Ecology, 94(2), 2013, pp. 403–413
? 2013 by the Ecological Society of America
The responses of soil and rhizosphere respiration to simulated
climatic changes vary by season
VIDYA SUSEELA1,4AND JEFFREY S. DUKES1,2,3
1Department of Forestry and Natural Resources, Purdue University, West Lafayette, Indiana 47907 USA
2Department of Biological Sciences, Purdue University, West Lafayette, Indiana 47907 USA
3Department of Biology, University of Massachusetts Boston, Boston, Massachusetts 02125 USA
terrestrial carbon storage and, thus, feed back to warming. To provide insight into how
warming and changes in precipitation regimes affect the rate and temperature sensitivity of Rs
and rhizosphere respiration (Rr) across the year, we subjected a New England old-field
ecosystem to four levels of warming and three levels of precipitation (ambient, drought, and
wet treatments). We measured Rsand heterotrophic respiration (Rh) monthly (in areas of the
plots with and without plants, respectively) and estimated Rrby calculating the difference in
respiration between Rsand Rh. Even in this mesic ecosystem, Rsand Rrresponded strongly to
the precipitation treatments. Drought reduced Rsand Rr, both annually and during the
growing season. Annual cumulative Rsresponded nonlinearly to precipitation treatments;
both drought and supplemental precipitation suppressed Rs compared to the ambient
treatment. Warming increased Rsand Rrin spring and winter when soil moisture was optimal
but decreased these rates in summer when moisture was limiting. Cumulative winter Rr
increased by about 200% in the high warming (;3.58C) treatment. The effect of climate
treatments on the temperature sensitivity of Rsdepended on the season. In the fall, the drought
treatment decreased apparent Q10relative to the other precipitation treatments. The responses
of Rsto warming and altered precipitation were largely driven by changes in Rr. We emphasize
the importance of incorporating realistic soil moisture responses into simulations of soil
carbon fluxes; the long-term effects of warming on carbon–climate feedback will depend on
future precipitation regimes. Our results highlight the nonlinear responses of soil respiration to
soil moisture and, to our knowledge, quantify for the first time the loss of carbon through
winter rhizosphere respiration due to warming. While this additional loss is small relative to
the cumulative annual flux in this system, such increases in rhizosphere respiration during the
non-growing season could have greater consequences in ecosystems where they offset or
reduce subsequent warming-induced gains in plant growth.
Responses of soil respiration (Rs) to anthropogenic climate change will affect
microbial respiration; precipitation; root respiration; temperature sensitivity; warming;
The rate of terrestrial carbon (C) storage depends on
the balance between C fixed by photosynthesis and
released to the atmosphere through plant and hetero-
trophic respiration (Friedlingstein et al. 2006). Thus, if
climate change alters rates of soil respiration without
offsetting changes in plant productivity, it will affect the
C budget. Although temperature and moisture are
clearly major drivers of Rs(Luo and Zhou 2006), the
seasonal responses of the two components of soil
respiration (heterotrophic [Rh] and rhizosphere [Rr]
respiration) to combined effects of warming and altered
precipitation are poorly understood (Reichstein and
Beer 2008). Wintertime processes have received the least
attention; few experimental studies have included winter
measurements of soil respiration, and we are not aware
of any studies that have quantified winter rhizosphere
respiration to warming. A recent warming and nitrogen
addition experiment at Harvard Forest reported a
winter flux of 2–17% of annual C and nitrogen flux,
emphasizing the importance of winter measurements for
predicting winter biogeochemical processes (Contosta et
al. 2011). Rhizosphere respiration during the non-
growing season is mostly for maintenance of roots,
and a majority of the C used for winter maintenance
respiration is derived from stored resources (Regier et al.
2010, Kuptz et al. 2011). C reserves can help plants,
especially mature trees, to survive through environmen-
tal stresses such as droughts, fires, or pest outbreaks
(Chapin et al. 1990). Increases in winter temperatures
due to climate change could deplete plant C reserves by
enhancing maintenance respiration. This could subse-
quently reduce reemergence or new root growth and
Manuscript received 27 January 2012; revised 15 August
2012; accepted 5 September 2012. Corresponding Editor: S. D.
4Present address: School of Agricultural, Forest, and
Environmental Sciences, Clemson University, South Caro-
lina 29634 USA. E-mail: firstname.lastname@example.org
productivity of plants in the spring, reducing C uptake
and offsetting positive effects of warmer spring temper-
atures on growth. Previous studies evaluating the effect
of warming on Rrhave focused on the snow-free season
(Scott-Denton et al. 2006, Schindlbacher et al. 2009),
and have not considered the effects of warming on
winter rhizosphere respiration and in turn on plant
carbon use efficiency (CUE; ratio of NPP to production
plus autotrophic respiration [Ryan et al. 1997]) and
ecosystem C balance.
Untangling the effects of multiple factors of climate
change on terrestrial C stocks is complex due to the
differential responses of Rhand Rr. Rhresults from the
microbial decomposition of a range of substrates,
including soil organic matter and plant litter with
varying ages and complexities (Trumbore 2006). Rr
not only includes root and associated mycorrhizal
respiration, but also the decomposition of labile root
exudates by microorganisms in the root (Kuzyakov and
Larionova 2005). Although several studies have report-
ed similar responses of Rr and Rh to warming
(Schindlbacher et al. 2009), many other experiments
have reported a greater response of Rrthan Rh(Boone et
al. 1998, Lavigne et al. 2003).Our lack of understanding
of the relative sensitivities of these responses limits our
ability to predict soil C loss in future climate scenarios.
Although Rs is often modeled as a function of
temperature, changes in soil moisture, photosynthesis,
and substrate availability may modify this temperature
response function (Davidson et al. 1998). Temperature
sensitivity of Rsderived from annual patterns may not
be accurate, due to seasonal changes in plant phenology
and belowground C allocation (Yuste et al. 2004).
Hence, representing Rsand Rras exponential functions
of temperature could over- or underestimate respiration,
reducing the realism of C budget projections under
future climate scenarios.
To date, most studies with multiple climate factors
such as warming and altered precipitation have been
conducted in semiarid ecosystems where soil water
availability is critical to many ecosystem processes (Liu
et al. 2009). Mesic systems are thought to be less
responsive to changes in soil water. However, Suseela et
al. (2012) found that soil heterotrophic respiration
responded more strongly to manipulations of soil
moisture than manipulations of temperature in a mesic
old field. Whereas Suseela et al. (2012) explored the
response of heterotrophic respiration (Rh) alone, the
present study, using the same plots at the Boston-Area
Climate Experiment (BACE), seeks to address the
following questions: (1) How do soil and rhizosphere
respiration (Rr) respond to gradients of warming and
precipitation at different temporal scales? and (2) What
are the roles of warming and precipitation in driving the
temperature sensitivities of Rs and Rr at different
temporal scales? We hypothesized that the rate and
temperature sensitivity of Rsto warming would strongly
depend on soil moisture. During dry periods in the
growing season, we expected warming to dry the soil,
and thus suppress Rsby limiting substrate availability.
We also hypothesized that the response of Rr to
warming and altered precipitation would depend on
the season. More specifically, we expected that warming
would increase Rrin winter.
MATERIALS AND METHODS
Measurements were taken from March 2009 to March
2010 at the Boston-Area Climate Experiment (BACE)
located in Waltham, Massachusetts, USA (42823.10N,
71812.90W). The study site was an old-field dominated
by grasses and forbs (;40 species; Hoeppner and Dukes
2012). We also planted seedlings of four tree species;
Acer rubrum, Quercus rubra, Betula lenta, and Pinus
strobus. Nearby, the city of Boston, Massachusetts has a
mean annual temperature of 9.58C and mean annual
precipitation of 1194 mm (NOAA National Climatic
Data Center Cooperative station ID 190535, years
1960–2008). The soil at the site is a Mesic Typic
Dystrudept with a loamy topsoil (45% sand, 46% silt,
9% clay; gravel content 7%) and a gravelly sandy loam
The BACE employs a factorial design, with precipi-
tation treatments applied to main plots and warming
treatments applied to subplots across three experimental
blocks (36 plots in total; Tharayil et al. 2011). The four
warming treatments included unwarmed controls, and
low, medium, and high warming levels, which warmed
the plant canopy by a maximum of ;18, ;2.78, and 48C,
respectively. For each 2 3 2 m plot, warming was
achieved using four ceramic infrared heaters, which were
mounted 1 m above the corners of each subplot. Heaters
of different wattages were used in the low (200 W),
medium (600 W), and high warming (1000 W)
treatments. Each precipitation plot contained four
subplots that were arranged linearly, with one subplot
assigned to each of the four warming treatments. Within
each main plot, canopy temperatures of the control and
high warming subplots were measured every 10 s using
infrared radiometers. Feedback control of all four
subplots was achieved by programming control software
(Labview, National Instruments, Austin, Texas, USA)
to limit the temperature difference between the un-
warmed and high subplots to a maximum of 48C, and by
controlling all heaters within each precipitation plot on
the same circuit. Main plots experienced either ambient
precipitation, a ‘‘drought’’ treatment, or a ‘‘wet’’
treatment. Drought treatments were located under
rainout shelters with polycarbonate slats (15 cm wide)
that excluded 50% of the rainfall. During the non-
freezing months, this water was diverted to storage tanks
and immediately applied to the wet section of the block
using overhead sprinklers. The wet treatment thus
VIDYA SUSEELA AND JEFFREY S. DUKES 404 Ecology, Vol. 94, No. 2
received 150% of ambient rainfall during the growing
season. By July 2008, all treatments were operational.
We measured soil and heterotrophic respiration
monthly, between 10:00 and 16:00 (local time), using a
LiCor 6400 attached to a 6400-09 soil CO2flux chamber
(Li-Cor, Lincoln, Nebraska, USA). In each plot,
measurements of Rs were taken inside two replicate
PVC rings that were 10 cm in diameter and 5 cm in
height, inserted to a depth of 2–3 cm in the ground; these
measurements captured respiration from plant roots and
microbes (shoots were removed from within the rings).
Measurements of Rhwere taken in each plot using a
similar PVC ring (10 cm diameter and 5 cm height)
installed inside a larger (25 cm diameter, 30 cm deep)
PVC ‘‘plant exclusion collar’’ that excluded plant roots
and organic matter inputs (Suseela et al. 2012). The CO2
efflux from this collar was mainly due to the microbial
decomposition of soil organic matter (Suseela et al.
2012). Most of the root growth at our site occurred in
the top 10 cm and few roots grew beyond 30 cm; thus,
we could rule out the possibility of meaningful amounts
of root respiration coming from the plant exclusion
collar. Along with respiration measurements, we simul-
taneously measured soil temperature (5 cm depth) in
each ring using a thermocouple attached to the LiCor
6400. We also measured volumetric soil moisture in the
top 10 cm of the soil column in all main plots and plant
exclusion collars using time domain reflectometry
Rhizosphere respiration (Rr) was estimated as the
difference in soil CO2efflux between Rsand corrected
Rh. Because soils inside the plant exclusion collars were
wetter than those in the surrounding plot, we corrected
Rhvalues for the difference in soil moisture. A best-fit
quadratic equation was obtained relating Rh to soil
moisture in the plant exclusion collar. This function was
then used to ‘‘correct’’ measured Rhvalues, based on
differences in moisture between soil inside and outside of
each plant exclusion collar. Soil temperatures also
differed slightly (P , 0.05) between the Rh and Rs
collars on four of the measurement dates; soils were
0.78C warmer in the plant exclusion collars in March
and May, and 0.6–0.88C cooler in August and Septem-
ber. Because these temperature differences were slight
(,18C) and transient, we did not correct for them. Soil
respiration and its temperature sensitivity in the low
warming treatment (which consistently warmed plots by
;18C) were statistically similar to the unwarmed
treatment, suggesting that temperature correction was
We periodically measured diel respiration (measure-
ments every 2 hours over 24-hour cycles) in both the
main plots and plant exclusion collars. We made diel
measurements only in the unwarmed and high warming
treatments, due to time constraints. The diel measure-
ments were used to calculate the annual, growing
season (April–September), non-growing season (Octo-
ber to March), and winter season (December to
February) cumulative Rs and Rr. We followed the
method of Bremer et al. (1998) for calculating
cumulative values. Briefly, respiration measured at the
BACE during the daytime was assumed to be the daily
maximum soil CO2efflux. We used diel measurements
to calculate the daily minimum efflux as a percentage of
maximum efflux. The daily minimum and maximum
efflux were used to calculate the average daily efflux.
We estimated cumulative flux as the product of average
daily flux and the number of days between each
In the main plots, Rs measurements from the two
replicate rings were averaged by plot for statistical
analysis. To test the main and interactive effects of
warming and altered precipitation on climate variables
(soil temperature and soil moisture) and on respiration,
both annually and seasonally, we used mixed model
restricted maximum likelihood estimation with repeated
measures (PROC MIXED; SAS version 9.2; SAS
Institute 2008). Warming and precipitation treatments
were assigned as fixed factors and block as a random
factor. We used Tukey’s HSD multiple comparison test
to identify differences among treatments.
We used an exponential function (Eq. 1; Zhou et al.
2007) to calculate the temperature sensitivity of Rsand
where Rxis the soil CO2efflux (lmol?m?2?s?1), T is the
soil temperature (8C) at 5 cm depth, a is the basal
respiration, and b is the temperature sensitivity of soil
CO2efflux. The respiratory quotient (Q10) is calculated
as Q10¼ e10b.
To assess the effect of moisture on Rswe fitted Rsand
volumetric soil moisture using a quadratic function (Eq.
2; Wan and Luo 2003):
Rs¼ y0þ ax þ bx2
where x is the volumetric moisture content (V/V) in the
top 10 cm of soil, and y0, a, and b are constants. To
evaluate the effect of both soil temperature and soil
moisture on Rs, we fitted Rs using a combined
exponential and quadratic function (Eq. 3; Mielnick
and Dugas 2000):
Rs¼ ðaebTÞ 2:12ðhv ? min hvÞðmax hv ? hvÞc
where hv is the volumetric moisture content (the
minimum volumetric water content of our data set was
0.3% and the maximum was 36.8%) and c is the
coefficient for soil moisture. We used Sigmaplot (version
12; Systat Software, San Jose, California, USA) for all
February 2013 405 SOIL RESPIRATION AND CLIMATE CHANGE
Soil temperature at 5 cm depth varied seasonally from
an average minimum of 0.298 6 0.58C (mean 6 SE) in
January to a maximum of 26.48 6 0.68C in August in the
unwarmed, ambient-precipitation treatment (Appendix
A: Fig. A1a). Warming treatments consistently altered
soil temperature throughout the summer and fall (P ,
0.0001; Appendix A: Table A1). However, the effect of
warming varied by month during spring (P ¼ 0.0137)
and winter (P , 0.0001). Warming increased soil
temperature by an annual average of 0.768, 2.38, and
3.18C in the low, medium, and high warming treatments,
respectively, relative to the unwarmed treatment (P ,
0.0001). From July to September, drought treatment
plots were warmer than the ambient (by 28C) and wet
plots (by 2.68C; P , 0.05).
Soil moisture fluctuated dramatically throughout the
year, corresponding to rain events (Appendix A: Fig.
A1b). The drought treatment had lower soil moisture
from April to October (except for June) compared to
both ambient and wet treatments (P , 0.05; Appendix
A: Table A2). The reduction in soil moisture in the
drought treatment varied from 19% (in April compared
to ambient treatment) to 80% (in August compared to
ambient treatment). High warming (þ;48C) reduced
soil moisture compared to the unwarmed treatment
from April to October and compared to the low
(þ;18C) warming treatment during May, September,
and October (Appendix A: Fig. A2). In May, medium
(þ;2.78C) warming reduced soil moisture relative to the
unwarmed and the low warming treatments.
Soil and rhizosphere respiration: cumulative measures
The annual cumulative Rs for the 12 treatments
ranged from 717–1103 g C?m?2?yr?1. Although warming
did not affect the annual cumulative Rs, warming had
opposing effects on cumulative Rsduring the growing (P
¼ 0.0037) and non-growing seasons (P , 0.0001).
During the growing season, both medium (P ¼ 0.0072)
and high warming (P¼0.0117; Fig. 1a) suppressed Rsby
17% and 16%, respectively, compared to the unwarmed
treatment. This trend was reversed during the non-
growing season, when medium and high warming
increased Rsby 16% and 37% respectively, compared
to the unwarmed treatment (P , 0.0001; Fig. 1b). Over
this period, high warming also increased cumulative
respiration compared to the low (P¼0.0028;þ24%) and
medium (P ¼ 0.0196; þ18%) warming treatments.
Drought reduced annual cumulative Rs (806 6 32 g
C?m?2?yr?1) compared to ambient (1061 6 21 g
C?m?2?yr?1; P , 0.0001) and wet treatments (975 6 24
g C?m?2?yr?1; P ¼ 0.0001; Appendix A: Table A3).
During the growing season, drought reduced cumulative
Rsby 28% and 23% compared to the ambient (P ,
respiration. Bars represent meansþSE (n¼9 samples). The dagger (?) in panel (c) marks the high-warming treatment, which only
marginally decreased Rrcompared to the unwarmed treatment (see Results: Soil and rhizosphere respiration: cumulative measures).
Cumulative growing-season (April–September) and non-growing-season (a, b) soil respiration and (c, d) rhizosphere
VIDYA SUSEELA AND JEFFREY S. DUKES 406 Ecology, Vol. 94, No. 2
0.0001) and wet (P , 0.0001) treatments, respectively.
The wet treatment showed a trend of decreasing annual
cumulative Rsin comparison with the ambient treatment
(P¼0.08;?8%) from April 2009 to March 2010. We also
calculated cumulative Rs from December 2008 to
November 2009 to check whether the above pattern
was consistent over time. During this period, supple-
mental precipitation suppressed cumulative Rsby 10%
compared to ambient (P ¼ 0.046).
Annual cumulative Rr varied with precipitation
treatments; drought reduced estimated annual Rrrelative
to the ambient (?29%; P¼0.0002) and wet precipitation
treatments (?20%; P ¼ 0.0094; Appendix A: Table A4).
Growing season cumulative Rr responded similarly;
drought decreased Rrcompared to ambient (?34%; P
, 0.0001) and supplemental precipitation treatments
(?29%; P¼0.0008). During the growing season medium
warming decreased Rrcompared to unwarmed plots (P¼
0.0067) and high warming marginally decreased Rr
compared to the unwarmed treatment (P ¼ 0.07; Fig.
1c). During the non-growing season, high warming
increased cumulative Rrcompared to low (þ54%; P ¼
0.0349) and unwarmed (þ81%; P ¼ 0.0008) treatments
(Fig. 1d). When winter months were analyzed separately,
the high warming treatment increased cumulative Rrby
243% and 202 % compared to no warming (P ¼0.0027)
and low warming treatments (P ¼ 0.0044), respectively
(Fig. 2a). Rhizosphere respiration also contributed a
higher proportion of cumulative winter Rsin the high
warming treatment (Fig. 2b).
Soil respiration (Rs) and rhizosphere respiration (Rr):
Across the year, soil respiration generally tracked soil
temperature, with the maximum efflux during summer
and the minimum in winter (Appendix A: Fig. A3a).
However, seasonal variability in precipitation modified
the soil respiration pattern (Appendix A: Fig. A3b, c).
The effect of warming and precipitation treatments on
Rsover the year (April 2009–March 2010) varied by
month (Table 1; Appendix A: Table A5; P , 0.0001).
During early and late winter, high warming increased
respiration (Fig. 3; Appendix A: Table A6). However,
during summer (June and August) warming decreased
Rs(Fig. 3a; Appendix A: Table A6). During summer
and fall, precipitation treatments altered Rs(Appendix
A: Fig. A4). The diel pattern of soil respiration did not
closely follow the diel patterns of soil temperature in
either August 2009 (Appendix A: Fig. A5) or April 2010
(Appendix A: Fig. A6).
Rhizosphere respiration was higher from April to
September, which coincided with the period of maxi-
mum plant activity. Rhizosphere respiration of the
lmol?m?2?s?1in July to no detectable respiration on
measurement dates in January and February. The effects
of warming (P , 0.0001; Table 2) and precipitation on
Rrvaried by month (P , 0.0001; Fig. 4; Appendix A:
Table A7). Drought decreased Rrin August (61%; P ¼
0.0142; Fig. 4b) and had a marginal effect in June (48%;
P ¼ 0.06) compared to ambient precipitation. During
September, drought decreased Rr compared to both
ambient (49%) and wet (55%) treatments (P , 0.0001).
Soil respiration (Rs) and rhizosphere respiration (Rr):
When we modeled Rsusing the exponential temper-
ature function (Eq. 1), temperature at 5 cm depth
sphere respiration Rr and (b) proportion of cumulative
respiration attributable to rhizosphere and heterotrophic
respiration. Values represent means (n ¼ 9 samples); þSE is
shown in panel (a).
(a) Cumulative winter (December–February) rhizo-
compared to ambient and wet precipitation treatments
during different months.
Decrease in soil respiration (Rs) in drought
February 2013407SOIL RESPIRATION AND CLIMATE CHANGE
explained more than 50% of the variation in Rsin the
ambient and wet precipitation treatments (Fig. 5, Table
3). However, the correlation of Rs with temperature
decreased to about 25% in the high and medium
warming, drought treatments. The annual Q10 of Rs
varied from 1.29 in the high warming, drought treatment
to 2.07 in the unwarmed, wet treatment.
Mixed model analysis of annual apparent Q10
revealed that warming and altered precipitation both
affected the apparent Q10of Rs(Appendix A: Table A8)
and Rr(Appendix A: Table A9 and Appendix B). The
high and medium warming treatments decreased appar-
ent Q10and coefficient b of Rsand Rrcompared to the
low and unwarmed treatments (P , 0.05). Drought
decreased apparent Q10and coefficient b of Rs(P ,
0.05) and Rr (P ¼ 0.05 and P ¼ 0.036, respectively)
relative to the wet treatment. Basal respiration (coeffi-
cient a) was not affected by either warming or
precipitation. Apparent Q10of Rsranged from 1.4–2.8
in the spring, but plummeted to about 1.0 during the
summer drought (Fig. 6). In the fall, the apparent Q10
ranged from 1.2 to 2.4. The effect of climatic treatments
on apparent Q10and the coefficients varied with seasons
Soil respiration: characterizing sensitivity to moisture
Most of the annual variability in Rswas explained by
soil temperature at 5 cm (Eq. 1; Appendix C: Table C1).
During spring, the combined moisture and temperature
(Melnick-Dugas) model provided the best fit to data
from the high warming treatments across all precipita-
tion treatments. In the other treatments, the exponential
temperature model explained most of the variation. The
soil moisture function (Eq. 3) performed best in summer
and the temperature function alone provided the best fit
in the fall.
Many studies have examined the effects of warming
on Rrduring the growing season, but to our knowledge,
this is the first study to examine the response of winter
Rrto warming. We speculate that the increased Rrwe
observed in winter, presumably maintenance respiration
(Rm), may not only deplete stored plant C, but also
reduce productivity in the following season. If increases
in non-growing-season Rmsuch as this were a general
phenomenon, they could alter ecosystem C budgets,
particularly in high latitude ecosystems with a long cold
season. Here, in a mesic ecosystem where water
availability might not be expected to drive ecosystem
processes, the effect of warming on Rs and Rr
nonetheless depended on precipitation. Seasonal vari-
ability in soil moisture strongly influenced temperature
sensitivity of Rs, and cumulative annual Rsresponded
nonlinearly to precipitation inputs.
Response of soil (Rs) and rhizosphere (Rr) respiration
to warming and altered precipitation
Many single-factor climate change experiments have
reported increases in soil respiration with warming
(Rustad et al. 2001, Mellillo et al. 2002). In contrast,
we found that the effect of warming on Rsand Rrvaried
by season in the BACE, with warming decreasing Rs
(Appendix A: Fig. A7) and Rr during summer, but
increasing them during spring and winter. These
contrasting effects of warming in different seasons are
most likely a consequence of differences in soil moisture
content. Drought also decreased annual cumulative Rs
winter and (b) winter alone from March 2009 to February 2010. Values represent means 6 SE. Key: W, warming; M, month.
Seasonal variation in soil respiration Rsaveraged by warming treatment (n ¼ 9 samples) in (a) spring, summer, fall,
warming compared to low and unwarmed treatments during
different measurement periods.
Increase in rhizosphere respiration (Rr) in high
High vs. low
Note: Empty cells indicate that there was no significant
difference in rhizosphere respiration between high and low
warming in June and November.
VIDYA SUSEELA AND JEFFREY S. DUKES 408 Ecology, Vol. 94, No. 2
and Rr, likely due to the soil moisture stress experienced
by plants and microbes (Shen et al. 2008). Dry soils
reduce the movement of substrates in the soil, lowering
the activity of root microbes, and stressing plants,
leading to reduced photosynthesis and belowground C
allocation. Drought also reduces the coupling between
plant photosynthesis and belowground processes by
reducing the movement of photosynthates to phloem
loading sites and also by impairing the phloem loading
itself, which potentially affects substrate availability and
hence C cycling (Ruehr et al. 2009). During winter and
spring, when more water was available for root and
microbial activity, warming increased Rs. However,
warming exacerbated soil moisture limitation of plant
and microbial activity during summer, further reducing
Response of non-growing season rhizosphere respiration
(Rr) to warming
Although many studies evaluating the effects of
warming on rhizosphere respiration have omitted the
winter months, information on temperature sensitivity
during these months can contribute to useful bench-
marks for ecosystem models. Ecosystem autotrophic
respiration in the dormant season (mostly Rm) can
constitute .25% of the annual C budget in forest
systems (Ryan et al. 1997), and model representations of
the warming response of respiration can strongly
influence projected climate feedbacks (Piao et al.
2010). In the BACE, warming treatments increased Rr
during the non-growing season (October to March). The
pattern of increase in Rr during winter in the high
warming treatment was consistent over four years
precipitation treatment (n ¼ 12 samples) Symbols represent means 6 SE.
Seasonal variation in Rrfrom March 2009 to March 2010 averaged by (a) warming treatment (n ¼ 9 samples) and (b)
February 2013 409 SOIL RESPIRATION AND CLIMATE CHANGE
(2008–2012) of treatment application (V. Suseela and
J. S. Dukes, unpublished data). The energy expended in
this process, presumably as Rm, would have reduced
carbohydrate reserves. We speculate that this C loss
could affect emergence and reduce growth of perennial
species the following year (Ogren et al. 1997, Regier et
al. 2010). Differences among treatments in cumulative
non-growing season Rrsuggest that the high (þ;48C)
warming treatment would sustain an ecologically
meaningful loss of C by increasing Rr in the non-
growing season (Fig. 1). During the growing season,
warming could offset this C loss through increases in
production due to the lengthening of the growing season
or increased nutrient availability (Melillo et al. 2011).
However, such increases are not universal (Zhao and
Running 2010) and can depend on precipitation regimes
(Berdanier and Klein 2011, Hoeppner and Dukes 2012).
Further research to characterize the implications of
increased Rrin the non-growing season for subsequent
growth and productivity could help to refine ecosystem
Nonlinear response of cumulative soil respiration to
Drought and wet treatments both decreased cumula-
tive Rs. The decrease in cumulative Rs in the wet
treatment resulted from equal reductions in Rrand Rh.
The reduction in microbial activity and respiration could
have resulted from limited diffusion of oxygen, as
previously discussed (Suseela et al. 2012). Unstudied
changes in microbial community structure could have
also affected respiration; a mechanistic understanding of
how these communities respond to warming and
changes in precipitation would help to better predict
the microbially mediated carbon-climate feedback (Da-
vidson et al. 2012, Zhou et al. 2012).
The reduction in Rrin the wet treatment could have
resulted from a nutrient feedback. Aboveground tree
biomass was higher in the wet than the drought
treatments (S. S. Hoeppner and J. S. Dukes, unpublished
data), suggesting that more water and nutrients were
available for plant production. Increased nitrogen
availability has been found to decrease soil respiration
by reducing belowground C allocation, as plants need
not provide more C to symbionts to obtain nutrients
(Janssens et al. 2010). Increased nutrient availability has
also been shown to decrease fine root biomass and soil
respiration (Haynes and Gower 1995, Jimenez et al.
2009). Reduced C allocation below ground may
consequently have reduced the associated mycorrhizal
and rhizosphere microbial respiration.
temperature (April 2009–March 2010) in (a) drought, (b)
ambient, and (c) wet precipitation treatments. Details are in
Exponential relationship between Rs and soil
TABLE 3. Exponential relationship between Rsand soil temperature T (8C; April 2009–March 2010) and apparent Q10values.
VIDYA SUSEELA AND JEFFREY S. DUKES410 Ecology, Vol. 94, No. 2
Effects of climate treatments on temperature sensitivity of
Rsand Rr: implications for climate-carbon feedbacks
The uncertainty in climate-carbon feedback projec-
tions could be reduced by more accurately representing
the sensitivity of soil respiration to warming (Jones et al.
2003, Friedlingstein et al. 2006, Luo 2007). Many global
scale models represent the relationship between respira-
tory processes and temperature using a constant Q10
value of 2 (Friedlingstein et al. 2006, Mahecha et al.
2010). However, as soil respiration involves heterotro-
phic and autotrophic (rhizosphere) components con-
trolled by plant and climatic factors, the differential
responses of Rhand Rrto warming and precipitation
changes may alter the exponential relationship between
soil respiration and temperature. Temperature sensitiv-
ity can be influenced by changes in environmental
constraints to decomposition such as soil moisture and
substrate availability (Davidson and Janssens 2006).
Warming and drought can also cause physiological
stress to plants, slowing photosynthesis and below-
ground C allocation. In our system, the annual apparent
Q10of both Rsand Rrbehaved similarly, suggesting that
Rr drove the response of the apparent temperature
sensitivity of soil respiration to warming and altered
precipitation. Heterotrophic respiration at the BACE
(Suseela et al. 2012) was less sensitive to warming than
Rs and Rr, a pattern previously observed in other
systems (Boone et al. 1998, Zhou et al. 2007). The
annual apparent Q10 of Rs and Rr decreased with
warming and drought. Warming has also decreased the
temperature sensitivity of Rsin other systems (Luo et al.
2001, Zhou et al. 2006, 2012).
The apparent Q10 of Rs responded differently to
warming and altered precipitation in different seasons
and was greater in spring and fall than in summer (Fig.
6). Greater substrate availability in spring (due to freeze
thaw cycles; Schimel and Clein 1996) and fall (due to
litter inputs) could contribute to the higher Q10values.
Also, a severe soil moisture limitation in the summer
resulted in a sharp decline in apparent Q10across all
precipitation treatments and eliminated treatment dif-
ferences due to both warming and precipitation. This
underscores the strong dependence of Rson moisture,
even in mesic systems. The trees in the old field at the
BACE were relatively small during this study. Nonethe-
less, in a soil warming experiment with mature
hardwood trees in the same region, Contosta et al.
(2011) reported similar seasonal differences in the effect
of warming on Rs, with the greatest effects in spring and
Our results suggest that the apparent temperature
sensitivity of Rs varies with season mainly due to
changes in soil moisture and seasonal variability in
substrate availability. This has important implications
for climate-carbon modeling, as mean annual tempera-
ture and mean annual precipitation may not explain the
temperature sensitivity of Rsand Rr(Wang et al. 2010).
Rather, our results suggest that, at least in this system,
intra-annual variation in precipitation and subsequent
changes in substrate availability modulate the apparent
Q10of Rsand Rr. Although many studies have used
Arrhenius kinetics to represent temperature sensitivity,
their application might be limited under conditions of
varying substrate availability (Reichstein and Janssens
2009). Underlying assumptions of the Q10concept, such
as abundant substrate and stable enzyme quantities,
may not be met under future global change scenarios
(and may not be met currently; Ise and Moorcroft
2006). Thus, representing Rs(or its components Rhand
Rr) as exponential functions of temperature without
incorporating other potentially limiting factors could
overestimate the C cycle feedback to climate change.
We thank Carol Goranson, Hollie Emery, Eric Bestrom,
Nick Smith, and Susanne Hoeppner for help with the fieldwork.
ambient, and (c) wet precipitation treatments for each warming
treatment. Values represent means 6 SE (n ¼ 3).
Seasonal apparent Q10values of (a) drought, (b)
February 2013 411SOIL RESPIRATION AND CLIMATE CHANGE
We also thank Brita Jessen, Nicole Fallon, and the many others
who helped build and maintain the infrastructure of the BACE.
This research was financially supported by the NSF (DEB-
0546670) and the U.S. Department of Energy’s Office of
Science (BER), through the Northeastern Regional Center of
the National Institute for Climatic Change Research. This is
publication No. 1230 of the Purdue Climate Change Research
Berdanier, A. B., and J. A. Klein. 2011. Growing season length
and soil moisture interactively constrain high elevation
aboveground net primary production. Ecosystems 14:963–
Boone, R. D., K. J. Nadelhoffer, J. D. Canary, and J. P. Kaye.
1998. Roots exert a strong influence on the temperature
sensitivity of soil respiration. Nature 396:570–572.
Bremer, D. J., J. M. Ham, C. E. Owensby, and A. K. Knapp.
1998. Responses of soil respiration to clipping and grazing in
a tallgrass prairie. Journal of Environmental Quality
Chapin, F. S., E. D. Schulze, and H. A. Mooney. 1990. The
ecology and economics of storage in plants. Annual Review
of Ecology and Systematics 21:423–447.
Contosta, A. R., S. D. Frey, and A. B. Cooper. 2011. Seasonal
dynamics of soil respiration and N mineralization in
chronically warmed and fertilized soils. Ecosphere 2:1–21.
Davidson, E. A., E. Belk, and R. D. Boone. 1998. Soil water
content and temperature as independent or confounded
factors controlling soil respiration in a temperate mixed
hardwood forest. Global Change Biology 4:217–227.
Davidson, E. A., and I. A. Janssens. 2006. Temperature
sensitivity of soil carbon decomposition and feedbacks to
climate change. Nature 440:165–173.
Davidson, E. A., S. Samanta, S. S. Caramori, and K. Savage.
2012. The dual Arrhenius and Michaelis-Menten kinetics
model for decomposition of soil organic matter at hourly to
seasonal time scales. Global Change Biology 18:371–384.
Friedlingstein, P., P. Cox, R. Betts, L. Bopp, et al. 2006.
Climate–carbon cycle feedback analysis: results from the
C4MIP model intercomparison. Journal of Climate 19:3337–
Haynes, B. E., and S. T. Gower. 1995. Belowground carbon
allocation in unfertilized and fertilized red pine plantations in
northern Wisconsin. Tree Physiology 15:317–325.
Hoeppner, S. S., and J. S. Dukes. 2012. Interactive responses of
old-field plant growth and composition to warming and
precipitation. Global Change Biology 18:1754–1768.
Ise, T., and P. R. Moorcroft. 2006. The global-scale temper-
ature and moisture dependencies of soil organic carbon
decomposition: an analysis using a mechanistic decomposi-
tion model. Biogeochemistry 80:217–231.
Janssens, I. A., et al. 2010. Reduction of forest soil respiration
in response to nitrogen deposition. Nature Geoscience 3:315–
Jimenez, E. M., F. H. Moreno, M. C. Penuela, S. Patino, and J.
Lloyd. 2009. Fine root dynamics for forests on contrasting
soils in the Colombian Amazon. Biogeosciences 6:2809–2827.
Jones, C. D., P. Cox, and C. Huntingford. 2003. Uncertainty in
climate-carbon-cycle projections associated with the sensitiv-
ity of soil respiration to temperature. Tellus Series B-
Chemical and Physical Meteorology 55:642–648.
Kuptz, D., F. Fleischmann, R. Matyssek, and T. E. E. Grams.
2011. Seasonal patterns of carbon allocation to respiratory
pools in 60-yr-old deciduous (Fagus sylvatica) and evergreen
(Picea abies) trees accessed via whole-tree stable carbon
isotope labeling. New Phytologist 191:160–172.
Kuzyakov, Y., and A. A. Larionova. 2005. Root and
rhizomicrobial respiration: A review of approaches to
estimate respiration by autotrophic and heterotrophic
organisms in soil. Journal of Plant Nutrition and Soil Science
Lavigne, M. B., R. Boutin, R. J. Foster, G. Goodine, P. Y.
Bernier, and G. Robitaille. 2003. Soil respiration responses to
temperature are controlled more by roots than by decompo-
sition in balsam fir ecosystems. Canadian Journal of Forest
Liu, W. X., Z. Zhang, and S. Q. Wan. 2009. Predominant role
of water in regulating soil and microbial respiration and their
responses to climate change in a semiarid grassland. Global
Change Biology 15:184–195.
Luo, Y. Q. 2007. Terrestrial carbon-cycle feedback to climate
warming. Annual Review of Ecology, Evolution, and
Luo, Y. Q., S. Q. Wan, D. F. Hui, and L. L. Wallace. 2001.
Acclimatization of soil respiration to warming in a tall grass
prairie. Nature 413:622–625.
Luo, Y., and X. Zhou. 2006. Soil respiration and the
environment. Academic Press/Elsevier, San Diego, Califor-
Mahecha, M. D., et al. 2010. Global convergence in the
temperature sensitivity of respiration at ecosystem level.
Melillo, J. M., et al. 2011. Soil warming, carbon-nitrogen
interactions, and forest carbon budgets. Proceedings of the
National Academy of Sciences USA 108:9508–9512.
Melillo, J. M., P. A. Steudler, J. D. Aber, K. Newkirk, H. Lux,
F. P. Bowles, C. Catricala, A. Magill, T. Ahrens, and S.
Morrisseau. 2002. Soil warming and carbon-cycle feedbacks
to the climate system. Science 298:2173–2176.
Mielnick, P. C., and W. A. Dugas. 2000. Soil CO2flux in a
tallgrass prairie. Soil Biology and Biochemistry 32:221–228.
Ogren, E., T. Nilsson, and L. G. Sundblad. 1997. Relationship
between respiratory depletion of sugars and loss of cold
hardiness in coniferous seedlings over-wintering at raised
temperatures: Indications of different sensitivities of spruce
and pine. Plant Cell and Environment 20:247–253.
Piao, S. L., S. Luyssaert, P. Ciais, I. A. Janssens, A. P. Chen, C.
Cao, J. Y. Fang, P. Friedlingstein, Y. Q. Luo, and S. P.
Wang. 2010. Forest annual carbon cost: a global-scale
analysis of autotrophic respiration. Ecology 91:652–661.
Regier, N., S. Streb, S. C. Zeeman, and B. Frey. 2010. Seasonal
changes in starch and sugar content of poplar (Populus
deltoides x nigra cv. Dorskamp) and the impact of stem
girdling on carbohydrate allocation to roots. Tree Physiology
Reichstein, M., and C. Beer. 2008. Soil respiration across scales:
the importance of a model-data integration framework for
data interpretation. Journal of Plant Nutrition and Soil
Reichstein, M., and I. A. Janssens. 2009. Semi-empirical
modeling of the response of soil respiration to environmental
factors in laboratory and field conditions. Pages 207–220 in
W. L. Kutsch, B. Michael, and A. Heinemeyer, editors. Soil
carbon dynamics: an integrated methodology. Cambridge
University Press, Cambridge, UK.
Ruehr, N. K., C. A. Offermann, A. Gessler, J. B. Winkler, J. P.
Ferrio, N. Buchmann, and R. L. Barnard. 2009. Drought
effects on allocation of recent carbon: from beech leaves to
soil CO2efflux. New Phytologist 184:950–961.
Rustad, L. E., J. L. Campbell, G. M. Marion, R. J. Norby,
M. J. Mitchell, A. E. Hartley, J. H. C. Cornelissen, and J.
Gurevitch. 2001. A meta-analysis of the response of soil
respiration, net nitrogen mineralization, and aboveground
plant growth to experimental ecosystem warming. Oecologia
Ryan, M. G., M. B. Lavigne, and S. T. Gower. 1997. Annual
carbon cost of autotrophic respiration in boreal forest
ecosystems in relation to species and climate. Journal of
Geophysical Research-Atmospheres 102:28871–28883.
VIDYA SUSEELA AND JEFFREY S. DUKES412 Ecology, Vol. 94, No. 2
SAS Institute. 2008. SAS version 9.2. SAS Institute, Cary, Download full-text
North Carolina, USA.
Schimel, J. P., and J. S. Clein. 1996. Microbial response to
freeze-thaw cycles in tundra and taiga soils. Soil Biology and
Schindlbacher, A., S. Zechmeister-Boltenstern, and R. Jandl.
2009. Carbon losses due to soil warming: Do autotrophic and
heterotrophic soil respiration respond equally? Global
Change Biology 15:901–913.
Scott-Denton, L. E., T. N. Rosenstiel, and R. K. Monson.
2006. Differential controls by climate and substrate over the
heterotrophic and rhizospheric components of soil respira-
tion. Global Change Biology 12:205–216.
Shen, W. J., G. D. Jenerette, D. F. Hui, R. P. Phillips, and H.
Ren. 2008. Effects of changing precipitation regimes on
dryland soil respiration and C pool dynamics at rainfall
event, seasonal and interannual scales. Journal of Geophys-
ical Research-Biogeosciences 113: G03024.
Suseela, V., R. Conant, M. Wallenstein, and J. S. Dukes. 2012.
Effects of soil moisture on the temperature sensitivity of
heterotrophic respiration vary seasonally in an old-field
climate change experiment. Global Change Biology 18:336–
Tharayil, N., V. Suseela, D. J. Triebwasser, C. M. Preston,
P. D. Gerard, and J. S. Dukes. 2011. Changes in the
structural composition and reactivity of Acer rubrum leaf
litter tannins exposed to warming and altered precipitation:
climatic stress-induced tannins are more reactive. New
Trumbore, S. 2006. Carbon respired by terrestrial ecosystems—
recent progress and challenges. Global Change Biology
Wan, S. Q., and Y. Q. Luo. 2003. Substrate regulation of soil
respiration in a tallgrass prairie: results of a clipping and
shading experiment. Global Biogeochemical Cycles 17.
Wang, X. H., S. L. Piao, P. Ciais, I. A. Janssens, M. Reichstein,
S. S. Peng, and T. Wang. 2010. Are ecological gradients in
seasonal Q(10) of soil respiration explained by climate or by
vegetation seasonality? Soil Biology and Biochemistry
Yuste, J. C., I. A. Janssens, A. Carrara, and R. Ceulemans.
2004. Annual Q(10) of soil respiration reflects plant
phenological patterns as well as temperature sensitivity.
Global Change Biology 10:161–169.
Zhao, M. S., and S. W. Running. 2010. Drought-induced
reduction in global terrestrial net primary production from
2000 through 2009. Science 329:940–943.
Zhou, J. Z., K. Xue, J. P. Xie, Y. Deng, L. Y. Wu, X. H. Cheng,
S. F. Fei, S. P. Deng, Z. L. He, J. D. Van Nostrand, and
Y. Q. Luo. 2012. Microbial mediation of carbon-cycle
feedbacks to climate warming. Nature Climate Change
Zhou, X. H., R. A. Sherry, Y. An, L. L. Wallace, and Y. Q.
Luo. 2006. Main and interactive effects of warming, clipping,
and doubled precipitation on soil CO2efflux in a grassland
ecosystem. Global Biogeochemical Cycles 20: GB1003.
Zhou, X., S. Q. Wan, and Y. Q. Luo. 2007. Source components
and interannual variability of soil CO2 efflux under
experimental warming and clipping in a grassland ecosystem.
Global Change Biology 13:761–775.
Seven figures depicting seasonal variation in soil temperature and moisture and response to precipitation and temperature
treatments, as well as nine tables showing results from mixed-model REML analysis of response to warming and precipitation
treatments (Ecological Archives E094-034-A1).
Detailed results of the seasonal variation in apparent Q10and coefficients of Rsand Rr(Ecological Archives E094-034-A2).
A table of the correlation of soil respiration with different environmental variables predicted using temperature, moisture, and
combined exponential and quadratic (Mielnick-Dugas) functions (Ecological Archives E094-034-A3).
February 2013413SOIL RESPIRATION AND CLIMATE CHANGE