Arbuscular mycorrhizal mycelial respiration in a moist
tropical forest
Andrew T. Nottingham1, Benjamin L. Turner2, Klaus Winter2, Marcel G. A. van der Heijden3,4,5 and
Edmund V. J. Tanner1
1Department of Plant Sciences, University of Cambridge, Downing Street, Cambridge CB2 3EA, UK; 2Smithsonian Tropical Research Institute, 0843-
03092 Balboa, Ancon, Panama; 3Institute of Ecological Science, Vrije Universiteit, De Boelelaan 1085, 1081 HV Amsterdam, The Netherlands; 4Ecological
Farming Systems, Research Station ART, Swiss Federal Research Institute Agroscope, Zurich, Switzerland; 5Plant?Microbe Interactions, Institute of
Environmental Biology, Faculty of Science, Utrecht University, 3508 TB, Utrecht, the Netherlands
Author for correspondence:
Andrew T. Nottingham
Tel: +44 (0) 1223 333900
Email: atn24@cam.ac.uk
Received: 24 December 2009
Accepted: 24 January 2010
New Phytologist (2010) 186: 957?967
doi: 10.1111/j.1469-8137.2010.03226.x
Key words: carbon cycle, fine roots, mycelia,
mycorrhizas, Pseudobombax septenatum,
soil CO2 efflux, soil microbes, tropical forest.
Summary
? Arbuscular mycorrhizal fungi (AMF) are widespread in tropical forests and repre-
sent a major sink of photosynthate, yet their contribution to soil respiration in such
ecosystems remains unknown.
? Using in-growth mesocosms we measured AMF mycelial respiration in two sepa-
rate experiments: (1) an experiment in a semi-evergreen moist tropical forest, and
(2) an experiment with 6-m-tall Pseudobombax septenatum in 4.5-m3 containers,
for which we also determined the dependence of AMF mycelial respiration on the
supply of carbon from the plant using girdling and root-cutting treatments.
? In the forest, AMF mycelia respired carbon at a rate of 1.4 t ha)1 yr)1, which
accounted for 14 ? 6% of total soil respiration and 26 ? 12% of root-derived res-
piration. For P. septenatum, 40 ? 6% of root-derived respiration originated from
AMF mycelia and carbon was respired < 4 h after its supply from roots.
? We conclude that arbuscular mycorrhizal mycelial respiration can be substantial
in lowland tropical forests. As it is highly dependent on the recent supply of carbon
from roots, a function of aboveground fixation, AMF mycelial respiration is there-
fore an important pathway of carbon flux from tropical forest trees to the atmo-
sphere.
Introduction
Arbuscular mycorrhizal fungi (AMF) form symbiotic associ-
ations with almost all trees in hyper-diverse tropical forests
(Alexander & Lee, 2005), which are more productive than
any other terrestrial ecosystem (Melillo et al., 1993). These
fungi are thought to rely solely on the host plant for provi-
sion of carbon (C) and may utilize as much as 5?20% of net
photosynthate (Pearson & Jakobsen, 1993), while simulta-
neously benefiting the host through protection against root
pathogenic fungi and improved uptake of nutrients, particu-
larly phosphorus, from strongly weathered soils (Smith &
Read, 1997; Alexander & Lee, 2005). Although presently
overlooked, AMF could therefore be an important compo-
nent of the carbon cycle in tropical forests.
Tropical forests not only fix more C through photosyn-
thesis than any other terrestrial ecosystem (Melillo et al.,
1993), they also release more through respiration, over half
of which is released from soils (Chambers et al., 2004). Soils
in global tropical forests release c. 24 Pg C yr)1 into the
atmosphere in soil CO2 efflux (Raich et al., 2002). Soil CO2
efflux arises from the activity of a wide variety of below-
ground organisms, but only two sources have been assessed
in tropical forests: (1) ?autotrophic respiration?, derived from
roots, arbuscular mycorrhizal fungi (AMF) and the meta-
bolism of root exudates by microorganisms inhabiting the
rhizosphere, and (2) ?heterotrophic respiration?, derived
from microbial decomposition of plant detritus and soil
organic matter (Hanson et al., 2000). Respiration from
roots and that from AMF are almost always assessed together
as a single source of respiration (Hanson et al., 2000; Subke
et al., 2006), yet their assessment as distinct sources is
required given that they may respond somewhat indepen-
dently to environmental change (Alberton et al., 2005).
The significance of AMF as a sink of plant C and a source
of soil CO2 efflux has been revealed in studies of temperate
New
Phytologist Research
 The Authors (2010)
Journal compilation  New Phytologist Trust (2010)
New Phytologist (2010) 186: 957?967 957
www.newphytologist.com
plant communities. The dependence of AMF respiration on
current photosynthetic activity was indicated by a pulse-
labelling experiment in temperate grassland where 4?6% of
photo-assimilates were respired by the mycelia of AMF
within 21 h of fixation (Johnson et al., 2002). The quantity
of plant C respired by AMF was indicated by the 16%
increase in soil CO2 efflux found when Lolium perenne roots
were colonized by AMF (Grimoldi et al., 2006) and by
the finding that 8% of soil CO2 efflux was derived from
the mycelia of AMF in a barley (Hordeum vulgare) field
(Moyano et al., 2007). To date, no studies have quantified
the flux of CO2 released by AMF in a tropical forest. Addi-
tionally, very few studies have quantified seasonal patterns
in root and free-living microbial sources of soil CO2 efflux
in tropical forest. Knowledge of the relative importance
of these two sources of CO2 is vital to understand the
components of net ecosystem production and the forest C
balance (Hanson et al., 2000; Kuzyakov, 2006; Subke et al.,
2006).
Here we report the results from two separate experiments
designed to quantify AMF mycelial respiration in tropical
forest soils in the Republic of Panama. First, we measured
the contribution of AMF mycelia, roots and free-living
microorganisms to soil CO2 efflux in a semi-evergreen
moist tropical forest over 9 months. Secondly, under more
controlled conditions, we measured AMF mycelial and root
respiration for 6-m-tall individuals of the tropical tree
Pseudobombax septenatum grown in large (c. 4.5-m3) con-
tainers. For P. septenatum, we also determined the depen-
dence of AMF mycelial respiration on C supplied from
recent photosynthate and root nonstructural carbohydrates
by measuring the reduction in CO2 efflux after girdling
trees and then after cutting around each mesocosm to sever
root and AMF mycelial connections to the tree. As AMF
have been shown to consume substantial portions of photo-
synthate in temperate studies (e.g. 5?20% in Pearson &
Jakobsen, 1993; 8% in Grimoldi et al., 2006) and are
considered abundant in the soils of hyper-diverse tropical
forests (Alexander & Lee, 2005), we hypothesized that
AMF mycelia would be a significant source of CO2 from
tropical forest soils.
Materials and Methods
In-growth mesocosm design
Mesh-walled in-growth mesocosms were used to partition
sources of belowground respiration into three components
according to the following size classes of in-growth: roots,
AMF mycelia, and free-living soil microorganisms (Fig. 1)
(Johnson et al., 2001). Treatments were fine-root and myc-
elia in-growth (2-mm mesh; ?FR + AMF?), mycelia in-
growth (35-lm mesh; ?AMF?) and soil-only controls (either
rotated 35-lm mesh or 1-lm mesh; ?CTL?; see the follow-
ing paragraph for explanation of why two different designs
were used). Mesh sizes were chosen according to typical size
classes of soil microorganisms (< 1 lm), AMF hyphae
(2?20 lm) and fine roots (< 2 mm) (Friese & Allen, 1991;
Coleman & Crossley, 2003). Mesocosms were inserted to
20 cm depth in the soil, because this is the source of the
majority of microbial soil CO2 efflux (e.g. c. 80% for a for-
est in Costa Rica; Veldkamp et al., 2003) and contains the
majority of fine roots (in a nearby forest Cavelier (1992)
estimated that > 90% of fine roots to 100 cm depth were in
the top 25 cm), which dominate root respiration (Pregitzer
et al., 1998; Desrochers et al., 2002).
Control mesocosms had 35-lm mesh and were rotated
180 every week for the P. septenatum experiment, and had
1-lm mesh and were static for the forest experiment. We
planned to use the rotated design for both experiments
(Johnson et al., 2001) because this guarantees that hyphae
are effectively excluded and minimizes problems associated
with water-logging of soils as a consequence of poor drain-
age through mesh with a small diameter. However, preli-
10 cm
20 cm
FR+AMF
(2 mm)
AMF
(35 ?m)
CTL
(35 ?m rotated or 1 ?m)
Undisturbed soil
(soil collar)
2 cm
12 cm
Aboveground
Belowground
16 cm
Drainage
20 cm
Fig. 1 In-growth mesocosms were made from PVC and nylon mesh. Treatments were 2-mm mesh (in-growth of fine roots (FR) and arbuscu-
lar mycorrhizal fungal (AMF) mycelia; ?FR + AMF?) and 35-lm mesh (in-growth of AMF mycelia; ?AMF?), and soil-only controls (no in-growth;
?CTL?) were rotated 35-lm mesh for Pseudobombax septenatum trees and 1-lm mesh for the forest. Four holes (6-cm-diameter) were cut into
the base and eight holes (four 6-cm-diameter and four 3-cm-diameter) were cut into the side; these holes were fitted with nylon mesh using
hot silicon glue and high-strength duct tape. Undisturbed soil CO2 efflux was measured over soil collars in the forest (undisturbed). The figure
is not to scale.
958 Research
New
Phytologist
 The Authors (2010)
Journal compilation  New Phytologist Trust (2010)
New Phytologist (2010) 186: 957?967
www.newphytologist.com
minary tests indicated that the rotated design was suitable
for soils in P. septenatum pots, but not for the forest site.
For P. septenatum, mesocosms could be rotated freely and
in a preliminary experiment there was no significant differ-
ence in soil CO2 efflux between rotated and nonrotated 35-
lm mesh in-growth free mesocosms when measured 1 wk
following rotation (P > 0.05, n = 6); whereas in the clay-
loam forest soils, either mesocosms could not be rotated or,
when they could, rotation resulted in major disturbance to
surrounding soils. Therefore, in the forest we used 1-lm
mesh mesocosms to exclude hyphae (e.g. Moyano et al.,
2007), which drained to field capacity (there was no signifi-
cant difference in soil moisture in mesocosms compared
with undisturbed forest soils during the wet season; Fig. 2),
and excluded c. 72% of hyphal growth. In P. septenatum
containers, rotated 35-lm mesh mesocosms excluded c.
79% of hyphal growth (compare hyphal length in CTL and
AMF mesocosms for both experiments in Table 1; the
hyphae present in CTL mesocosms may have included dead
hyphae already in soils before the start of the experiment).
Experimental design: forest
The forest under study is mature (> 60 yr old) secondary
seasonal moist tropical forest in the Republic of Panama.
The site receives a mean annual rainfall of c. 2455 mm,
with a dry season from January to April (in 2008, 6.5% of
annual rainfall fell during these 4 months), and has an
average monthly temperature of c. 27C, based on measure-
ments from nearby Barro Colorado Island where the
monthly means varied by < 1C during the year (Windsor,
1990). For a detailed description of forest composition, the
reader is referred to site ?15? in Pyke et al. (2001), which is
located just a few kilometers from our study site. Although
the soils at our site have not been classified in detail (e.g. US
Soil Taxonomy), preliminary data indicate that they are
Alfisols (udalfs). The soils have a clay loam texture and are
derived from marine sediments; total organic C, total nutri-
ents and pH are listed in Table 2. A preliminary analysis of
spores from these soils showed a relatively high abundance
of spores from the genus Glomus, which is abundant in
nearby forest in Panama (Husband et al., 2002).
Ten plots were randomly located in a 10-ha area of forest.
Each plot measured 1 m2 and contained four areas for mea-
surement of soil CO2 efflux: three mesocosms (FR + AMF,
AMF and CTL) and one soil collar (undisturbed soil) ? a
total of 30 mesocosms and 40 sampling areas for soil CO2
efflux. In the forest plots we excavated 16-cm-diameter
holes at mesocosm locations and we collected soils at depths
0?5, 5?10, 10?15 and 15?20 cm. Each soil section was
kept separate and all visible root material was removed by
hand. Each 5-cm soil profile, still moist, was replaced inside
mesocosms at a bulk density similar to that of forest soils.
Soil-filled mesocosms were then reinserted into the forest at
the same locations where soils were removed and any gap
remaining around each mesocosm was refilled using the
same soil from the appropriate depth. Soils were removed
from the forest for a total of only 4 days and were kept at
field moisture, but were disturbed during root removal,
which therefore almost eliminated CO2 efflux attributable
to decomposition of observable dead roots. At the start of
the experiment in May 2007, each mesocosm received
11.1 g (dry mass) of litter standing crop, equivalent to aver-
age litter standing crop across all plots measured in May
2007. Two weeks before measurements began in September
2007, mesocosm litter standing crop was harvested for a
separate experiment and replaced with a further 7.3 g of lit-
ter standing crop, equivalent to average litter standing crop
across all plots measured in September 2007.
Fig. 2 Forest soil temperature (a) and soil moisture (b) for meso-
cosms and undisturbed soil in the tropical forest. Soil temperature
was measured at 5 cm depth and volumetric soil moisture was mea-
sured at 0?6 cm depth. Data were collected immediately after soil
CO2 efflux data from September 2007 to May 2008. Missing data
points resulted from equipment malfunction; soil moisture deter-
mined in June 2008 using gravimetric methods did not differ signifi-
cantly between treatments (not shown in figure). Data are
means ? 1 SE (n = 10 for undisturbed soil and for in-growth of fine
roots (FR) and arbuscular mycorrhizal fungal (AMF) mycelia
(?FR + AMF?); n = 6 for in-growth of AMF mycelia (?AMF?); n = 9
for no in-growth control (?CTL?)).
New
Phytologist Research 959
 The Authors (2010)
Journal compilation  New Phytologist Trust (2010)
New Phytologist (2010) 186: 957?967
www.newphytologist.com
Experimental design: Pseudobombax septenatum
Six plants of Pseudobombax septenatum (Bombacaceae) were
planted in large containers (1.8 m diameter and 1.8 m tall)
in 2004. In October 2007 the trees measured 6.0 ? 0.4 m
tall with diameter at 1.3 m of 14.7 ? 1.0 cm. Pseudo-
bombax septenatum is a fast-growing light-demanding tropi-
cal tree that occurs commonly in secondary lowland
tropical forest. It is fully deciduous, losing its leaves at the
start of the dry season in January and regrowing them with
the first rains in April ?May; all trees were in full leaf
throughout the experiment.
Containers were filled with soil collected from 0 to
20 cm depth in a nearby orchard in 2004. Mesocosm soil
for the P. septenatum experiment were different from the
container soil because we originally intended to use differ-
ences in d13C signatures of C3 plants and C4 soils (Ineson
et al., 1996) to partition sources of soil CO2 efflux in fur-
ther detail. However, the technique did not work because
the d13C signatures of the C3 and C4 sources were not suf-
ficiently different. Soil for the mesocosms was collected
from 0 to 20 cm depth from a recently reforested pasture
(see Wilsey et al., 2002 for details of this site) in December
2006. It was sieved (5 mm) and stored at field moisture for
a few days and then mixed with sand (70 : 30, soil:sand);
the sand had minimal organic C and CaCO3 and had a
similar pH to that of soil (sand pH 6 and soil pH 5.5).
Total organic nutrients in container and mesocosm soils
were similar (Table 2) and a preliminary analysis of spores
in both soils showed a relatively high abundance of spores
from the genus Glomus.
Mesocosms were inserted into the P. septenatum contain-
ers midway between the tree trunk and container edge in
May 2007 immediately following leaf flush at the onset of
the rainy season. Two replicates of each of the three
mesocosm treatments (2 ? CTL, 2 ? AMF and 2 ? FR +
AMF) were installed in each container, resulting in a total
of six mesocosms in each container. Soil CO2 efflux data
were averaged for each pair of mesocosms with a container,
with the resulting average value representing a single experi-
mental replication to give n = 6 (containers) per treatment
(or n = 3 when including the girdling effect; see methods
section ?soil CO2 efflux?). During intensive measurements
(c. 6 months following mesocosm installation; 3 d before
and 12 d after a girdling treatment), containers were cov-
ered with large tarpaulin sheets at night and during daytime
rain showers to prevent water-logging but were watered
with c. 50 l of water each day per container at the end of
measurements, water being applied equally to mesocosm
and surrounding soils using a hose. Tarpaulin sheets were
removed at least 2 h before measurements to allow soil
surface CO2 gradients to equilibrate.
Fine-root biomass and AMF hyphal length
Fine-root biomass (for all mesocosm soils) and AMF hyphal
length (for a representative subsample of mesocosm soils)
were determined for all mesocosms at the end of each exper-
Table 1 Bulk density and abundance of fine roots (FR) and arbuscular mycorrhizal fungi (AMF) for mesocosm soil and undisturbed forest soil
measured at the end of each experiment
Forest Pseudobombax septenatum
Bulk density
(g cm)3)
Fine-root biomass
(g kg)1)
AMF hyphal
length (m g)1)
Bulk density
(g cm)3)
Fine-root
biomass (g kg)1)
AMF hyphal
length (m g)1)
CTL 0.54 ? 0.07a 0.00 ? 0.00a 0.15 ? 0.02a 0.91 ? 0.02 0.00 ? 0.00a 0.28 ? 0.04a
AMF 0.57 ? 0.06a 0.00 ? 0.00a 0.54 ? 0.07b 0.92 ? 0.02 0.00 ? 0.00a 1.32 ? 0.21b
FR + AMF 0.63 ? 0.05a 3.93 ? 0.52b 0.46 ? 0.04b 0.90 ? 0.02 1.35 ? 0.04b 1.47 ? 0.30b
Undisturbed soil 0.76 ? 0.05b 3.80 ? 0.47b
Significant differences between treatments according to split-plot ANOVA are indicated by different lowercase letters (where P < 0.05). Data
are means and error bars are 1 SE (for P. septenatum, n = 6; for the forest: control (CTL), n = 9; AMF, n = 6; FR + AMF, n = 10; undisturbed
soil, n = 10; CTL and AMF mesocosms with significant root in-growth are not reported).
Table 2 Total carbon (C), nitrogen (N) and phosphorus (P), and pH in Pseudobombax septenatum and forest soils
C (g kg)1) N (g kg)1) P (mg kg)1) C:N C:P N:P pH
P. septenatum mesocosms 12.5 1.2 167 10 75 7 5.7
P. septenatum containers 12.7 1.1 256 11 50 4 5.7
Forest 40.1 3.5 278 11 144 13 6.3
All soils were sampled from 0 to 10 cm depth. Soils in P. septenatum mesocosms were a soil:sand mixture (70 : 30) and the values reported
are for the mixture; the mesocosm soil was different from the soil in the containers (see Materials and Methods section for details). Data are
single analyses of pooled subsamples.
960 Research
New
Phytologist
 The Authors (2010)
Journal compilation  New Phytologist Trust (2010)
New Phytologist (2010) 186: 957?967
www.newphytologist.com
iment. For P. septenatum, two out of 12 AMF mesocosms
contained fine roots and for the forest, four AMF and one
CTL mesocosm contained fine roots; all these mesocosms
were eliminated from subsequent analyses. Hyphae were
extracted using an aqueous extraction and membrane filter
technique and stained with trypan blue, and hyphal length
was determined using the grid line intersect method where
AMF hyphae were distinguished from non-AMF hyphae by
the presence of nonregular septa, dichotomous branching,
irregular wall thickness and ?or connection to chlamydosp-
ores (Rillig et al., 2002).
Soil CO2 efflux
Soil CO2 efflux was measured using a Li-8100 soil CO2 flux
system (Li-Cor, Lincoln, NE, USA). Undisturbed soil CO2
efflux in the forest was measured over the 20-cm-diameter,
12-cm-deep PVC soil collars, which were inserted to 2 cm
depth at least 1 month before measurements began and
remained fixed throughout the experiment (Fig. 1). To
enable measurements of mesocosm CO2 efflux, a soil collar
was made from PVC, 13 cm deep, that fitted tightly to the
rim of each mesocosm at the base (16 cm diameter) and to
the Li-8100 soil chamber at the top (20 cm diameter). The
Li-8100 was calibrated to account for the additional vol-
ume. Immediately following all soil CO2 efflux measure-
ments, soil temperature and volumetric soil moisture were
measured at 5 and 0?6 cm depths, respectively. Volumetric
soil moisture was measured using a thetaprobe (Delta-T
Devices, Cambridge, UK), which was calibrated to soil types
following the procedure described by the manufacturer.
For the forest site, CO2 efflux was measured between
07:00 and 13:00 h for all treatments and controls in all 10
plots every 2 wk from 17 September 2007 to 17 June 2008
and at least 12 h after any rainfall event. No measurements
were made for 3 months immediately following mesocosm
installation to allow any increased soil CO2 efflux following
soil aeration to subside and to allow time for in-growth of
fine roots and mycelia; Cavelier (1989) reported that it took
5 months (including the whole of the dry season) for 9-cm-
diameter ?ingrowth cores? to reach maximum root biomass
(< 2-mm-diameter roots) in a nearby forest.
For P. septenatum, mesocosm CO2 efflux was measured
between 07:00 and 12:00 h each day for 2 wk when the
trees were in full leaf (24 October to 12 November 2007);
c. 6 months following installation of mesocosms. On the
day of girdling (27 October), soil CO2 efflux data were col-
lected in the morning, three trees were girdled around mid-
day, and a further set of CO2 efflux data were collected in
the late afternoon. Trees were stratified according to the
average FR + AMF mesocosm CO2 efflux measured in the
morning and randomly allocated to girdled or nongirdled.
Two weeks following girdling, one of each mesocosm treat-
ment was removed from each container, thus cutting all
root and AMF mycelial connections. Mesocosm CO2 efflux
was measured immediately before removal and then
10 min, 120 min, 240 min and 24 h afterwards. We mea-
sured CO2 efflux in paired cut and uncut mesocosms before
and 24 h after removal of cut mesocosms.
Calculations and statistics
The percentage contributions of fine roots (root), AMF
mycelia (mycelial) and free-living microorganisms (micro-
bial) to the CO2 efflux (mg C m
)2 h)1) from root-ingrowth
(FR + AMFefflux) mesocosms were calculated as follows:
Roots?%? ? ?FR + AMFefflux  AMFefflux?=FR
? AMFefflux  100
Eqn 1
Mycelial?%? ? ?AMFefflux  CTLefflux?=FR
? AMFefflux  100
Eqn 2
Microbial?%? ? CTLefflux=FR ? AMFefflux  100 Eqn 3
For forest soils, FR + AMFefflux is CO2 efflux measured
for FR + AMF mesocosms, AMFefflux is CO2 efflux mea-
sured for AMF mesocosms and CTLefflux is CO2 efflux
measured for CTL mesocosms (see Heinemeyer et al., 2007
for a similar approach). Soil CO2 efflux measured for soil
collars (undisturbed soilefflux) was multiplied by the results
from Eqns 1?3 to estimate the absolute contributions of
fine roots, AMF mycelia and free-living microorganisms.
Thus, of total soil CO2 efflux (undisturbed soilefflux), ?roots?
is the portion derived from fine roots and rhizosphere-
dwelling microorganisms (Eqn 1), ?AMF? is the portion
derived from AMF mycelia and hyphosphere-dwelling
microorganisms (Eqn 2), and ?microbial? is the portion
derived from free-living soil microorganisms (Eqn 3). For
P. septenatum, Eqns 1 and 2 only were used to determine
the contribution of AMF mycelia to ?root-derived CO2
efflux?. Root-derived CO2 efflux (root + mycelial) is the res-
piration by roots and any organism using C recently derived
from roots (i.e. mycorrhizal fungi and rhizosphere microor-
ganisms; Kuzyakov, 2006). For each sampling date and for
each plot (for the forest) or tree (for P. septenatum), we cal-
culated respiration components and calculated averages and
standard errors according to the variation in calculated
components among all plots ? trees.
Variations in CO2 efflux, soil temperature and soil mois-
ture were compared between undisturbed soil (in forest),
FR + AMF, AMF and CTL treatments and partitioned
root, mycelial and microbial components using split-plot
repeated measures ANOVA (n = 10 per treatment for the
forest). For P. septenatum the analysis was initially com-
pleted with ?girdle effect? as an additional fixed main plot
factor (n = 3). However, girdled and nongirdled mesocosms
were finally pooled because there was no girdling effect
New
Phytologist Research 961
 The Authors (2010)
Journal compilation  New Phytologist Trust (2010)
New Phytologist (2010) 186: 957?967
www.newphytologist.com
for soil CO2 efflux (therefore n = 6 per treatment for
P. septenatum). Dependent variables were CO2 efflux, soil
temperature and soil moisture and were individually
analyzed with mesocosm treatment and time as the fixed
main plot factors, which were nested with either individual
tree (P. septenatum) or plot (forest). For P. septenatum, root-
cutting effects on FR + AMF and AMF mesocosm CO2
efflux were analyzed using the same method with time as
the fixed main plot factor, which was nested within individ-
ual tree. Analyses were performed with all data included
(September 2007 to June 2008) and then separately for late
wet season 2007 (September?December 2007), dry season
2008 (January?April 2008) and early wet season 2008
(May?June 2008). Treatment effects on fine-root biomass
and AMF hyphal length were analyzed using split-plot
ANOVA, with fine-root biomass or AMF hyphal length as
the fixed factor, which was nested within either individual
tree (P. septenatum) or plot (forest). We investigated the
relationships between CO2 efflux and soil temperature and
soil moisture for each mesocosm by respectively fitting
exponential and log-normal models using SigmaPlot
(version 10; Systat Inc., Chicago, IL, USA). Pair-wise com-
parisons were performed using Tukey post-hoc analyses and
significant interactions were determined at P ? 0.05. Before
analysis, data were tested for normality using Ryan?Joiner
tests and nonnormal data were log-transformed. All data are
presented as mean ? 1 SE. All statistical analyses were
performed using minitab (version 15; Minitab Inc., State
College, PA, USA).
Results
In-growth mesocosms and the soil abiotic
environment
Mesocosms generally excluded fine roots and AMF where
appropriate (Table 1). Soil temperature did not differ sig-
nificantly among mesocosm treatments and undisturbed
soils over the entire year in the forest (P > 0.05; or when
only dry or wet season data were included P > 0.05;
Fig. 2a) or in P. septenatum soils (P > 0.05; average soil
temperature over 2 wk was 29 ? 1C for soils in all meso-
cosms). Bulk density did not differ significantly among
mesocosm treatments in either experiment (P > 0.05 for
all comparisons), but was significantly lower in forest meso-
cosms compared with undisturbed soil (P = 0.05; Table 1).
Moisture in forest soils did not differ significantly among
treatments over the entire year (P > 0.05; Fig. 2b). How-
ever, during the dry season (January?April 2008), soil mois-
ture was significantly lower in FR + AMF mesocosms and
undisturbed soil when compared with AMF mesocosms
(P < 0.05 for both comparisons) and soil-only controls
(P < 0.01 for both comparisons). Dry season soil moisture
was not significantly different between FR + AMF meso-
cosms and undisturbed soil (P > 0.05) or between AMF
and CTL mesocosms (P > 0.05). Soil moisture in P. septen-
atum mesocosms decreased in the order CTL > AMF >
FR + AMF, with a significant difference between CTL
and FR + AMF (P < 0.05); average values of volumetric
soil moisture in mesocosms measured over 2 wk were
0.311 ? 0.018, 0.304 ? 0.017, and 0.266 ? 0.026 m3
H2O m
)3 soil, respectively.
The physical structure of soils was affected by mesocosms
during the dry season in the forest only. Small cracks
appeared between soils and the inner circumference of all
mesocosm treatments during the dry season, which disap-
peared soon after the first rains of the wet season. Cracking
severity was scored on a scale of 1?5 for all mesocosms on 1
March 2008 and there was no significant difference in
cracking severity among mesocosm treatments (P > 0.05).
Tropical forest soil CO2 efflux
For the tropical forest site, soil CO2 efflux varied signifi-
cantly over the year, from 115 ? 9 mg C m)2 h)1 in
March 2008 during the dry season to 248 ? 11 mg C m)2
h)1 in June 2008 in the early wet season (P < 0.001; see
undisturbed soil in Fig. 3). Soil and mesocosm CO2 efflux
decreased in the order FR + AMF > undisturbed soil >
AMF > CTL. The CO2 efflux from FR + AMF mesocosms
followed the same pattern as undisturbed soil CO2 efflux,
except during the dry season, when CO2 efflux was much
higher from FR + AMF mesocosms compared with undis-
turbed soils (Fig. 3a).
Mycelial CO2 efflux tended to increase during the year
but there was no significant seasonal variation (Fig. 3b;
P > 0.05); average values were 1 ? 8 mg C m)2 h)1 during
the late wet season of 2007, 19 ? 9 mg C m)2 h)1 during
the dry season of 2008 and 32 ? 15 mg C m)2 h)1 during
the early wet season of 2008. During the dry and early wet
seasons of 2008, mycelia contributed c. 14 ? 6% and 14 ?
6% of FR + AMF mesocosm CO2 efflux or 23 ? 10% and
29 ? 13% of ?root-derived? (root + mycelial) CO2 efflux,
respectively.
Root CO2 efflux did not vary significantly throughout
the year, with an average of 64 ? 8 mg C m)2 h)1, consti-
tuting c. 58 ? 7, 63 ? 9 and 77 ? 10% of FR + AMF mes-
ocosm CO2 efflux during the late wet season of 2007, the
dry season of 2008 and the early wet season of 2008, respec-
tively (P > 0.05; Fig. 3b).
Microbial CO2 efflux varied significantly throughout the
year (P < 0.001), from 105 ? 4 mg C m)2 h)1 during the
late wet season in 2007, to 51 ? 4 mg C m)2 h)1 during
the dry season of 2008, to 107 ? 4 mg C m)2 h)1 during
the early wet season of 2008. Microbial CO2 efflux consti-
tuted 64 ? 2, 38 ? 3 and 46 ? 2% of FR + AMF mesocosm
CO2 efflux during the late wet season of 2007, the dry season
of 2008 and the early wet season of 2008, respectively.
962 Research
New
Phytologist
 The Authors (2010)
Journal compilation  New Phytologist Trust (2010)
New Phytologist (2010) 186: 957?967
www.newphytologist.com
Mycelial, root, and microbial sources of CO2 efflux were
significantly different from each other over the year
(P < 0.001), but when data were grouped according to dry
and wet seasons there was no significant difference between
root and microbial CO2 effluxes during the dry season
(P > 0.05).
Soil CO2 efflux was not significantly affected by soil tem-
perature for any of the treatments, either overall or for the
dry or wet season alone. There was no significant effect of
soil moisture on soil CO2 efflux for any treatment, although
it is notable that the highest rates of soil CO2 efflux occurred
at the beginning of the wet season (see May ? June; Fig. 3).
Pseudobombax septenatum fine-root and AMF CO2
efflux
For P. septenatum, soil CO2 efflux decreased in the order
FR + AMF > AMF > CTL, with significant differences among
treatments (P < 0.001). Average soil CO2 efflux measured
above FR + AMF, AMF and CTL mesocosms was 611 ?
28, 389 ? 24 and 239 ? 11 mg C m)2 h)1, respectively
(Fig. 4a). Using Eqns 1 and 2 we calculated that root CO2
efflux was 222 ? 19 mg C m)2 h)1 and mycelial CO2
efflux was 147 ? 17 mg C m)2 h)1. Thus, root-derived
(root + mycelial) CO2 efflux was 40 ? 6% from mycelia
and 60 ? 6% from roots.
Girdling had no significant effect on CO2 efflux or soil
moisture; there were no significant differences in CO2 efflux
or soil moisture for FR + AMF, AMF, or CTL mesocosms
between girdled and nongirdled trees (P > 0.05 for all com-
parisons; data not shown). Root-cutting, on the other hand,
caused a large and very significant reduction (c. 75%) in
CO2 efflux in < 10 min for both fine roots and mycelia
(P < 0.001; Fig. 4b). We estimate that c. 28% and 44% of
this reduction, for FR + AMF and AMF mesocosms,
respectively, was attributable to the loss of CO2 diffusing
from the soil in the large containers (the quantity of CO2
diffusing from soil in the large containers was estimated by
the reduction in CO2 efflux following removal of CTL mes-
ocosms; 173 ? 9 mg C m)2 h)1). We calculate that sever-
ing roots and hyphae reduced respiration to negligible levels
after 4 h for roots and after 2 h for mycelia (the point at
which FR + AMF and AMF = CTL in Fig. 4b).
There were no significant relationships between soil CO2
efflux and either soil temperature or soil moisture for all
mesocosms and for both girdled and nongirdled trees in the
P. septenatum experiment (for all cases R2 < 0.01, P > 0.05;
data not shown).
Discussion
AMF mycelia as a source of CO2 efflux from tropical
soils
The extraradical mycelia of AMF contributed c. 14 ? 6% of
soil CO2 efflux in the tropical forest during dry and wet sea-
sons in 2008, which was equivalent to 26 ? 12% of annual
?autotrophic? or ?root-derived? CO2 efflux. The slight trend
of increased AMF mycelial respiration during the year was
attributed to the gradual colonization of mesocosms by
AMF hyphae. At the end of the experiment there was no
significant difference in hyphal length between AMF and
FR + AMF mesocosms or in fine-root biomass between
FR + AMF mesocosms and undisturbed soils (Table 1).
We therefore suggest that measurements made during the
later stages of the forest experiment accurately represent
AMF mycelial respiration in undisturbed forest soils.
(a)
(b)
Fig. 3 Tropical forest soil CO2 efflux partitioned into microbial,
arbuscular mycorrhizal fungi (AMF) and fine-root (FR) components.
(a) Soil CO2 efflux from undisturbed soils measured over 1 yr from
June 2007 to June 2008 and from control (CTL), AMF and
FR + AMF mesocosms measured over 9 months from September
2007 to June 2008. (b) Partitioned components of soil CO2 efflux
over 9 months from September 2007 to June 2008. The ?microbial?,
?mycelial? and ?root? components of undisturbed soil CO2 efflux
were calculated using values for mesocosm CO2 efflux in (a) and
Eqns 1?3. In (b) undisturbed soil = fine root + AMF + microbial.
Of the respiration components shown in (b), undisturbed soil and
microbial varied significantly over time according to repeated
measures ANOVA (where P < 0.05). Data are means and error bars
are 1 SE (n = 10 for undisturbed soil and FR + AMF ? root; n = 6 for
AMF ? mycelial; n = 9 for CTL ? microbial).
New
Phytologist Research 963
 The Authors (2010)
Journal compilation  New Phytologist Trust (2010)
New Phytologist (2010) 186: 957?967
www.newphytologist.com
There are very few published estimates of mycorrhizal
mycelial respiration in natural ecosystems with which to
compare our estimates in this tropical forest. Using mesh
in-growth cores in a temperate agricultural system, Moyano
et al. (2007) estimated an average annual AMF mycelial
respiration of 15 mg C m)2 h)1, while in a temperate
coniferous forest Heinemeyer et al. (2007) estimated an
average annual ectomycorrhizal (EM) mycelial respiration
of 26 mg C m)2 h)1. The similarity between our estimate
of 26 ? 12 mg C m)2 h)1 for lowland tropical forest (aver-
aged over the dry and early wet seasons in 2008) and that of
Heinemeyer et al. (2007) is surprising given the large differ-
ences in soils, vegetation and mycorrhizal type (e.g. greater
mycelial respiration may be expected for EM fungi because
they have a mycelial network an order of magnitude greater
than that of AMF; Smith & Read, 1997).
Our finding that AMF mycelia are a significant source of
soil CO2 efflux at the forest site was further supported by
measurements for P. septenatum under more controlled con-
ditions. In this case, there were greater absolute CO2 efflux
values (mycelia respired 147 ? 17 mg C m)2 h)1; Fig. 4)
and higher contributions of mycelia to root-derived CO2
efflux (40 ? 6% compared with 26 ? 12% for the forest
during the wet season).
The higher AMF mycelial respiration for P. septenatum
compared with the forest probably reflects the higher inci-
dent radiation for P. septenatum, which is a fast-growing
light-demanding species. Light intensity has a major influ-
ence on the respiration of AMF mycelia by altering rates of
photosynthetic activity and allocation of plant C to the
fungi (Heinemeyer et al., 2006). The pot-grown P. septena-
tum plants were well illuminated, which will have allowed
high rates of C fixation and high growth rates, provided that
there were sufficient nutrients; the high rates of AMF myce-
lial respiration for P. septenatum may thus reflect high allo-
cation of plant C to AMF in order to acquire those
nutrients. Differences in the allocation of plant C to AMF
according to plant nutrient demand could also result in root
colonization by different mycorrhizal species with different
rates of respiration (e.g. according to their parasitic or
mutualistic traits; Johnson et al., 1997), which can be
addressed in future studies that use molecular tools to iden-
tify mycorrhizal communities.
In both experiments our estimates of AMF mycelial res-
piration may be subject to several minor sources of error.
First, our experimental design did not account for respira-
tion by any soil macrofauna and saprophytic fungi that were
excluded from control mesocosms. However, for P. septena-
tum, the rapid decrease in AMF mycelial respiration to neg-
ligible levels after severing all in-growth indicates that the
vast majority of respiration was indeed from AMF mycelia
alone (Fig. 4b). Secondly, the lower bulk density of meso-
cosm soils could have caused preferential in-growth of roots
and hyphae, although the absence of a significant difference
in fine roots between forest mesocosms and undisturbed
forest soils (P > 0.05; Table 1) suggests that this was negli-
gible. Thirdly, higher soil moisture in AMF compared with
CTL mesocosms during the dry season in the forest
(Fig. 2b) may have resulted in higher AMF mycelial in-
growth and overestimation of mycelial respiration, but our
data show no difference in hyphal length between AMF and
CTL mesocosms at the end of the experiment (Table 1).
Fourthly, we may have overestimated fine-root and AMF
mycelial respiration, because roots and hyphae inside meso-
cosms were potentially younger and therefore had higher
metabolic rates than roots and hyphae in undisturbed soils.
However, fine roots and hyphae turn over relatively rapidly
(e.g. AMF hyphae can turn over every 5?6 d; Staddon
et al., 2003; and fine roots in a lowland tropical forest can
turn over 0.4?0.7 times per year; Silver et al., 2005), so it is
likely that this artefact would be greatly diminished towards
the end of the experiment.
Fifthly, while a significant portion of our ?root? CO2
efflux estimate may be from rhizosphere-dwelling microor-
ganisms, which contributed 50?60% of root-derived CO2
efflux in a study using Lolium perenne seedlings (Kuzyakov,
Fig. 4 Soil CO2 efflux measured above mes-
ocosms (control (CTL), arbuscular mycorrhi-
zal fungi (AMF) and fine roots (FR) + AMF)
installed within Pseudobombax septenatum
containers: (a) following installation (at day
0), and (b) following cutting of fine-root and
hyphal connections to trees and removal
from containers (at hour 0). In view of the
lack of effect of girdling (performed on day
183) on CO2 efflux, data are pooled for gir-
dled and nongirdled trees. Data are means
and error bars are 1 SE (n = 6).
964 Research
New
Phytologist
 The Authors (2010)
Journal compilation  New Phytologist Trust (2010)
New Phytologist (2010) 186: 957?967
www.newphytologist.com
2002), it is not clear whether a portion of ?mycelial respira-
tion? is from hyphosphere-dwelling microorganisms; no
study has reported significant respiration originating from
these microorganisms. The presence of mycorrhizal hyphae
can lead to either a small increase or a small decrease in the
abundance of other microorganisms (Johansson et al.,
2004), which suggests that hyphosphere-dwelling microor-
ganisms contribute a negligible net flux of CO2 from soils.
Sixthly, growth of hyphae within control mesocosms poten-
tially resulted in an underestimation of AMF mycelial CO2
efflux by as much as 21?28% (c. 72 and 79% of hyphae
were excluded from CTL mesocosms in the forest and P.
septenatum containers, respectively; Table 1). Lastly, we
quantified AMF respiration from the extraradical mycelia
only, yet AMF biomass inside roots is significant and may
comprise up to 20% of root weight (Smith & Read, 1997),
which suggests that the entire contribution to root-derived
CO2 efflux by AMF is greater than our estimated contribu-
tion by mycelia.
The dependence of AMF mycelial CO2 efflux on plant
C supply
Remarkably, girdling of P. septenatum had no significant
effect on respiration of fine roots or mycelia, which is sur-
prising given that the majority of FR + AMF mesocosm
CO2 efflux was derived from fine roots and mycelia, which
utilize plant-derived C. Pseudobombax septenatum has a well-
defined separation of xylem and phloem tissue and we are
confident that phloem tissue was completely severed by gir-
dling. By contrast, girdling of Pinus sylvestris caused a 37%
reduction in root CO2 efflux over the first 5 d (Ho?gberg
et al., 2001) and isotopic labeling studies have measured
time lags of just 1?4 d between photosynthetic C fixation
and release of that same C within soil CO2 efflux (Ekblad &
Ho?gberg, 2001). We hypothesize that, in the short term,
respiration of roots and mycelia was maintained by root
carbohydrate reserves; this interpretation is supported by the
rapid drop to virtually zero of root and AMF mycelial CO2
efflux following root and mycelia cutting when microcosms
were removed from the surrounding soil, and the later exca-
vation of a large tap root (c. 0.04 m3). The same mechanism
was suggested following no response of girdling on soil CO2
efflux for a plantation of Eucalyptus grandis ? E. urophylla
clones in Brazil (Binkley et al., 2006).
The very rapid decrease in respiration by mycelia severed
from roots is consistent with the findings of Johnson et al.
(2002), who detected a 13C label in mycelial respiration 9?
14 h after plant uptake, and with those of So?derstro?m &
Read (1987), who measured significant reduction of EM
mycelial respiration within hours of severing hyphae in a
laboratory study. The negligible rates of CO2 efflux once
hyphae were cut also showed that microbial decomposition
of dead hyphae, which have been shown to turn over rap-
idly (Staddon et al., 2003), did not contribute a significant
source of CO2. Our findings are consistent with growth and
maintenance respiratory losses being the major source of
AMF mycelial CO2 efflux, in contrast to slow decomposi-
tion of hyphal residues, probably as a result of the abun-
dance of recalcitrant compounds such as chitin (Rillig,
2004) that make a negligible contribution to CO2 efflux.
Seasonal dynamics in components of tropical forest
soil CO2 efflux
In addition to demonstrating the considerable contribution
of AMF mycelia to soil CO2 efflux, our data also provide
new insight into how fine roots and microorganisms affect
seasonal variation in tropical forest soil CO2 efflux. Fine-
root CO2 efflux at the forest site did not vary significantly
throughout the year and maintained an average rate of
64 ? 8 mg C m)2 h)1 (note that this estimate includes an
unknown portion of respiration by rhizosphere-dwelling
microorganisms). This suggests that fine-root CO2 efflux
was not limited by water during the dry season despite sig-
nificant drying of soils (Fig. 2b) and significant reductions
in soil water potential (see ?pipeline? plots in Santiago et al.,
2004; which are just a few km from our site), a hypothesis
supported by a 7-yr throughfall exclusion experiment in an
Amazonian forest (Brando et al., 2008). The near-surface
roots probably have a sufficient supply of water during the
dry season from their connections with deeper roots of the
same trees. We showed that the major seasonal shift in soil
CO2 efflux at the forest site (from 115 ? 9 mg C m
)2 h)1
during the dry season to 248 ? 11 mg C m)2 h)1 at the
start of the wet season) was attributable to significant varia-
tion in microbial CO2 efflux (free-living microbial rather
than rhizomicrobial; Fig. 3b). The major seasonal con-
straints on microbial activity are litterfall and precipitation
(Vasconcelos et al., 2004; Valentini et al., 2008), so we pro-
pose that the seasonality of soil CO2 efflux in our semi-ever-
green moist tropical forest was a consequence of the
decomposition of litter leachate following rainfall, seasonal
priming of soil C stimulated by input of labile litter C
(Sayer et al., 2007), and rewetting effects (Jarvis et al.,
2007).
Our estimates of seasonal variation in root and microbial
activity depend largely on the assumption that any effects of
mesocosms on the soil abiotic environment were consistent
across all treatments at all times (CTL, AMF and
FR + AMF), which may not be entirely the case in the dry
season. This ?dry season effect? of higher CO2 efflux from
FR + AMF mesocosms compared with undisturbed soils
(Fig. 3a) was probably caused by cracking of the soil in the
mesocosms, which was consistent across all mesocosm treat-
ments. The only effect that was not consistent across all
mesocosms was higher soil moisture in the control and
AMF mesocosms during the dry season (Fig. 2b), which
New
Phytologist Research 965
 The Authors (2010)
Journal compilation  New Phytologist Trust (2010)
New Phytologist (2010) 186: 957?967
www.newphytologist.com
may have resulted in slight underestimation of root respira-
tion and overestimation of microbial and AMF mycelial res-
piration.
It was also assumed that CO2 production from roots and
mycelia beneath control mesocosms was negligible, an
assumption supported by studies showing that 90% of fine
roots were in the top 25 cm of soil in nearby forest (Cave-
lier, 1992); estimates of microbial CO2 efflux may have
been overestimated if this assumption was not met. We
conclude that our estimates of root, AMF mycelial and
microbial components of CO2 efflux (Fig. 3b) may be sub-
ject to minor errors during the dry season because of differ-
ences in mesocosm soil moisture (Fig. 2b) but are accurate
during the wet season, when most annual CO2 efflux
occurred and when mesocosm effects were small and likely
to have been consistent across all treatments.
Mycorrhizal mycelial respiration in moist tropical for-
ests and environmental change
The large release of CO2 from AMF in tropical forest soils
to the atmosphere has implications for ecosystem carbon
storage under future climate scenarios. Mycorrhizal fungi
have a physiology and chemistry very different to those of
plant roots (Smith & Read, 1997) and therefore may
respond independently to changes in climate and plant pro-
duction. A meta-analysis of laboratory and field CO2
enrichment studies using temperate plants has shown that
plants consistently allocate more C to mycorrhizal fungi
under elevated CO2 and nutrient-limiting conditions
(Alberton et al., 2005). Therefore, in a world with elevated
concentrations of atmospheric CO2, there may be signifi-
cant increases in the allocation of C to AMF in tropical for-
ests, which generally grow on nutrient-poor soils (Vitousek
& Sanford, 1986), and which already appear to be increas-
ing in productivity, probably in response to raised concen-
trations of atmospheric CO2 (Phillips et al., 2008). The
fate of plant C allocated to mycorrhizal fungi could have
a substantial impact on the future C balance of tropical
forests, whether stabilized in soils or, as we have shown,
released as a major component of the CO2 efflux from soils.
Conclusions
We have shown that the extraradical mycelia of AMF are a
major source of CO2 efflux from soils in a semi-evergreen
moist tropical forest, releasing 1.4 ? 0.6 t C ha)1 yr)1.
Our study provides the first evidence in a tropical system
that AMF mycelia are an important pathway of C flux from
tropical forest trees to the atmosphere, by rapidly returning
plant-derived C to the atmosphere (Fig. 4b) through their
high rates of respiration (Figs 3, 4). Mycorrhizal fungi are
therefore an important component of the C cycle in tropi-
cal forests and require more attention. Further information
is now required on mycorrhizal respiration in different
tropical forests where nutrient limitation of plant growth
and community composition of mycorrhiza may vary (e.g.
Husband et al., 2002 cf. Lovelock et al., 2003), and on
how increased concentrations of atmospheric CO2 will
affect the allocation of C from plants to mycorrhizal fungi
in tropical forests.
Acknowledgements
We thank Jorge Aranda, Lucas Cernusak, Ludo Luckerhoff,
Scott Mangan, Catherine Potvin, Tania Romero, Emma
Sayer, Michael Tobin and Didimo Urena for their support.
We thank David Wardle and three anonymous reviewers
for comments on the manuscript. The project was funded
by a NERC grant (NER ?S ?A ?2004 ?12241A) and a Smith-
sonian Tropical Research Institute Short-Term Fellowship
to ATN.
References
Alberton O, Kuyper TW, Gorissen A. 2005. Taking mycocentrism
seriously: mycorrhizal fungal and plant responses to elevated CO2. New
Phytologist 167: 859?868.
Alexander IJ, Lee SS. 2005. Mycorrhizas and ecosystem processes in
tropical rain forest: implications for diversity. In: Burslem DFRP,
Pinard MA, Hartley SE, eds. Biotic interactions in the tropics: their role in
the maintenance of species diversity. Cambridge, UK: Cambridge
University Press, 165?203.
Binkley D, Ernesto JLS, Takahashi N, Ryan MG. 2006. Tree-girdling to
separate root and heterotrophic respiration in two eucalyptus stands in
Brazil. Oecologia 148: 447?454.
Brando PM, Nepstad DC, Davidson EA, Trumbore SE, Ray D, Camargo
P. 2008. Drought effects on litterfall, wood production and
belowground carbon cycling in an amazon forest: results of a throughfall
reduction experiment. Philosophical Transactions of the Royal Society of
London. Series B, Biological Sciences 363: 1839?1848.
Cavelier J 1989. Root biomass, production and the effect of fertilization in
two tropical forests. PhD thesis, University of Cambridge, UK.
Cavelier J. 1992. Fine-root biomass and soil properties in a semideciduous
and a lower montane rain-forest in Panama. Plant and Soil 142: 187?
201.
Chambers JQ, Tribuzy ES, Toledo LC, Crispim BF, Higuchi N, dos
Santos J, Araujo AC, Kruijt B, Nobre AD, Trumbore SE. 2004.
Respiration from a tropical forest ecosystem: partitioning of sources and
low carbon use efficiency. Ecological Applications 14: S72?S88.
Coleman DC, Crossley DA. 2003. Fundamentals of soil ecology. London,
UK: Academic Press.
Desrochers A, Landhausser SM, Lieffers VJ. 2002. Coarse and fine root
respiration in aspen (Populus tremuloides). Tree Physiology 22: 725?732.
Ekblad A, Ho?gberg P. 2001. Natural abundance of C-13 in CO2 respired
from forest soils reveals speed of link between tree photosynthesis and
root respiration. Oecologia 127: 305?308.
Friese CF, Allen MF. 1991. The spread of VA mycorrhizal fungal hyphae
in the soil: inoculum types and external hyphal architecture. Mycologia
83: 409?418.
Grimoldi AA, Kavanova M, Lattanzi FA, Schaufele R, Schnyder H. 2006.
Arbuscular mycorrhizal colonization on carbon economy in perennial
ryegrass: quantification by 13CO2 ? 12CO2 steady-state labelling and gas
exchange. New Phytologist 172: 544?553.
966 Research
New
Phytologist
 The Authors (2010)
Journal compilation  New Phytologist Trust (2010)
New Phytologist (2010) 186: 957?967
www.newphytologist.com
Hanson PJ, Edwards NT, Garten CT, Andrews JA. 2000. Separating root
and soil microbial contributions to soil respiration: a review of methods
and observations. Biogeochemistry 48: 115?146.
Heinemeyer A, Hartley IP, Evans SP, De la Fuente JAC, Ineson P. 2007.
Forest soil CO2 flux: uncovering the contribution and environmental
responses of ectomycorrhizas. Global Change Biology 13: 1786?1797.
Heinemeyer A, Ineson P, Ostle N, Fitter AH. 2006. Respiration of the
external mycelium in the arbuscular mycorrhizal symbiosis shows strong
dependence on recent photosynthates and acclimation to temperature.
New Phytologist 171: 159?170.
Ho?gberg P, Nordgren A, Buchmann N, Taylor AFS, Ekblad A, Hogberg
MN, Nyberg G, Ottosson-Lofvenius M, Read DJ. 2001. Large-scale
forest girdling shows that current photosynthesis drives soil respiration.
Nature 411: 789?792.
Husband R, Herre EA, Turner SL, Gallery R, Young JPW. 2002.
Molecular diversity of arbuscular mycorrhizal fungi and patterns of host
association over time and space in a tropical forest. Molecular Ecology 11:
2669?2678.
Ineson P, Cotrufo MF, Bol R, Harkness DD, Blum H. 1996.
Quantification of soil carbon inputs under elevated CO2: C3 plants in a
C4 soil. Plant and Soil 187: 345?350.
Jarvis P, Rey A, Petsikos C, Wingate L, Rayment M, Pereira J, Banza J,
David J, Miglietta F, Borghetti M et al. 2007. Drying and wetting of
mediterranean soils stimulates decomposition and carbon dioxide
emission: the ??Birch effect??. Tree Physiology 27: 929?940.
Johansson JF, Paul LR, Finlay RD. 2004. Microbial interactions in the
mycorrhizosphere and their significance for sustainable agriculture. Fems
Microbiology Ecology 48: 1?13.
Johnson NC, Graham JH, Smith FA. 1997. Functioning of mycorrhizal
associations along the mutualism-parasitism continuum. New Phytologist
135: 575?586.
Johnson D, Leake JR, Ostle N, Ineson P, Read DJ. 2002. In situ (CO2)-
C-13 pulse-labelling of upland grassland demonstrates a rapid pathway
of carbon flux from arbuscular mycorrhizal mycelia to the soil. New
Phytologist 153: 327?334.
Johnson D, Leake JR, Read DJ. 2001. Novel in-growth core system
enables functional studies of grassland mycorrhizal mycelial networks.
New Phytologist 152: 555?562.
Kuzyakov Y. 2002. Separating microbial respiration of exudates from root
respiration in non-sterile soils: a comparison of four methods. Soil
Biology & Biochemistry 34: 1621?1631.
Kuzyakov Y. 2006. Sources of CO2 efflux from soil and review of
partitioning methods. Soil Biology & Biochemistry 38: 425?448.
Lovelock CE, Andersen K, Morton JB. 2003. Arbuscular mycorrhizal
communities in tropical forests are affected by host tree species and
environment. Oecologia 135: 268?279.
Melillo JM, McGuire AD, Kicklighter DW, Moore B, Vorosmarty CJ,
Schloss AL. 1993. Global climate-change and terrestrial net primary
production. Nature 363: 234?240.
Moyano FE, Kutsch WL, Schulze ED. 2007. Response of mycorrhizal,
rhizosphere and soil basal respiration to temperature and photosynthesis
in a barley field. Soil Biology & Biochemistry 39: 843?853.
Pearson JN, Jakobsen I. 1993. Symbiotic exchange of carbon and
phosphorus between cucumber and 3 arbuscular mycorrhizal fungi. New
Phytologist 124: 481?488.
Phillips OL, Lewis SL, Baker TR, Chao KJ, Higuchi N. 2008. The
changing Amazon forest. Philosophical Transactions of the Royal Society of
London. Series B, Biological Sciences 363: 1819?1827.
Pregitzer KS, Laskowski MJ, Burton AJ, Lessard VC, Zak DR. 1998.
Variation in sugar maple root respiration with root diameter and soil
depth. Tree Physiology 18: 665?670.
Pyke CR, Condit R, Aguilar S, Lao S. 2001. Floristic composition across a
climatic gradient in a neotropical lowland forest. Journal of Vegetation
Science 12: 553?566.
Raich JW, Potter CS, Bhagawati D. 2002. Interannual variability in
global soil respiration, 1980?94. Global Change Biology 8: 800?812.
Rillig MC. 2004. Arbuscular mycorrhizae and terrestrial ecosystem
processes. Ecology Letters 7: 740?754.
Rillig MC, Wright SF, Shaw MR, Field CB. 2002. Artificial climate
warming positively affects arbuscular mycorrhizae but decreases
soil aggregate water stability in an annual grassland. Oikos 97:
52?58.
Santiago LS, Kitajima K, Wright SJ, Mulkey SS. 2004. Coordinated
changes in photosynthesis, water relations and leaf nutritional traits of
canopy trees along a precipitation gradient in lowland tropical forest.
Oecologia 139: 495?502.
Sayer EJ, Powers JS, Tanner EVJ. 2007. Increased litterfall in tropical
forests boosts the transfer of soil CO2 to the atmosphere. PLoS ONE 2:
e1299.
Silver WL, Thompson AW, McGroddy ME, Varner RK, Dias JD,
Silva H, Crill PM, Keller M. 2005. Fine root dynamics and trace gas
fluxes in two lowland tropical forest soils. Global Change Biology 11:
290?306.
Smith SE, Read DJ. 1997. Mycorrhizal symbiosis. London, UK: Academic
Press.
So?derstro?m B, Read DJ. 1987. Respiratory activity of intact and excised
ectomycorrhizal mycelial systems growing in unsterilized soil. Soil
Biology & Biochemistry 19: 231?236.
Staddon PL, Ramsey CB, Ostle N, Ineson P, Fitter AH. 2003. Rapid
turnover of hyphae of mycorrhizal fungi determined by AMS
microanalysis of C-14. Science 300: 1138?1140.
Subke JA, Inglima I, Cotrufo MF. 2006. Trends and methodological
impacts in soil CO2 efflux partitioning: a metaanalytical review. Global
Change Biology 12: 921?943.
Valentini CMA, Sanches L, de Paula SR, Vourlitis GL, Nogueira
JD, Pinto OB, Lobo FD. 2008. Soil respiration and aboveground
litter dynamics of a tropical transitional forest in northwest Mato
Grosso, Brazil. Journal of Geophysical Research-Biogeosciences 113:
G00B10.
Vasconcelos SS, Zarin DJ, Capanu M, Littell R, Davidson EA, Ishida FY,
Santos EB, Araujo MM, Aragao DV, Rangel-Vasconcelos LGT et al.
2004. Moisture and substrate availability constrain soil trace gas fluxes
in an eastern Amazonian regrowth forest. Global Biogeochemical Cycles
18: GB2009.1?GB2009.10.
Veldkamp E, Becker A, Schwendenmann L, Clark DA, Schulte-Bisping
H. 2003. Substantial labile carbon stocks and microbial activity in
deeply weathered soils below a tropical wet forest. Global Change Biology
9: 1171?1184.
Vitousek PM, Sanford RL. 1986. Nutrient cycling in moist tropical forest.
Annual Review of Ecology and Systematics 17: 137?167.
Wilsey BJ, Parent G, Roulet NT, Moore TR, Potvin C. 2002. Tropical
pasture carbon cycling: relationships between C source ? sink strength,
above-ground biomass and grazing. Ecology Letters 5: 367?376.
Windsor D. 1990. Climate and moisture availability in a tropical forest,
long term record for Barro Colorado island, Panama. Smithsonian
Contributions to the Earth Sciences 29: 1?145.
New
Phytologist Research 967
 The Authors (2010)
Journal compilation  New Phytologist Trust (2010)
New Phytologist (2010) 186: 957?967
www.newphytologist.com