0.32 ‰ (mean ± 1SE; n =6) for the September DOC measurement.
M. vimineum has a C4-photosynthetic value (n =13; mean ± 1SE = -14.31 ± 0.14 ‰). The
difference in the C isotope composition between M. vimineum and native C is sufficient
to discriminate sources (Staddon 2004). The amount of C derived from M. vimineum was
calculated as follows(sensu Ineson et al. 1996): CM. vimineum derived = Cpool × (δ13
Cinvaded -
δ13
Cuninvaded)/( δ13
CM. vimineum - δ13
Cuninvaded), where Cpool is the measured size of the pool
114
(total, POM, Mineral, DOC, Microbial C, mineralizable, or litter), δ13
Cinvaded is the δ13
C
value of the pool in plots where M. vimineum is present, δ13
Cuninvaded is the δ13
C value of
the pool in plots where M. vimineum is absent, and δ13
CM. vimineum is the value for M.
vimineum itself.
13C-glucose pulse-chase
To determine whether or not the presence of M. vimineum was associated with more
rapidly cycling carbon pools, we conducted a 13
C pulse-chase in November 2006. The
pulse-chase was conducted by making additions of 13
C-labeled glucose and tracking its
mineralization as 13
CO2. This was accomplished by placing two PVC collars (15.4 cm
dia., inserted 5 cm into the soil) in three plots where M. vimineum was present and 3 plots
where it was absent. Water was then added to each core 24 h prior to sampling in order to
alleviate any water stress between plots. Soil CO2 efflux rates were determined using a
closed-chamber approach (e.g. Bradford et al. 2001), where CO2 concentrations were
determined at the start and end of a 45 min capping period. We conducted a pilot study to
determine the appropriate capping time and found that headspace CO2 concentrations
increase linearly from 0 to 45 min; flux rate estimates only began to decrease after 60
min. Although this closed-chamber approach is likely to have caveats associated with it
(i.e. underestimated CO2 efflux rates), it is nonetheless a widely used approach across an
array of ecosystem types (Nay et al. 1994; Franzluebbers et al. 2002). Similar capping
times and protocols as described here have been employed by other researchers (Iqbal et
al. 2008; Mo et al. 2008). Headspace samples were taken with 20 mL SGE gas syringes,
transported to the laboratory in 12 mL Exetainers, and then CO2 concentrations were
115
determined using an IRGA (Li-Cor Biosciences, Lincoln, NE, USA, Model LI-7000). A
second sample was analyzed using continuous flow, IRMS to determine the stable C
isotope composition (see above for details). The initial headspace sampling provided the
natural abundance values for the isotope mixing equations. After this initial sampling, 1 L
of 2.5 mM 13
C-labeled glucose solution (99 atom %) was added to the collars and
permitted to drain. The capping procedure was repeated post addition at 2, 5, 24, 48 and
72 h, permitting a negative exponential ‗decay‘ of 13
C label to be tracked in the soil CO2
efflux, from which cumulative mineralization rates were estimated. The contribution of
13C-labeled glucose to soil respiration was estimated using isotope mixing equations
similar to those described above.
Statistical analysis
Linear mixed-effects models were used to analyze the effect of M. vimineum
presence/absence (Pinheiro & Bates 2000). For this approach the presence of M.
vimineum and month (when applicable) were treated as fixed effects. Pair and plot were
treated as random effects with plot nested within pair. The pair identity was included as a
blocking variable to account for spatial heterogeneity among paired plots where M.
vimineum was present and absent. If the removal of this blocking variable resulted in a
more parsimonious model, as determined by AIC, then it was dropped from the analysis.
Plot identity was included as a random effect to account for repeated sampling across
months. When reported as such, data were loge-transformed to conform to assumptions of
homoscedasticity (verified using model checking). All analyses were conducted using the
freeware statistical package R (http://cran.r-project.org/). In all cases we considered
116
results statistically significant at P < 0.05 and marginally significant at P < 0.10. Notably,
it is acceptable practice to consider changes in soil C pools at P < 0.10 to be statistically
significant because of the large spatial variation associated with them (e.g. Carney et al.
2007). However, we took a more conservative stance in this work.
Results
Of the standard measures taken at six time points spanning from September 2006 to July
2007, we observed an average decrease in mineralizable C of 35% in the presence of M.
vimineum (F1,5=14.05; P<0.05; Fig. 3.1a). Average declines in mineralizable C of 36, 18,
36, 39, 43, and 37% during the months of September, November, January, February,
May, and July, respectively, were noted when the invader was present (Fig. 3.1a). A
significant effect of sampling month was also detected (F5,50=2.59; P<0.05) but the
relative difference between sites where M. vimineum was present and where it was absent
remained the same across all six sampling points (interaction: F5,50=1.29; P<0.28).
Furthermore, of the initial mineralizable C (i.e. CO2 evolved during the first of 60
incubation days), M. vimineum derived inputs accounted for approximately 10% on
average across all sampling months (Fig. 3.1a). A high of approximately 20% was found
for the May sampling and a low of approximately 4% was found for the November
sampling (Fig. 3.1a). The presence of M. vimineum was not associated with a change in
SIR biomass (F1,10=3.10; P=0.11; Fig. 3.1b).
The presence of M. vimineum did not affect soil CO2 efflux (F1,10=1.66; P=0.23;
Fig. 3.2), but did interact with soil moisture (F5,50=3.30; P<0.05), temperature
(F5,50=2.55; P<0.05), and pH (F5,50=4.17; P<0.01). Soil CO2 efflux did vary across
117
sampling month (F5,50=63.8; P<0.001; Fig. 3.2) with greater efflux during the growing
season (i.e. September, May, and July) and lower efflux when plant species were not
actively growing (i.e. November, January, and February). Soil moisture tended to be
seasonally more stable in the presence of M. vimineum than in its absence. Soil
temperature tended to be lower in the presence of M. vimineum during the months of
November, January, and February but during the months of September, May, and July
sites where it was present had similar or higher temperatures to sites where it was absent.
In general, pH was typical higher in the presence of M. vimineum. This was dependent on
sampling month with higher pH values for soils where M. vimineum was present during
September, January, and May but similar pH values regardless of M. vimineum
presence/absence for the months of November, February, and July.
Several soil C and N pools were intensively sampled in September and/or
November 2006. Of the soil C pools, we noted a significant decrease of 64% and 41%
(September and November, respectively) in CFE microbial biomass when M. vimineum
was present (P<0.05; Table 3.2). Regardless of the invader‘s status microbial biomass
was marginally greater post-senescence (F1,10=3.69; P=0.08). We also noted marginally
significant declines of 22% and 29%, respectively in both total soil organic C (P=0.09)
and POM-associated C (P=0.08; Table 3.2). The marginally significant decrease in total
soil C in the presence of M. vimineum was likely driven by the decrease in POM C given
that total C is the sum of POM and mineral-associated C. No significant change in
mineral-associated soil C was noted, although the pool size was lower in invaded plots
(P=0.13; Table 3.2). Microstegium vimineum presence also was not associated with a
significant effect on DOC but again this pool was lower where the invader was present
118
(P=0.19; Table 3.2). There were no obvious differences in any of the measured N pools
(P>0.12 in all cases).
Inputs derived from M. vimineum contributed between 1 and 29% of the C found
in the belowground pools of invaded plots (Table 3.2). These inputs did not, however,
compensate fully for the observed declines in several native plant-derived C pools (Table
3.2). Specifically we noted significant declines in both native-derived CFE microbial
biomass C (P<0.05) and POM-associated C (P<0.05), as well as marginally significant
declines in native-derived total soil organic C (P =0.07) and DOC (P =0.10) in the
presence of M. vimineum (Table 3.2). The presence of M. vimineum was associated with a
72% decline in native derived CFE microbial biomass C in September and a 43% decline
in November (Table 3.2). Overall, native-derived CFE microbial biomass C tended to
increase post-senescence in November (F1,10=5.40; P<0.05). Microstegium vimineum‘s
presence was also associated with a 25% decline in native derived DOC (P=0.10; Table
3.2) and this was independent of sampling month (F1,10=0.00; P=0.93). Native derived
POM-associated soil C was 34% lower and total soil organic C was 24% lower in the
presence of M. vimineum but there was no statistical support for changes in the mineral-
associated soil C, albeit this pool was lower on average in invaded plots (Table 3.2).
Of the remaining measures taken in September and November 2006, no
statistically significant differences in earthworm biomass m-2
were detected between
invaded and uninvaded plots (F1,10=0.04; P=0.84), nor in fungal-to-bacterial ratios
(F1,10=0.00; P=0.99). In contrast, the cumulative amount of 13
C-glucose mineralized to
13CO2 was lower in sites where M. vimineum was absent and higher in sites where it was
present (F1,2=19.6; P<0.05; Fig. 3.3). Analysis of the time-series data showed that 13
C-
119
glucose mineralized to 13
CO2 generally followed a pattern of exponential decay across the
72 h sampling period. The significant difference in cumulative values between sites was
likely due to higher rates of 13
C-glucose mineralized to 13
CO2 observed in the plots
invaded by M. vimineum during the 2 h and 5 h sampling periods (Fig. 3.3).
Discussion
We evaluated whether the abundant invader, M. vimineum, impacted belowground C
pools. We expected that if it impacted belowground C pools, the mechanism could likely
be framed around the theories of preferential substrate utilization or priming. If
preferential substrate utilization was the mechanism then our expectation (H1) was that
there would be either no change or an increase in belowground C pools, meaning that soil
C turnover rates would likely remain the same or decrease (Fontaine et al. 2004a;
Fontaine & Barot 2005; Bradford et al. 2008c). In contrast, if a priming effect (H2) is the
potential mechanism then our expectation was that M. vimineum‘s presence would be
associated with a decrease in the size of soil C pools and an increase in their turnover
rates (Fontaine et al. 2004b). We observed that the presence of M. vimineum was
associated with increased C turnover rates and declines in several belowground C pools.
Our findings suggest that M. vimineum invasion is altering soil C cycling via a priming
effect. We do, however, recognize that our study was observational and so causation
could be attributed to other response variables. Given that both moisture and temperature
were neither consistently higher nor lower in invaded sites and earthworm biomass was
similar across invaded and uninvaded plots we discuss below why priming may be the
120
most plausible explanation and contrast our results with other studies that appear to have
found preferential substrate utilization manifesting in grass-invaded forest soils.
We found a marginal decrease in both total organic and POM C associated with
the invasion but mineral-associated C was not lower (Table 3.2). The decrease in total
soil C appeared to be largely an effect of the decrease in POM C, which is part of the
total C pool. The reason that a decrease in POM C was observed while no statistically
measurable decrease in mineral-associated C was observed may be due to the greater
stability of the mineral pool and its slower turnover rates (Schlesinger & Lichter 2001;
Grandy & Robertson 2007; Bradford et al. 2008c). These declines in both POM and total
soil C represent an average 29 and 22% decrease under M. vimineum. If expressed on an
areal basis, this would be equivalent to a decline of ~375 g m-2
of total C which
represents a decrease of ~25 g m-2
year-1
, conservatively assuming M. vimineum has been
present at this site for 15 years. We recognize that we are basing some of our inferences
on statistically marginal differences (see Table 3.1). However, we think this is legitimate
given that a power analysis indicated the likelihood of detecting a significant effect
associated with a 22% total or 29% POM C decline was extremely low (0.18 and 0.29,
respectively) and even lower power would be associated with smaller percentage declines
(e.g. a 10% decline is associated with a power of 0.08). In both instances to achieve the
recommended power of 0.80 would have required a sample size of > 40 and > 20
replicates for total and POM C, respectively (Thomas 1997).
The lack of power to detect significant changes when measuring soil C pools,
given their large spatial variation at fine scales, is a common issue (Throop & Archer
2008). Recent work by Saby et al. (2008) indicates that the ability to detect changes in
121
soil C pools is largely site specific and that annual changes in these pools are extremely
difficult to detect. This may have prompted the acceptance of studies conducted at fine
spatial and temporal scales to deem changes in soil C pools at P<0.10 significant (e.g.
Carney et al., 2007). We followed the more conventional approach of P<0.05 and
recognize this may be overly conservative. Given the widespread occurrence of M.
vimineum invasion and a call for increased emphasis on understory invaders of intact
forest systems (Martin et al. 2009), clearly there is a need for more studies such as ours
that assess the impacts of invasion on forest soil C pools. Use of fractionation and stable
isotope approaches will increase the ability to resolve changes in soil C stocks and
cycling. In conducting future studies, investigators need be aware that the power to detect
statistically significant changes in soil C pools is generally below recommended values
when logistically feasible replicate numbers are used. The failure to detect statistically
significant effects should not simply be interpreted as a null effect of grass invasion on
forest soil C pools. Undoubtedly, more studies and then meta-analysis of their collective
results will indicate the significance, direction, and size of the effect on soil C associated
with grass invasion of forest understories. The use of meta-analysis to explore the result
of multiple studies using relatively small sample sizes is successfully and widely-adopted
in medical fields (Gurevitch & Hedges 1999).
After accounting for the contribution of M. vimineum-derived C in invaded plots
there was relatively little change in the contribution of native-derived C to the mineral
pool (Table 3.2). Microstegium vimineum-derived C only accounted for ~1% of mineral-
associated C, which again is indicative of the greater stability of this pool and its slower
turnover time (Schlesinger & Lichter 2001; Grandy & Robertson 2007; Bradford et al.
122
2008c). It may also suggest that a lesser proportion of M. vimineum-derived C when
compared to native-derived sources is microbially processed to the extent that it becomes
stabilized in the mineral pool and more C derived from this invader may be lost as CO2.
This is simply speculation based in part on the knowledge that mineral-associated C may
often be largely microbially-derived (Grandy & Robertson 2007). Alternatively, it may
be that M. vimineum has not been present (i.e. <15 years) at this site long enough to affect
the mineral-associated C pool to an extent that marked differences can be resolved;
similar explanations have been suggested by others (e.g. Litton et al., 2008).
In contrast to the mineral-associated C pool, C derived from native sources found
in total organic and POM C were lower in the presence of M. vimineum. Although the
POM C pool was smaller in the presence of M. vimineum, the invader accounted for ~6%
of this pool (compared to 1% for the mineral-associated C pool). Thus, most of the
decline in the POM C pool is from the loss of native-derived sources. Declines in native-
derived POM C may indicate that native-derived C is cycling faster in plots where the
invader is present due to invader inputs priming its decomposition. Furthermore, native-
derived POM C decreased by ~34% and given that M. vimineum-derived C only accounts
for ~6% of this pool then further declines in this pool may be expected. This outcome
will be largely contingent on whether or not the declines in native-derived POM C have
stabilized and/or inputs derived from M. vimineum remain the same. Future research will
need to address these possibilities if we are to understand the long term impacts of this
invader on soil C pools.
Total DOC (i.e. both native and M. vimineum derived) was not significantly lower
in the presence of M. vimineum. However, DOC derived from native sources was
123
marginally lower in the presence of M. vimineum. Interestingly, DOC derived from M.
vimineum was greater on average pre-senescence than post-senescence (Table 3.2). This
may indicate that rhizodeposition from M. vimineum is an important source of C inputs to
soils, despite their small root to shoot ratios (unpub. data). This possibility remains to be
tested.
Like DOC, DON was not significantly impacted by the presence of M. vimineum
and we found no significant difference in inorganic N pools. This last result contrasts
with other studies conducted in the northeastern U.S. which showed that the presence of
M. vimineum was associated with changes in available NO3- and nitrification rates
(Kourtev et al. 1999; Ehrenfeld 2003; Kourtev et al. 2003). We found that M. vimineum
was not associated with a statistically significant change in available NO3- although it is
worth noting that pH, which typically increases when NO3- is the dominant form of plant
available N (Nye 1981; Ehrenfeld 2003), was significantly greater in the invader‘s
presence. One possible reason for the contrasting results between our study and studies
conducted in northeastern forests may be the association between non-native earthworm
biomass and M. vimineum. Non-native earthworm biomass is often associated with
altered N availability and mineralization in soils (Scheu 1987; Kourtev et al. 1999;
Eriksen-Hamel & Whalen 2008) and may have contributed to changes in N pools in
northeastern US forests invaded by M. vimineum (Bohlen et al. 2004; Marhan & Scheu
2006; Eisenhauer et al. 2007). Given that there was no difference in earthworm biomass
between our plots, we were able to explore the effect of M. vimineum in the presence of a
uniform earthworm gradient.
124
Extractable microbial biomass C, both total and native-derived, was found to be
significantly lower in plots invaded by M. vimineum. Microbial biomass C was
determined using a modified CFE method and is expected to relate to the total active and
inactive biomass (Wardle & Ghani 1995; Bradford et al. 2008c). Declines in microbial
biomass C have often been associated with declines in soil C pools (Wardle & Ghani
1995; Bradford et al. 2008a; Bradford et al. 2008b). Since microbial biomass C is likely
to be one of the more responsive belowground C pools then its decrease may be an
indicator of an overall decline in soil C associated with M. vimineum invasion. We
observed that a greater proportion of microbial biomass C was derived from M. vimineum
pre-senescence than post-senescence (Table 3.2). Under actively growing M. vimineum,
the microbial community derived ~30% of its C from the invader (Table 3.2). This is a
remarkable percentage because it is much larger than the aboveground biomass of M.
vimineum at the study site (relative to native plant biomass), and M. vimineum also has a
rather superficial root system. That C derived from M. vimineum in the microbial biomass
was much higher pre-senescence may suggest that the microbial community is attaining
the bulk of M. vimineum-derived C via root exudates rather than foliar litter. Root
exudates are in part composed of an array of low-molecular weight compounds, such as
glucose, which are expected to be highly labile (van Hees et al. 2005). Such compounds
have been shown to cause priming effects when applied to soils and may be the direct
link between M. vimineum invasion and associated declines in soil C (Dalenberg & Jager
1981; Blagodatskaya et al. 2007; Carney et al. 2007). Further work is necessary to
establish this linkage between M. vimineum root exudates, the microbial community, and
declines in soil C pools.
125
Although the presence of M. vimineum was associated with a decrease in
microbial biomass C, no change in the fungal-to-bacterial dominance of the microbial
community was found. This result was somewhat unexpected because a shift toward
bacterial dominance might be expected, due to this invasion, since M. vimineum
represents inputs of higher chemical quality (Table 3.1) and its presence was also
associated with a more neutral pH. Both factors are expected to favor bacteria over fungi
(Bardgett & McAlister 1999; Six et al. 2006; van der Heijden et al. 2008). Changes in
fungal-to-bacterial dominance may not have had long enough to occur (Bardgett et al.
1996; Bardgett & McAlister 1999) but some studies have noted changes across short time
intervals (Carney et al. 2007). Another possibility is that fungal-to-bacterial dominance
may be too coarse an estimate of microbial community structure in this case. A finer
measure of the community may have shown differences associated with M. vimineum
which have been reported elsewhere for this species (Ehrenfeld et al. 2001; Kourtev et al.
2002). Indeed, evidence is growing that specific taxa within fungal and bacterial
communities are associated with specific ecological strategies, such as r- versus K-
strategists (Fierer et al. 2007; Lauber et al. 2008). Shifts in the community at this level
would not necessarily be detected simply by determining fungal-to-bacterial ratios. Such
a shift (below the Kingdom level) might explain why we observed a decrease in the size
of the microbial biomass pool but saw no concomitant decline in SIR biomass (Table 3.2;
Fig. 3.1b). SIR biomass differs from CFE biomass because it may be more indicative of
physiologically active microbial biomass rather than absolute biomass (Wardle & Ghani
1995; Bradford et al. 2008c). If true, then on a per unit biomass basis, the microbial
community associated with M. vimineum was more active.
126
The mineralizable C pool, a measure of microbially available C (Bradford et al.
2008c), was significantly lower when associated with M. vimineum invasion (Fig. 3.1a).
The decrease in the size of this pool under M. vimineum may indicate that it is cycling
more rapidly. The soil CO2 efflux data also support the idea that C is cycling more
rapidly under M. vimineum invasion. Indeed, soil CO2 efflux did not decrease under M.
vimineum but yet declines in mineralizable C as well as other soil C pools were observed
(Fig. 3.2; Table 3.2). This discrepancy may be explained by more rapid turnover of
belowground C pools (assuming autotrophic respiration is unchanged). That is, if soil C
pools were simply smaller and turnover rates were unchanged then soil CO2 efflux would
be smaller in invaded plots. Instead, soil CO2 efflux rates were equivalent in the presence
and absence of M. vimineum, indicating that labile C pools under M. vimineum may be
turning over more rapidly (i.e. C atoms have reduced residence times in the soil in
invaded plots). This scenario is synonymous with depletion of fast-cycling soil C pools
under experimental soil warming, where soil CO2 efflux rates in control and warmed
plots are equivalent when plots reach equilibrium due to faster cycling of active soil C
pools in the warmed plots (Kirschbaum 2004; Knorr et al. 2005; Bradford et al. 2008a).
The possibility that an increase in autotrophic respiration is associated with M. vimineum
presence seems unlikely given that, if this were the case, soil CO2 efflux in invaded plots
should have been lower than in uninvaded plots once M. vimineum had senesced, but it
was not. Although, if after M. vimineum senesced decomposition increased due to the
invader‘s litter inputs then the lack of a difference in CO2 efflux between plots might be
attributed to increased autotrophic respiration pre-senescence and increased heterotrophic
respiration post-senescence. To help decipher the mechanism we conducted an in situ
127
pulse-chase with 13
C-glucose. The more rapid mineralization of this added substrate
supports our inference that soil C is turning over more rapidly in plots invaded by M.
vimineum. Specifically, when M. vimineum was present more glucose was mineralized to
CO2 during the first 72 h after addition (Fig. 3.3). Notably the difference in cumulative
values was largely driven by higher rates of glucose mineralization during the first five
hours after glucose addition (Fig. 3.3).
The results of our study show that the presence of M. vimineum was associated
with declines in several belowground C pools and these declines were likely associated
with more rapid cycling. Microstegium vimineum represents a relatively minor
quantitative detritus input (< 3% of litter layer biomass; Table 3.1) at our site but the
higher chemical quality of M. vimineum litter (Table 3.1), in addition to its root exudates,
appears to stimulate soil C decomposition through a possible priming effect. Other
studies of invasive plant effects on belowground C pools, similar in context to our own
(Stock et al. 1995; Mack et al. 2001; Mack & D'Antonio 2003; Bradley et al. 2006), have
observed relatively little change in soil C pools (Hook et al. 2004; Valery et al. 2004;
Koutika et al. 2007; Litton et al. 2008). We speculate that what appear to be
inconsistencies between our study results and those of others may actually be consistent
in the context of preferential substrate utilization versus priming effects. That is, the
impact that an invasive plant species has on belowground C pools may be dependent on
the quality and quantity of its C inputs relative to the quality and quantity of native
inputs. If invasive plant species‘ inputs are of high chemical quality (i.e. low C:N or low
Lignin:N) but low quantity relative to native inputs then a priming effect may occur
leading to a decrease in belowground carbon pools; however, if quantity is high then
128
preferential substrate utilization may instead occur. One key example of this is work
conducted by Litton et al. (2008). They found that grass invasion in Hawaiian forests
increased soil CO2 efflux, represented a very large increase in litter inputs (i.e. 16 to 44
fold increase), but did not impact soil C pools. Such results are at least indicative of
preferential substrate utilization. In our own study we found that the presence of M.
vimineum did not change soil CO2 efflux, its litter inputs only represented an ~3%
increase in total litter inputs, and we noted declines in several soil C pools. These results
are indicative of a priming effect and similar to other field based studies which have
found and concluded the same (Carney et al. 2007). Furthermore, when considering both
Litton et al.‘s (2008) study and our own there is some support for the proposition that the
strength of preferential substrate utilization versus priming may be dependent on the
relative increase in inputs (Blagodatskaya & Kuzyakov 2008; Bradford et al. 2008c).
This would be in agreement with Blagodatskya & Kuzyakov (2008) who found that
priming effects tended to decrease as relative inputs increase. It is also important to
recognize that effects may not be consistent across all belowground C pools, with both C
input rates and the nutrient content of these inputs affecting distinct soil C pools
differentially (Bradford et al. 2008c). By incorporating priming and preferential substrate
utilization theories with our current understanding of invasive plant species, we may be
better able to understand and predict how invasive plants are likely to impact soil C
cycling and hence long-term soil fertility.
In conclusion, we have demonstrated that at our study site M. vimineum invasion
is associated with declines in the size of several belowground C pools and C inputs
derived from M. vimineum do not offset the loss of native, plant-derived soil C. This loss
129
of C appears to be the result of a priming of the microbial community from higher quality
M. vimineum foliar litter inputs and root exudates, which leads to increased soil C
turnover rates where the invader is present. Whether our results can be generalized to M.
vimineum invasions at other similar sites is not yet known. Assuming they are
generalizable, then the results of this study indicate that invasions by M. vimineum may
have long term implications for forest soil fertility, carbon storage, and potentially the
flow of energy through detrital pathways to aboveground components of terrestrial
foodwebs.
Acknowledgements
We thank C.A. Davies for his assistance with the 13
C-glucose pulse-chase, N. Fierer and
C. Lauber for assistance with qPCR, and B.A. Snyder and M. Callaham for assistance
with earthworm sampling. This research was supported by the Andrew W. Mellon
Foundation and by National Science Foundation grants to the Coweeta LTER Program.
References
Allison S.D. & Vitousek P.M. (2004) Rapid nutrient cycling in leaf litter from invasive
plants in Hawai'i. Oecologia, 141, 612-619.
Ball B.A., Hunter M.D., Kominoski J.S., Swan C.M. & Bradford M.A. (2008)
Consequences of non-random species loss for decomposition dynamics:
Experimental evidence for additive and non-additive effects. Journal of Ecology,
96, 303-313.
130
Bardgett R.D., Hobbs P.J. & Frostegard A. (1996) Changes in soil fungal:bacterial
biomass ratios following reductions in the intensity of management of an upland
grassland. Biology and Fertility of Soils, 22, 261-264.
Bardgett R.D. & McAlister E. (1999) The measurement of soil fungal:bacterial biomass
ratios as an indicator of ecosystem self-regulation in temperate meadow
grasslands. Biology and Fertility of Soils, 29, 282-290.
Baruch Z. & Goldstein G. (1999) Leaf construction cost, nutrient concentration, and net
CO2 assimilation of native and invasive species in Hawaii. Oecologia, 121, 183-
192.
Batten K.M., Six J., Scow K.M. & Rillig M.C. (2005) Plant invasion of native grassland
on serpentine soils has no major effects upon selected physical and biological
properties. Soil Biology & Biochemistry, 37, 2277-2282.
Blagodatskaya E. & Kuzyakov Y. (2008) Mechanisms of real and apparent priming
effects and their dependence on soil microbial biomass and community structure:
critical review. Biology and Fertility of Soils, 45, 115-131.
Blagodatskaya E.V., Blagodatsky S.A., Anderson T.H. & Kuzyakou Y. (2007) Priming
effects in Chernozem induced by glucose and n in relation to microbial growth
strategies. Applied Soil Ecology, 37, 95-105.
Blumenthal D.M. (2006) Interactions between resource availability and enemy release in
plant invasion. Ecology Letters, 9, 887-895.
Bohlen P.J., Scheu S., Hale C.M., McLean M.A., Migge S., Groffman P.M. & Parkinson
D. (2004) Non-native invasive earthworms as agents of change in northern
temperate forests. Frontiers in Ecology and the Environment, 2, 427-435.
131
Bradford M.A., Davies C.A., Frey S.D., et al. (2008a) Thermal adaptation of soil
microbial respiration to elevated temperature. Ecology Letters, 11, 1316-1327.
Bradford M.A., Fierer N., Jackson R.B., Maddox T.R. & Reynolds J.F. (2008b)
Nonlinear root-derived carbon sequestration across a gradient of nitrogen and
phosphorous deposition in experimental mesocosms. Global Change Biology, 14,
1113-1124.
Bradford M.A., Fierer N. & Reynolds J.F. (2008c) Soil carbon stocks in experimental
mesocosms are dependent on the rate of labile carbon, nitrogen and phosphorous
inputs to soils. Functional Ecology, 22, 964-974.
Bradford M.A., Wookey P.A., Ineson P. & Lappin-Scott H.M. (2001) Controlling factors
and effects of chronic nitrogen and sulphur deposition on methane oxidation in a
temperate forest soil. Soil Biology & Biochemistry, 33, 93-102.
Bradley B.A., Houghtonw R.A., Mustard J.F. & Hamburg S.P. (2006) Invasive grass
reduces aboveground carbon stocks in shrublands of the western US. Global
Change Biology, 12, 1815-1822.
Carney K.M., Hungate B.A., Drake B.G. & Megonigal J.P. (2007) Altered soil microbial
community at elevated CO2 leads to loss of soil carbon. Proceedings of the
National Academy of Sciences of the United States of America, 104, 4990-4995.
D'Antonio C.M. & Corbin J. (2003) Effects of plant invaders on nutrient cycling: Using
models to explore the link between invasion and development of species effects.
In: Models in ecosystem science (eds. Canham CD, Cole JJ & Lauenroth WK),
pp. 363-384. Princeton University Press, Princeton, NJ, USA
132
Dalenberg J.W. & Jager G. (1981) Priming effect of small glucose additions to C14
-
labeled soil. Soil Biology & Biochemistry, 13, 219-223.
Dijkstra F.A. & Cheng W.X. (2007) Interactions between soil and tree roots accelerate
long-term soil carbon decomposition. Ecology Letters, 10, 1046-1053.
Ehrenfeld J.G. (2003) Effects of exotic plant invasions on soil nutrient cycling processes.
Ecosystems, 6, 503-523.
Ehrenfeld J.G., Kourtev P. & Huang W.Z. (2001) Changes in soil functions following
invasions of exotic understory plants in deciduous forests. Ecological
Applications, 11, 1287-1300.
Eisenhauer N., Straube D. & Scheu S. (2008) Efficiency of two widespread non-
destructive extraction methods under dry soil conditions for different ecological
earthworm groups. European Journal of Soil Biology, 44, 141-145.
Eisenhauer N., Partsch S., Parkinson D. & Scheu S. (2007) Invasion of a deciduous forest
by earthworms: Changes in soil chemistry, microflora, microarthropods and
vegetation. Soil Biology & Biochemistry, 39, 1099-1110.
Eriksen-Hamel N.S. & Whalen J.K. (2008) Earthworms, soil mineral nitrogen and forage
production in grass-based hayfields. Soil Biology & Biochemistry, 40, 1004-1010.
Fierer N., Bradford M.A. & Jackson R.B. (2007) Toward an ecological classification of
soil bacteria. Ecology, 88, 1354-1364.
Fierer N., Craine J.M., McLauchlan K. & Schimel J.P. (2005a) Litter quality and the
temperature sensitivity of decomposition. Ecology, 86, 320-326.
133
Fierer N., Jackson J.A., Vilgalys R. & Jackson R.B. (2005b) Assessment of soil microbial
community structure by use of taxon-specific quantitative PCR assays. Applied
and Environmental Microbiology, 71, 4117-4120.
Fierer N. & Schimel J.P. (2002) Effects of drying-rewetting frequency on soil carbon and
nitrogen transformations. Soil Biology & Biochemistry, 34, 777-787.
Fierer N. & Schimel J.P. (2003) A proposed mechanism for the pulse in carbon dioxide
production commonly observed following the rapid rewetting of a dry soil. Soil
Science Society of America Journal, 67, 798-805.
Fontaine S., Bardoux G., Abbadie L. & Mariotti A. (2004a) Carbon input to soil may
decrease soil carbon content. Ecology Letters, 7, 314-320.
Fontaine S., Bardoux G., Benest D., Verdier B., Mariotti A. & Abbadie L. (2004b)
Mechanisms of the priming effect in a savannah soil amended with cellulose. Soil
Science Society of America Journal, 68, 125-131.
Fontaine S. & Barot S. (2005) Size and functional diversity of microbe populations
control plant persistence and long-term soil carbon accumulation. Ecology
Letters, 8, 1075-1087.
Fontaine S., Mariotti A. & Abbadie L. (2003) The priming effect of organic matter: A
question of microbial competition? Soil Biology & Biochemistry, 35, 837-843.
Franzluebbers K., Franzluebbers A.J. & Jawson M.D. (2002) Environmental controls on
soil and whole-ecosystem respiration from a tallgrass prairie. Soil Science Society
of America Journal, 66, 254-262.
Grandy A.S. & Robertson G.P. (2007) Land-use intensity effects on soil organic carbon
accumulation rates and mechanisms. Ecosystems, 10, 58-73.
134
Gurevitch J. & Hedges L.V. (1999) Statistical issues in ecological meta-analyses.
Ecology, 80, 1142-1149.
Harguindeguy N.P., Blundo C.M., Gurvich D.E., Diaz S. & Cuevas E. (2008) More than
the sum of its parts? Assessing litter heterogeneity effects on the decomposition of
litter mixtures through leaf chemistry. Plant and Soil, 303, 151-159.
Hook P.B., Olson B.E. & Wraith J.M. (2004) Effects of the invasive forb Centaurea
maculosa on grassland carbon and nitrogen pools in Montana, USA. Ecosystems,
7, 686-694.
Ineson P., Cotrufo M.F., Bol R., Harkness D.D. & Blum H. (1996) Quantification of soil
carbon inputs under elevated CO2: C3 plants in a C4 soil. Plant and Soil, 187, 345-
350.
Iqbal J., Hu R.G., Du L.J., Lan L., Shan L., Tao C. & Ruan L.L. (2008) Differences in
soil CO2 flux between different land use types in mid-subtropical China. Soil
Biology & Biochemistry, 40, 2324-2333.
Kirschbaum M.U.F. (2004) Soil respiration under prolonged soil warming: are rate
reductions caused by acclimation or substrate loss? Global Change Biology, 10,
1870-1877.
Knorr W., Prentice I.C., House J.I. & Holland E.A. (2005) Long-term sensitivity of soil
carbon turnover to warming. Nature, 433, 298-301.
Kourtev P.S., Ehrenfeld J.G. & Haggblom M. (2002) Exotic plant species alter the
microbial community structure and function in the soil. Ecology, 83, 3152-3166.
135
Kourtev P.S., Ehrenfeld J.G. & Haggblom M. (2003) Experimental analysis of the effect
of exotic and native plant species on the structure and function of soil microbial
communities. Soil Biology & Biochemistry, 35, 895-905.
Kourtev P.S., Ehrenfeld J.G. & Huang W.Z. (1998) Effects of exotic plant species on soil
properties in hardwood forests of New Jersey. Water Air and Soil Pollution, 105,
493-501.
Kourtev P.S., Huang W.Z. & Ehrenfeld J.G. (1999) Differences in earthworm densities
and nitrogen dynamics in soils under exotic and native plant species. Biological
Invasions, 1, 237-245.
Koutika L.S., Vanderhoeven S., Chapuis-Lardy L., Dassonville N. & Meerts P. (2007)
Assessment of changes in soil organic matter after invasion by exotic plant
species. Biology and Fertility of Soils, 44, 331-341.
Lal R. (2004) Soil carbon sequestration to mitigate climate change. Geoderma, 123, 1-22.
Lal R., Follett F., Stewart B.A. & Kimble J.M. (2007) Soil carbon sequestration to
mitigate climate change and advance food security. Soil Science, 172, 943-956.
Lauber C.L., Strickland M.S., Bradford M.A. & Fierer N. (2008) The influence of soil
properties on the structure of bacterial and fungal communities across land-use
types. Soil Biology & Biochemistry, 40, 2407-2415.
Lee K.E. (1985) Earthworms: Their ecology and relationship with soils and land use.
Academic Press, New York, New York, USA.
Liao C.Z., Peng R.H., Luo Y.Q., et al. (2008) Altered ecosystem carbon and nitrogen
cycles by plant invasion: a meta-analysis. New Phytologist, 177, 706-714.
136
Litton C.M., Sandquist D.R. & Cordell S. (2006) Effects of non-native grass invasion on
aboveground carbon pools and tree population structure in a tropical dry forest of
Hawaii. Forest Ecology and Management, 231, 105-113.
Litton C.M., Sandquist D.R. & Cordell S. (2008) A non-native invasive grass increases
soil carbon flux in a Hawaiian tropical dry forest. Global Change Biology, 14,
726-739.
Mack M.C. & D'Antonio C.M. (2003) Exotic grasses alter controls over soil nitrogen
dynamics in a Hawaiian woodland. Ecological Applications, 13, 154-166.
Mack M.C., D'Antonio C.M. & Ley R.E. (2001) Alteration of ecosystem nitrogen
dynamics by exotic plants: a case study of C4 grasses in Hawaii. Ecological
Applications, 11, 1323-1335.
Mack R.N., Simberloff D., Lonsdale W.M., Evans H., Clout M. & Bazzaz F.A. (2000)
Biotic invasions: causes, epidemiology, global consequences, and control.
Ecological Applications, 10, 689-710.
Marhan S. & Scheu S. (2006) Mixing of different mineral soil layers by endogeic
earthworms affects carbon and nitrogen mineralization. Biology and Fertility of
Soils, 42, 308-314.
Martin P.H., Canham C.D. & Marks P.L. (2009) Why forests appear resistant to exotic
plant invasions: intentional introductions, stand dynamics, and the role of shade
tolerance. Frontiers in Ecology and the Environment, 7, 142-149.
Mo J., Zhang W., Zhu W., Gundersen P., Fang Y., Li D. & Wang H. (2008) Nitrogen
addition reduces soil respiration in a mature tropical forest in southern China.
Global Change Biology, 14, 403-412.
137
Morrison J.A., Lubchansky H.A., Mauck K.E., McCartney K.M. & Dunn B. (2007)
Ecological comparison of two co-invasive species in eastern deciduous forests:
Alliaria petiolata and Microstegium vimineum. Journal of the Torrey Botanical
Society, 134, 1-17.
Nay S.M., Mattson K.G. & Bormann B.T. (1994) Biases of chamber methods for
measuring soil CO2 efflux demonstrated with a laboratory apparatus. Ecology, 75,
2460-2463.
Nuzzo V., Maerz J.C. & Blossey B. (2009) Earthworm invasion as the driving force
behind plant invasion and community change in northeastern North American
forests. Conservation Biology
Nye P.H. (1981) Changes of pH across the rhizosphere induced by roots. Plant and Soil,
61, 7-26.
Pattison R.R., Goldstein G. & Ares A. (1998) Growth, biomass allocation and
photosynthesis of invasive and native Hawaiian rainforest species. Oecologia,
117, 449-459.
Pinheiro J.C. & Bates D.M. (2000) Mixed-effects models in S and S-plus. Springer
Verlag, New York, New York, USA.
Scheu S. (1987) The influence of earthworms (lumbricidae) on the nitrogen dynamics in
the soil litter system of a deciduous forest. Oecologia, 72, 197-201.
Schlesinger W.H. & Lichter J. (2001) Limited carbon storage in soil and litter of
experimental forest plots under increased atmospheric CO2. Nature, 411, 466-469.
138
Six J., Frey S.D., Thiet R.K. & Batten K.M. (2006) Bacterial and fungal contributions to
carbon sequestration in agroecosystems. Soil Science Society of America Journal,
70, 555-569.
Soil Survey Staff, Natural Resources Conservation Service, United States Department of
Agriculture. Web Soil Survey. Available online at
http://websoilsurvey.nrcs.usda.gov/ accessed [4/15/2009].
Staddon P.L. (2004) Carbon isotopes in functional soil ecology. Trends in Ecology &
Evolution, 19, 148-154.
Stock W.D., Wienand K.T. & Baker A.C. (1995) Impacts of invading N2-fixing acacia
species on patterns of nutrient cycling in 2 cape ecosystems - evidence from soil
incubation studies and N15
natural-abundance values. Oecologia, 101, 375-382.
Strickland M.S., Osburn E., Lauber C., Fierer N. & Bradford M.A. (2009) Litter quality
is in the eye of the beholder: initial decomposition rates as a function of inoculum
characteristics. Functional Ecology, 23, 627-636.
Thomas L. (1997) Retrospective power analysis. Conservation Biology, 11, 276-280.
Throop H.L. & Archer S.R. (2008) Shrub (Prosopis velutina) encroachment in a
semidesert grassland: spatial-temporal changes in soil organic carbon and
nitrogen pools. Global Change Biology, 14, 2420-2431.
Valery L., Bouchard V. & Lefeuvre J.C. (2004) Impact of the invasive native species
Elymus athericus on carbon pools in a salt marsh. Wetlands, 24, 268-276.
van der Heijden M.G.A., Bardgett R.D. & van Straalen N.M. (2008) The unseen
majority: soil microbes as drivers of plant diversity and productivity in terrestrial
ecosystems. Ecology Letters, 11, 296-310.
139
van Hees P.A.W., Jones D.L., Finlay R., Godbold D.L. & Lundstomd U.S. (2005) The
carbon we do not see - the impact of low molecular weight compounds on carbon
dynamics and respiration in forest soils: a review. Soil Biology & Biochemistry,
37, 1-13.
Vitousek P.M. & Walker L.R. (1989) Biological invasion by Myrica faya in Hawaii -
plant demography, nitrogen-fixation, ecosystem effects. Ecological Monographs,
59, 247-265.
Vitousek P.M., Walker L.R., Whiteaker L.D., Muellerdombois D. & Matson P.A. (1987)
Biological invasion by Myrica faya alters ecosystem development in Hawaii.
Science, 238, 802-804.
Wardle D.A. & Ghani A. (1995) Why is the strength of relationships between pairs of
methods for estimating soil microbial biomass often so variable. Soil Biology &
Biochemistry, 27, 821-828.
Wu J., Brookes P.C. & Jenkinson D.S. (1993) Formation and destruction of microbial
biomass during the decomposition of glucose and ryegrass in soil. Soil Biology &
Biochemistry, 25, 1435-1441.
140
Table 3.1 Litter chemistry and percentage of native and M. vimineum derived inputs into M. vimineum invaded plots. Microstegium
vimineum litter was divided into leaf material and stem material because of expected differences in their chemical recalcitrance. The
proportion of M. vimineum stem and stem material was used to calculate whole litter chemistry for M. vimineum. Native litter was a
mixture of several species found at this site. For litter chemistry the mean ± 1 S.E. is reported (n=3 analytical repeats). For the
percentage of litter inputs the mean ± 1 S.E. is reported for all 6 invaded plots.
Litter Source Litter input
(%)
C (%) N (%) Lignin (%) Cellulose (%) Hemi-cellulose
(%)
Native 97.25 ± 0.98 48.52 ± 0.72 0.87 ± 0.01 25.58 ± 1.81 18.66 ± 0.23 12.40 ± 0.52
M. vimineum* 2.75 ± 0.98 42.50 1.10 17.97 25.17 21.03
Leaves 42.33 ± 1.81 1.34 ± 0.06 20.04 ± 2.38 20.54 ± 0.76 20.76 ± 0.54
Stems 42.83 ± 0.09 0.62 ± 0.03 13.99 ± 2.08 34.06 ± 1.09 21.54 ± 0.64
*Leaves represented 65.74 ± 0.02% of M. vimineum litter and stems represented 34.26 ± 0.02% (n=3).
141
Table 3.2 The mean ± 1S.E. of the Total (native + M. vimineum) and Native forest derived total soil organic C, POM associated C,
mineral associated C, DOC, and microbial biomass C for plots where M. vimineum is absent and where it is present (n = 6). Where
applicable, values are shown both before and after M. vimineum senesced (i.e. September and November samplings, respectively).
Also, shown is the mean percentage ± 1S.E. that M. vimineum derived C contributed to the total C in each pool. Analyses compared
both Total (native + M. vimineum) pool sizes between plots where M. vimineum was present/absent and Native (native only) C pool
sizes. In uninvaded plots Total = Native and so a dash is shown in the table for these entries. Total C, POM C, and Mineral C pools
were only measured pre-senescence. F-values, degrees of freedom, and P-values are reported for the main effect of M. vimineum
presence/absence. Where measures were taken both pre- and post-senescence (e.g. DOC) these values are reported for the main effects
across the two sampling points. Units for Total, POM, and mineral C pools are mg g dry wt soil-1
and µg g dry wt soil-1
for DOC and
microbial biomass C. Where significant (P<0.05) P-values are reported in bold. Marginally significant (P<0.10) P-values are
italicized.
142
Table 3.2 Cont.
Pre-senescence Post-senescence
M. vimineum
absent
M. vimineum
present
%
M. vimineum
derived C
M. vimineum
absent
M. vimineum
present
%
M. vimineum
derived C
Fd.f P-value
Total C
Total 22.21 ± 3.20 17.24 ± 3.56 NA NA NA 4.561,5 0.086
Native − 16.80 ± 3.50 2.87 ± 0.64 NA NA NA 5.281,5 0.070
POM C
Total 8.46 ± 1.24 5.99 ± 1.22 NA NA NA 4.791,5 0.080
Native − 5.60 ± 1.16 6.01 ± 1.34 NA NA NA 6.831,5 <0.05
Mineral C
Total 13.75 ± 2.41 11.25 ± 2.44 NA NA NA 3.371,5 0.126
Native − 11.13 ± 2.43 1.29 ± 0.51 NA NA NA 3.511,5 0.120
DOC
Total 102.97 ± 12.96 87.10 ± 15.55 113.13 ± 18.32 87.46 ± 21.68 2.341,5 0.186
Native − 77.16 ± 14.62 9.80 ± 2.12 − 85.09 ± 21.47 3.12 ± 1.48 4.071,5 0.100
Microbial biomass C
Total*
68.56 ± 22.61 25.03 ± 10.48 74.50 ± 16.24 44.20 ± 6.86 2.341,10 <0.05
Native*
− 19.11 ± 9.57 29.41 ± 7.50 − 42.55 ± 7.34 5.25 ± 3.63 10.081,10 <0.05
*Data were loge transformed.
143
Figure 3.1 Mineralizable C (a) and SIR microbial biomass (b) for plots where M.
vimineum was present and where it was absent (n = 6). Panel a shows the mean
cumulative values ± 1S.E. for mineralizable C across each 60-day incubation.
Mineralizable C is expected to represent microbially-available, labile C. Also shown in
(a) is the percentage ± 1S.E. of CO2-C derived from M. vimineum on the first day of each
60-day incubation for sites where M. vimineum was present (sub-set bar plot). Panel b
shows the mean ± 1S.E. for SIR microbial biomass which may be an indicator of
physiologically active microbial biomass. Microstegium vimineum was active during the
sampling months of Sep., May, and Jul. (note it is an annual). * = P<0.05; ** = P<0.01;
*** = P<0.001; ns = not significant.
Figure 3.2 Soil CO2 efflux (mean ± 1S.E.) measured in situ across all sampling dates for
plots where M. vimineum was present and where it was absent (n = 6). This is a measure
of both heterotrophic and autotrophic respiration. Microstegium vimineum was active
during the sampling months of Sep., May, and Jul. (note it is an annual). Symbols are the
same as in Fig. 3.1.
Figure 3.3 Results of the 13
C-glucose pulse-chase experiment conducted during the
November sampling. The panel insert shows the cumulative amount of 13
C-glucose
mineralized to 13
CO2-C (mean ± 1S.E.) across 72 h for plots where M. vimineum is
present and where it is absent (n = 6). The main panel shows the time course of glucose
mineralization. A marginally significant effect of M. vimineum presence was detected at 2
h (F1,2 = 10.5, P = 0.084 ) and 5 h (F1,2 = 9.98, P = 0.087) but no effect was detected at 24
h (F1,2 = 0.13, P = 0.76), 48 h (F1,2 = 0.01, P = 0.94), or 72h (F1,2 = 0.11, P = 0.77).
Symbols are the same as in Fig. 3.1.
144
Figure 3.1
Sep. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May Jun. Jul.
Month
SIR
(
g C
O2-C
g d
ry w
t so
il-1
h-1
)
0
1
2
3
4
5
a
b
M. vimineum presence *Sample month *Interaction ns
M. vimineum presence nsSample month ***Interaction ns
%
0
10
20
30M. vimineum derived
Min
eral
izab
le C
(mg
CO
2-C
g d
ry w
t so
il-1
)
0.0
0.2
0.4
0.6
0.8
1.0
1.2M. vimineum absent
M. vimineum present
145
Figure 3.2
Sep. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May Jun. Jul.
Month
Soil
CO
2 e
fflu
x (
g C
O2-C
m-2
day
-1)
0
10
20
30
40
50M. vimineum absent
M. vimineum present
M. vimineum presence nsSample month ***Interaction ns
146
Figure 3.3
Time since glucose added (h)
0 20 40 60 80
Glu
co
se m
inera
lizat
ion
(mg
13C
O2-C
m-2
h-1
)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
Absent Present
Cu
mu
lati
ve
Min
eral
izat
ion
(mg
13C
O2-C
m-2
)
0
2
4
6
8
10M. vimineum presence*
P=0.08
ns
nsns
P=0.09
M. vimineum absentM. vimineum present
147
CHAPTER 4
TESTING THE FUNCTIONAL SIGNIFICANCE OF MICROBIAL COMMUNITY
COMPOSITION4
4 Strickland, M.S., C. Lauber, N. Fierer, M.A. Bradford. Ecology. 90(2):441-451. Reprinted here with
permission of publisher.
148
Abstract
A critical assumption underlying terrestrial ecosystem models is that soil microbial
communities, when placed in a common environment, will function in an identical
manner regardless of the composition of that community. Given high species diversity in
microbial communities and the ability of microbes to adapt rapidly to new conditions,
this assumption of functional redundancy seems plausible. We test the assumption by
comparing litter decomposition rates in experimental microcosms inoculated with distinct
microbial communities. We find that rates of carbon dioxide production from litter
decomposition were dependent upon the microbial inoculum, with differences in the
microbial community alone accounting for substantial (~20%) variation in total carbon
mineralized. Communities that shared a common history with a given foliar litter
exhibited higher decomposition rates when compared to communities foreign to that
habitat. Our results suggest that the implicit assumption in ecosystem models (i.e.,
microbial communities in the same environment are functionally equivalent) is incorrect.
To predict accurately how biogeochemical processes will respond to global change may
require consideration of the community composition and/or adaptation of microbial
communities to past resource environments.
Keywords: bacteria, carbon mineralization, community composition, decomposition,
functional diversity, functional equivalence, functional redundancy, fungi, leaf litter,
microorganisms
149
Introduction
Many of the ecosystem processes critical to the functioning of ecosystems, such as
decomposition, are carried out in whole or in part by microorganisms. Ecosystem models
designed to explain and predict how the rates of these ecosystem processes change in
response to changing environmental conditions, such as temperature and moisture,
typically treat the microbial community as a single, homogenously functioning entity
(Parton et al. 1983). Yet it is increasingly being recognized that microbial community
composition plays a role in determining ecosystem process rates (Hendrix et al. 1986,
Schimel and Gulledge 1998, Reed and Martiny 2007). However, even in those models
(Hunt and Wall 2002) that explicitly consider microbial community composition (e.g.,
bacterial vs. fungal dominance), the assumption is still that the environment ultimately
controls ecosystem process rates. That is, changes in microbial community composition
exert proximate control on process rates but the contemporary environment ultimately
structures the community. This hypothesis, which we refer to as ―functional
equivalence,‖ has recently been challenged (Reed and Martiny 2007) given the finding
that historical contingencies, and not just the contemporary environment, structure
microbial communities both at local and regional scales (Martiny et al. 2006, Ramette
and Tiedje 2007). The counter-hypothesis, which we refer to as ―functional
dissimilarity,‖ is significant because it proposes that the responses of ecosystem
processes to global changes are not ultimately fixed by environmental conditions (Balser
and Firestone 2005). On the contrary, it proposes that microbial community composition
(i.e., the whole community genotype) may, in combination with the environment,
ultimately (not just proximally) determine ecosystem process rates.
150
The hypothesis of functional equivalence does appear intuitive. For example,
given that there are many thousands of species of microorganisms in a given soil (Fierer
et al. 2007) and that microbes can rapidly adapt to new environments (Goddard and
Bradford 2003), one might assume that this equates with a high degree of functional
redundancy (Andrén and Balandreau 1999, Behan-Pelletier and Newton 1999, Wall and
Virginia 1999). Any spatial or temporal change in community composition would then
produce a negligible change in ecosystem processes regulated by microbes (Bell et al.
2005, Franklin and Mills 2006, Wertz et al. 2006, Cardinale et al. 2007, Jiang 2007,
Verity et al. 2007). There is an expectation that the greatest functional redundancy will be
observed for biogeochemical processes performed by a broad array of soil
microorganisms, such as the mineralization of carbon compounds during litter
decomposition, rather than for processes such as nitrogen fixation (Schimel 1995, Bell et
al. 2005) that are performed only by specific microbial groups. Work by Balser and
Firestone (2005) tends to support this divide. They observed functional differences for
microbial processes such as nitrogen mineralization but not for carbon mineralization.
Indeed, equivalence in organic carbon degradation has been suggested by others among
both local and regional heterotrophic microbial communities in both aquatic
(Langenheder et al. 2005, 2006) and terrestrial habitats (Balser and Firestone 2005, Ayres
et al. 2006, Wertz et al. 2006).
There is evidence from both plant and animal communities that changes in
community composition impact ecosystem process rates (Tilman et al. 1996, 2001,
Naeem and Wright 2003, Taylor et al. 2006, Cardinale et al. 2007). Similarly, evidence is
now accumulating that microbial community composition influences ecosystem process
151
rates, albeit the identification of causation, as opposed to correlation, is not always
conclusive (Schimel and Gulledge 1998, Behan-Pelletier and Newton 1999, Wall and
Virginia 1999, Reed and Martiny 2007). To identify causation both common garden and
reciprocal transplant approaches are required, where both ecosystem processes and
community composition are monitored across time in different environments (Reed and
Martiny 2007).
Here we used a long-term (300-day) experimental litter–soil system that combines
both a common garden and a reciprocal transplant approach to test between the following
competing hypotheses for litter mineralization: (1) different soil microbial communities
are functionally equivalent vs. (2) different soil microbial communities are functionally
dissimilar. The hypothesis of ―functional equivalence‖ is based on the expectation that
the high diversity and/or adaptive ability of soil microbes confer functional redundancy
across different microbial communities (i.e., only the environment impacts function). The
competing hypothesis of ―functional dissimilarity‖ is based on the expectation that
historical contingencies play a role in shaping the functioning of microbial communities.
To test between the hypotheses, we collected soil inocula from three locations in the
continental United States where we might expect the microbial community composition
to differ either because of geographic distance (a surrogate for history) and/or
contemporary habitat differences (Martiny et al. 2006). Litters from the dominant plant
species at each of the three locations were also collected, sterilized, and then milled to
generate three, distinct, common garden environments. We inoculated each litter in a
factorial design with each of the three soils. If different inocula yielded distinct rates of
litter decomposition then the hypothesis of functional dissimilarity was supported, and if
152
rates were similar then this supported the hypothesis of functional equivalence. We posed
a sub-hypothesis under the hypothesis of functional dissimilarity that stated that soil
inocula that shared a common history with a litter (i.e., soils collected from the same
location as the litter) would mineralize that litter more rapidly than soil inocula without a
shared history (the ―home field advantage‖ hypothesis [Gholz et al. 2000]). Our findings
support the hypotheses of functional dissimilarity and home field advantage, suggesting
that differences in microbial community composition may contribute, as ultimate and not
just proximal controls, to spatial and temporal variation in ecosystem carbon dynamics.
Methods
Microcosm design
Litter and soil were collected from three sites within the continental United States (see
Plate 4.1). Grass litter (Hordeum murinum) was collected from Sedgwick Reserve,
California (34°42′ N, 120°03′ W), pine (Pinus taeda) from Duke Forest, North Carolina
(35°58′ N, 79°05′ W), and rhododendron (Rhododendron maximum) from the Coweeta
Long Term Ecological Research site, North Carolina (35°00′ N, 83°30′ W). Pine and
rhododendron were collected as recent litterfall and grass litter as standing-dead material.
Litter was oven-dried (65°C) and then milled (2 mm). Litter was sterilized by autoclaving
(121°C, 20 min) twice in succession and again 24 h later. Mineral soil (0–5 cm) was
collected together with the litter. Soils were passed through a 2-mm sieve, homogenized,
and then stored at 5°C until use as an inoculum. We refer to each soil inoculum as
rhododendron, pine, or grass inoculum and each litter type as rhododendron, pine, or
grass litter.
153
To 1 g of litter we added 0.5 g dry mass equivalent of soil inoculum and mixed
them by vortexing in a 50-mL plastic centrifuge tube. The mixture was adjusted to and
maintained at 50% water-holding capacity, which is favorable for microbial activity.
Tubes were incubated at 20°C and 100% humidity during the 300-d experiment. Our
design was a 3 × 3 combinatorial setup (i.e., all litters crossed with all soil inocula) plus
additional ―no-litter‖ soils. In all, 24 replicates per treatment were constructed, permitting
destructive harvest of replicates on incubation days 25, 99, and 300 for chemical and
microbial analyses (see Clone library construction and sequencing). The soil inoculum
accounted for an average of <5% of the total CO2 flux across all treatment combinations,
and soil carbon (much of which is not likely to be readily bioavailable) accounted for
<5% of total carbon in each microcosm (see Appendix F).
Determination of carbon mineralization rates and litter chemistry
Carbon mineralization rates were determined across 300 days using a static incubation
procedure as described in Fierer et al. (2003). Carbon mineralization rates were
determined on eight replicate samples per treatment combination at 22 time points during
the course of the incubation, and carbon dioxide production rates were corrected for the
contribution of the corresponding soil inoculum by subtracting the carbon mineralization
rates of the ―no-litter‖ soils. Since samples were harvested on days 25, 99, and 300 in
order to assess the composition of the microbial community (see Clone library
construction and sequencing below), we decided that those harvested microcosms should
correspond to the same eight replicates per treatment measured up to a given harvest date.
Specifically, those microcosms harvested on days 25, 99, or 300 were the same
154
microcosms on which we measured carbon mineralization rates during days 2–25, 26–99,
or 100–300 of the incubation, respectively. This resulted in three incubation periods
corresponding to days 2–25, 26–99, and 100–300. Since a treatment within each of these
incubation periods was composed of different replicates, we measured all microcosms
regardless of their specified incubation period on days 25 and 99 in order to insure that
each set of replicates had similar carbon mineralization rates. It should be noted that
microcosms from all three incubation periods (i.e., three sets of eight replicates) were
measured on day 25, but only microcosms from the last two incubation periods (i.e., two
sets of eight replicates) were measured on day 99, since the first set of microcosms had
already been destructively harvested. No significant difference in mineralization rates
was detected for samples from differing incubation periods measured on the same day
(F2, 357 = 0.12; P = 0.89). Since no differences in mineralization rates were detected
between samples measured on the same day but differing in incubation period we felt
confident in using the entire data set to calculate cumulative carbon mineralized across
the entire 300-d period of this experiment (See Statistical analyses below).
Total carbon, total nitrogen, and C:N ratios for each inocula and litter were
determined using an NA1500 CHN analyzer (Carlo Erba Strumentazione, Milan, Italy).
Clone library construction and sequencing
For each of the nine unique litter–soil inoculum combinations, DNA was extracted from
samples at three time points (days 25, 99, and 300 of the incubation periods) using the
PowerSoil DNA isolation kit (MoBio Laboratories, Carlsbad, California, USA), after
grinding 1 g of each sample with a mortar and pestle in liquid nitrogen. We monitored
155
microbial community composition at each of these three time points in order to determine
whether microbial community composition ever converged, an approach suggested by
Reed and Martiny (2007). The DNA was extracted from three replicate samples per
litter–soil inoculum combination at each time point, and these were pooled prior to
polymerase chain reaction (PCR) amplification. This yielded a total of 27 unique DNA
samples from which the clone libraries were constructed.
The DNA was also extracted and clone libraries were constructed from an
additional 12 samples (two of the grass litter–grass soil inoculum treatments and two of
the rhododendron litter–rhododendron soil inoculum treatments at each of the three time
points). This was done to assess variability in microbial community composition within a
given treatment at a given time point. In all cases, the replicate samples (the clone
libraries constructed from the same litter–soil inoculum–time point combination) were
essentially identical with respect to microbial community composition, and therefore we
present only the results from the 27 clone libraries generated by the different treatments
(nine litter type/soil inoculum combinations at three time points).
Clone libraries were constructed by amplifying the DNA samples with the
universal primer pair 515f (5′-GTGCCAGCMGCCGCGGTAA-3′) and 1391r (5′-
GACGGGCGGTGWGTRCA-3′). This primer pair amplifies small subunit rRNA genes
from all three domains (Archaea, Bacteria, and Eukarya) (Baker et al. 2003, Cullen and
MacFarlane 2005, Martiny et al. 2006) yielding a PCR product that is 850–1100 base
pairs (bp) in length. Although this primer set may not necessarily amplify all taxonomic
groups equally, any amplification bias should be consistent across all of the samples.
Each 50-μL PCR reaction contained 1× PCR buffer, 2.5 mmol/L MgCl2, 0.2 mmol/L of
156
each dNTP, 0.2 μmol/L of each primer, 0.5 units of Taq DNA polymerase (Invitrogen,
Carlsbad, California, USA), and 2 μL of template DNA. Each DNA sample was
amplified with three replicate PCR amplifications (30 cycles each, annealing temperature
of 52°C) with the amplicons from each sample pooled prior to cloning. Amplicons were
cleaned using the Promega WizardSV Gel Extraction and PCR Clean Up kit (Promega,
Madison, Wisconsin, USA).
Amplicons were cloned using TOPO TA for Sequencing kit (Invitrogen)
following the manufacturer's instructions. Plates were grown overnight at 37°C before
plasmid amplification and sequencing. A total of 32 clones were sequenced per sample at
Agencourt Bioscience (Beverly, Massachusetts, USA). Despite the small size of these
libraries, they were sufficiently large to allow us to compare quantitatively community
structure at a general level of phylogenetic resolution (see Sequence analyses).
Sequence analyses
Sequences were binned into major taxonomic groups (i.e., metazoan, bacteria, and fungi)
using the BLAST algorithm (Altschul et al. 1997) against the complete European
Ribosomal RNA database (http://www.psb.ugent.be/rRNA/index.html). Sequences with
expect (E) values greater than 1 × 10−100
to nearest neighbors were not included in the
analyses. Of the clones sequenced, ~3% of the sequences were not included in the
analyses as they were either of low quality, chimeric, or were neither bacterial nor fungal.
The taxonomic classification of the fungal sequences was determined by BLAST search
against the European Ribosomal RNA database. The bacterial sequences were aligned
against the Greengenes database (DeSantis et al. 2006b) using the NAST aligner
157
(DeSantis et al. 2006a) and classified into taxonomic groups with the Greengenes
―classify‖ utility (http://greengenes.lbl.gov/cgi-bin/nph-classify.cgi). Only those
sequences sharing >90% similarity to reference sequences over a 600-bp region were
classified. Taxonomic characteristics of the microbial communities are provided in Table
4.1. Sequences were deposited in GenBank under the accession numbers EU729748–
EU730583 and EU725929–EU726201 for bacteria and fungi, respectively.
The bacterial and fungal sequences were combined together and aligned against
bacterial and fungal guide sequences downloaded from the Silva ssu RNA database
(http://silva.mpi-bremen.de/). Sequences were aligned using MUSCLE (Edgar 2004) and
a neighbor-joining tree containing all bacterial and fungal sequences from the constructed
libraries. In addition, an outgroup (Haloferax volcanii) was constructed in MEGA
(Tamura et al. 2007).
We used the weighted UniFrac algorithm (Lozupone et al. 2006) to determine
whether soil inocula affected the composition of the microbial communities on each litter
type. UniFrac provides an overall estimate of the phylogenetic distance between each pair
of communities by examining the fraction of the total branch length within a single
phylogenetic tree that is unique to either of the two communities (as opposed to being
shared by both; Fierer et al. 2007).
Statistical analyses
Statistical analyses of mineralization data were performed using S-Plus 7.0 (Insightful,
Seattle, Washington, USA). To examine treatment effects on mineralization rates across
time, a linear mixed-effects model was used in which time, inoculum source, and litter
158
type (all discrete variables) were treated as fixed effects and permitted to interact.
Microcosm identity was included as a random effect to account for the repeated sampling
across time of microcosms. We analyzed the cumulative carbon mineralization rates
using ANOVA with litter type, inoculum source, and block as discrete variables;
inoculum source and litter type were permitted to interact. Cumulative carbon
mineralization rates were determined by integrating values under the curve of the 300-d
sampling period. Since replicates were destructively harvested for molecular analyses, to
generate cumulative values we blocked replicates within a treatment by mineralization
rates and generated the cumulative values from samples within a block. This approach
enabled us to determine whether litters crossed with their home (as opposed to foreign)
soil inoculum resulted in the highest, cumulative carbon mineralization, permitting us to
explore our ―home field advantage‖ hypothesis. Pairwise comparisons of means were
explored using the Tukey method to assess differences between inocula within the same
litter type. For statistical significance we assumed an α level of 0.05 and, when reported
as such, data were log10-transformed to conform to assumptions of homoscedasticity
(verified using model checking).
Results
Carbon mineralization
Across the course of the incubation we noted that in most cases, and regardless of the
inoculum–litter combination, that carbon mineralization dynamics followed an
approximately similar three-stage pattern. This consisted of an initial peak in
mineralization rates during the first 25 d, a secondary peak lasting from about day 30 to
159
as much as day 150, and finally a decline phase characterized by a gradual decrease in
mineralization rates over time (Fig. 4.1).
When examining each litter in turn, for rhododendron litter we found a significant
two-way interaction between inoculum and time on carbon mineralization rates for each
incubation period (P < 0.0001 in all cases; Table 4.2, Fig. 4.1A). The interactions, across
all incubation periods (i.e., days 2–25, 26–99, and 100–300), showed that mineralization
dynamics were different between at least two of the inocula and that these differences
between the inocula were dependent upon time. For pine litter a significant two-way
interaction between inoculum and time was found for the first two incubation periods (P
< 0.0001 in both cases; Table 4.2, Fig. 4.1B). In the third incubation period (days 126–
300), the interaction was not significant but the main effects of inoculum (P < 0.0001)
and time (P < 0.0001) were (thus, relative differences between inocula were time
independent). For grass litter we found a significant interaction between inoculum and
time on carbon mineralization rates for incubation periods one, two, and three (P < 0.05,
P < 0.0001, and P < 0.001, respectively; Table 4.2, Fig. 4.1C).
To examine further the dynamics we saw in the time course data, we looked at
cumulative mineralization across the 300-d experiment (Fig. 4.2). This examination
showed a significant two-way interaction term of inoculum and litter on cumulative
carbon mineralization rates (P < 0.001; Table 4.3). We investigated this result by looking
at the inoculum effect for each litter type individually to test our sub-hypothesis of ―home
field advantage.‖
For each litter type we found a significant effect of inoculum on cumulative
carbon mineralization (P < 0.001 in all cases). Post hoc analyses on rhododendron litter
160
showed that all three inocula differed, with the rhododendron inoculum yielding a higher
cumulative value than either the pine or grass inoculum (Fig. 4.2A). For pine litter, all
three inocula again differed except in this instance the pine inoculum yielded a higher
cumulative value than either of the other two inocula (Fig. 4.2B). For grass litter, both
grass and rhododendron inocula yielded a higher cumulative value than did the pine
inoculum (Fig. 4.2C).
Microbial community
For the UniFrac analyses we examined each litter type separately (building a separate tree
for each litter type), combining the sequences for each soil inoculum from the three time
points. By doing this, we could examine the overall influence of the inoculum source on
microbial community composition in each litter type across the entire incubation period.
We found that, as expected, litter type influenced microbial community composition
(data not shown). However, within each litter type, we also found a significant influence
(P < 0.01) of soil inoculum on microbial community composition (Fig. 4.3), as
determined by both the UniFrac significance test and the phylogenetic (P) test (Martin
2002). In most instances, the inoculum effects on the microbial communities were
apparent at even the coarsest levels of taxonomic resolution (e.g., shifts in
bacterial:fungal ratios or shifts in the dominance of particular bacterial phyla; Table 4.1).
Notably, the resulting microbial community composition from the different inocula was
most dissimilar in the grass litter microcosms and most similar in the rhododendron litter
microcosms. That is, the more chemically complex the litter habitat, the more similar the
resulting community compositions from the different inocula (Fig. 4.3, Table 4.1).
161
Furthermore, microbial communities on the same litter habitat never converged
compositionally across the course of this experiment (Fig. 4.3, Table 4.1).
Discussion
The three-stage pattern observed in mineralization dynamics across time (Fig. 4.1) is
consistent with a recently proposed mathematical model (Moorhead and Sinsabaugh
2006) that incorporates three guilds of microbial decomposers. However, we found that
the magnitude and characteristics of each stage within a given litter habitat were strongly
determined by the inoculum source (Fig. 4.1, Table 4.2). Indeed, the main effect of
inoculum explained 20% of the variation in cumulative carbon mineralization across litter
types, and its interaction with litter type explained an additional 25% of the variation
(Fig. 4.2, Table 4.3). When examining the inoculum effect for each litter separately, it
explained between 22% and 86% of the variation in cumulative carbon mineralization
(Fig. 4.2, Table 4.3). The significant inoculum effects and inoculum interactions with
time refute the hypothesis that microbial communities are functionally equivalent. These
results are contrary to those studies (Balser and Firestone 2005, Langenheder et al. 2005,
2006, Ayres et al. 2006, Wertz et al. 2006) that did not observe differences in carbon
mineralization rates between different microbial communities, and these results
demonstrate that microbial communities may impact processes that are generally
presumed to be carried out equivalently by a wide array of organisms (Schimel 1995).
Indeed, the competing hypothesis of functional dissimilarity was supported across the
course of the entire 300-day experiment. Given that the composition of the communities
in a given litter habitat, from different inocula, were consistently distinct across time (Fig.
162
4.3, Table 4.1), differences in function in the same litter habitat may be attributed to the
composition of the microbial community (Reed and Martiny 2007).
When the rhododendron and pine inocula were combined with their ―home‖
litter resources, the combinations yielded significantly greater cumulative mineralization
than instances in which the inocula were ―foreign‖ to the litter (Fig. 4.2A, B, Table 4.3).
The grass inoculum yielded cumulative mineralization values that were equivalent to the
rhododendron inoculum and higher than the pine inoculum only with grass litter (Fig.
4.2C, Table 4.3). These results support the ―home field advantage‖ hypothesis (Gholz et
al. 2000, Ayres et al. 2006), suggesting that soil microbial communities display some
adaptation (or pre-adaptation) to their past resource environment.
If adaptation to the past resource environment explains the ―home field
advantage‖ phenomenon that we observed (Fig. 4.2), this implies that decomposer
communities may be selected for, at least in part, by the specific characteristics of past
litter inputs as both Gholz et al. (2000) and Ayres et al. (2006) have suggested. More
specifically, litter characteristics may lead to changes in the composition of a given
microbial community, as observed in this study (Fig. 4.3, Table 4.1). Indeed, we chose
three litter resources that presented a gradient in litter chemical complexity and
recalcitrance, as indicated by both C:N ratios (Appendix F) and by in situ decomposition
rates reported for identical or very similar litter types (Andren et al. 1992, Sanchez 2001,
Ball et al. 2008), with rhododendron representing the most, and grass the least, complex
and recalcitrant litter. Potentially then, these differences in resource quality could account
for the strong home field advantage observed for both the rhododendron and pine inocula
on their respective home litters and the weaker home field advantage observed for the
163
grass inoculum on the grass litter (Fig. 4.2, Table 4.3). Notably, Langenheder et al.
(2005) observed that a microbial community inoculum from a lake with high-quality
dissolved organic carbon (DOC) yielded respiration rates lower than inocula from low-
quality DOC environments, when exposed to common garden environments with low-
quality DOC. Also, soils pre-exposed to a pesticide or herbicide may degrade that
compound more rapidly in subsequent applications (Taylor et al. 1996, Laha and Petrova
1997). Together, these results suggest that the ability of microbial communities to
decompose carbon compounds may be contingent on past exposure to similar compounds
(Taylor et al. 1996, Laha and Petrova 1997) and that the more chemically complex litter
carbon is, the more likely community function will be contingent on history and not
simply on chemical quality alone.
Our study cannot unambiguously disentangle whether the functional
differences between the inocula were the result of differences in the abundance of
different taxa or the degradative capacities of individuals from the same taxa from
different inocula. If only the former mechanism applied then communities of more
similar composition would function more similarly. We did not observe this relationship;
instead, those communities that developed on rhododendron litter were most similar in
composition (Fig. 4.3) and less similar with regards to their function (Figs. 4.1 and 2.2).
This observation was investigated statistically and the similarity plots are shown in
Appendix G. Intriguingly, these data suggest that the manner in which microbial
communities respond compositionally to environmental variation may be a poor predictor
of their impacts on ecosystem function and communities that are more similar in
taxonomic composition may not necessarily be more similar with regards to their
164
functional capabilities. Or to rephrase that statement using the definitions from the
biodiversity–ecosystem functioning literature, functional response groups may not
correlate with functional effect groups (Naeem and Wright 2003). This may make
prediction of the manner in which changing microbial community composition will
impact ecosystem functioning highly uncertain. However, at least from the functional
response group perspective, the microbes behaved like animal and plant communities.
That is, the harsher environmental filter (i.e., rhododendron > pine > grass litter)
generated communities of more similar phylogenetic composition (Fig. 4.3).
Ecosystem models are generally designed without incorporation of the
potential effects of variation in microbial community composition or the functional
capacities of individuals from the same taxonomic level on ecosystem processes (Zak et
al. 2006). Yet, under similar environmental conditions, our results show that the effects
of different microbial inocula can account for >85% of the variation in litter carbon
mineralization (Table 4.3) and significantly alter the temporal dynamics of this process
(Fig. 4.1). Before exploring the consequences of these findings and their possible
implications for predicting ecosystem carbon dynamics, we propose two necessary areas
of research. First we should resolve whether the taxonomic composition of microbial
communities or the past exposure of communities to specific resource conditions
predominately impacts ecosystem function through time. That is, does community
composition impact future functioning and/or is functioning a product of the selection
environment (Loehle and Pechmann 1988, Goddard and Bradford 2003)? Second, and
building on ideas relating to adaptation, future work should resolve whether microbial
communities from different selection environments functionally converge when placed in
165
a similar environment. Although our experiment ran for 300 days, to reliably test this
possibility we would need an experimental design in which microbial communities
sourced from differing environments were exposed to continuous inputs of the same
resource. By using such an approach we may be able to determine how rapidly microbial
communities adapt to new resource conditions and whether different communities
functionally converge. Such studies might then inform us as to how rapid land use
conversion may impact microbially mediated biogeochemical processes and would be
vital to our understanding of the degree to which ecosystem functions are resistant and
resilient to environmental change.
Our findings suggest that soil microbial communities of differing
composition are functionally dissimilar. We propose future work to explore further the
assumption of functional equivalence for soil microbial communities. If the weight of
evidence supports the alternate hypothesis of functional dissimilarity, then the historical
factors that shape microbial community functioning must be explored in ecosystem
models to predict accurately how the carbon cycle will respond to global change.
Acknowledgements
We thank Tom Maddox in the Analytical Chemistry Laboratory of the Odum School of
Ecology, University of Georgia, for element analyses. We also thank two anonymous
reviewers who helped to significantly improve an earlier version of this manuscript. The
authors acknowledge funding from the Andrew W. Mellon Foundation and from the
National Science Foundation to the Coweeta LTER.
166
References
Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and D. J.
Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of
protein database search programs. Nucleic Acids Research 25:3389–3402.
Andrén, O. and J. Balandreau. 1999. Biodiversity and soil functioning—From black box
to can of worms. Applied Soil Ecology 13:105–108.
Andrén, O., E. Steen, and K. Rajkai. 1992. Modeling the effects of moisture on barley
straw and root decomposition in the field. Soil Biology and Biochemistry
24:727–736.
Ayres, E., K. M. Dromph, and R. D. Bardgett. 2006. Do plant species encourage soil
biota that specialise in the rapid decomposition of their litter. Soil Biology
and Biochemistry 38:183–186.
Baker, G. C., J. J. Smith, and D. A. Cowan. 2003. Review and re-analysis of domain-
specific 16S primers. Journal of Microbiological Methods 55:541–555.
Ball, B. A., M. D. Hunter, J. S. Kominoski, C. M. Swan, and M. A. Bradford. 2008.
Consequences of non-random species loss for decomposition dynamics:
experimental evidence for additive and non-additive effects. Journal of
Ecology 96:303–313.
Balser, T. C. and M. K. Firestone. 2005. Linking microbial community composition and
soil processes in a California annual grassland and mixed-conifer forest.
Biogeochemistry 73:395–415.
167
Behan-Pelletier, V. and G. Newton. 1999. Computers in biology—linking soil
biodiversity and ecosystem function—the taxonomic dilemma. BioScience
49:149–153.
Bell, T., J. A. Newman, B. W. Silverman, S. L. Turner, and A. K. Lilley. 2005. The
contribution of species richness and composition to bacterial services. Nature
436:1157–1160.
Cardinale, B. J., J. P. Wright, M. W. Cadotte, I. T. Carroll, A. Hector, D. S. Srivastava,
M. Loreau, and J. J. Weis. 2007. Impacts of plant diversity on biomass
production increase through time because of species complementarity.
Proceedings of the National Academy of Sciences (USA) 104:18123–18128.
Cullen, L. E. and C. MacFarlane. 2005. Comparison of cellulose extraction methods for
analysis of stable isotope ratios of carbon and oxygen in plant material. Tree
Physiology 25:563–569.
DeSantis, T., P. Hugenholtz, K. Keller, E. L. Brodie, N. Larsen, Y. M. Piceno, R. Phan,
and G. L. Andersen. 2006a. NAST: a multiple sequence alignment server for
comparative analysis of 16S rRNA genes. Nucleic Acids Research
34:W394–399.
DeSantis, T. Z., P. Hugenholtz, N. Larsen, M. Rojas, E. L. Brodie, K. Keller, T. Huber,
D. Dalevi, P. Hu, and G. L. Andersen. 2006b. Greengenes, a chimera-
checked 16S rRNA gene database and workbench compatible with ARB.
Applied and Environmental Microbiology 72:5069–5072.
Edgar, R. C. 2004. MUSCLE: multiple sequence alignment with high accuracy and high
throughput. Nucleic Acids Research 32:1792–1797.
168
Fierer, N., A. S. Allen, J. P. Schimel, and P. A. Holden. 2003. Controls on microbial CO2
production: a comparison of surface and subsurface soil horizons. Global
Change Biology 9:1322–1332.
Fierer, N., M. Breitbart, J. Nulton, P. Salamon, C. Lozupone, R. Jones, M. Robeson, R.
A. Edwards, B. Felts, S. Rayhawk, R. Knight, F. Rohwer, and R. B. Jackson.
2007. Metagenomic and small-subunit rRNA analyses of the genetic
diversity of bacteria, archaea, fungi, and viruses in soil. Applied and
Environmental Microbiology 73:7059–7066.
Franklin, R. B. and A. L. Mills. 2006. Structural and functional responses of a sewage
microbial community to dilution-induced reductions in diversity. Microbial
Ecology 52:280–288.
Gholz, H. L., D. A. Wedin, S. M. Smitherman, M. E. Harmon, and W. J. Parton. 2000.
Long-term dynamics of pine and hardwood litter in contrasting
environments: toward a global model of decomposition. Global Change
Biology 6:751–765.
Goddard, M. R. and M. A. Bradford. 2003. The adaptive response of a natural microbial
population to carbon- and nitrogen-limitation. Ecology Letters 6:594–598.
Hendrix, P. F., R. W. Parmelee, D. A. Crossley, D. C. Coleman, E. P. Odum, and P. M.
Groffman. 1986. Detritus food webs in conventional and no-tillage
agroecosystems. BioScience 36:374–380.
Hunt, H. W. and D. H. Wall. 2002. Modelling the effects of loss of soil biodiversity on
ecosystem function. Global Change Biology 8:33–50.
169
Jiang, L. 2007. Negative selection effects suppress relationships between bacterial
diversity and ecosystem functioning. Ecology 88:1075–1085.
Laha, S. and K. P. Petrova. 1997. Biodegradation of 4-nitrophenol by indigenous
microbial populations in Everglades soils. Biodegradation 8:349–356.
Langenheder, S., E. S. Lindstrom, and L. J. Tranvik. 2005. Weak coupling between
community composition and functioning of aquatic bacteria. Limnology and
Oceanography 50:957–967.
Langenheder, S., E. S. Lindstrom, and L. J. Tranvik. 2006. Structure and function of
bacterial communities emerging from different sources under identical
conditions. Applied and Environmental Microbiology 72:212–220.
Loehle, C. and J. H. K. Pechmann. 1988. Evolution—the missing ingredient in systems
ecology. American Naturalist 132:884–899.
Lozupone, C., M. Hamady, and R. Knight. 2006. UniFrac—an online tool for comparing
microbial community diversity in a phylogenetic context. BMC
Bioinformatics 7:371.
Martin, A. P. 2002. Phylogenetic approaches for describing and comparing the diversity
of microbial communities. Applied and Environmental Microbiology
68:3673–3682.
Martiny, J. B. H., et al 2006. Microbial biogeography: putting microorganisms on the
map. Nature Reviews Microbiology 4:102–112.
Moorhead, D. L. and R. L. Sinsabaugh. 2006. A theoretical model of litter decay and
microbial interaction. Ecological Monographs 76:151–174.
170
Naeem, S. and J. P. Wright. 2003. Disentangling biodiversity effects on ecosystem
functioning: deriving solutions to a seemingly insurmountable problem.
Ecology Letters 6:567–579.
Parton, W. J., D. W. Anderson, C. V. Cole, and J. W. B. Stewart. 1983. Simulation of soil
organic matter formation and mineralization in semiarid agroecosystems.
Special publication. University of Georgia, College of Agriculture
Experiment Stations. Athens, Georgia, USA.
Pinheiro, J. C. and D. M. Bates. 2000. . Mixed-effects models in S and S-PLUS. Springer
Verlag. New York, New York, USA.
Ramette, A. and J. M. Tiedje. 2007. Multiscale responses of microbial life to spatial
distance and environmental heterogeneity in a patchy ecosystem.
Proceedings of the National Academy of Sciences (USA) 104:2761–2766.
Reed, H. E. and J. B. H. Martiny. 2007. Testing the functional significance of microbial
composition in natural communities. Fems Microbiology Ecology 62:161–
170.
Sanchez, F. G. 2001. Loblolly pine needle decomposition and nutrient dynamics as
affected by irrigation, fertilization, and substrate quality. Forest Ecology and
Management 152:85–96.
Schimel, J. P. 1995. Ecosystem consequences of microbial diversity and community
structure. Pages 239–244. in Chapin III, F. S. and C. Korner, editors. Arctic
and Alpine biodiversity: patterns, causes, and ecosystem consequences.
Springer-Verlag. Berlin, Germany.
171
Schimel, J. P. and J. Gulledge. 1998. Microbial community structure and global trace
gases. Global Change Biology 4:745–758.
Tamura, K., J. Dudley, M. Nei, and S. Kumar. 2007. MEGA4: Molecular evolutionary
genetics analysis (MEGA) software. Version 4.0. Molecular Biology and
Evolution 24:1596–1599.
Taylor, B. W., A. S. Flecker, and R. O. Hall. 2006. Loss of a harvested fish species
disrupts carbon flow in a diverse tropical river. Science 313:833–836.
Taylor, G. E., K. B. Schaller, J. D. Geddes, M. S. Gustin, G. B. Lorson, and G. C. Miller.
1996. Microbial ecology, toxicology and chemical fate of methyl
isothiocyanate in riparian soils from the Upper Sacramento River.
Environmental Toxicology and Chemistry 15:1694–1701.
Tilman, D., P. B. Reich, J. Knops, D. Wedin, T. Mielke, and C. Lehman. 2001. Diversity
and productivity in a long-term grassland experiment. Science 294:843–845.
Tilman, D., D. Wedin, and J. Knops. 1996. Productivity and sustainability influenced by
biodiversity in grassland ecosystems. Nature 379:718–720.
Verity, P. G., C. P. Brussaard, J. C. Nejstgaard, M. A. van Leeuwe, C. Lancelot, and L.
K. Medlin. 2007. Current understanding of Phaeocystis ecology and
biogeochemistry, and perspectives for future research. Biogeochemistry
83:311–330.
Wall, D. H. and R. A. Virginia. 1999. Controls on soil biodiversity: insights from
extreme environments. Applied Soil Ecology 13:137–150.
172
Wertz, S., V. Degrange, J. I. Prosser, F. Poly, C. Commeaux, T. Freitag, N. Guillaumaud,
and X. Le Roux. 2006. Maintenance of soil functioning following erosion of
microbial diversity. Environmental Microbiology 8:2162–2169.
Zak, D. R., C. B. Blackwood, and M. P. Waldrop. 2006. A molecular dawn for
biogeochemistry. Trends in Ecology and Evolution 21:288–295.
173
Table 4.1 Taxonomic description of the microbial communities of each litter type/soil inoculum combination at each of the three time
points.
L
itte
r ty
pe
Soil
Inocu
lum
Incu
bat
ion D
ay
% B
acte
ria
(all
gro
ups)
% F
ungi
(all
gro
ups)
% P
rote
obac
teri
a
% A
cidobac
teri
a
% A
ctin
obac
teri
a
% C
FB
% F
irm
icute
s
% O
ther
bac
teri
a
% A
scom
yco
ta
% O
ther
Fungi
Rhod. Grass 25 62.1 37.9 24.1 13.8 6.9 17.2 0.0 0.0 37.9 0.0
Pine 25 51.6 48.4 25.8 9.7 3.2 9.7 0.0 3.2 38.7 9.7
Rhod. 25 36.7 63.3 23.3 9.4 3.0 0.0 0.0 1.0 45.2 18.1
Grass 99 58.6 41.4 37.9 13.8 0.0 3.4 0.0 3.4 34.5 6.9
Pine 99 40.0 60.0 6.7 20.0 6.7 6.7 0.0 0.0 46.7 13.3
Rhod. 99 47.3 52.7 31.6 11.4 2.2 1.1 0.0 1.0 44.5 8.2
Grass 300 46.4 53.6 28.6 7.1 3.6 3.6 0.0 3.6 50.0 3.6
Pine 300 76.7 23.3 23.3 30.0 0.0 23.3 0.0 0.0 20.0 3.3
Rhod. 300 82.7 17.3 30.0 36.6 3.6 4.4 0.0 8.0 11.6 5.7
Pine Grass 25 72.4 27.6 55.2 17.2 0.0 0.0 0.0 0.0 27.6 0.0
Pine 25 78.3 21.7 56.5 21.7 0.0 0.0 0.0 0.0 21.7 0.0
Rhod. 25 71.4 28.6 46.4 25.0 0.0 0.0 0.0 0.0 21.4 7.1
Grass 99 85.2 14.8 37.0 40.7 3.7 0.0 3.7 0.0 14.8 0.0
Pine 99 70.4 29.6 40.7 14.8 0.0 14.8 0.0 0.0 29.6 0.0
Rhod. 99 87.1 12.9 51.6 35.5 0.0 0.0 0.0 0.0 12.9 0.0
Grass 300 68.4 31.6 57.9 5.3 0.0 0.0 0.0 5.3 31.6 0.0
Pine 300 75.0 25.0 41.7 8.3 0.0 12.5 0.0 12.5 25.0 0.0
Rhod. 300 93.3 6.7 50.0 40.0 0.0 0.0 0.0 3.3 6.7 0.0
174
Table 4.1 Cont.
Taxonomic description was determined by the 30-32 clones sequenced per library. Firmicutes refers to the ‗Bacillus-Clostridium‘
group and CFB refers to the ‗Cytophaga-Flavobacterium-Bacteriodes’ group. The ―other bacteria‖ category consists primarily of
sequences identified as either Planctomycetes or Verrucomicrobia. The ―other fungi‖ category is evenly divided between sequences
identified as Basidiomycota or Zygomycota.
Lit
ter
type
Soil
Inocu
lum
Incu
bat
ion D
ay
% B
acte
ria
(all
gro
ups)
% F
ungi
(all
gro
ups)
% P
rote
obac
teri
a
% A
cidobac
teri
a
% A
ctin
obac
teri
a
% C
FB
% F
irm
icute
s
% O
ther
bac
teri
a
% A
scom
yco
ta
% O
ther
Fungi
Grass Grass 25 87.1 12.9 32.7 4.0 3.8 41.7 1.4 3.6 8.9 4.0
Pine 25 89.7 10.3 55.2 6.9 3.4 17.2 3.4 3.4 6.9 3.4
Rhod. 25 86.2 13.8 65.5 3.4 13.8 0.0 3.4 0.0 13.8 0.0
Grass 99 92.7 7.3 21.6 1.1 4.8 47.2 3.5 14.6 7.3 0.0
Pine 99 96.8 3.2 35.5 0.0 12.9 35.5 9.7 3.2 0.0 3.2
Rhod. 99 87.5 12.5 68.8 6.3 12.5 0.0 0.0 0.0 12.5 0.0
Grass 300 93.0 7.0 22.5 1.2 1.1 25.3 32.8 10.1 7.0 0.0
Pine 300 96.8 3.2 22.6 0.0 3.2 67.7 0.0 3.2 3.2 0.0
Rhod. 300 100.0 0.0 76.7 6.7 3.3 3.3 0.0 10.0 0.0 0.0
175
Table 4.2 Linear mixed-effects model results for the effect of inoculum and time on
carbon mineralization rates.
Litter type
Incu
bat
ion
per
iod
(day
s)
Inoculum effect Inoculum Time effect
Num
d.f.
Den
d.f.
F P Num
d.f.
Den
d.f.
F P
Rhododendron 2-25 2 21 6.12 <0.01 16 168 5.08 <0.0001
26-99 2 21 285.33 <0.0001 10 105 18.73 <0.0001
100-300 2 21 51.85 <0.0001 12 126 6.34 <0.0001
Pine 2-25 2 21 3.2 0.061 16 168 3.65 <0.0001
26-99 2 21 59.16 <0.0001 10 105 4.03 <0.0001
100-300 2 21 93.63 <0.0001 12 126 ns ns
Grass 2-25 2 21 6.61 <0.01 16 168 1.85 <0.05
26-99 2 21 4.40 <0.05 10 105 6.22 <0.0001
100-300 2 21 13.01 <0.001 12 126 3.15 <0.001
Notes: Results were analyzed based on litter type and incubation period. F and P values
are reported for significant (P < 0.05) model terms; ns: not significant. All data were
log10-transformed to correct for heteroscedasticity. F-values and P-values are estimated in
linear mixed-effects models using, in this case, restricted maximum likelihood (REML)
estimates and are not calculated using least squares methods (Pinheiro and Bates 2000).
Hence, means and sums of squares values are not calculated.
176
Table 4.3 ANOVA tables with percentage sums of squares explained (%SS) for the
effects of inoculum across all litter types, and for the effect of inoculum within each litter
type, on cumulative carbon mineralization.
Litter Type Source of
Variation
d.f. SS % SS F P
All Inoculum 2 5.6 19.6 179.8 <0.001
Litter 2 13.2 46.1 422.2 <0.001
Block 7 1.8 6.2 16.3 <0.001
Inoculum × Litter 4 7.1 24.9 114.12 <0.001
Error 56 0.9 3.0
Rhododendron Inoculum 2 10.2 86.2 145.0 <0.001
Block 7 1.1 9.7 4.6 <0.01
Error 14 0.5 4.2
Pine Inoculum 2 2.4 82.5 114.5 <0.001
Block 7 0.4 12.4 4.9 <0.01
Error 14 0.1 5.1
Grass Inoculum 2 0.1 22.3 17.6 <0.001
Block 7 0.4 68.9 15.6 <0.001
Error 14 0.1 8.9
All data were log10-transformed.
177
Figure legends
Figure 4.1 Time course of carbon mineralization for each litter type and soil inoculum
combination. In all cases across all time periods there was a significant inoculum by time
interaction on CO2 production rates (P<0.05), except in the case of pine litter between
incubation days 126 and 300 where there were significant main effects of inoculum
source and time only (P<0.0001 and P<0.0001, respectively, see Table 2.2). In (A)
mineralization dynamics for rhododendron litter are shown and in this instance
rhododendron inoculum is native to this litter type. In (B) pine litter is represented and
pine inoculum is native to this litter type. In (C) grass litter is represented and grass
inoculum is native to this litter type. Means are shown for carbon mineralization rates ± 1
SE (n=8).
Figure 4.2 Cumulative carbon mineralized for each litter type by inoculum across the
300 day experiment. The inoculum effects for each litter type were highly significant
(P<0.001). Letters denote significant differences between treatments within each litter
type using a Tukey post-hoc pair-wise comparison and those letters in bold denote the
home litter by home inoculum combination, as do grey bars. In (A) cumulative carbon
mineralization values are shown for rhododendron litter with rhododendron inoculum
being native to this litter type. In (B) pine litter is represented and in (C) grass litter is
represented. Means are shown ± 1 SE (n=8).
Figure 4.3 Relative UniFrac distances between the decomposer communities developing
from the three soil inocula in rhododendron litter (A), pine litter (B), and grass litter (C).
The ―home‖ soil inoculum is highlighted in bold type. The sequence data from the three
sampling dates are combined for these analyses to illustrate overall differences in the
178
microbial communities. For each litter type, the three inocula always yielded distinct
communities (P<0.01), as determined by the UniFrac significance test and the
Phylogenetic (P) test (see text for details).
179
Figure 4.1
(A) Rhododendron litter
0
10
20
30
40
(B) Pine litter
Ca
rbo
n m
ine
raliz
atio
n r
ate
(
g C
-CO
2 h
-1)
0
5
10
15
20
25
Rhododendron inoculum
Pine inoculum
Grass inoculum
0 25 50 75 100 125 150 175 200 225 250 275 300
(C) Grass litter
0
10
20
30
40
50
Time (incubation day)
180
Figure 4.2
(A) Rhododendron litter
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
(B) Pine litter
Cu
mu
lative
ca
rbo
n m
ine
raliz
atio
n (
g C
-CO
2)
0.00
0.01
0.02
0.03
0.04
0.05
0.06
(C) Grass Litter
Inoculum source
Rhododendron Pine Grass
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
a
b
c
a
b
c
a a
b
181
Figure 4.3
182
Plate legends
Plate 4.1. Sites (U.S.) where litter and soil inocula were collected: (top left) Sedgwick
Reserve, California, (lower left) Duke Forest, North Carolina, and (right) Coweeta Long
Term Ecological Research site, North Carolina (Photo credits: B. A. Ball).
183
Plate 4.1
184
CHAPTER 5
LITTER QUALITY IS IN THE EYE OF THE BEHOLDER: INITIAL
DECOMPOSITION RATES AS A FUNCTION OF INOCULUM
CHARACTERISTICS5
5 Strickland, M.S., E. Osburn, C. Lauber, N. Fierer, M.A. Bradford. Functional Ecology. 23:627-636.
Reprinted here with permission of publisher.
185
Abstract
1. The chemical composition of plant litter is commonly considered to indicate its
quality as a resource for decomposer organisms. Litter quality, defined in this way, has
been shown to be a major determinant of litter decomposition rates both within and
across terrestrial ecosystems. Notably, the structure of the microbial community that is
directly responsible for primary decomposition is rarely considered as an empirical
predictor of litter decay rates.
2. Microbial communities are generally assumed to perceive litters of the same chemical
composition to be of equivalent resource quality but evidence from field studies suggests
that these same communities may adapt to the prevalent litter types at a given site. Here,
we tested this assumption by assessing how microbial communities sourced from
different forest- and herbaceous-dominated ecosystems perceive the quality of novel,
foliar litters derived from a tree (Rhododendron maximum) and from a grass (Panicum
virgatum) species. Based on chemical composition, we would expect R. maximum litter
to be of lower quality than P. virgatum litter.
3. We used an experimental litter-soil system which employs a ‗common garden‘
approach and measured rates of CO2 production across 50 days; higher rates of
production were assumed to indicate higher quality (i.e. more easily degradable) litter.
4. We found that communities sourced from habitats under differing plant cover
perceived litter quality differently. Those communities sourced from herbaceous habitats
perceived the grass litter to be of higher quality than the tree litter, whereas communities
from forest habitats decomposed both litter types similarly. Within a litter type,
186
differences in both community composition and nutrient availability of the source habitat
were related to decomposition rates.
5. Our results suggest that litter quality cannot necessarily be predicted solely from
chemical characteristics; instead the perceived quality is dependent on the quality of past
resource inputs a community has experienced and the structure of those microbial
communities responsible for the initial stages of litter decomposition.
Key-words: Bacteria, Carbon mineralization, Fungi, Resource history, Soil properties
Introduction
The decomposition of plant litter by soil microbial and faunal communities is one of the
most heavily researched areas in modern ecology and this process is one of the primary
pathways in the terrestrial carbon cycle (Raich & Schlesinger 1992; Couteaux et al. 1995;
Aerts 1997). Research on litter decomposition has shown that, at the global scale,
decomposition rates of plant litter are strongly related to mean annual temperature (MAT)
and precipitation (MAP) (Meentemeyer 1978; Vitousek et al. 1994; Aerts 1997). Within
individual ecosystem or biome types, litter quality becomes a better determinant of
decomposition rates than climate (Meentemeyer 1978; Aerts 1997). At the ecosystem
scale, litter quality is most often related to the chemical characteristics of the litter, for
example carbon-to-nitrogen ratios and/or lignin content (Aber et al. 1990; Aerts 1997).
Support for this relationship is derived from a multitude of litter decomposition studies
which often employ experimental set-ups in which a range of litter types are decomposed
across a variety of different biomes (Vitousek et al. 1994; Gholz et al. 2000; Hobbie et
al. 2006; Hobbie et al. 2007; Parton et al. 2007). Based on this evidence most models of
187
ecosystem carbon dynamics rely heavily on the effects of climate (i.e. MAT and MAP) as
well as estimates of quality based on litter chemistry (Parton 1983; Heal 1997).
Models of decomposition based on climate and litter chemistry effectively ignore
the potential influence of microbial community structure on rates of litter decomposition
(Zak et al. 2006; Reed & Martiny 2007). This omission is partly due to the difficulty
associated with manipulating microbial communities in situ and is also largely based on
the assumption that given their abundance, diversity, and ubiquity soil microbial
communities are, for the most part, redundant with regards to their functional attributes
(Cardinale et al. 2007; Jiang 2007; Verity et al. 2007). This assumption has been
questioned (Schimel 1995; Martiny et al. 2006; Zak et al. 2006; Ramette & Tiedje 2007;
Reed & Martiny 2007) but there are surprisingly few studies directly testing how distinct
microbial communities may perceive the decomposability of a given litter type. Indeed,
we do know that microbial communities are not homogeneous, varying across space and
time in composition due to environmental factors and/or historical contingencies (Andrén
& Balandreau 1999; Behan-Pelletier & Newton 1999; Wall & Virginia 1999; Fierer &
Jackson 2006; Martiny et al. 2006; Ramette & Tiedje 2007; Reed & Martiny 2007).
Furthermore recent work has demonstrated both implicitly and explicitly that structural
differences in the microbial community can influence ecosystem function (Balser &
Firestone 2005; Reed & Martiny 2007; Strickland et al. 2009). Such results demonstrate
that, like comparisons between individual species of microorganisms (Newell 1984;
Dowson et al. 1989; Osono & Takeda 2002), whole microbial communities also differ
with regards to their functional capacity and these differences may influence the process
of litter decomposition.
188
The specific mechanisms which lead to functional dissimilarity (see Strickland et
al. 2009) in litter decomposition amongst microbial communities remains poorly
understood. One possible reason that functional dissimilarity may arise is due to a
microbial community‘s past exposure to litters of differing chemistry (Hunt et al. 1988;
Gholz et al. 2000). Implicit evidence of this phenomenon has been shown for both the
degradation of subsequent applications of pesticides and litter residues (Taylor et al.
1996; Laha & Petrova 1997; Cookson et al. 1998) as well as in reciprocal litter transplant
experiments (Hunt et al. 1988; Gholz et al. 2000; Castanho & de Oliveira 2008; Vivanco
& Austin 2008). However, it is unknown whether or not the past exposure of a microbial
community to litter inputs of differing chemistry also relates to that community‘s ability
to decompose novel litters which themselves differ chemically. Such knowledge will
improve our ability to understand, and possibly predict, the factors which influence litter
decomposition rates.
To test whether or not past exposure of microbial communities to high or low
quality litters influences the decomposition of novel litters (to those communities) which
differ in litter chemistry, we used a ‗common garden‘ experimental approach. We paired
one of 12 soil inocula, six sourced from herbaceous habitats and six sourced from forest
habitats, with one of two litters which differed in their litter chemistry. We hypothesized
that soil microbial inocula from forested habitats would perceive the lower-nutrient,
higher-lignin litters, taken from a tree species, to be of higher ‗quality‘ than would
incoula sourced from habitats dominated by herbaceous cover. Similarly, we reasoned
that the higher-nutrient, lower-lignin litters, taken from a grass species, would be
perceived of equivalent ‗quality‘ by all inocula. This hypothesis was based on the
189
expectation that microbial communities which develop in forest habitats will be adapted
(or pre-adapted) to decompose lower-nutrient, higher-lignin litters due to a history of
exposure to them, which will have acted as a selection pressure and/or an environmental
filter (sensu Lambers 1998). Hunt et al. (1988) present field data which may be explained
by this hypothesis. To test the hypothesis our microcosm, experimental design permits
effects associated with the microbial community to be separated from co-varying factors
in the field. Henceforth, we use the term ‗perceived quality‘. This is, essentially, a biotic
definition of litter quality (see Fierer et al. 2005) because a litter‘s decomposition rate in
our experiment is a function of how it is ―perceived‖ by the microbial community.
Methods
Microcosm design
The tree foliar litter was taken from Rhododendron maximum collected from the Coweeta
Long Term Ecological Research (LTER) site, North Carolina, USA (35º00‘N, 83º30‘W)
as recent, senesced litterfall. The herbaceous litter was taken from Panicum virgatum and
was collected from the University of Georgia, Georgia, USA (33º53‘N, 83º21‘W) as
recent, standing dead material. We chose these species because they have litters of low
and high chemical quality, respectively (Table 5.1). That is, R. maximum litter had a C:N
ratio which was nearly twice as great as, and a lignin concentration greater than twice that
of, P. virgatum litter. Neither litter was present at the sites from where the inocula were
sourced. This design ensured that effects of differing litter chemistries were not
confounded by prior exposure to the litter species used (i.e. ‗home-field advantage‘,
sensu Gholz et al. 2000). Litters were air-dried, passed twice through a Wiley mill (2 mm
190
mesh), and then sterilised by autoclaving twice in succession and again 24 h later (121°C,
20 min), before being dried to constant mass at 65C. The milling was done to increase
surface area for microbial colonization, to make the material between replicates more
homogenous, and to remove the influence of physical litter structure on decomposition
rates. For example, the slow decomposition rate of R. maximum foliar litters in the field
(e.g. Ball et al. 2008) is thought to result from both its chemical composition and
physically-tough leaves.
Soils for use as inocula were collected from, or adjacent to, the Calhoun
Experimental Forest (CEF), which is managed by the US Department of Agriculture and
located in the Piedmont region (approximately 34.5°N, 82°W) of northwestern South
Carolina, USA (Gaudinski et al. 2001; Callaham et al. 2006). Twelve soil inocula were
collected from four land-uses (cultivated, pasture, pine, and hardwood) within this region,
each land-use being represented three times (4 land-uses 3 locations = 12 inocula). This
provided six inocula from forest habitats (pine and hardwood) and six from habitats
dominated by herbaceous vegetation (cultivated and pasture). Each inoculum was a
subsample from soil samples composed of 10 individual A horizon soil cores (8 cm dia.,
0 - 7.5 cm depth), which were collected from a 100 m2 plot within each of the 12 sites
using a stratified random sampling approach. Soils were sieved (4 mm), homogenized,
and stored at +5ºC until use.
Microcosms were constructed by adding 0.5 g dry mass equivalent of soil
inoculum to 1 g of litter, which was then mixed by vortexing in a 50 mL plastic
centrifuge tube. The mixture was adjusted to and maintained at 50% water-holding
capacity, which is favourable for microbial activity. Tubes were incubated at 20°C and
191
100% humidity during the experiment. Our design was a 2 12 combinatorial set-up (i.e.
all litters crossed with all soil inocula) plus additional ―no-litter‖ soils. This resulted in
two inocula treatments, herbaceous and forest. The herbaceous inocula were composed of
three inocula sourced from cultivated sites and three sourced from pasture sites giving a
total of 6 herbaceous replicates. The forest inocula were composed of three inocula
sourced from hardwood stands and three inocula sourced from pine stands giving a total
of 6 forest replicates. Six analytical repeats were constructed for each inoculum × litter
combination giving 144 experimental units (i.e. 12 inocula × 2 litters × 6 analytical
repeats). To avoid pseudoreplication the mean was taken across the 6 analytical repeats
for a given inoculum × litter combination resulting in 6 herbaceous and 6 forest replicates
per litter type. The soil inocula accounted for an average of ~ 5% of the total carbon (C)
dioxide (CO2) flux across all treatment combinations and soil C (much of which is not
likely to be bioavailable) accounted for <20% of total C in each microcosm (see Table
5.1, see Appendix H).
Determination of decomposition rates and litter chemistry
Litter decomposition rates were estimated by periodic measurement of CO2 production
rates from the microcosms. This was done across 50 days using a 24 h static incubation
procedure described in Fierer et al. (2003). Before the start of the incubation period, soil-
litter mixtures were incubated for 10 days to allow microbial colonization of the litters.
Following this period CO2 production rates were measured on days 1, 5, 10, 20, 30, 40,
and 50 and were corrected for soil contributions by subtracting the production rates
192
measured from the corresponding ‗no-litter‘ soils. Cumulative CO2 production was
calculated by integrating values under the curve for the incubation period.
Total percentage C, nitrogen (N), non-fibrous material, hemi-cellulose, cellulose,
lignin, and pH, were determined for both litter types (Table 5.1). Total C and N were
determined using an NA1500 CHN Analyzer (Carlo Erba Strumentazione, Milan, Italy).
Non-fibrous material, hemi-cellulose, cellulose, and lignin concentrations were
determined using an Ankom A200 Fiber Analyzer (Ankom, Macedon, USA). Litter pH
was determined in water (2:1 ratio of water-to-litter) using a benchtop pH meter.
Determination of inoculum source edaphic characteristics
For each site from which an inoculum was sourced, we determined a series of edaphic
factors. Carbon and nutrient pools, microbial biomass, pH, soil texture, bulk density, and
cation concentrations were calculated from three analytical replicates per soil sample
from each inoculum source.
Total, POM (particulate organic material) associated, and mineral associated C
and N were determined using an NA1500 CHN Analyser (Carlo Erba Strumentazione,
Milan, Italy). Carbon and N associated with POM and mineral material were determined
using the fractionation method described in Bradford et al. (2008). Microbial biomass C
and N, and dissolved organic C (DOC), were determined using the method described in
Fierer and Schimel (2003). Ammonium (NH4+) and nitrate (NO3
-) were determined
colorimetrically following Fierer and Schimel (2002). Extractable phosphorus (P) was
measured on an Alpkem auto-analyzer (OI Analytical, College Station, TX) using
Murphy-Riley chemistry after extraction with Mehlich I double-acid (H2SO4 –HCl) using
193
a 1:4 mass:volume ratio (Kuo, 1996). Soil pH was determined in water (1:1 ratio of
water-to-soil) using a benchtop pH meter. Silt and clay contents were measured using a
simplified version of the hydrometer method as described by Gee and Orr (2002). Soil
bulk density was calculated after correcting for the mass and volume of roots and stones
(Culley 1993). Exchangeable cations (Ca+, Mg
+, and K
+) were measured by atomic
absorption in the presence of LaCl3 after extraction with 1 M NH4OAc (pH 7) using a
1:10 mass:volume ratio (Sumner & Miller 1996). Edaphic data for each inoculum source
are provided in Appendix H.
Determination of the starting community composition of the soil inocula
Soils for community analyses were collected and treated the same as those collected for
edaphic characterization. A DNA-based rather than an RNA-based assessment of
microbial communities was used because it probably better reflects the potential pool of
microbes that might develop on our microcosm litter environments from the soil incolua;
also DNA-based assessments are less variable and hence provide a more integrated
assessment of differences in microbial community inocula. DNA was isolated from 3
replicate sub-samples of fresh soil per plot using the MoBio Power Soil DNA Extraction
kit (MoBio Laboratories, Carlsbad, CA) with the modifications described in Lauber et al.
(2008). Each DNA sample was amplified in triplicate using both bacterial and fungal
specific primer sets described in Lauber et al. (2008) in order to assess bacterial and
fungal community composition in each sample. For each plot, we constructed an
individual clone library and sequenced 80-100 cloned amplicons per library following
standard protocols. Additional details on the PCR conditions, clone library construction,
194
and sequencing are provided in Lauber et al. (2008). To classify the fungal sequences, we
used the BLAST algorithm (Altschul et al. 1990) to compare the fungal sequences to an
in house data base of 100 fungal 18S sequences derived from the AFTOL data set
(Lutzoni et al. 2004). Bacterial sequences were aligned against the Greengenes data base
(http://greengenes.lbl.gov/cgi-bin/nph-index.cgi ) using the NAST alignment utility
(DeSantis et al. 2006a) and fungal sequences were aligned using MUSCLE 3.6 (Edgar
2004a). The sequences were chimera-checked using utilities available on the Greengenes
website (DeSantis et al. 2006b). Sequences which were of poor quality or suspected to
be chimeric were eliminated from the analysis (less than 9% of the sequences).
Sequences were aligned using MUSCLE 3.6 (Edgar 2004b) and a neighbor-joining
phylogenetic tree containing either all of the bacterial or all of the fungal sequences was
generated with PHYLIP 4.0. The phylogenetic distance between the bacterial and fungal
communities within each soil inoculum was determined using the weighted-normalized
UniFrac algorithm (Lozupone & Knight 2005). For full details and results concerning the
composition of microbial communities, see Lauber et al. (2008). Briefly, Lauber et al.
(2008) found that bacterial community composition was related to both soil pH and soil
texture whereas fungal community composition was related to both extractable P and soil
C:N ratios.
Statistical analyses
Statistical analyses of cumulative CO2 production were performed in S-Plus 7.0
(Insightful Corp., Seattle, USA), using ANOVA with litter type and inoculum source as
discrete variables that were permitted to interact. Using this approach, a significant
195
inoculum or inoculum by litter type interaction would support our hypothesis that
resource history influences the perception of litter quality whereas only a litter type effect
would refute this hypothesis. We further explored our results using correlation and
regression approaches (see below). For statistical significance we assumed an -level of
0.05. When reported as such, data were loge-transformed to conform to assumptions of
homoscedasticity (verified using model checking).
Relationships between cumulative CO2 production, edaphic characteristics of the
source environment, and the starting community composition of the inocula were
assessed using Plymouth Routines in Multivariate Ecological Research v5 software
(Primer v5, Lutton, UK). Mantel tests were performed to determine if there were
significant correlations between community function on either R. maximum or P.
virgatum litter, edaphic characteristics of the source environment, and the starting whole
fungal or whole bacterial community composition of the inoculum (individual taxa were
examined using regression analyses, see below). Correlations between these factors were
considered significant when P<0.05.
Linear regression analyses were performed to look at the relationships between
cumulative CO2 production on either R. maximum or P. virgatum and the starting edaphic
characteristics of the inocula. This was only conducted for those edaphic characteristics
which Mantel tests showed were significantly correlated to cumulative CO2 production.
Regression analyses were also used to look at relationships between cumulative CO2
production and the dominant bacterial and fungal taxa in the starting inocula. We were
able to look at the relationships between cumulative CO2 production and the dominant
bacterial and fungal taxa by performing phylogenetic independent analyses which
196
recorded the distribution of taxa for each site and normalized these values to represent the
percentage of each taxon within a given inoculum. Regression analyses, like ANOVAs,
were performed using S-Plus 7.0.
Results
Litter decomposition
During this 50 day incubation, between 2.4 and 3.1% of R. maximum litter C was lost as
CO2 and between 3.1 and 4.2% of P. virgatum litter C was lost as CO2. When examining
cumulative CO2 production, we found a significant inoculum by litter type interaction
(P<0.01; Fig. 5.1). This interaction suggested that the quality of past resources influences
how the community inocula perceive the quality of the two litters. This difference is
clearly represented in Fig. 5.1 - cumulative CO2 production differs from one litter type to
the other for the different inocula. Specifically, cumulative CO2 production was nearly
identical for the forest inoculum on either litter resource suggesting that there was no
difference in perceived quality. In contrast, the herbaceous inocula perceived the P.
virgatum litter to be of higher quality than the R. maximum litter. To further explore this
interaction we examined each litter type separately. We did not detect a significant
inoculum effect on cumulative CO2 production for P. virgatum litter (F1,10 = 2.40;
P=0.15) but did detect a significant inoculum effect on cumulative CO2 production for R.
maximum litter (F1,10 = 25.80; P<0.001). This suggested that the interaction between
inoculum and litter type was driven by minimal differences in CO2 production on P.
virgatum litter while CO2 production did differ between inocula on R. maximum litter. To
further explain these results and to assess other cover-type related differences in the
197
perception of litter quality, we explored relationships between cumulative CO2
production and the edaphic and microbial community characteristics of each inoculum
(see next).
Relationship of litter decomposition rates to inoculum characteristics
When examining how cumulative CO2 production from R. maximum litter was related to
the microbial community composition of the inocula, we found significant relationships
with both fungal and bacterial community composition (P<0.05 for fungi and P<0.01 for
bacteria; Table 5.2). Regression analyses showed that the relative abundances of
Sordariomycete, Leotiomycete, Chaetothyriomycete, and Agaricales (the dominant
fungal taxa in these inocula; Lauber et al. (2008)) were significantly related to cumulative
CO2 production (Fig. 5.2a-d). The relative abundances of Leotiomycete and
Sordariomycete were negatively related to cumulative CO2 production (Fig. 5.2a-b). The
relative abundances of Agaricales and Chaetothyriomycete were positively related (Fig.
5.2c-d). Of the bacterial taxa, the relative abundances of Acidobacteria in the starting
inocula were positively related to cumulative CO2 production on R. maximum (Fig. 5.2e).
We also found that site specific differences in the soil chemical and physical
characteristics from which an inoculum was sourced were related to cumulative CO2
production on R. maximum (Table 5.2). More specifically, we found that NO3-, NH4
+,
extractable P, K+, and percentage clay were all significantly related to cumulative CO2
production on R. maximum (P<0.01 for NO3- and NH4
+, P<0.05 for P, K
+, and %clay;
Table 5.2). All except for NH4+ were negatively related to cumulative CO2 production on
R. maximum (Fig. 5.3).
198
When examining how cumulative CO2 production from P. virgatum litter was
related to the microbial community composition of the inocula, we found that only the
composition of the bacterial community was significantly related (P<0.05; Table 5.2),
while the fungal community composition was not (Table 5.2). Regression analyses of
bacterial taxa showed that only the relative abundance of Actinobacteria was positively
related to cumulative CO2 production (Fig. 5.2f). Of the soil chemical and physical
characteristics, only K+ concentration was positively related to cumulative CO2
production (P<0.05; Table 5.2, Fig. 5.3f).
Discussion
We initially hypothesized that resource quality history would influence how microbial
communities perceived litters which differed in chemical quality. Our expectation was
that microbial communities which develop in forest habitats will be adapted (or pre-
adapted) to decompose lower chemical quality litters due to their history of exposure to
lower quality litters (Hunt et al. 1988; Gholz et al. 2000). We also expected that
microbial communities, regardless of their previous history, would be equally well
adapted to decompose high chemical quality litters (Hunt et al. 1988). As hypothesized,
we found that those microbial communities sourced from forest habitats perceived the
low chemical quality R. maximum litter to be of higher quality than did the communities
sourced from herbaceous habitats (Fig. 5.1). This result is similar to results from
reciprocal litter transplant studies which have shown that litters often decompose more
rapidly when placed in their native forests (Cookson et al. 1998; Castanho & de Oliveira
2008; Vivanco & Austin 2008). Also confirming our original hypothesis was the
199
observation that communities sourced from either habitat generally perceived the high
chemical quality P. virgatum litter equally (Fig. 5.1), a result also observed in studies
which have found no differences in decomposition rates across sites for the same litter
type (Prescott et al. 2000; Ayres et al. 2006).
Initial litter decomposition rates with inocula sourced from herbaceous sites
confirmed the expectation (Aber et al. 1990; Aerts 1997) that litter chemistry is a good
predictor of decomposition rates. That is, the R. maximum litter with its greater C:N ratio
and higher lignin content decomposed more slowly than the P. virgatum litter with its
lower C:N ratio and lower lignin content (Fig. 5.1). In contrast to the inocula sourced
from the herbaceous sites, the decomposition rates observed with the forest inocula were
similar for the two litter types (Fig. 5.1). This result agrees with those studies which have
found that litter chemistry is not a strong predictor of litter decomposition rates (Hunt et
al. 1988; Gholz et al. 2000) and provides empirical support for the suggestion from these
previous studies that the decomposer community characteristics affect litter
decomposition rates. Overall, our results imply that an understanding of the microbial
community and/or its past resource environment may increase our ability to predict
decomposition dynamics. Indeed, the community or the resource history of that
community may account for some of the unexplained variation in decomposition models
where climate and litter quality are used as explanatory variables (e.g. Gholz et al. 2000,
Parton et al. 2007). Our microcosm experiment demonstrates the potential for both
decomposer communities and the resource history of the decomposer community to
influence decomposition rates, but it remains to be determined if our results are relevant
to decomposition dynamics observed in the field.
200
To further explore how differences in microbial community structure and resource
quality history impact the perception of litter quality we assessed the relationship
between cumulative CO2 production of both litter types and the biological, chemical, and
physical characteristics of the sites from which each inoculum was sourced. When
examining the biological factors of the inocula sources which were related to cumulative
CO2 production from either R. maximum or P. virgatum litters, we found that both fungal
and bacterial community composition was related to the cumulative CO2 production from
R. maximum litter but only the bacterial community composition was significantly related
to the cumulative CO2 production from P. virgatum litter (Table 5.2). The fact that the
fungal community was related to cumulative CO2 production on R. maximum but not on
P. virgatum litter may be due to the ability of fungi to degrade litters of lower-nutrient
and higher-lignin content, or at least out-compete bacteria in such environments (Six et
al. 2006). When more closely examining the fungal community, we found that four
fungal taxa (Sordariomycete, Leotiomycete, Chaetothyriomycete, and Agaricales) of the
inocula sources were related to the cumulative CO2 production on R. maximum litter. We
found that those inocula sources which had comparably higher relative abundances of
either Sordariomycetes or Leotiomycetes had lower cumulative CO2 production on R.
maximum litter than sites with lower abundances of these groups (Fig. 5.2a, 5.2b).
Conversely, those inocula sourced from sites having a higher relative abundance of
Chaetothyriomycete or Agaricales tended to have higher cumulative CO2 production on
R. maximum litter than did sites with lower abundances of either of these fungal taxa
(Fig. 5.2c, 5.2d). This finding is similar to observations which have demonstrated
differences in the litter degrading abilities of fungal phyla (Osono & Takeda 2002). Our
201
observations suggest that such differences may also occur within phyla given that within
the phylum Ascomycota the abundance of one taxon (Chaetothyriomycete) was
positively related, while two other taxa (Sordariomycetes and Leotiomycetes) were
negatively related, to cumulative CO2 production from R. maximum litter (Fig. 5.2a-d).
When considering the bacterial community, we found that those inocula sourced
from sites with higher relative abundances of Acidobacteria and Actinobacteria were
positively related to cumulative CO2 production on R. maximum litter and P. virgatum
litter, respectively (Fig. 5.2e, 5.2f). Acidobacteria have recently been associated with
resource environments of lower quality (Fierer et al. 2007) and this may explain why a
positive relationship was observed on R. maximum litter which has both a high C:N ratio
and lignin content (Fig. 5.2e). When considering the Actinobacteria, relatively little is
known concerning their ecological attributes. However, due to their filamentous growth
form and a presumed ability to effectively scavenge nutrients (Goodfellow & Williams
1983; Steger et al. 2007), those inocula which started with a higher relative abundance of
Actinobacteria may be better at decomposing P. virgatum litter across the course of our
experiment.
Many more chemical and physical factors of the inocula sources were related to
cumulative CO2 production on R. maximum than on P. virgatum (Table 5.2). Of those
factors that significantly correlated with CO2 production on R. maximum litter we found
that all but one was negatively related (Fig. 5.3a-e). Specifically, we observed that NO3-,
extractable P, K+, and percentage clay were negatively related to cumulative CO2
production on R. maximum while NH4+ was positively related. Only K
+ was related to
cumulative CO2 production on P. virgatum and this relationship was positive (Table 5.2,
202
Fig. 5.3f). Soil physical and chemical factors are often products of management. In the
region where our inocula were sourced NO3-, extractable P, and K
+ are generally higher
in cultivated and pasture sites (than forests) due to annual fertilizer inputs (Richter et al.
2000). Percentage clay is generally higher in cultivated sites due to long-term intensive
agriculture and weathering (Richter & Markewitz 2001; Callaham et al. 2006) and NH4+
is generally higher in forest sites due to lower levels of nitrification (Adams 1986). These
relationships between chemical and physical variables and the source environments of the
soil microbial inocula make it difficult to determine whether the chemistry of the litter
inputs in the source environments, the nutrient status of the environments, or both were
important factors in structuring the function of the microbial inocula in our litter
microcosms. Notably, differences between the inocula in their chemical and physical
characteristics, and microbial community composition, did not follow land use
differences exactly (Table S1 and Lauber et al. 2008). This may suggest that litter
chemistry (i.e. herbaceous vs. woody plant litters) was the over-riding cause of
differences in the functioning of our inocula on the R. maximum and P. virgatum litters.
Our results provide a foundation upon which future studies can investigate how
historical factors that may shape microbial communities may influence their function. In
future work, investigation of how microbial community composition develops on
different litter types and across time may facilitate a more detailed understanding of the
relationships between the taxonomic composition of microbial communities and their
function (Osono 2005; Trinder et al. 2008; Strickland et al. 2009). Such studies might
address whether compositionally distinct communities, when faced with similar
environments, develop in the same manner. They should consider which characteristics
203
(e.g. lignin content) of litter chemistry likely influence this development both at earlier
and later stages of decomposition. They should also ask how long differences in the
functioning of microbial communities might persist once different communities are
exposed to the same environment. Given the rapid generation times of microbes it seems
plausible that microbial community functioning might rapidly converge in the same
environment (Strickland et al. 2009). Such information will add to our understanding of
the process of decomposition and may also enable prediction of how a given microbial
community will respond to environmental change.
Our study involved a short-term, laboratory, common garden approach to
determine the significance of microbial community structure and its resource history for
the initial decomposition rate of leaf litters of differing chemical composition. There are a
number of criticisms that can be leveled at our approach. These include the fact that we
do not know whether microfauna, which affect the activities of microbial decomposers,
were present in our inocula or not. Certainly, the functional symbiosis of plants and
mycorrhizal fungi, which may play a crucial role in decomposition (Talbot et al. 2008),
was absent. In addition autoclaving, as any sterilization technique, impacts soil and
potentially litter properties (Berns et al. 2008). However, in our study such non-target
effects would have been consistent within a given litter type or inoculum and hence
would not have been able to explain our most significant finding. That is litter chemistry
impacting decomposition rates of litters inoculated with soils from herbaceous but not
forested environments.
What our approach did provide was a tightly-controlled experiment which
demonstrated that the past resource environment of a microbial community impacted its
204
ability to decompose litters differing in their chemistry. We cannot conclude to what
extent this effect influences decomposition rates in the field, but it does offer one
explanation for why different litter chemistries do not always lead to expected differences
in decomposition rates in situ (Hunt et al. 1988; Gholz et al. 2000; Castanho & de
Oliveira 2008; Vivanco & Austin 2008). Our work suggests that the differential
perception of litter chemistry by different microbial inocula may be explained by the
starting, community composition of microbial inocula and/or the resource status of the
environment from which they were derived. To fully understand decomposition dynamics
in the environment across space and time may require recognition and further exploration
of the role of microbial community composition and that community‘s resource history as
a driving variable.
Acknowledgements
We gratefully acknowledge funding from the Andrew W. Mellon Foundation, and from
the National Science Foundation to the Coweeta LTER. We thank Tom Maddox in the
Analytical Chemistry Laboratory of the Odum School of Ecology, Univ. of Georgia, for
element analyses; Dan Richter and Mac Callaham for establishing the permanent plots in
the Calhoun Experimental Forest; and Mr. Jack Burnett for allowing us to collect samples
on his property.
205
References
Aber, J.D., Melillo, J.M. & McClaugherty, C.A. (1990) Predicting long-term patterns of
mass-loss, nitrogen dynamics, and soil organic-matter formation from initial fine
litter chemistry in temperate forest ecosystems. Canadian Journal of Botany-
Revue Canadienne De Botanique, 68, 2201–2208.
Adams, J.A. (1986) Nitrification and ammonification in acid forest litter and humus as
affected by peptone and ammonium-n amendment. Soil Biology and
Biochemistry, 18, 45–51.
Aerts, R. (1997) Climate, leaf litter chemistry and leaf litter decomposition in terrestrial
ecosystems: a triangular relationship. Oikos, 79, 439–449.
Altschul, S.F., Gish, W., Miller, W., Myers, E.W. & Lipman, D.J. (1990) Basic local
alignment search tool. Journal of Molecular Biology, 215, 403–410.
Andrén, O. & Balandreau, J. (1999) Biodiversity and soil functioning – from black box
to can of worms? Applied Soil Ecology, 13, 105–108.
Ayres, E., Dromph, K.M. & Bardgett, R.D. (2006) Do plant species encourage soil biota
that specialise in the rapid decomposition of their litter? Soil Biology and
Biochemistry, 38, 183–186.
Ball, B.A., Hunter, M.D., Kominoski, J.S., Swan, C.M. & Bradford, M.A. (2008)
Consequences of non-random species loss for decomposition dynamics:
experimental evidence for additive and non-additive effects. Journal of Ecology,
96, 303–313.
206
Balser, T.C. & Firestone, M.K. (2005) Linking microbial community composition and
soil processes in a California annual grassland and mixed-conifer forest.
Biogeochemistry, 73, 395–415.
Behan-Pelletier, V. & Newton, G. (1999) Computers in biology – linking soil
biodiversity and ecosystem function – the taxonomic dilemma. Bioscience, 49,
149–153.
Berns, A.E., Philipp, H., Narres, H.D., Burauel, P., Vereecken, H. & Tappe, W. (2008)
Effect of γ-sterilization and autoclaving on soil organic matter structure as studied
by solid state nmr, uv and fluorescence spectroscopy. European Journal of Soil
Science, 59, 540–550.
Bradford, M.A., Fierer, N., Jackson, R.B., Maddox, T.R. & Reynolds, J.F. (2008)
Nonlinear root-derived carbon sequestration across a gradient of nitrogen and
phosphorous deposition in experimental mesocosms. Global Change Biology, 14,
1113–1124.
Callaham, M.A., Richter, D.D., Coleman, D.C. & Hofmockel, M. (2006) Long-term
land-use effects on soil invertebrate communities in southern piedmont soils,
USA. European Journal of Soil Biology, 42, S150–S156.
Cardinale, B.J., Wright, J.P., Cadotte, M.W., Carroll, I.T., Hector, A., Srivastava, D.S.,
Loreau, M. & Weis, J.J. (2007) Impacts of plant diversity on biomass production
increase through time because of species complementarity. Proceedings of the
National Academy of Sciences USA, 104, 18123–18128.
207
Castanho, C.D. & de Oliveira, A.A. (2008) Relative effect of litter quality, forest type
and their interaction on leaf decomposition in south-east Brazilian forests. Journal
of Tropical Ecology, 24, 149–156.
Cookson, W.R., Beare, M.H. & Wilson, P.E. (1998) Effects of prior crop residue
management on microbial properties and crop residue decomposition. Applied
Soil Ecology, 7, 179–188.
Couteaux, M.M., Bottner, P. & Berg, B. (1995) Litter decomposition, climate and litter
quality. Trends in Ecology and Evolution, 10, 63–66.
Culley, J.L.B. (1993) Density and compressibility. Lewis Publishers, Boca Raton,
Florida, USA.
DeSantis, T., Hugenholtz, P., Keller, K., Brodie, E.L., Larsen, N., Piceno, Y.M., Phan, R.
& Andersen, G.L. (2006a) Nast: a multiple sequence alignment server for
comparative analysis of 16s rRNA genes. Nucleic Acids Research, 34, W394–
W399.
DeSantis, T.Z., Hugenholtz, P., Larsen, N., Rojas, M., Brodie, E.L., Keller, K., Huber,
T., Dalevi, D., Hu, P. & Andersen, G.L. (2006b) Greengenes, a chimera-checked
16s rRNA gene database and workbench compatible with arb. Applied and
Environmental Microbiology, 72, 5069–5072.
Dowson, C.G., Springham, P., Rayner, A.D.M. & Boddy, L. (1989) Resource
relationships of foraging mycelial systems of phanerochaete-velutina and
hypholoma-fasciculare in soil. New Phytologist, 111, 501–509.
Edgar, R.C. (2004) Muscle: multiple sequence alignment with high accuracy and high
throughput. Nucleic Acids Research, 32, 1792–1797.
208
Fierer, N. & Jackson, R.B. (2006) The diversity and biogeography of soil bacterial
communities. Proceedings of the National Academy of Sciences USA, 103, 626–
631.
Fierer, N. & Schimel, J.P. (2002) Effects of drying-rewetting frequency on soil carbon
and nitrogen transformations. Soil Biology & Biochemistry, 34, 777–787.
Fierer, N. & Schimel, J.P. (2003) A proposed mechanism for the pulse in carbon dioxide
production commonly observed following the rapid rewetting of a dry soil. Soil
Science Society of America Journal, 67, 798–805.
Fierer, N., Allen, A.S., Schimel, J.P. & Holden, P.A. (2003) Controls on microbial CO2
production: a comparison of surface and subsurface soil horizons. Global Change
Biology, 9, 1322–1332.
Fierer, N., Craine, J.M., McLauchlan, K. & Schimel, J.P. (2005) Litter quality and the
temperature sensitivity of decomposition. Ecology, 86, 320–326.
Gaudinski, J.B., Trumbore, S.E., Davidson, E.A., Cook, A.C., Markewitz, D. & Richter,
D.D. (2001) The age of fine-root carbon in three forests of the eastern united
states measured by radiocarbon. Oecologia, 129, 420–429.
Gee, G.W. & Or, D. (2002) Particle-size analysis. Methods of Soil Analysis, Part 4:
Physical Methods (eds JH Dane & GC Topp), pp. 255–293. Soil Science Society
of America, Inc., Madison, Wisconsin, USA.
Gholz, H.L., Wedin, D.A., Smitherman, S.M., Harmon, M.E. & Parton, W.J. (2000)
Long-term dynamics of pine and hardwood litter in contrasting environments:
toward a global model of decomposition. Global Change Biology, 6, 751–765.
209
Goodfellow, M. & Williams, S.T. (1983) Ecology of actinomycetes. Annual Review of
Microbiology, 37, 189–216.
Heal, O.W., Anderson, J.M. & Swift, M.J. (1997) Plant Litter Quality and
Decomposition: An Historical Overview. CAB International, Wallingford, UK.
Hobbie, S.E., Reich, P.B., Oleksyn, J., Ogdahl, M., Zytkowiak, R., Hale, C. &
Karolewski, P. (2006) Tree species effects on decomposition and forest floor
dynamics in a common garden. Ecology, 87, 2288–2297.
Hobbie, S.E., Ogdahl, M., Chorover, J., Chadwick, O.A., Oleksyn, J., Zytkowiak, R. &
Reich, P.B. (2007) Tree species effects on soil organic matter dynamics: the role
of soil cation composition. Ecosystems, 10, 999–1018.
Hunt, H.W., Ingham, E.R., Coleman, D.C., Elliott, E.T. & Reid, C.P.P. (1988) Nitrogen
limitation of production and decomposition in prairie, mountain meadow, and
pine forest. Ecology, 69, 1009–1016.
Jiang, L. (2007) Negative selection effects suppress relationships between bacterial
diversity and ecosystem functioning. Ecology, 88, 1075–1085.
Laha, S. & Petrova, K.P. (1997) Biodegradation of 4-nitrophenol by indigenous
microbial populations in everglades soils. Biodegradation, 8, 349–356.
Lambers, H., Chapin, F.S. III & Pons, T.L. (1998) Plant physiological ecology. Springer-
Verlag, New York, USA.
Lauber, C.L., Strickland, M.S., Bradford, M.A. & Fierer, N. (2008) Soil properties that
structure bacterial and fungal communities across a land-use gradient. Soil
Biology and Biochemistry, 40, 2407–2415.
210
Lozupone, C. & Knight, R. (2005) Unifrac: a new phylogenetic method for comparing
microbial communities. Applied and Environmental Microbiology, 71, 8228–
8235.
Lutzoni, F., Kauff, F., Cox, C.J., McLaughlin, D., Celio, G., Dentinger, B., Padamsee,
M., Hibbett, D., James, T.Y., Baloch, E., Grube, M., Reeb, V., Hofstetter, V.,
Schoch, C., Arnold, A.E., Miadlikowska, J., Spatafora, J., Johnson, D.,
Hambleton, S., Crockett, M., Shoemaker, R., Hambleton, S., Crockett, M.,
Shoemaker, R., Sung, G.H., Lucking, R., Lumbsch, T., O'Donnell, K., Binder, M.,
Diederich, P., Ertz, D., GueiDan, C., Hansen, K., Harris, R.C., Hosaka, K., Lim,
Y.W., Matheny B., Nishida, H., Pfister, D., Rogers, J., Rossman, A., Schmitt, I.,
Sipman, H., Stone, J., Sugiyama, J., Yahr, R. & Vilgalys, R. (2004) Assembling
the fungal tree of life: progress, classification and evolution of subcellular traits.
American Journal of Botany, 91, 1446–1480.
Martiny, J.B.H., Bohannan, B.J.M., Brown, J.H., Colwell, R.K., Fuhrman, J.A., Green,
J.L., Horner-Devine, M.C., Kane, M., Krumins, J.A., Kuske, C.R., Morin, P.J.,
Naeem, S., Ovreas, L., Reysenbach, A.L., Smith, V.H. & Staley, J.T. (2006)
Microbial biogeography: putting microorganisms on the map. Nature Reviews
Microbiology, 4, 102–112.
Meentemeyer, V. (1978) Macroclimate and lignin control of litter decomposition rates.
Ecology, 59, 465–472.
Newell, K. (1984) Interaction between 2 decomposer basidiomycetes and a collembolan
under sitka spruce – grazing and its potential effects on fungal distribution and
litter decomposition. Soil Biology and Biochemistry, 16, 235–239.
211
Osono, T. & Takeda, H. (2002) Comparison of litter decomposing ability among diverse
fungi in a cool temperate deciduous forest in Japan. Mycologia, 94, 421–427.
Osono, T. (2005) Colonization and succession of fungi during decomposition of Swida
controversa leaf litter. Mycologia, 97, 589–597.
Parton, W., Silver, W.L., Burke, I.C., Grassens, L., Harmon, M.E., Currie, W.S., King,
J.Y., Adair, E.C., Brandt, L.A., Hart, S.C. & Fasth B. (2007) Global-scale
similarities in nitrogen release patterns during long-term decomposition. Science,
315, 361–364.
Parton, W.J., Anderson, D.W., Cole, C.V. & Stewart, J.W.B. (1983) Simulation of Soil
Organic Matter Formation and Mineralization in Semiarid Agroecosystems. The
University of Georgia, College of Agriculture Experiment Stations, Special Publ.,
Athens, Georgia.
Prescott, C.E., Zabek, L.M., Staley, C.L. & Kabzems, R. (2000) Decomposition of
broadleaf and needle litter in forests of British Columbia: Influences of litter type,
forest type, and litter mixtures. Canadian Journal of Forest Research-Revue
Canadienne De Recherche Forestiere, 30, 1742–1750.
Raich, J.W. & Schlesinger, W.H. (1992) The global carbon-dioxide flux in soil
respiration and its relationship to vegetation and climate. Tellus Series B-
Chemical and Physical Meteorology, 44, 81–99.
Ramette, A. & Tiedje, J.M. (2007) Multiscale responses of microbial life to spatial
distance and environmental heterogeneity in a patchy ecosystem. Proceedings of
the National Academy of Sciences USA, 104, 2761–2766.
212
Reed, H.E. & Martiny, J.B.H. (2007) Testing the functional significance of microbial
composition in natural communities. FEMS Microbiology Ecology, 62, 161–170.
Richter, D.D. & Markewitz, D. (2001) Understanding soil change: Soil sustainability
over millenia, centuries, and decades. Cambridge University Press, New York.
Richter, D.D., Markewitz, D., Heine, P.R., Jin, V., Raikes, J., Tian, K. & Wells, C.G.
(2000) Legacies of agriculture and forest regrowth in the nitrogen of old-field
soils. Forest Ecology and Management, 138, 233–248.
Schimel, J.P. (1995) Ecosystem Consequences of Microbial Diversity and Community
Structure. Springer-Verlag, Berlin, Germany.
Six, J., Frey, S.D., Thiet, R.K. & Batten, K.M. (2006) Bacterial and fungal contributions
to carbon sequestration in agroecosystems. Soil Science Society of America
Journal, 70, 555–569.
Steger, K., Jarvis, A., Vasara, T., Romantschuk, M. & Sundh, I. (2007) Effects of
differing temperature management on development of actinobacteria populations
during composting. Research in Microbiology, 158, 617–624.
Strickland, M.S., Lauber, C., Fierer, N. & Bradford, M.A. (2009) Testing the functional
significance of microbial community composition. Ecology , 90, 441-451.
Sumner, M.E. & Miller, W.P. (1996) Cation exchange capacity and exchange
coefficients. Methods of Soil Analysis, Part 3. Chemical Methods (ed. DL
Sparks), pp. 1201–1230. Soil Science Society of America, Madison, WI.
Talbot, J.M., Allison, S.D. & Treseder, K.K. (2008) Decomposers in disguise:
mycorrhizal fungi as regulators of soil C dynamics in ecosystems under global
change. Functional Ecology, 22, 955-963.
213
Taylor, G.E., Schaller, K.B., Geddes, J.D., Gustin, M.S., Lorson, G.B. & Miller, G.C.
(1996) Microbial ecology, toxicology and chemical fate of methyl isothiocyanate
in riparian soils from the upper sacramento river. Environmental Toxicology and
Chemistry, 15, 1694–1701.
Trinder, C.J., Johnson, D. & Artz, R.R.E. (2008) Interactions among fungal community
structure, litter decomposition and depth of water table in a cutover Peatland.
FEMS Microbiology Ecology, 64, 433–448.
Verity, P.G., Brussaard, C.P., Nejstgaard, J.C., van Leeuwe, M.A., Lancelot, C. &
Medlin, L.K. (2007) Current understanding of phaeocystis ecology and
biogeochemistry, and perspectives for future research. Biogeochemistry, 83, 311–
330.
Vitousek, P.M., Turner, D.R., Parton, W.J. & Sanford, R.L. (1994) Litter decomposition
on the Mauna-Loa environmental matrix, Hawaii – patterns, mechanisms, and
models. Ecology, 75, 418–429.
Vivanco, L. & Austin, A.T. (2008) Tree species identity alters forest litter decomposition
through long-term plant and soil interactions in Patagonia, Argentina. Journal of
Ecology, 96, 727–736.
Wall, D.H. & Virginia, R.A. (1999) Controls on soil biodiversity: Insights from extreme
environments. Applied Soil Ecology, 13, 137–150.
Zak, D.R., Blackwood, C.B. & Waldrop, M.P. (2006) A molecular dawn for
biogeochemistry. Trends in Ecology and Evolution, 21, 288–295.
214
Table 5.1 Initial %C, %N, C:N ratios, %Lignin, and pH of the litters.
Litter
Type
%C %N C:N %Non-
fibrous
%Hemi-
cellulose
%Cellulose %Lignin pH
R.
maximum
48.9 ±
0.24
0.42 ±
0.01
116 ±
1.70
60.91 ±
1.23
8.81 ±
0.22
17.73 ±
0.22
12.54 ±
1.15
4.5 ±
0.01
P.
virgatum
41.8 ±
0.07
0.62 ±
0.02
68 ±
2.32
38.14 ±
0.25
27.53 ±
0.52
29.08 ±
0.18
5.24 ±
0.38
5.2 ±
0.03
Values shown are means ± 1 SE (n=3).
215
Table 5.2 Mantel test results showing correlations between either inoculum source
edaphic factors or initial inoculum community composition and cumulative CO2
production on either R. maximum or P. virgatum litter. Spearman‘s correlation
coefficients relate the calculated Euclidian distance between cumulative CO2 production
and edaphic factors; UniFrac distance is used for the community factors. Significant
correlation coefficients are reported in bold (P<0.05).
R. maximum litter P. virgatum Litter
Spearman's
correlation
coefficient
P-value Spearman's
correlation
coefficient
P-value
Edaphic factors
DOC (mg g dry weight soil-1
) 0.104 ns -0.063 ns
NO3- (μg g dry weight soil
-1)* 0.750 <0.01 0.044 ns
NH4+
(μg g dry weight soil-1
)* 0.480 <0.01 -0.20 ns
Total C (mg g dry weight soil-1
) 0.089 ns -0.137 ns
POM C (mg g dry weight soil-1
) 0.034 ns -0.108 ns
Mineral C (mg g dry weight soil-1
) -0.072 ns -0.029 ns
Total N (mg g dry weight soil-1
) -0.013 ns -0.012 ns
POM N (mg g dry weight soil-1
) 0.069 ns -0.073 ns
Mineral N (mg g dry weight soil-1
) 0.062 ns 0.203 ns
Extractable P (μg g dry weight soil-
1)*
0.314 <0.05 -0.032 ns
Ca+
(mg g dry weight soil-1
) 0.026 ns 0.034 ns
Mg+
(mg g dry weight soil-1
) -0.03 ns 0.120 ns
K+
(mg g dry weight soil-1
)* 0.251 <0.05 0.379 <0.05
C:N 0.094 ns 0.114 ns
C:P 0.152 ns -0.14 ns
N:P 0.101 ns -1.38 ns
Sand (%) 0.134 ns 0.032 ns
Silt (%) -0.145 ns -0.196 ns
Clay (%)* 0.244 <0.05 0.189 ns
pH -0.061 ns -0.014 ns
SIR (μg g dry weight soil-1
) 0.088 ns -0.077 ns
Microbial C (μg g dry weight soil-1
) 0.060 ns -0.145 ns
Microbial N (μg g dry weight soil-1
) 0.136 ns -0.167 ns
Community factors
Bacteria 0.417 <0.01 0.356 <0.05
Fungi 0.327 <0.05 0.017 ns
*Data were loge transformed for both litters, except for K+ which was only transformed
for R. maximum.
216
Figure legends
Figure 5.1 Cumulative CO2 production from microcosms consisting of microbial inocula
sourced from four land-uses representing two cover-types combined with either R.
maximum or P. virgatum litters. Land-uses representing herbaceous cover-types are the
cultivated and pasture inocula and land-uses representing forest cover-types are the pine
and hardwood inocula. Values are means±1SE (n = 6). ANOVA results are based on
cover-type. A significant interaction between inoculum and litter was detected for
cumulative CO2 production (F1,20=12.71; P<0.01). Main effects of litter type and
inoculum were significant (F1,20=10.39; P<0.01) and not significant (F1,20=0.583;
P=0.454), respectively.
Figure 5.2 Relationships between cumulative CO2 production and the relative
abundances of three fungal taxa (Sordariomycete, Leotiomycete, and Agaricales) and two
bacterial taxa (Acidobacteria and Actinobacteria) on R. maximum (a-e) and P. virgatum
litter (f). Circles, inverted triangles, diamonds, and squares represent cultivated, pasture,
pine, and hardwood land-uses, respectively. Each individual plot within a given land-use
is indicated by a different shade (e.g. Cultivated plot 1 (Cu1 in the legend) is represented
by a white circle).
Figure 5.3 Relationships between cumulative CO2 production and the edaphic
characteristics of the site from which the inocula were sourced on R. maximum (a-e) and
P. virgatum litter (f). Symbols are as in Fig. 5.2.
217
Figure 5.1
Inoculum ns
Litter **
Inoculum x Litter **
Litter type
P. virgatum R. maximum
Cum
ula
tive
CO
2-C
(m
g g
lit
ter-1
)
0
2
4
6
8
10
12
14
16
18
20
22Cultivated inoculum
Pasture inoculum
Pine inoculum
Hardwood inoculum
218
Figure 5.2
Source habitat
Cu1
Cu2
Cu3
Pa1
Pa2
Pa3
Cu1
Cu2
Cu3
Pa1
Pa2
Pa3
Pi1
Pi2
Pi3
Ha1
Ha2
Ha3
Pi1
Pi2
Pi3
Ha1
Ha2
Ha3
Source habitat
Cu1
Cu2
Cu3
Pa1
Pa2
Pa3
Cu1
Cu2
Cu3
Pa1
Pa2
Pa3
Pi1
Pi2
Pi3
Ha1
Ha2
Ha3
Pi1
Pi2
Pi3
Ha1
Ha2
Ha3
Cu1
Cu2
Cu3
Pa1
Pa2
Pa3
Cu1
Cu2
Cu3
Pa1
Pa2
Pa3
Pi1
Pi2
Pi3
Ha1
Ha2
Ha3
Pi1
Pi2
Pi3
Ha1
Ha2
Ha3
Acidobacteria (%)0 10 20 30 40 50 60
Agaricales (%)0 20 40 60 80 100
Cu
mu
lati
ve C
O2-C
(m
g g
lit
ter-1
)
Leotiomycete (%)0 10 20 30 40 50
11
12
13
14
15
16
Actinobacteria (%)0 10 20 30 40 50
12
13
14
15
16
17
18
19
R. maximum
r2=0.42
P<0.05
Sordariomycete (%)0 20 40 60 80
11
12
13
14
15
16
R. maximum
r2=0.46
P<0.01
R. maximum
r2=0.36
P<0.05
R. maximum
r2=0.41
P<0.05
P. virgatum
r2=0.45
P<0.01
Chaetothyriomycete (%)
0 10 20 30
11
12
13
14
15
16
R. maximum
r2=0.57
P<0.01
a
b
c f
e
d
219
Figure 5.3
NH4
+ (mg g dry weight soil
-1)
0 2 4 6 8
11
12
13
14
15
16
R. maximum
r2=0.54
P<0.01
NO3
- (mg g dry weight soil
-1)
0 10 20 30 40 50
11
12
13
14
15
16
Extractable P (mg g dry weight soil-1
)
0 5 10 15 20
Cum
ula
tive
CO
2-C
(m
g g
lit
ter-1
)
11
12
13
14
15
16
Clay (%)
0 5 10 15 20 25 30
K+
(mg g dry weight soil-1
)
0.02 0.06 0.10 0.14 0.18
12
13
14
15
16
17
18
19
K+
(mg g dry weight soil-1
)
0.02 0.06 0.10 0.14 0.18
R. maximum
r2=0.69
P<0.001
R. maximum
r2=0.63
P<0.01
R. maximum
r2=0.63
P<0.01
R. maximum
r2=0.33
P<0.05
P. virgatum
r2=0.30
P<0.05
a
b
c
d
e
f
220
CHAPTER 6
CONCLUSIONS
The role that soil microbial community structure plays in regulating ecosystem processes
has been thought redundant, or dependent on binary distinctions of microbes (e.g.
fungal:bacterial dominance). In part the presumed functional redundancy of these
communities began with the famous quote, ―Everything is everywhere, the environment
selects‖ (Beijerinck 1913, Baas-Becking 1934). Whether this was the intention of
Beijerinck / Baas-Becking or not, this statement has been taken to suggest that either due
to their incredible diversity or adaptive abilities that microbial community function is
largely driven by the environment. Others have taken a more mechanistic approach to the
role soil microbial communities‘ play. Often this has relied on slotting components of
these communities into one of two groups such as zymogenous vs. autochthonous,
copiotrophs vs. oligotrophs, r-selected vs. K-selected, and bacteria vs. fungi
(Winogradsky 1924, Koch 2001, Fierer et al. 2007, van der Heijden et al. 2008).
Although such distinctions made an incredibly diverse community of organisms more
manageable, these binary distinctions overlook a grand spectrum of organisms and their
respective traits. Ultimately this has left us with the question, ―How is soil microbial
community structure related to ecosystem function?‖
My doctoral work was intended to begin addressing this question by examining
how microbial communities are related to ecosystem carbon (C)-processes. In the first of
these studies (Chapter 2) I set out to describe the role that soil microbial communities
221
played in the mineralization of glucose. Glucose is a low molecular weight organic
compound found in root exudates, as well as leachates derived from plant litter, and
likely represents a broadly available form of C to microbes (van Hees et al. 2005). I
expected that the mineralization of this compound would likely be related to some
component of the microbial community such as its size, activity, or fungal:bacterial
dominance. Surprisingly, glucose mineralization was related to none of these variables
but was related to the land-use that these communities were found in and more
specifically land-use associated variation in soil phosphorus. This may well be an
indication of the land-use legacy that these communities have endured (Richter and
Markewitz 2001). That is, the land-use regimes found at the Calhoun Experimental Forest
are for the most part products of 100 years of extensive cotton agriculture but upon
cessation of these practices have been managed differently (Callaham et al. 2006).
Specifically, that these communities share a similar history prior to cotton agriculture and
at the time were likely broadly similar with regards to their structure and function.
However, post-cotton these communities were exposed to differing land-use/management
regimes which ultimately influenced these characteristics. In light of this finding it
appears that at least microbial community function was influenced by its environment
and further collaborative work at these sites indicated that community composition below
the level of fungal:bacterial dominance was also influenced (Lauber et al. 2008).
Results from my examination of Microstegium vimineum invasion of hardwood
forests revealed that changes in the soil microbial community altered C-processes
(Chapter 3). Like my work in the Calhoun, I found that there was little change in the
fungal:bacterial dominance of the microbial community but in spite of this, these
222
communities were functionally distinct. This functional distinction resulted in more rapid
C-cycling associated with those communities where M. vimineum was present. This study
also suggested that the function of the soil microbial communities was primed by inputs
from the invasive plant. This outcome was likely contingent on the chemical quality of
M. vimineum in comparison to the chemical quality of native-derived plant C (Kuzyakov
et al. 2000, Litton et al. 2008). This may indicate that the functional response of soil
microbial communities is in part related to their contemporary environment. This
invasion may also indicate the sort of events which lead to functionally distinct microbial
communities. Like the research at the Calhoun, the forest microbial community in this
study shared a similar historical legacy but the invasion of M. vimineum altered this
legacy for a spatial subset of the community. Whether this will lead to permanent
functional and/or compositional divergence in these communities is as yet unknown. It
does though suggest one mechanism, the historic resource environment, which is likely to
influence the function of soil microbial communities with regards to C-processes.
The influence of the historic resource environment was resolved in the first of
my lab-based experimental studies (Chapter 4). In this work I found that microbial
communities were not functionally redundant and that communities paired with their
‗home‘ litter resource typically exhibited the highest litter mineralization rates for that
litter. This suggested that microbial communities which are pre-exposed to a given litter
substrate are subsequently better at decomposing it. This phenomenon has been termed
‗home-field advantage‘ (Hunt et al. 1988, Gholz et al. 2000, Ayres et al. 2009).
Interestingly, I also found that communities sourced from different environments in
combination with the most chemically-complex litter resource were compositionally the
223
most similar but functionally the most dissimilar. This may imply that the contemporary
environment is a structuring force with regards to the composition of microbial
communities but that the historical environment is the dominant structuring force with
regards to function.
In my second lab-based experiment (Chapter 5) I found further confirmation of
the influence of the historic environment on function. I paired two novel litter resources
with microbial communities sourced from either forest or herbaceous habitats. We found
that the mineralization of each litter (i.e. the perception of litter quality) was dependent
on the environment that the microbial community was sourced from. Specifically,
communities sourced from forest environments perceived no differences in quality
between the chemically-complex and -simple litters. In contrast, communities sourced
from herbaceous habitats perceived the chemically-simple litter to be of greater quality
than the chemically-complex one. This indicates that the function of soil microbial
communities might be further generalized with regards to their past exposure to different
C-compounds (Table 6.1).
Madsen (2005) argued that one of the major unknowns in microbial ecology was
the link between microbial communities and C-processes. Although my dissertation work
does not link specific soil microorganisms to specific C-processes, it does suggest that the
functioning of soil microbial communities is shaped by a history of differing C-inputs
(Fig. 6.1). This provides us with one possible mechanism which leads to functional
dissimilarity in soil microbial communities (Fig. 6.1). Although we may never be able to
determine if everything is everywhere, we can say that in part the environment does
224
select with regards to function but that this selection is contingent on both the
contemporary and historical environment of the soil microbial community.
Future Directions
One of the unanswered questions from my work is whether or not soil microbial
communities from differing historic environments are likely to converge functionally,
given enough time, when placed in the same contemporary environment. This is an area
of research which I am currently exploring. I plan to first determine if functional
convergence occurs. Second, if it does how long it takes and, finally, whether
convergence is associated with loss of function in alternate environments (i.e. a
‗functional trade-off‘). The results from this work will likely inform us as to the impact
that a changing environment is likely to have on the role that soil microbial communities
play in regulating ecosystem processes.
Other avenues of future research involve determining the legacy of historical
inputs on the function of soil microbial communities. Specifically, if communities share a
common history but a dissimilar contemporary environment then is function still broadly
similar amongst these communities? For example, in the Calhoun microbial communities
likely shared a common history but are now found in distinct contemporary
environments. So, we can ask if the function of these Calhoun communities is more
similar to each other than to historically-distinct communities, or whether it is the
contemporary environment which exerts the strongest influence on function.
Finally, I believe it is important to determine whether it is the input history of C-
compounds alone which impacts soil microbial community function. The structure of soil
225
microbial communities is influenced by an array of factors besides input history, such as
soil pH and nutrient status. However, whether or not these structuring forces also impact
the function of microbial communities has for the most part not been explored. I believe
that controlled-experimentation should be first used to explore these questions, but
ultimately that there is a need to explore the significance of these findings for the
regulation of ecosystem processes in situ.
References
Ayres, E., H. Steltzer, B. L. Simmons, R. T. Simpson, J. M. Steinweg, M. D. Wallenstein,
N. Mellor, W. J. Parton, J. C. Moore, and D. H. Wall. 2009. Home-field
advantage accelerates leaf litter decomposition in forests. Soil Biology &
Biochemistry 41:606-610.
Baas-Becking, L. G. M. 1934. Geobiologie of Inleiding Tot de Milieukunde. Van
Stockkum & Zoon,, The Hague.
Beijerinck, M. W. 1913. De infusies en de ontdekking der backteriën. in Müller, editor.
Jaarboek van de Koninklijke Akademie van Wetenschappen, Amsterdam.
Callaham, M. A., D. D. Richter, D. C. Coleman, and M. Hofmockel. 2006. Long-term
land-use effects on soil invertebrate communities in Southern Piedmont soils,
USA. European Journal of Soil Biology 42:S150-S156.
Fierer, N., M. A. Bradford, and R. B. Jackson. 2007. Toward an ecological classification
of soil bacteria. Ecology 88:1354-1364.
226
Gholz, H. L., D. A. Wedin, S. M. Smitherman, M. E. Harmon, and W. J. Parton. 2000.
Long-term dynamics of pine and hardwood litter in contrasting environments:
toward a global model of decomposition. Global Change Biology 6:751-765.
Hunt, H. W., E. R. Ingham, D. C. Coleman, E. T. Elliott, and C. P. P. Reid. 1988.
Nitrogen Limitation of Production and Decomposition in Prairie, Mountain
Meadow, and Pine Forest. Ecology 69:1009-1016.
Koch, A. L. 2001. Oligotrophs versus copiotrophs. Bioessays 23:657-661.
Kuzyakov, Y., J. K. Friedel, and K. Stahr. 2000. Review of mechanisms and
quantification of priming effects. Soil Biology & Biochemistry 32:1485-1498.
Lauber, C. L., M. S. Strickland, M. A. Bradford, and N. Fierer. 2008. The influence of
soil properties on the structure of bacterial and fungal communities across land-
use types. Soil Biology & Biochemistry 40:2407-2415.
Litton, C. M., D. R. Sandquist, and S. Cordell. 2008. A non-native invasive grass
increases soil carbon flux in a Hawaiian tropical dry forest. Global Change
Biology 14:726-739.
Madsen, E. L. 2005. Identifying microorganisms responsible for ecologically significant
biogeochemical processes. Nature Reviews Microbiology 3:439-446.
Richter, D. D., and D. Markewitz. 2001. Understanding Soil Change: Soil sustainability
over millenia, centuries, and decades. Cambridge University Press, New York,
New York, USA.
van der Heijden, M. G. A., R. D. Bardgett, and N. M. van Straalen. 2008. The unseen
majority: soil microbes as drivers of plant diversity and productivity in terrestrial
ecosystems. Ecology Letters 11:296-310.
227
van Hees, P. A. W., D. L. Jones, R. Finlay, D. L. Godbold, and U. S. Lundstomd. 2005.
The carbon we do not see - the impact of low molecular weight compounds on
carbon dynamics and respiration in forest soils: a review. Soil Biology &
Biochemistry 37:1-13.
Winogradsky, S. 1924. Sur la microflora autochtone de la terre arable. Comptes Rendus
Hebdomadaire des Séances de l'Academie des Sciences, Paris 178:1236-1239.
228
Table 6.1 Generalized functional implications for soil microbial communities with
distinct histories of high and low chemical quality inputs as they relate to C-cycling. The
realization and strength of these generalized outcomes will likely be contingent on the
resource history of one microbial community relative to another. For example, a
microbial community from an unfertilized grassland may be considered to have a high
quality resource history when compared to an old growth forest but is likely low quality
when compared to a fertilized grassland.
Resource History
High quality Low quality
1. Mineralization of low molecular
weight organic compounds
(LMWOC) such as glucose is high.
Possibly because LMWOC
compounds are major C inputs in
these systems or because
communities are not C-limited.
2. Perception of litter quality is
related to the litter‘s chemical
recalcitrance.
3. Increasing labile C inputs may lead
to an increase or no change in soil
C pools due to preferential
substrate utilization.
4. Relatively poor decomposers of an
array of litter resources.
1. Mineralization of LMWOC
compounds is low. Possibly because
LMWOC compounds are minor C
inputs in these systems or because
communities are C-limited.
2. Perception of litter quality is not or
less strongly related to the litter‘s
chemical recalcitrance.
3. Increasing labile C inputs may lead
to a decrease in soil C-pools due to a
priming effect.
4. Relatively good decomposers of an
array of litter resources but this may
be confounded by home-field
advantage.
229
Figure Legends
Figure 6.1 A schematic showing how alterations in the quality of C-compounds may
impact the function of soil microbial communities across time. Community X represents
the ancestral community. Community X1, 2 … represent communities distinct with regards
to their function. Each node (●) represents environmental change leading to an alteration
in the quality of C inputs a given subset of the ancestral community is exposed to. In this
example, subsets of Community X are impacted by an environmental change which alters
the quality of C inputs they are exposed to leading to two functionally dissimilar
communities, Community X1 and X2. Communities X1 and X2 are again impacted by an
environmental change leading to the functionally distinct communities of X1.1, X1.2, X2.1,
and X2.2. Finally, Communities X1.1 and X1.2 are impacted by the same change
experienced by the ancestral community (i.e. X1) leading to convergence of these two
communities and a return to the functioning of the ancestral community. On the other
hand communities X2.1 and X2.2 experience a similar perturbation but never functionally
converge. Environmental change may be as dramatic as land-use change or fertilizer
inputs or even as minor as a change from one tree species to another within a forest. The
question marks represent a change in input quality to those found one community back
and is related to whether or not microbial communities will converge functionally.
230
Figure 6.1
Community X
?
Community X1
Community X2.1
Community X2.2
Community X2
Community X1.1
Community X1.2
Community X1
Community X2.1.1
Community X2.2.1
?
Time
231
APPENDIX A
REFERENCES USED TO CONSTRUCTUCT FIGURE 1.1
References
Allison, S. D., C. Nielsen, and R. F. Hughes. 2006. Elevated enzyme activities in soils
under the invasive nitrogen-fixing tree Falcataria moluccana. Soil Biology &
Biochemistry 38:1537-1544.
Amer, G. A. M., D. V. Singh, R. Aggarwal, and P. Dureja. 1998. Microbial antagonism
to Neovossia indica, causing Karnal bunt of wheat. Current Science 75:1393-
1396.
Ananyeva, N. D., E. A. Susyan, O. V. Chernova, I. Y. Chernov, and O. L. Makarova.
2006. The ratio of fungi and bacteria in the biomass of different types of soil
determined by selective inhibition. Microbiology 75:702-707.
Anderson, T. H., and H. J. Weigel. 2003. On the current debate about soil biodiversity.
Landbauforschung Volkenrode 53:223-233.
Andersson, B. E., S. Lundstedt, K. Tornberg, Y. Schnurer, L. G. Oberg, and B.
Mattiasson. 2003. Incomplete degradation of polycyclic aromatic hydrocarbons in
soil inoculated with wood-rotting fungi and their effect on the indigenous soil
bacteria. Environmental Toxicology and Chemistry 22:1238-1243.
Bååth, E., and T. H. Anderson. 2003. Comparison of soil fungal/bacterial ratios in a pH
gradient using physiological and PLFA-based techniques. Soil Biology &
Biochemistry 35:955-963.
232
Bååth, E., A. Frostegård, T. Pennanen, and H. Fritze. 1995. Microbial Community
Structure and pH Response in Relation to Soil Organic-Matter Quality in Wood-
Ash Fertilized, Clear-Cut or Burned Coniferous Forest Soils. Soil Biology &
Biochemistry 27:229-240.
Bailey, V. L., J. L. Smith, and H. Bolton. 2003. Novel antibiotics as inhibitors for the
selective respiratory inhibition method of measuring fungal : bacterial ratios in
soil. Biology and Fertility of Soils 38:154-160.
Baldrian, P., J. Trogl, J. Frouz, J. Snajdr, V. Valaskova, V. Merhautova, T. Cajthaml, and
J. Herinkova. 2008. Enzyme activities and microbial biomass in topsoil layer
during spontaneous succession in spoil heaps after brown coal mining. Soil
Biology & Biochemistry 40:2107-2115.
Bardgett, R. D., P. J. Hobbs, and A. Frostegård. 1996. Changes in soil fungal:bacterial
biomass ratios following reductions in the intensity of management of an upland
grassland. Biology and Fertility of Soils 22:261-264.
Bardgett, R. D., and E. McAlister. 1999. The measurement of soil fungal : bacterial
biomass ratios as an indicator of ecosystem self-regulation in temperate meadow
grasslands. Biology and Fertility of Soils 29:282-290.
Bardgett, R. D., and L. R. Walker. 2004. Impact of coloniser plant species on the
development of decomposer microbial communities following deglaciation. Soil
Biology & Biochemistry 36:555-559.
Barret, M., P. Frey-Klett, M. Boutin, A. Y. Guillerm-Erckelboudt, F. Martin, L. Guillot,
and A. Sarniguet. 2009. The plant pathogenic fungus Gaeumannomyces graminis
var. tritici improves bacterial growth and triggers early gene regulations in the
233
biocontrol strain Pseudomonas fluorescens Pf29Arp. New Phytologist 181:435-
447.
Bautista-Banos, S., A. N. Hernandez-Lauzardo, M. G. Velazquez-del Valle, M.
Hernandez-Lopez, E. A. Barka, E. Bosquez-Molina, and C. L. Wilson. 2006.
Chitosan as a potential natural compound to control pre and postharvest diseases
of horticultural commodities. Crop Protection 25:108-118.
Beary, T. P., R. Boopathy, and P. Templet. 2002. Accelerated decomposition of
sugarcane crop residue using a fungal-bacterial consortium. International
Biodeterioration & Biodegradation 50:41-46.
Bell, J. M., J. L. Smith, V. L. Bailey, and H. Bolton. 2003. Priming effect and C storage
in semi-arid no-till spring crop rotations. Biology and Fertility of Soils 37:237-
244.
Bharadwaj, D. P., P. O. Lundquist, P. Persson, and S. Alstrom. 2008. Evidence for
specificity of cultivable bacteria associated with arbuscular mycorrhizal fungal
spores. Fems Microbiology Ecology 65:310-322.
Bianciotto, V., D. Minerdi, S. Perotto, and P. Bonfante. 1996. Cellular interactions
between arbuscular mycorrhizal fungi and rhizosphere bacteria. Protoplasma
193:123-131.
Blagodatskaya, E. V., and T. H. Anderson. 1999. Adaptive responses of soil microbial
communities under experimental acid stress in controlled laboratory studies.
Applied Soil Ecology 11:207-216.
234
Boonchan, S., M. L. Britz, and G. A. Stanley. 2000. Degradation and mineralization of
high-molecular-weight polycyclic aromatic hydrocarbons by defined fungal-
bacterial cocultures. Applied and Environmental Microbiology 66:1007-1019.
Boyle, S. A., R. R. Yarwood, P. J. Bottomley, and D. D. Myrold. 2008. Bacterial and
fungal contributions to soil nitrogen cycling under Douglas fir and red alder at
two sites in Oregon. Soil Biology & Biochemistry 40:443-451.
Brant, J. B., E. W. Sulzman, and D. D. Myrold. 2006. Microbial community utilization of
added carbon substrates in response to long-term carbon input manipulation. Soil
Biology & Biochemistry 38:2219-2232.
Brye, K. R., N. A. Slaton, and R. J. Norman. 2006. Soil physical and biological properties
as affected by land leveling in a clayey aquert. Soil Science Society of America
Journal 70:631-642.
Brye, K. R., N. A. Slaton, M. C. Savin, R. J. Norman, and D. M. Miller. 2003. Short-term
effects of land leveling on soil physical properties and microbial biomass. Soil
Science Society of America Journal 67:1405-1417.
Busse, M. D., F. G. Sanchez, A. W. Ratcliff, J. R. Butnor, E. A. Carter, and R. F. Powers.
2009. Soil carbon sequestration and changes in fungal and bacterial biomass
following incorporation of forest residues. Soil Biology & Biochemistry 41:220-
227.
Byss, M., D. Elhottova, J. Triska, and P. Baldrian. 2008. Fungal bioremediation of the
creosote-contaminated soil: Influence of Pleurotus ostreatus and Irpex lacteus on
polycyclic aromatic hydrocarbons removal and soil microbial community
composition in the laboratory-scale study. Chemosphere 73:1518-1523.
235
Chen, M. M., B. D. Chen, and P. Marschner. 2008. Plant growth and soil microbial
community structure of legumes and grasses grown in monoculture or mixture.
Journal of Environmental Sciences-China 20:1231-1237.
de Boer, W., L. B. Folman, R. C. Summerbell, and L. Boddy. 2005. Living in a fungal
world: impact of fungi on soil bacterial niche development. FEMS Microbiology
Reviews 29:795-811.
de Vries, F. T., J. Bloem, N. van Eekeren, L. Brusaard, and E. Hoffland. 2007. Fungal
biomass in pastures increases with age and reduced N input. Soil Biology &
Biochemistry 39:1620-1630.
de Vries, F. T., E. Hoffland, N. van Eekeren, L. Brussaard, and J. Bloem. 2006.
Fungal/bacterial ratios in grasslands with contrasting nitrogen management. Soil
Biology & Biochemistry 38:2092-2103.
Denef, K., D. Roobroeck, M. Wadu, P. Lootens, and P. Boeckx. 2009. Microbial
community composition and rhizodeposit-carbon assimilation in differently
managed temperate grassland soils. Soil Biology & Biochemistry 41:144-153.
Deveau, A., B. Palin, C. Delaruelle, M. Peter, A. Kohler, J. C. Pierrat, A. Sarniguet, J.
Garbaye, F. Martin, and P. Frey-Klett. 2007. The mycorrhiza helper Pseudomonas
fluorescens BBc6R8 has a specific priming effect on the growth, morphology and
gene expression of the ectomycorrhizal fungus Laccaria bicolor S238N. New
Phytologist 175:743-755.
Eiland, F., M. Klamer, A. M. Lind, M. Leth, and E. Bååth. 2001. Influence of initial C/N
ratio on chemical and microbial composition during long term composting of
straw. Microbial Ecology 41:272-280.
236
Elfstrand, S., K. Hedlund, and A. Martensson. 2007. Soil enzyme activities, microbial
community composition and function after 47 years of continuous green
manuring. Applied Soil Ecology 35:610-621.
Entry, J. A., J. J. Fuhrmann, R. E. Sojka, and G. E. Shewmaker. 2004. Influence of
irrigated agriculture on soil carbon and microbial community structure.
Environmental Management 33:S363-S373.
Frey, S. D., M. Knorr, J. L. Parrent, and R. T. Simpson. 2004. Chronic nitrogen
enrichment affects the structure and function of the soil microbial community in
temperate hardwood and pine forests. Forest Ecology and Management 196:159-
171.
Frey-Klett, P., and J. Garbaye. 2005. Mycorrhiza helper bacteria: a promising model for
the genomic analysis of fungal-bacterial interactions. New Phytologist 168:4-8.
Frey-Klett, P., J. Garbaye, and M. Tarkka. 2007. The mycorrhiza helper bacteria
revisited. New Phytologist 176:22-36.
Frostegård, A., and E. Bååth. 1996. The use of phospholipid fatty acid analysis to
estimate bacterial and fungal biomass in soil. Biology and Fertility of Soils 22:59-
65.
Gastaldo, R. A., and J. R. Staub. 1999. A mechanism to explain the preservation of leaf
litter lenses in coals derived from raised mires. Palaeogeography
Palaeoclimatology Palaeoecology 149:1-14.
Grayston, S. J., and C. E. Prescott. 2005. Microbial communities in forest floors under
four tree species in coastal British Columbia. Soil Biology & Biochemistry
37:1157-1167.
237
Han, W. Y., S. J. Kemmitt, and P. C. Brookes. 2007. Soil microbial biomass and activity
in Chinese tea gardens of varying stand age and productivity. Soil Biology &
Biochemistry 39:1468-1478.
Hanel, L. 2003. Soil nematodes in cambisol agroecosystems of the Czech Republic.
Biologia 58:205-216.
Kanerva, T., A. Palojarvi, K. Ramo, and S. Manninen. 2008. Changes in soil microbial
community structure under elevated tropospheric O3 and CO2. Soil Biology &
Biochemistry 40:2502-2510.
Klein, D. A., T. McLendon, M. W. Paschke, and E. F. Redente. 1995. Saprophytic
fungal-bacterial biomass variations in successional communities of a semiarid
steppe ecosystem. Biology and Fertility of Soils 19:253-256.
Klein, D. A., T. McLendon, M. W. Paschke, and E. F. Redente. 1996. Nitrogen
availability and fungal-bacterial responses in successional semiarid steppe soils.
Arid Soil Research and Rehabilitation 10:321-332.
Klein, D. A., and M. W. Paschke. 2000. A soil microbial community structural-functional
index: the microscopy-based total/active/active fungal/bacterial (TA/AFB)
biovolumes ratio. Applied Soil Ecology 14:257-268.
Klein, D. A., M. W. Paschke, and E. F. Redente. 1998. Assessment of fungal-bacterial
development in a successional shortgrass steppe by direct integration of
chloroform-fumigation extraction (FE) and microscopically derived data. Soil
Biology & Biochemistry 30:573-581.
Kooijman, A. M., M. M. Kooijman-Schouten, and G. B. Martinez-Hernandez. 2008.
Alternative strategies to sustain N-fertility in acid and calcaric beech forests: Low
238
microbial N-demand versus high biological activity. Basic and Applied Ecology
9:410-421.
Lipson, D. A., C. W. Schadt, and S. K. Schmidt. 2002. Changes in soil microbial
community structure and function in an alpine dry meadow following spring snow
melt. Microbial Ecology 43:307-314.
Macdonald, L. M., E. Paterson, L. A. Dawson, and A. J. S. McDonald. 2006. Defoliation
and fertiliser influences on the soil microbial community associated with two
contrasting Lolium perenne cullivars. Soil Biology & Biochemistry 38:674-682.
Mamilov, A. S., B. A. Byzov, D. G. Zvyagintsev, and O. M. Dilly. 2001. Predation on
fungal and bacterial biomass in a soddy-podzolic soil amended with starch, wheat
straw and alfalfa meal. Applied Soil Ecology 16:131-139.
Manucharova, N. A., A. M. Yaroslavtsev, E. G. Kornyushenko, A. L. Stepanov, A. V.
Smagin, D. G. Zvyagintsev, and Sudnitsyn, II. 2007. Methane production and
growth of microorganisms under different moisture conditions in soils with added
chitin and without it. Eurasian Soil Science 40:860-865.
Mechri, B., A. Echbili, M. Issaoui, M. Braham, S. Ben Elhadj, and M. Hammami. 2007.
Short-term effects in soil microbial community following agronomic application
of olive mill wastewaters in a field of olive trees. Applied Soil Ecology 36:216-
223.
Moller, J., M. Miller, and A. Kjoller. 1999. Fungal-bacterial interaction on beech leaves:
influence on decomposition and dissolved organic carbon quality. Soil Biology &
Biochemistry 31:367-374.
239
Morillo, J. A., M. Aguilera, B. Antizar-Ladislao, S. Fuentes, A. Ramos-Cormenzana, N.
J. Russell, and M. Monteoliva-Sanchez. 2008. Molecular microbial and chemical
investigation of the bioremediation of two-phase olive mill waste using
laboratory-scale bioreactors. Applied Microbiology and Biotechnology 79:309-
317.
Morris, S. J., and R. E. J. Boerner. 1998. Interactive influences of silvicultural
management and soil chemistry upon soil microbial abundance and nitrogen
mineralization. Forest Ecology and Management 103:129-139.
Neely, C. L., M. H. Beare, W. L. Hargrove, and D. C. Coleman. 1991. Relationships
between Fungal and Bacterial Substrate-Induced Respiration, Biomass and Plant
Residue Decomposition. Soil Biology & Biochemistry 23:947-954.
Newton, A. C., I. K. Toth, P. Neave, and L. J. Hyman. 2004. Bacterial inoculum from a
previous crop affects fungal disease development on subsequent nonhost crops.
New Phytologist 163:133-138.
Pankhurst, C. E., G. R. Stirling, R. C. Magarey, B. L. Blair, J. A. Holt, M. J. Bell, and A.
L. Garside. 2005. Quantification of the effects of rotation breaks on soil biological
properties and their impact on yield decline in sugarcane. Soil Biology &
Biochemistry 37:1121-1130.
Pennanen, T., A. Frostegård, H. Fritze, and E. Bååth. 1996. Phospholipid fatty acid
composition and heavy metal tolerance of soil microbial communities along two
heavy metal-polluted gradients in coniferous forests. Applied and Environmental
Microbiology 62:420-428.
240
Piotrowska-Seget, Z., R. Engel, E. Nowak, and J. Kozdroj. 2008. Successive soil
treatment with captan or oxytetracycline affects non-target microorganisms.
World Journal of Microbiology & Biotechnology 24:2843-2848.
Postma, J., M. T. Schilder, J. Bloem, and W. K. van Leeumen-Haagsma. 2008. Soil
suppressiveness and functional diversity of the soil microflora in organic farming
systems. Soil Biology & Biochemistry 40:2394-2406.
Powell, J. R., D. J. Levy-Booth, R. H. Gulden, W. L. Asbil, R. G. Campbell, K. E.
Dunfield, A. S. Hamill, M. M. Hart, S. Lerat, R. E. Nurse, K. P. Pauls, P. H.
Sikkema, C. J. Swanton, J. T. Trevors, and J. N. Klironomos. 2009. Effects of
genetically modified, herbicide-tolerant crops and their management on soil food
web properties and crop litter decomposition. Journal of Applied Ecology 46:388-
396.
Rai, J. P. N. 1992. Effects of Long-Term 2,4-D Application on microbial-populations and
biochemical processes in cultivated soil. Biology and Fertility of Soils 13:187-
191.
Rajapaksha, R., M. A. Tobor-Kaplon, and E. Bååth. 2004. Metal toxicity affects fungal
and bacterial activities in soil differently. Applied and Environmental
Microbiology 70:2966-2973.
Rousk, J., and E. Bååth. 2007. Fungal and bacterial growth in soil with plant materials of
different C/N ratios. FEMS Microbiology Ecology 62:258-267.
Rousk, J., P. C. Brookes, and E. Bååth. 2009. Contrasting soil pH sffects on fungal and
bacterial growth suggest functional redundancy in carbon mineralization. Applied
and Environmental Microbiology 75:1589-1596.
241
Rousk, J., and N. M. Nadkarni. 2009. Growth measurements of saprotrophic fungi and
bacteria reveal differences between canopy and forest floor soils. Soil Biology &
Biochemistry 41:862-865.
Sakamoto, K., and Y. Oba. 1994. Effect of fungal to bacterial biomass ratio on the
relationship between CO2 evolution and total soil microbial biomass. Biology and
Fertility of Soils 17:39-44.
Schaefer, M., S. O. Petersen, and J. Filser. 2005. Effects of Lumbricus terrestris,
Allolobophora chlorotica and Eisenia fetida on microbial community dynamics in
oil-contaminated soil. Soil Biology & Biochemistry 37:2065-2076.
Scheu, S., and D. Parkinson. 1995. Successional changes in microbial biomass,
respiration and nutrient status during litter decomposition in an aspen and pine
forest. Biology and Fertility of Soils 19:327-332.
Seneviratne, G., J. S. Zavahir, W. Bandara, and M. Weerasekara. 2008. Fungal-bacterial
biofilms: their development for novel biotechnological applications. World
Journal of Microbiology & Biotechnology 24:739-743.
Sengelov, G., G. A. Kowalchuk, and S. J. Sorensen. 2000. Influence of fungal-bacterial
interactions on bacterial conjugation in the residuesphere. FEMS Microbiology
Ecology 31:39-45.
Smith, R. S., R. S. Shiel, R. D. Bardgett, D. Millward, P. Corkhill, P. Evans, H. Quirk, P.
J. Hobbs, and S. T. Kometa. 2008. Long-term change in vegetation and soil
microbial communities during the phased restoration of traditional meadow
grassland. Journal of Applied Ecology 45:670-679.
242
Sohlenius, B. 1996. Structure and composition of the nematode fauna in pine forest soil
under the influence of clear-cutting - Effects of slash removal and field layer
vegetation. European Journal of Soil Biology 32:1-14.
Tate, R. L. 1991. Microbial biomass measurement in acidic soil - effect of fungal -
bacterial-activity ratios and of soil amendment. Soil Science 152:220-225.
Thakuria, D., N. C. Talukdar, C. Goswami, S. Hazarika, M. C. Kalita, and G. D.
Bending. 2009. Evaluation of rice-legume-rice cropping system on grain yield,
nutrient uptake, nitrogen fixation, and chemical, physical, and biological
properties of soil. Biology and Fertility of Soils 45:237-251.
Thiele-Bruhn, S., and I. C. Beck. 2005. Effects of sulfonamide and tetracycline
antibiotics on soil microbial activity and microbial biomass. Chemosphere
59:457-465.
Thiet, R. K., S. D. Frey, and J. Six. 2006. Do growth yield efficiencies differ between soil
microbial communities differing in fungal: bacterial ratios? Reality check and
methodological issues. Soil Biology & Biochemistry 38:837-844.
Tischer, S., H. Tanneberg, and G. Guggenberger. 2008. Microbial parameters of soils
contaminated with heavy metals: Assessment for ecotoxicological monitoring.
Polish Journal of Ecology 56:471-479.
Velvis, H. 1997. Evaluation of the selective respiratory inhibition method for measuring
the ratio of fungal:bacterial activity in acid agricultural soils. Biology and Fertility
of Soils 25:354-360.
243
Vong, P. C., S. Piutti, S. Slezack-Deschaumes, E. Benizri, and A. Guckert. 2008. Sulphur
immobilization and arylsulphatase activity in two calcareous arable and fallow
soils as affected by glucose additions. Geoderma 148:79-84.
Wallenstein, M. D., S. McNulty, I. J. Fernandez, J. Boggs, and W. H. Schlesinger. 2006.
Nitrogen fertilization decreases forest soil fungal and bacterial biomass in three
long-term experiments. Forest Ecology and Management 222:459-468.
Weyman-Kaczmarkowa, W., and Z. Pedziwilk. 2000. The development of fungi as
affected by pH and type of soil, in relation to the occurrence of bacteria and soil
fungistatic activity. Microbiological Research 155:107-112.
Williams, C. A., L. V. Hills, and F. F. Krause. 1996. Preserved organic matter and
miospores in buried Middle Devonian (Givetian) paleosols: Indicators of
weathering, oxidation and maturity. Catena 28:1-19.
Williamson, W. M., D. A. Wardle, and G. W. Yeates. 2005. Changes in soil microbial
and nematode communities during ecosystem decline across a long-term
chronosequence. Soil Biology & Biochemistry 37:1289-1301.
Yeates, G. W., and H. vanderMeulen. 1996. Recolonization of methyl-bromide sterilized
soils by plant and soil nematodes over 52 months. Biology and Fertility of Soils
21:1-6.
Zhang, W. J., W. Y. Rui, C. Tu, H. G. Diab, F. J. Louws, J. P. Mueller, N. Creamer, M.
Bell, M. G. Wagger, and S. Hu. 2005. Responses of soil microbial community
structure and diversity to agricultural deintensification. Pedosphere 15:440-447.
244
APPENDIX B
REFERENCES USED TO CONSTRUCTUCT FIGURE 1.2
References
Bertilsson, S., L. A. Hansson, W. Graneli, and A. Philibert. 2003. Size-selective
predation on pelagic microorganisms in Arctic freshwaters. Journal of Plankton
Research 25:621-631.
Chrzanowski, T. H., and J. P. Grover. 2008. Element content of Pseudomonas
fluorescens varies with growth rate and temperature: A replicated chemostat study
addressing ecological stoichiometry. Limnology and Oceanography 53:1242-
1251.
Chrzanowski, T. H., and M. Kyle. 1996. Ratios of carbon, nitrogen and phosphorus in
Pseudomonas fluorescens as a model for bacterial element ratios and nutrient
regeneration. Aquatic Microbial Ecology 10:115-122.
Cooney, M. J., I. W. Marison, W. M. vanGulik, and U. vonStockar. 1996. Calorimetric
and stoichiometric analysis of growth of Kluyveromices fragilis in continuous
culture: Nitrogen limitation imposed upon carbon-limited growth. Applied
Microbiology and Biotechnology 44:643-653.
Cuhel, R. L., and J. B. Waterbury. 1984. Biochemical-Composition and Short-Term
Nutrient Incorporation Patterns in a Unicellular Marine Cyanobacterium,
Synechococcus (Wh7803). Limnology and Oceanography 29:370-374.
245
Fagerbakke, K. M., M. Heldal, and S. Norland. 1996. Content of carbon, nitrogen,
oxygen, sulfur and phosphorus in native aquatic and cultured bacteria. Aquatic
Microbial Ecology 10:15-27.
Fukuda, R., H. Ogawa, T. Nagata, and I. Koike. 1998. Direct determination of carbon and
nitrogen contents of natural bacterial assemblages in marine environments.
Applied and Environmental Microbiology 64:3352-3358.
Hart, S. C., C. A. Gehring, P. C. Selmants, and R. J. Deckert. 2006. Carbon and nitrogen
elemental and isotopic patterns in macrofungal sporocarps and trees in semiarid
forests of the south-western USA. Functional Ecology 20:42-51.
Heldal, M., D. J. Scanlan, S. Norland, F. Thingstad, and N. H. Mann. 2003. Elemental
composition of single cells of various strains of marine Prochlorococcus and
Synechococcus using X-ray microanalysis. Limnology and Oceanography
48:1732-1743.
Kaiser, K., and R. Benner. 2008. Major bacterial contribution to the ocean reservoir of
detrital organic carbon and nitrogen. Limnology and Oceanography 53:99-112.
Lovdal, T., E. F. Skjoldal, M. Heldal, S. Norland, and T. F. Thingstad. 2008. Changes in
morphology and elemental composition of Vibrio splendidus along a gradient
from carbon-limited to phosphate-limited growth. Microbial Ecology 55:152-161.
Makino, W., and J. B. Cotner. 2004. Elemental stoichiometry of a heterotrophic bacterial
community in a freshwater lake: implications for growth- and resource-dependent
variations. Aquatic Microbial Ecology 34:33-41.
246
Norland, S., K. M. Fagerbakke, and M. Heldal. 1995. Light-Element Analysis of
Individual Bacteria by X-Ray-Microanalysis. Applied and Environmental
Microbiology 61:1357-1362.
Shannon, S. P., T. H. Chrzanowski, and J. P. Grover. 2007. Prey food quality affects
flagellate ingestion rates. Microbial Ecology 53:66-73.
Taylor, A. F. S., P. M. Fransson, P. Hogberg, M. N. Hogberg, and A. H. Plamboeck.
2003. Species level patterns in C13
and N15
abundance of ectomycorrhizal and
saprotrophic fungal sporocarps. New Phytologist 159:757-774.
Trudell, S. A., P. T. Rygiewicz, and R. L. Edmonds. 2004. Patterns of nitrogen and
carbon stable isotope ratios in macrofungi, plants and soils in two old-growth
conifer forests. New Phytologist 164:317-335.
Wallander, H., L. O. Nilsson, D. Hagerberg, and U. Rosengren. 2003. Direct estimates of
C : N ratios of ectomycorrhizal mycelia collected from Norway spruce forest
soils. Soil Biology & Biochemistry 35:997-999.
247
APPENDIX C
CANDIDATE MODEL SET
Table C.1 List of candidate models used to describe cumulative glucose mineralized
(13
CO2). The first 30 models and the intercept only model were the initial set of candidate
models used. All models were compared as the secondary set of candidate models.
# Model
13CO2 = 1
1 13CO2 = MicC
2 13CO2 = SIR
3 13CO2 = F:B
4 13CO2 = MicC + SIR
5 13CO2 = MicC + F:B
6 13CO2 = SIR + F:B
7 13CO2 = MicC + SIR + F:B
8 13CO2 = Seas. + MicC
9 13CO2 = Seas. + SIR
10 13CO2 = Seas. + F:B
11 13CO2 = Seas. + MicC + SIR
12 13CO2 = Seas. + MicC + F:B
13 13CO2 = Seas. + SIR + F:B
14 13CO2 = Seas. + MicC + SIR + F:B
15 13CO2 = Land
16 13CO2 = Land + Seas.
17 13CO2 = Land + MicC
18 13CO2 = Land + SIR
19 13CO2 = Land + F:B
20 13CO2 = Land + MicC + SIR
21 13CO2 = Land + MicC + F:B
22 13CO2 = Land + SIR + F:B
23 13CO2 = Land + MicC + SIR + F:B
24 13CO2 = Land + Seas. + MicC
25 13CO2 = Land + Seas. + SIR
26 13CO2 = Land + Seas. + F:B
27 13CO2 = Land + Seas. + MicC + SIR
28 13CO2 = Land + Seas. + MicC + F:B
29 13CO2 = Land + Seas. + SIR + F:B
30 13CO2 = Land + Seas. + MicC + SIR + F:B
31 13CO2 = Moist.*Temp.
32 13CO2 = Moist.*Temp. + Seas.
33 13CO2 = Moist.*Temp. + MicC
34 13CO2 = Moist.*Temp. + SIR
248
Appendix C (cont.)
# Model
35 13CO2 = Moist.*Temp. + F:B
36 13CO2 = Moist.*Temp. + MicC + SIR
37 13CO2 = Moist.*Temp. + MicC + F:B
38 13CO2 = Moist.*Temp. + SIR + F:B
39 13CO2 = Moist.*Temp. + MicC + SIR + F:B
40 13CO2 = Moist.*Temp. + Seas. + MicC
41 13CO2 = Moist.*Temp. + Seas. + SIR
42 13CO2 = Moist.*Temp. + Seas. + F:B
43 13CO2 = Moist.*Temp. + Seas. + MicC + SIR
44 13CO2 = Moist.*Temp. + Seas. + MicC + F:B
45 13CO2 = Moist.*Temp. + Seas. + SIR + F:B
46 13CO2 = Moist.*Temp. + Seas. + MicC + SIR + F:B
47 13CO2 = C:N
48 13CO2 = C:N + Seas.
49 13CO2 = C:N + MicC
50 13CO2 = C:N + SIR
51 13CO2 = C:N + F:B
52 13CO2 = C:N + MicC + SIR
53 13CO2 = C:N + MicC + F:B
54 13CO2 = C:N + SIR + F:B
55 13CO2 = C:N + MicC + SIR + F:B
56 13CO2 = C:N + Seas. + MicC
57 13CO2 = C:N + Seas. + SIR
58 13CO2 = C:N + Seas. + F:B
59 13CO2 = C:N + Seas. + MicC + SIR
60 13CO2 = C:N + Seas. + MicC + F:B
61 13CO2 = C:N + Seas. + SIR + F:B
62 13CO2 = C:N + Seas. + MicC + SIR + F:B
63 13CO2 = S+C
64 13CO2 = S+C + Seas.
65 13CO2 = S+C + MicC
66 13CO2 = S+C + SIR
67 13CO2 = S+C + F:B
68 13CO2 = S+C + MicC + SIR
69 13CO2 = S+C + MicC + F:B
70 13CO2 = S+C + SIR + F:B
71 13CO2 = S+C + MicC + SIR + F:B
72 13CO2 = S+C + Seas. + MicC
73 13CO2 = S+C + Seas. + SIR
74 13CO2 = S+C + Seas. + F:B
75 13CO2 = S+C + Seas. + MicC + SIR
76 13CO2 = S+C + Seas. + MicC + F:B
77 13CO2 = S+C + Seas. + SIR + F:B
249
Appendix C (cont.)
# Model
78 13CO2 = S+C + Seas. + MicC + SIR + F:B
79 13CO2 = pH
80 13CO2 = pH + Seas.
81 13CO2 = pH + MicC
86 13CO2 = pH + SIR + F:B
87 13CO2 = pH + MicC + SIR + F:B
88 13CO2 = pH + Seas. + MicC
89 13CO2 = pH + Seas. + SIR
90 13CO2 = pH + Seas. + F:B
91 13CO2 = pH + Seas. + MicC + SIR
92 13CO2 = pH + Seas. + MicC + F:B
93 13CO2 = pH + Seas. + SIR + F:B
94 13CO2 = pH + Seas. + MicC + SIR + F:B
95 13CO2 = P
96 13CO2 = P + Seas.
97 13CO2 = P + MicC
98 13CO2 = P + SIR
99 13CO2 = P + F:B
100 13CO2 = P + MicC + SIR
101 13CO2 = P + MicC + F:B
102 13CO2 = P + SIR + F:B
103 13CO2 = P + MicC + SIR + F:B
104 13CO2 = P + Seas. + MicC
105 13CO2 = P + Seas. + SIR
106 13CO2 = P + Seas. + F:B
107 13CO2 = P + Seas. + MicC + SIR
108 13CO2 = P + Seas. + MicC + F:B
109 13CO2 = P + Seas. + SIR + F:B
110 13CO2 = P + Seas. + MicC + SIR + F:B
SIR = substrate induced respiration, a measure of microbial activity; MicC = microbial
biomass C as determined via sCFE, a measure of the size of the microbial biomass; F:B =
fungal-to-bacterial dominance as determined via qPCR, a measure of community
composition; Seas. = Season, either winter or summer when the pulse-chase experiment
was conducted. Land = land-use. Moist.*Temp. = the interaction between moisture and
temperature. C:N = the soil C:N ratio. S+C = silt plus clay content of the soil. pH = Soil
pH. P = extractable phosphorous.
250
APPENDIX D
FIGURE SHOWING SOIL RESPIRATION DURING THE PULSE-CHASE
EXPERIMENT
Figure D.1 Mean soil respiration ±1 S.E. per land-use (n=3) during the summer (a) and
winter (b) across the 72h pulse-chase. The x-axis denotes the time since glucose was
added. Glucose was added at time zero as indicated by the arrow.
(b) Summer (June)
Time since glucose added (h)
-2 0 2 4 6 20 40 60 80
CO
2-C
(m
g m
-2 h
-1)
0
50
100
150
200
250
300
350
400
(a) Winter (December)
0
20
40
60
80
100
120
140
160
180Cultivated
Pasture
Pine
Hardwood
Glucose added
Glucose added
CO
2-C
(m
g m
-2 h
-
1)
251
APPENDIX E
PARAMETER ESTIMATES
Tables showing the restricted maximum likelihood (REML) estimates for models from
Tables 2.2 and 2.3 which are within the 95% confidence interval.
Table E.1 Restricted maximum likelihood (REML) estimates for models from Table 2.2
which are within the 95% confidence interval.
Model Factor Estimate SE d.f. t P 13CO2 = Land*
random effects Intercept 0.05 Residual 0.35
fixed effects
Intercept 3.50 0.15 12 23.85 < 0.001 Land-Ha -0.72 0.21 8 -3.46 <0.01 Land-Pa 0.38 0.21 8 1.85 0.10 Land-Pi -0.35 0.21 8 -1.69 0.13 13CO2 = 1*
random effects
Intercept 0.41
Residual 0.36
fixed effects
Intercept 3.33 0.14 12 24.09 < 0.001 13CO2 = MicC*
random effects
Intercept 0.42
Residual 0.35
fixed effects
Intercept 3.48 0.19 11 18.06 < 0.001 MicC -0.02 0.02 11 -1.13 0.28 13CO2 = F:B*
random effects
Intercept 0.40
Residual 0.37
fixed effects
Intercept 3.35 0.14 11 23.73 < 0.001 F:B -0.39 0.82 11 -0.48 0.64 13CO2 = SIR*
random effects
Intercept 0.40
Residual 0.37
fixed effects
Intercept 3.29 0.27 11 12.08 < 0.001 SIR 0.32 1.78 11 0.18 0.86
252
Table E.2 Restricted maximum likelihood (REML) estimates for models from Table 3
which are within the 95% confidence interval.
Model Factor Estimate SE d.f. t P 13CO2 = P*
random effects
Intercept 0.02
Residual 0.35
fixed effects
Intercept 2.81 0.12 12 23.72 <0.001 P 0.48 0.09 10 5.48 <0.001 13CO2 = P + MicC*
random effects
Intercept 0.09
Residual 0.34
fixed effects
Intercept 2.91 0.16 11 18.31 <0.001 P 0.48 0.09 10 5.22 <0.001
Mic C -0.02 0.02 11 -0.95 0.36 13CO2 = P + SIR*
random effects
Intercept 0.00
Residual 0.36
fixed effects
Intercept 2.68 0.23 11 11.88 <0.001 P 0.48 0.09 10 5.43 <0.001
SIR 0.98 1.45 11 0.68 0.51 13CO2 = P + Seas.*
random effects
Intercept 0.00
Residual 0.36
fixed effects
Intercept 2.77 0.14 11 19.68 <0.001 P 0.48 0.09 10 5.42 <0.001
Seas. (Winter) 0.09 0.15 11 0.64 0.53
253
APPENDIX F
TABLE SHOWING THE INITIAL %C, %N, AND THE C:N RATIOS OF THE
LITTERS AND THE SOIL INOCULA
Table F.1
Source %C %N C:N
Rhododendron
Litter 48.9 ± 0.24 0.42 ± 0.01 116 ± 1.70
Pine Litter 47.3 ± 0.67 0.66 ± 0.01 96.7 ± 1.81
Grass Litter 43.4 ± 0.27 0.49 ± 0.02 66.3 ± 1.60
Rhododendron
Inoc. 4.5 ± 0.16 0.18 ± 0.01 25.3 ± 1.42
Pine Inoc. 2.7 ± 0.24 0.13 ± 0.01 21.5 ± 0.57
Grass Inoc. 3.3 ± 0.54 0.32 ± 0.03 10.1 ± 0.73
Values shown are means ± 1 SE (n=3).
254
APPENDIX G
BRAY-CURTIS SIMILARITY IN MICROBIAL COMMUNITY FUNCTION
Table G.1 Bray-Curtis similarity (%) matrix between the cumulative mineralization for
each inoculum for rhododendron litter during the entire incubation period (All periods)
and for each incubation period (Days 2-25, 26-99, and 100-300). Similarities were
calculated from the mean of eight replicates per treatment. The home inoculum
(Rhododendron) is in bold.
Incubation
period
Inoculum Source Rhododendron Pine Grass
All periods Rhododendron 100.00
Pine 65.13 100.00
Grass 33.23 58.41 100.00
Days 2-25 Rhododendron 100.00
Pine 93.81 100.00
Grass 99.88 93.69 100.00
Days 26-99 Rhododendron 100.00
Pine 54.23 100.00
Grass 39.56 79.73 100.00
Days 100-300 Rhododendron 100.00
Pine 67.15 100.00
Grass 22.79 40.56 100.00
255
Table G.2 Bray-Curtis similarity (%) matrix between the cumulative mineralization for
each inoculum for pine litter during the entire incubation period (All periods) and for
each incubation period (Days 2-25, 26-99, and 100-300). Similarities were calculated
from the mean of eight replicates per treatment. The home inoculum (Pine) is in bold.
Incubation
period
Inoculum Source Rhododendron Pine Grass
All periods Rhododendron 100.00
Pine 84.42 100.00
Grass 76.56 62.36 100.00
Days 2-25 Rhododendron 100.00
Pine 99.83 100.00
Grass 98.28 98.45 100.00
Days 26-99 Rhododendron 100.00
Pine 57.11 100.00
Grass 84.15 45.00 100.00
Days 100-300 Rhododendron 100.00
Pine 95.38 100.00
Grass 65.89 70.04 100.00
256
Table G.3 Bray-Curtis similarity (%) matrix between the cumulative mineralization for
each inoculum for grass litter during the entire incubation period (All periods) and for
each incubation period (Days 2-25, 26-99, and 100-300). Similarities were calculated
from the mean of eight replicates per treatment. The home inoculum (Grass) is in bold.
Incubation
period
Inoculum Source Rhododendron Pine Grass
All periods Rhododendron 100.00
Pine 93.67 100.00
Grass 97.70 91.38 100.00
Days 2-25 Rhododendron 100.00
Pine 93.00 100.00
Grass 86.12 93.05 100.00
Days 26-99 Rhododendron 100.00
Pine 89.48 100.00
Grass 84.05 94.48 100.00
Days 100-300 Rhododendron 100.00
Pine 95.14 100.00
Grass 83.94 79.24 100.00
257
Figure G.1 Bray-Curtis similarity (%) between the decomposer community function
across all incubation periods for the three soil inocula in rhododendron litter (A), pine
litter (B), and grass litter (C). The ―home‖ soil inoculum is highlighted in bold type.
Pine
Group average
DS
DC
DD
Sa
mp
les
100908070605040
Similarity
Resemblance: S17 Bray Curtis similarity
Grass
Group average
SD
SC
SS
Sa
mp
les
100908070605040
Similarity
Resemblance: S17 Bray Curtis similarity
Grass Inoculum
Rhodo. Inoculum
Grass Inoculum
Rhodo. Inoculum
Pine Inoculum
A
B
C
Rhodo
Group average
CS
CC
CD
Sa
mp
les
100806040
Similarity
Resemblance: S17 Bray Curtis similarity
Rhodo. Inoculum
Pine Inoculum
40 60 80 100
Similarity (%)
Pine Inoculum
Grass Inoculum
258
APPENDIX H
EDAPHIC CHARACTERISTICS FOR THE CALHOUN PLOTS
Table H.1 Edaphic characteristics measured at each plot. Cultivated and pasture sites were classified as herbaceous-cover and pine
and hardwood plots were classified as forest-cover. Means of three analytical repeats are shown.
Land-use Plot DOC (μg g
dry wt soil-1)
NO3- (μg
g dry wt
soil-1)
NH4+ (μg
g dry wt
soil-1)
Total C
(mg g
dry
weight
soil-1)
POM C
(mg g
dry
weight
soil-1)
Mineral
C (mg g
dry
weight
soil-1)
Total N
(mg g
dry
weight
soil-1)
POM N
(mg g
dry
weight
soil-1)
Mineral
N (mg g
dry
weight
soil-1)
Total P
(μg g dry
wt soil-1)
Cultivated Cu1 166.91 11.54 0.64 15.64 6.03 9.61 0.95 0.24 0.71 17.13
Cultivated Cu2 129.94 18.04 0.58 10.96 3.57 7.39 0.88 0.19 0.69 16.74
Cultivated Cu3 183.88 46.45 0.75 10.42 4.17 6.25 0.69 0.16 0.53 18.59
Pasture Pa1 72.95 10.09 0.84 9.14 5.21 3.93 0.68 0.30 0.39 9.91
Pasture Pa2 59.59 9.50 0.95 9.22 5.38 3.84 0.64 0.27 0.37 18.86
Pasture Pa3 54.95 26.34 1.71 11.27 7.67 3.61 0.78 0.43 0.35 18.94
Pine Pi1 38.64 0.04 5.47 8.21 4.52 3.69 0.33 0.12 0.21 1.02
Pine Pi2 73.56 0.02 4.05 8.18 4.66 3.53 0.56 0.24 0.31 13.45
Pine Pi3 114.33 0.00 0.77 9.52 5.20 4.32 0.30 0.12 0.18 2.30
Hardwood Ha1 146.66 0.17 7.54 13.88 7.07 6.80 0.77 0.25 0.52 1.07
Hardwood Ha2 242.11 0.00 3.23 24.94 14.73 10.21 0.90 0.45 0.45 3.55
Hardwood Ha3 96.95 10.33 4.13 16.10 7.93 8.17 0.82 0.27 0.55 0.74
Top Related