Global Change Biology (2011) 17, 1637–1645, doi: 10.1111/j.1365-2486.2010.02327.

Fungal community responses to precipitation

C H R I S T I N E V. H A W K E S *, S T E P H A N I E N . K I V L I N w , J E N N I F E R D . R O C C A *, VA L E R I E
H U G U E T z, M E R E D I T H A . T H O M S E N § and K E N W Y N B L A K E S U T T L E }
*Section of Integrative Biology, University of Texas at Austin, Austin, TX 78712, USA, wDepartment of Ecology and Evolutionary
Biology, University of California, Irvine, CA 92697, USA, zDepartment of Ecology and Evolutionary Biology, Rice University,
Houston, TX 77005, USA, §Department of Biology, University of Wisconsin-La Crosse, La Crosse, WI 54601, USA, }Division of
Biology, Imperial College London, London SW7 2AZ, UK

Abstract
Understanding how fungal communities are affected by precipitation is an essential aspect of predicting soil
functional responses to future climate change and the consequences of those responses for the soil carbon cycle.
We tracked fungal abundance, fungal community composition, and soil carbon across 4 years in long-term field
manipulations of rainfall in northern California. Fungi responded directly to rainfall levels, with more abundant,
diverse, and consistent communities predominating under drought conditions, and less abundant, less diverse, and
more variable communities emerging during wetter periods and in rain-addition treatments. Soil carbon storage itself
did not vary with rainfall amendments, but increased decomposition rates foreshadow longer-term losses of soil
carbon under conditions of extended seasonal rainfall. The repeated recovery of fungal diversity and abundance
during periodic drought events suggests that species with a wide range of environmental tolerances coexist in this
community, consistent with a storage effect in soil fungi. Increased diversity during dry periods further suggests that
drought stress moderates competition among fungal taxa. Based on the responses observed here, we suggest that
there may be a relationship between the timescale at which soil microbial communities experience natural
environmental fluctuations and their ability to respond to future environmental change.
Keywords: climate change, fungi, grassland, northern California, rainfall, seasonality

Received 31 May 2010; revised version received 26 July 2010 and accepted 25 August 2010

isms in aerobic soils (Thorn, 1997); the biomass of

Introduction
saprophytic and mycorrhizal fungi ranges from 50 to
Water availability is a key determinant of carbon fluxes in 900 kg C ha1 (Zhu & Miller, 2003). Saprophytic fungi
terrestrial ecosystems that is rapidly being altered by decompose plant litter, incorporate carbon and nutri-
climate change. Variation in mean annual precipitation, ents into fungal biomass, and move carbon between
both interannual and spatial, acts as a primary influence surface and deeper soils (Frey et al., 2003). Mycorrhizal
on ecosystem processes such as respiration (Raich & fungi are a direct sink for plant photosynthate (Allen,
Potter, 1995; Curiel-Yuste et al., 2007; Vargas et al., 2010), 1991; Johnson et al., 2002) and in some cases contribute
decomposition (Aerts, 1997; Epstein et al., 2002; Adair more to soil respiration than roots (Heinemeyer et al.,
et al., 2008), and plant productivity (Lauenroth & Sala, 2007). Turnover of mycorrhizal hyphae can also be the
1992; Sala et al., 1988). Future precipitation projections primary pathway for carbon inputs to the soil organic
and predicted ecosystem responses remain highly vari- matter pool (Fitter et al., 2000; Godbold et al., 2006).
able (Solomon et al., 2007). In particular, belowground Other fungal tissues, such as chitin and glomalin, are
responses to climate change may have large effects on more recalcitrant and represent the potential capacity
ecosystem processes and represent a large source of for long-term storage of fungal carbon in soil (Treseder
uncertainty for ecosystem feedbacks to climate change & Allen, 2000; Rillig et al., 2001; Steinberg & Rillig,
(Solomon et al., 2007). We know little about how the 2003). Moreover, soil hyphal abundance is linearly
belowground components of ecosystems respond to cur- related to increases in soil aggregation, soil organic
rent precipitation patterns, limiting our ability to predict carbon, and soil organic nitrogen (Wilson et al., 2009).
their responses to departures from historical trends. Fungal communities in soil can respond directly to
Soil fungi play a critical role in belowground carbon changes in soil moisture. When water is limiting, con-
cycling. Fungi are typically the dominant microorgan- straints on substrate diffusion (Skopp et al., 1990) may
force soil hyphal networks to expand. Expansion may
Correspondence: Christine V. Hawkes, tel. 1 1 512 471 6049, also occur for mycorrhizal networks during drought –
fax 1 1 512 471 3878, e-mail: chawkes@mail.utexas.edu mycorrhizas are known to improve plant drought tol-

r 2010 Blackwell Publishing Ltd 1637

1638 C . V. H A W K E S et al.

erance partly through increased rates of water move- active in response to water additions that extended the
ment from soil into host plants (Augé, 2001). At very rainy season into late spring. Winter rain additions were
low soil water, fungal hyphae in soil may stop growing expected to have less of an effect unless they created
or contract if the hyphae cannot maintain structural anaerobic conditions. Because the treatments were super-
integrity. Current evidence for fungal responses to imposed on ambient conditions, we also expected rela-
moisture is equivocal. Total fungal biomass in soil tively greater responses to water manipulations during
increased linearly with soil moisture across agricultural drought years compared with wet years.
sites (Frey et al., 1999). Yet within individual sites,
mycorrhizal hyphae in soils and roots have been shown
to increase, decrease, or remain unchanged in response Materials and methods
to drought (Miller et al., 1995, Lutgen et al., 2003,
Staddon et al., 2003, Clark et al., 2009, Querejeta et al., Site and experimental design
2009). Compositional turnover in the fungal community In April 2005, we began collecting soils from 70 m2 circular
may also occur if different fungi are better at growth plots (n 5 18) in an open meadow site at the University of
under limited or optimal water conditions. To the best California Angelo Coast Range Reserve in Branscomb, CA
of our knowledge, however, changes in fungal commu- (Mendocino County). Mean annual precipitation is 216 cm and
nity composition with altered precipitation have not mean temperatures range from a winter minimum of 16 1C to a
been examined. summer maximum of 31 1C. Since 2001, the plots have been
Alternatively, altered rainfall may indirectly affect fun- subjected to one of three rainfall treatments (n 5 6 plots per
treatment): control (no rain addition), winter rain addition, or
gi via changes in the plant community. Plant community
spring rain addition (Suttle et al., 2007). Rain additions were
composition, diversity, and productivity can change with
applied either from January to March (winter addition) or
precipitation patterns (Knapp et al., 2002; Suttle et al., April to June (spring addition), delivering 14–16 mm of water
2007) and this could impact both mycorrhizal and sapro- from automated sprinklers for one hour shortly after dawn
phytic fungi. Different plant species can host distinct every third day and resulting in a seasonal supplement of
mycorrhizal communities (e.g., Vandenkoornhuyse 44 cm of water (  20% increase). Water for the rain-addition
et al., 2003) and these can be affected by local community treatments was from an adjacent mountain spring, filtered
properties such as neighborhood composition, diversity, through a 40 mm mesh. Control plots received no additional
establishment order, and plant age (Husband et al., 2002; water; all plots were open to ambient rainfall. Vegetation at the
Hausmann & Hawkes, 2009, Hausmann & Hawkes, 2010, site is a mixture of annual and perennial grasses and forbs,
Kivlin & Hawkes, in press). Saprophytic fungi can also be with annual species dominant and no woody plants present.
During the study, plant biomass aboveground was typically
affected by changes in the plant community that affect
higher in plots subjected to spring water additions than in
resource availability, such as substrate inputs and chemi-
other treatments, but without consistent differences in species
cal characteristics (Wardle & Nicholson, 1996; Waldrop composition among treatments or years (Suttle et al., 2007,
et al., 2006). Thus, soil fungi may track rainfall, the plant K. B. Suttle and M. Thomsen, unpublished results).
community, or both. We collected soils from each plot in late winter (April) and
Understanding responses of soil fungi to climate late spring (June) for 2005–2006, and early winter (January)
change and the mechanisms underlying those responses and late spring (June) for 2007–2008. At each of these eight
will allow us to better resolve both the ecosystem impacts sampling time points, two soil cores (2 cm wide  15 cm deep)
of and feedbacks to climate change. We examined how were taken from each plot and homogenized before subsam-
soil fungi responded to seasonal changes in precipitation pling for immediate measurement of soil hyphal abundance,
in a grassland ecosystem in northern California. This ergosterol, and glomalin, and storage for later measurement of
fungal community composition (80 1C freezer) and total soil
region is characterized by a strongly Mediterranean
organic carbon (air dried). To quantify fungal root colonization
climate, where highly seasonal rainfall patterns could
we collected two to three whole individual plants per plot of
experience boundary shifts with climate change (e.g., the dominant species in June of each year. We used only roots
Chou et al., 2008) that directly affect carbon fluxes (Xu that were attached to the plant to be certain of their identity.
& Baldocchi, 2004). Specifically, we analyzed soils col- Roots were separated from the plant, washed, and stored in
lected across seasons and years from a large-scale field 95% isopropyl alcohol. Colonization was averaged across
manipulation of rainfall in northern California grassland, species for plot-level measurements.
where the winter rainy season was either intensified
above ambient levels, extended seasonally beyond ambi-
ent levels, or left unamended. We measured responses of Fungal abundance
fungal community abundance and composition, as well For quantifying intraradical fungal colonization, roots were
as soil carbon pools and decomposition. We predicted stained with acid fuchsin. Root length colonization by hyphae,
that fungal hyphae in soil would expand and be more arbuscules, and vesicles was quantified by microscopy using

r 2010 Blackwell Publishing Ltd, Global Change Biology, 17, 1637–1645

 F U N G A L C L I M AT E R E S P O N S E 1639

the magnified intersections method for 100 fields of view per Cloned sequences were initially identified to the phyla level
sample at  200 (McGonigle et al., 1990). Aseptate and septate with BLAST searches of GenBank (Altschul et al., 1997). We used
hyphae were counted separately in roots as indicators of AM Simultaneous Alignment and Tree Estimation (SATé, Liu et al.,
and non-AM fungal hyphae; the latter were assumed to be 2009) to infer sequence alignments and phylogenetic trees
saprotrophs, pathogens, or endophytes. For extraradical fun- simultaneously for each phylum, combining unknown se-
gal abundance, hyphae were extracted from 5 g of fresh soil in quences with 1000 representative fungal sequences from Silva
5% sodium hexametaphosphate following Brundrett et al. (Pruesse et al., 2007) and best matches from the GenBank
(1994). Soil hyphae were visualized by staining with acid database. Phylotypes were defined as monophyletic groups with
fuchsin on nylon filters and quantified by grid-line intersect sequence divergence less than 3%, which represented the aver-
microscopy with 50 grids at  200 (Brundrett et al., 1994). age sequence divergence present in known monophyletic clades
In June 2008, we measured ergosterol and Bradford reactive in these phylogenies. A total of 143 phylotypes were defined
soil proteins. Ergosterol is a membrane sterol found in fungi (102 Ascomycota, 24 Basidiomycota, seven Chytridiomycota,
that can be used to estimate total living fungal biomass as well seven basal lineages, and three Glomeromycota).
as undecomposed fungal structures (Zhao et al., 2005). Brad- OTUs were defined from the phylotypes based on distinct
ford reactive soil proteins can be used as an indicator of T-RFLP patterns. Digestion of the cloned sequences in silico
glomalin-related soil proteins produced by AM fungi (Rillig, (GENEIOUS PRO v. 4.8.3, Drummond et al., 2009) suggested that
2004), with the understanding that other heat-stable soil pro- two enzymes, HaeIII and HinfI, (Promega) would generate
teins also may be detected (Rosier et al., 2006). Ergosterol was sufficient polymorphic fragments to distinguish among most
extracted from 5 g fresh soil with cyclohexane following phylotypes. All samples were amplified in triplicate with
Nylund & Wallendar (1992). Extracts were analyzed by HPLC fluorescently tagged primers (HEX-ITS1F and 6-FAM-ITS4)
(SRI Model 202, Torrance, CA, USA) with Nova-Pak C18 using the PCR conditions described above; triplicate reactions
reverse phase column (Waters Corp., Milford, MA, USA). were pooled before digestion. The tagged products were
Bradford-reactive soil proteins were extracted from 1 g fresh digested and sequenced; runs were only accepted that had
soil with 20 mM sodium citrate and autoclaving (Wright et al., total fluorescence of 200–5000 RFU and a minimum peak size
1996; Wright & Upadhyaya, 1996; Rosier et al., 2006). A single of 25 RFU. T-RFLP profiles not meeting these criteria were
90-min autoclave cycle (121 1C) was used for easily extractable rerun with adjusted template concentrations. We used internal
and six subsequent cycles were used for total extractable standards of known size replicated in each run; the size of
Bradford reactive soil proteins. Samples were quantified for these standards did not change between plates. T-RFLP pro-
Bradford reactive soil proteins in triplicate with Bio-Rad files were keyed to OTU based on the clone profiles using
protein dye (Bio-Rad Laboratories, Hercules, CA, USA) at TRAMP-R (Fitzjohn & Dickie, 2007). An OTU was considered
590 nm on a DTX880 Multimode Plate Reader (Beckman present if all four identifying fragments were present within
Coulter Inc., Fullerton, CA, USA). 3 BP (Manhattan distance) of the expected fragment length. The
four enzyme fragments were sufficient to distinguish all but
eight phylotypes, resulting in a total of 138 OTUs. The col-
Fungal community composition lapsed phylotypes were all sister taxa at terminal nodes in the
Ascomycota phylogeny.
For each sample, nucleic acids were extracted from 0.5 g of
homogenized soil using a CTAB phenol–chloroform protocol
with bead beating (Griffiths et al., 2000). We amplified samples Soil organic carbon
in duplicate with the primers ITS1F (Gardes & Bruns, 1993)
In June 2008, we measured total soil carbon. For analysis of total
and ITS4 (White et al., 1990) in a reaction mixture of  50 ng of
soil organic carbon, air dried soils were sieved to 2 mm, ground
template DNA, 12.5 pmol of each primer, 0.8 mg mL1 BSA,
to a fine powder, and acidified with 0.73 M H2SO3 to remove
250 mM dNTPs (Promega, Madison, WI, USA), 10  ExTaq
carbonates (Heron et al., 1997). Soil samples were then run in
Mg2 1 buffer (Takara Bio Inc., Shiga, Japan), and 1 U ExTaq
duplicate for total organic carbon on an Apollo 9000 Analyzer
DNA polymerase (Takara Bio Inc.). Cycling parameters were
with Boat Sampler (Teledyne Tekmar, Mason, OH, USA).
94 1C for 5 min followed by 35 cycles of 94 1C for 30 s, 51 1C for
1 min, 72 1C for 1 min, and a final extension of 72 1C for 5 min
(Anderson et al., 2003). Duplicate reactions were pooled to
Decomposition incubation experiment
reduce PCR bias.
We created an experiment-wide clone library to use as a key We used a lab experiment to examine how field rainfall
for terminal-restriction fragment length polymorphism (T-RFLP) treatments affect microbial decomposition rates independently
of samples following Hausmann and Hawkes (2009). Briefly, the of direct moisture effects and how those rates were affected by
clone library was generated by combining PCR products from temperature. Specifically, we exposed soils from the field
all samples, cloning (TOPO-TA Kit, Invitrogen, Carlsbad, CA, experiment to winter and spring temperatures with and with-
USA), and sequencing (ABI 3130 Genetic Analyzer, Applied out litter addition, holding soil moisture constant.
Biosystems, Foster City, CA, USA) until we reached at least 95% Field soils were collected from the top 10 cm at two random
of the estimated sequence diversity based on rarefaction analysis positions in each plot in January 2007 using a 2 cm diameter
(ESTIMATES, Collwell, 2009). All sequences were deposited in corer, 6 months after the end of the 2006 spring rain addition
GenBank under the accession numbers HM239678–HM240255. treatment and before the start of the 2007 winter rain addition

r 2010 Blackwell Publishing Ltd, Global Change Biology, 17, 1637–1645

1640 C . V. H A W K E S et al.

treatments. Field soil moisture levels were the same across all fungal root colonization, we also examined bivariate correla-
treatments at this time (21.2%  0.68, P 5 0.242). This repre- tions of soil and root hyphae with annual aboveground
sents the maximum amount of time between treatment appli- biomass production from the current and previous year.
cations, and thus represents a conservative measure of the Fungal community composition was analyzed as a function
treatment effects, which are likely to be larger during active of treatment and date using multiple response permutation
application when there are pronounced moisture differences procedures (MRPP), a Monte Carlo approach that compares
among treatments. dissimilarities among groups (McCune & Grace, 2002; Mielke
Soils from each plot were sieved to 2 mm and  20 g was & Berry, 2007). To understand the direction and magnitude of
added to glass canning jars (237 mL) within 1 week of collec- differences among fungal communities tested with MRPP, we
tion. To examine decomposition rates, half the jars received used nonmetric multidimensional scaling (NMS) with Bray-
0.37 g of a common leaf litter that was 13C-labeled (2.23 at.% Curtis dissimilarity matrices. To examine the influence of plant
13
C), equivalent to adding  8 mg C g1 soil. Litter was from biomass and ambient rainfall on root colonization, soil hyphal
Avena barbata, a common grass at this site, which had been abundance, and fungal richness, we looked for treatment-level
raised with periodic exposure to a 13CO2 atmosphere, har- bivariate correlations across sampling dates. We used precipi-
vested, dried, ground to 1 mm particle size, and thoroughly tation levels measured on site and the Palmer Drought Sever-
mixed. Using labeled litter allowed us to track the movement ity Index for the northern California coastal region (http://
of carbon from the litter pool to the CO2 pool. Jars were sealed www.ncdc.noaa.gov). SPSS v. 17 for Windows was used for
with an airtight lid with a 13 mm butyl rubber septum (Bellco ANOVA and correlation statistics. All MRPP and NMS tests
Glass, Vineland, NJ, USA) for gas sampling. were run in PC-ORD v. 5.32 (McCune & Mefford, 2006).
Jars were incubated for 4 weeks at two temperatures, a
winter treatment (8 1C day, 5 1C night) and a spring treatment
(14 1C day, 7 1C night), based on averages of hourly tempera- Results
ture data taken at the site from 2000 to 2006. The headspace
was sampled periodically over 4 weeks in order to allow us to Fungal abundance
calculate average CO2 production rates. Specifically, we
sampled the headspace when the jars were initially closed, Soil hyphal abundance varied across precipitation treatments
every 2 days for the first week, and every 4–6 days thereafter, (Po0.001), with more hyphae in spring rain addition plots
for a total of seven sampling times over 4 weeks. At each time, during June 2005, 2007, and 2008 (Fig. 1), though there was
12 mL was removed from the jar headspace with an airtight variation across time in the overall abundance of hyphae in soil
syringe, stored in a 12 mL glass tube with septa (Labco Ltd, (Po0.001) and the nature of the precipitation treatment effect
High Wycombe, UK), and replaced in the jar with CO2-free air. was dependent on sampling date (P 5 0.035). The abundance of
Gas samples were analyzed on a GC-MS (Thermo Fisher soil hyphae was significantly inversely correlated with drought
Finnigan MAT GCQ, Waltham, MA, USA) with a DB5 column severity in spring rain additions (r 5 –0.783, P 5 0.021), but
to quantify 12C- and 13C-CO2. Total microbial activity was not winter rain additions (r 5 –0.484, P 5 0.225) or controls
quantified by total CO2 production and litter decomposition (r 5 –0.655, P 5 0.078). Soil hyphal abundance was independent
was quantified as the difference in 13CO2 production between of plant biomass in both the contemporary and previous year.
litter addition and control jars. Precipitation treatment did not affect root colonization by
mycorrhizal aseptate hyphae (P 5 0.754), arbuscules (P 5 0.857),
vesicles (P 5 0.329), or septate hyphae (P 5 0.352) (Table 1). There
Plant measurements
Aboveground plant biomass and litter were measured season-
ally (except in January 2008) within 4 weeks of the hyphal
measurements from plots adjacent to those sampled for fungi.
All vegetation and litter were collected from 900 cm2 subplots,
dried to constant weight at 60 1C, and weighed. These mea-
surements were used for treatment-level correlations between
fungal and plant patterns.

Statistics
We used repeated measures ANOVA to examine the effects of
precipitation treatment over time on fungal abundance (soil
hyphal abundance, fungal root colonization) and fungal OTU
richness. Univariate ANOVAs were used to examine precipita-
tion treatment effects on ergosterol, Bradford reactive soil
proteins, total organic carbon and both CO2 and 13CO2 pro-
duction in the decomposition experiment. To look for indirect Fig. 1 Abundance of soil hyphae at each sampling point.
effects of plant production on soil hyphal abundance and *Significant differences among rainfall treatments.

r 2010 Blackwell Publishing Ltd, Global Change Biology, 17, 1637–1645

 F U N G A L C L I M AT E R E S P O N S E 1641

Table 1 Percent root colonization by aseptate mycorrhizal hyphae, arbuscules, vesicles, and septate nonmycorrhizal hyphae

Aseptate hyphae Arbuscules Vesicles Septate hyphae

Sample date Rain addition Mean 95% CI Mean 95% CI Mean 95% CI Mean 95% CI

June 2005 Ambient 92.2 90.3–93.9 13.0 9.5–17.0 10.4 6.9–14.5 42.9 34.2–51.9
Winter 90.1 88.4–91.7 13.9 9.8–18.5 7.9 6.5–9.4 45.6 42.4–48.8
Spring 89.0 87.0–90.9 12.1 9.0–15.6 8.4 4.8–13.0 39.7 39.2–40.2
June 2006 Ambient 79.8 64.9–91.3 8.4 5.5–11.7 12.9 6.7–20.8 15.9 13.9–18.1
Winter 78.0 76.4–79.5 8.3 6.4–10.5 4.6 1.7–8.7 9.4 4.9–15.2
Spring 87.5 78.4–94.4 6.3 3.3–10.1 10.4 5.8–16.2 13.0 12.4–13.6
June 2007 Ambient 88.7 86.9–90.5 8.4 6.9–9.9 8.9 7.0–11.0 12.7 9.9–15.8
Winter 94.5 93.4–95.5 11.3 8.0–15.2 4.4 0.9–10.5 13.5 5.1–25.0
Spring 92.8 90.2–95.0 11.1 9.0–13.4 10.8 7.6–14.5 13.1 9.9–16.6
June 2008 Ambient 88.0 79.0–94.7 9.7 2.2–21.8 11.4 4.1–21.7 26.7 17.1–37.7
Winter 85.9 79.2–91.5 4.5 1.5–9.1 10.8 6.6–15.7 18.0 10.4–27.2
Spring 86.7 81.8–90.9 6.8 0.2–21.4 9.4 4.5–15.9 17.2 11.2–24.2

All data were arcsine square root transformed for analysis; backtransformed means are presented with asymmetric 95% confidence
intervals. Measurements were made on three replicate plots per treatment per time point.

were annual changes in root colonization by both aseptate

(Po0.001) and septate hyphee (Po0.001) but not arbuscules
(P 5 0.150) or vesicles (P 5 0.414). Differences between years
were as great as 30% for septate hyphae and smaller for aseptate
hyphae (o12%). Changes in septate hyphae through time were
significantly positively correlated with background rainfall in
winter (r 5 0.953, P 5 0.047) and spring (r 5 0.958, P 5 0.042) rain
addition treatments, but not in control plots (r 5 0.822, P 5 0.178).
There was no relationship between ambient rainfall and root
colonization by mycorrhizal hyphae, arbuscules, or vesicles over
time. Root colonization was not significantly correlated with
plant biomass in the contemporary or previous year.

Fungal community composition

Fig. 2 Fungal community composition in control and rainfall
Fungal community composition varied significantly with the treatments across sample dates based on nonmetric multidimen-
interaction of treatment and date (Po0.001) (Fig. 2). Overall, sional scaling of ITS rDNA. Data are averages  1 SE. NMS,
composition was very similar across treatments in June 2006, nonmetric multidimensional scaling.
2007, and 2008 when rainfall during the month before sam-
pling was extremely low (30, 31, and 5 mm, respectively). The was not significantly correlated with plant biomass in either
fungal community in spring rainfall addition plots shifted the current or previous year.
away from control plots at two time points, June 2005 and
January 2007, when rainfall during the month before sampling
was 4200 mm; a similar trend was observed for January 2008 Soil carbon pools
when rainfall was  190 mm, though this time point repre-
sented a break from longer-term drought conditions and Total organic carbon pools were the same in all precipitation
composition was more variable. Fungal richness paralleled treatments (P 5 0.872) (Table 2). Similarly, ergosterol did not
the shifts in fungal composition, with no effect of rainfall differ across precipitation treatments (P 5 0.922), nor did
treatment (P 5 0.945), but significant variation across time Bradford reactive soil proteins pools (easily extractable
(Po0.001) that was inversely correlated to background rainfall P 5 0.926; total P 5 0.984) (Table 2).
in control plots (r 50.944, P 5 0.005), winter rain additions
(r 50.921, P 5 0.009), and spring rain additions (r 50.836,
Microbial activity and decomposition
P 5 0.038). Richness was highest in June 2006 and June 2008
(111 and 122 taxa), lowest in January 07 (51 taxa), and inter- Total microbial activity doubled in soils from spring precipita-
mediate at the remaining dates (75–90 taxa). Fungal richness tion treatments (P 5 0.004), but only when litter was added

r 2010 Blackwell Publishing Ltd, Global Change Biology, 17, 1637–1645

1642 C . V. H A W K E S et al.

Table 2 Ergosterol (mg g1 soil), easily extractable and total Bradford reactive soil proteins (mg g1 soil), and total soil organic
carbon (mg g1 soil)

Bradford reactive
soil protein – easily Bradford reactive Soil total organic
Ergosterol extractable soil protein – total carbon

Rain addition Mean SD Mean SD Mean SD Mean SD

Ambient 1.03 0.23 2.49 0.74 5.53 2.26 4.22 0.51

Winter 1.06 0.18 2.43 1.18 5.66 2.39 4.53 1.16
Spring 1.09 0.36 2.66 1.09 5.45 1.17 4.33 1.20

Measurements were made in June 2008 on soils from six replicate plots per treatment.

(Fig. 3a). Similarly, decomposition was 1.5 times greater in

spring rain additions (P 5 0.020) (Fig. 3b). Not surprisingly,
total microbial activity doubled under warmer spring tem-
peratures compared with cooler winter temperatures
(P 5 0.005), but decomposition rates showed only a marginally
significant difference between seasonal temperatures
(P 5 0.051). There were no interactions between precipitation
and temperature (P40.165).

Discussion
Soil fungi were highly responsive to ambient rainfall
within and across years, leading to changes in fungal
abundance, composition, and function. These changes
were correlated to rainfall and drought stress, but not
current soil moisture or plant biomass, consistent with
direct responses of fungi to short-term rainfall patterns.
Essentially, fungal communities under low rainfall were
more diverse and abundant, but also more similar and
repeatable through time; increased rainfall led to less Fig. 3 Total microbial activity (a) as indicated by CO2 produc-
diverse, less abundant, and more variable fungal com- tion, and litter decomposition (b) as indicated by 13CO2 produc-
munities. That the fungal responses were rapid, rever- tion from 13C-labeled litter.
sible, and repeatable implies an inherently plastic
community composed of species with a range of envir- ing the current winter rainy season. The impacts of
onmental tolerances. The retention of broadly different spring rainfall additions on soil fungi were also highly
physiological tolerances has also been found in soil contingent on the background inputs of ambient rainfall
bacterial communities under fluctuating redox conditions – expansion of fungal hyphae in soil only occurred in
(Pett-Ridge & Firestone, 2005), perhaps representing a spring rain addition plots during drought periods
storage effect in microbes where temporal fluctuations (PDSIo2.5, moderate drought). Positive relationships
allow stable coexistence of competitors (Chesson, 2000). of fungal abundance with precipitation or soil moisture
Higher fungal diversity under drought conditions further have been found across agricultural sites (Frey et al.,
suggests that stress periodically moderates fungal com- 1999) and across time in semiarid sites (Querejeta et al.,
petition (Chase & Leibold, 2003). Understanding the 2009; Trent et al., 1994) and forests (Bååth & Söderström,
mechanisms maintaining soil microbial community 1982). Seasonal variation in hyphal abundance has also
diversity and function under current climatic conditions been observed (Miller et al., 1995; Lutgen et al., 2003;
will be critical for predicting soil responses to future Clark et al., 2009). Variability in fungal responses may be
climate change (Allison & Martiny, 2008). based on relative levels of water available for metabolic
Soil fungi were affected by an experimental extension activity, so that fungi only increase with rainfall in drier
of the rainy season into the normally dry month of June, soils where water is limiting (Raich & Potter, 1995).
but not by increasing the amount of precipitation dur- Though not detected in the current study, soil hyphal

r 2010 Blackwell Publishing Ltd, Global Change Biology, 17, 1637–1645

 F U N G A L C L I M AT E R E S P O N S E 1643

growth could also be affected by rainfall-driven served in our study, it is also possible that increasing
changes in plant production or composition. rainfall can lead to increased soil hyphal abundance with
In contrast to soil hyphae, differences in the composi- concomitant increases in soil aggregation, humification,
tion and richness of fungi in spring rain addition plots and organic carbon (Martin-Neto et al., 1998; Wilson et al.,
were evident only when monthly rainfall was 2009). Microbial community composition has been
4200 mm. These patterns are consistent with a stress shown to affect ecosystem processes in other studies in
threshold, below which biotic interactions such as com- similar grasslands (e.g., Hawkes et al., 2005). Under-
petition determine coexistence (Chase & Leibold, 2003). standing the links between composition and function
Species interactions and feedbacks can be as or more in microbial communities remains a key challenge in our
important than direct responses to climate change over ability to assess soil resilience with climate change
extended time scales (Suttle et al., 2007). In the current (Allison & Martiny, 2008).
study, sufficient rainfall may allow good resource com- The responses of soil fungi to altered rainfall clearly
petitors to dominate or may alter trophic interactions depend on how those changes relate to current rainfall
via the abundance of soil fauna (e.g., Lindberg et al., patterns. The frequency of environmental fluctuations
2002). Rainfall responses probably differ among micro- and the range of species responses to those fluctuations
bial groups more generally, as another study at the same are predicted to affect coexistence and stability in com-
site found little to no cumulative change in soil bacterial munities, often in unexpected ways (Ives et al., 1999,
or archaeal community composition among treatments Ripa & Ives, 2003, Ruokolainen et al., 2009). Adaptation
between Dec 2005 and July 2007, despite wide seasonal and extinction processes will also likely depend on the
swings in composition across this period (Cruz-Marti- frequency of unfavorable environmental conditions
nez et al., 2009). Moreover, during the only time points (Bell et al., 2009). If repeated exposure to unfavorable
where bacteria and archaea communities differed conditions selects for microbial taxa that are tolerant to
among rainfall treatments (April and July 2006), fungal a range of environments, we might expect the baseline
communities were the same (June 2006). Fluctuations in frequency and magnitude of environmental fluctua-
other environmental factors such as temperature, tions to drive resistance and resilience of the microbial
resources, and oxygen levels also affect coexistence, community. Thus, if the characteristics of environmen-
diversity, and structure of microbial populations and tal change fall within the range of fluctuations typically
communities (Descamps-Julien & Gonzalez, 2005; Pett- experienced at a site, the microbial community response
Ridge & Firestone, 2005; Long et al., 2007; Matthews & should be more predictable than if they fall outside this
Gonzalez, 2007; Hiltunen et al., 2008). We still lack a range. Further support for this idea is provided by
comprehensive mechanistic framework for these functional and compositional responses of soil micro-
changes, however, which would facilitate better predic- bial communities reciprocally transplanted between
tive modeling of soil microbial responses to altered habitats where the environmental extremes of one fall
environmental conditions. within the range experienced by the other (Waldrop &
Compositional shifts in fungi translated into functional Firestone, 2006). Linking microbial community dy-
shifts that could be important for soil carbon cycling. namics to the temporal structure of environmental
While no differences were observed in carbon storage variation (Ruokolainen et al., 2009) may provide us with
related to soil fungi over the time frame of the current a framework for testing general response patterns to
study, CO2 fluxes to the atmosphere were increased by environmental changes in the future.
spring rain addition, implying likely emergence of long-
er-term differences in carbon storage among treatments.
Microbial community activity and decomposition in- Acknowledgements
creased in soils from spring rain plots collected in We thank the following for their support and assistance: P. Bennett,
January 2007, when a shift in fungal community compo- E. Brault, K. Cruz, M. Firestone, D. Herman, R. Linder, J. Miner,
sition was observed. Hyphae, soil moisture, and soil M. Power, A. Rosling, P. Steel, and the UC Angelo Coast Range
Reserve. Funding for this project was provided to C. V. H. by an
bacterial and archaeal community composition (Cruz- NSF Research Starter Grant (DEB0528416) and to K. B. S. by an EPA
Martinez et al., 2009) were the same across treatments at STAR Fellowship and Canon National Parks Science Scholarship.
this time, strongly supporting a role for fungal commu-
nity composition. Because we only incubated soils at a
single time point, we do not know whether this pattern is References
maintained consistently or only when composition shifts. Adair EC, Parton WJ, Del Grosso SJ et al. (2008) Simple three-pool model accurately
describes patterns of long-term litter decomposition in diverse climates. Global
Similar differences in CO2 flux rates were observed in the
Change Biology, 14, 2636–2660.
field for May and July 2004 (data not shown), but we did Aerts R (1997) Climate, leaf litter chemistry and leaf litter decomposition in terrestrial
not measure composition at that time. Though not ob- ecosystems: a triangular relationship. Oikos, 79, 439–449.

r 2010 Blackwell Publishing Ltd, Global Change Biology, 17, 1637–1645

1644 C . V. H A W K E S et al.

Allen MF (1991) The Ecology of Mycorrhizae. Cambridge University Press, Cambridge, UK. Heron G, Barcelona MJ, Andersen ML, Christensen TH (1997) Determination of
Allison SD, Martiny JBH (2008) Resistance, resilience, and redundancy in microbial nonvolatile organic carbon in aquifer solids after carbonate removal by sulfurous
communities. Proceedings of the National Academy of Sciences, 105, 11512–11519. acid. Groundwater, 35, 6–11.
Altschul SF, Thomas LS, Alejandro AS, Jingui Z, Webb M, David JL (1997) Gapped Hiltunen T, Laakso J, Kaitala V, Suomalainen L-R, Pekkonen M (2008) Temporal
BLAST and PSI-BLAST: a new generation of protein database search programs. variability in detritus resource maintains diversity of bacterial communities. Acta
Nucleic Acids Research, 25, 3389–3402. Oecologica, 33, 291–299.
Anderson IC, Campbell CD, Prosser JI (2003) Potential bias of fungal 18S rDNA and Husband R, Herre EA, Turner SL, Gallery R, Young JPW (2002) Molecular diversity of
internal transcribed spacer polymerase chain reaction primers for estimating fungal arbuscular mycorrhizal fungi and patterns of host association over time and space
biodiversity in soil. Environmental Microbiology, 5, 36–47. in a tropical forest. Molecular Ecology, 11, 2669–2678.
Augé RM (2001) Water relations, drought and vesicular–arbuscular mycorrhizal Ives AR, Gross K, Klug JL (1999) Stability and variability in competitive communities.
symbiosis. Mycorrhiza, 11, 3–42. Science, 286, 542–544.
Bååth E, Söderström B (1982) Seasonal and spatial variation in fungal biomass in a Johnson D, Leake JR, Read DJ (2002) Transfer of recent photosynthate into mycorrhizal
forest soil. Soil Biology and Biochemistry, 14, 353–358. mycelium of an upland grassland: short-term respiratory losses and accumulation
Bell CW, Acosta-Martinez V, McIntyre NE, Cox S, Tissue DT, Zak JC (2009) Linking of C-14. Soil Biology & Biochemistry, 34, 1521–1524.
microbial community structure and function to seasonal differences in soil moisture Kivlin SN, Hawkes CV (in press) Differentiating between effects of invasion and
and temperature in a Chihuahuan desert grassland. Microbial Ecology, 58, 827–842. diversity: impacts of aboveground plant communities on root fungal communities.
Brundrett M, Melville L, Peterson L (1994) Practical Methods in Mycorrhiza Research. New Phytologist.
Mycologue Publications, Guelph, Canada. Knapp AK, Fay PA, Blair JM et al. (2002) Rainfall variability, carbon cycling, and plant
Chase JM, Leibold MA (2003) Ecological Niches: Linking Classical and Contemporary species diversity in a mesic grassland. Science, 298, 2202–2205.
Approaches. University of Chicago Press, Chicago. Lauenroth WK, Sala OE (1992) Long-term forage production of North American
Chesson P (2000) Mechanisms of maintenance of species diversity. Annual Review of shortgrass steppe. Ecological Applications, 2, 397–403.
Ecology and Systematics, 31, 343–366. Lindberg N, Engtsson JB, Persson T (2002) Effects of experimental irrigation and
Chou WW, Silver WL, Jackson RD, Thompson AW, Allen-Diaz B (2008) The sensitivity drought on the composition and diversity of soil fauna in a coniferous stand.
of annual grassland carbon cycling to the quantity and timing of rainfall. Global Journal of Applied Ecology, 39, 924–936.
Change Biology, 14, 1382–1394. Liu K, Raghavan S, Nelesen S, Linder CR, Warnow T (2009) Rapid and accurate large-
Clark NM, Rillig MC, Nowak RS (2009) Arbuscular mycorrhizal fungal abundance in scale coestimation of sequence alignments and phylogenetic trees. Science, 324,
the Mojave Desert: seasonal dynamics and impacts of elevated CO2. Journal of Arid 1561–1564.
Environments, 73, 834–843. Long ZT, Petchey OL, Holt RD (2007) The effects of immigration and environmental
Collwell RK (2009) EstimateS: Statistical estimation of species richness and shared species variability on the persistence of an inferior competitor. Ecology Letters, 10,
from samples. Available at http://purl.oclc.org/estimates 574–585.
Cruz-Martinez K, Suttle KB, Brodie EL, Power ME, Andersen GL, Banfield JF (2009) Lutgen ER, Muir-Clairmont D, Graham J, Rillig MC (2003) Seasonality of arbuscular
Despite strong seasonal responses, soil microbial consortia are more resilient to mycorrhizal hyphae and glomalin in a western Montana grassland. Plant and Soil,
long-term changes in rainfall than overlying grassland. Isme Journal, 3, 738–744. 257, 71–83.
Curiel-Yuste J, Baldocchi DD, Gershenson A, Goldstein A, Misson L, Wong S (2007) Martin-Neto L, Rosell R, Sposito G (1998) Correlation of spectroscopic indicators of
Microbial soil respiration and its dependency on carbon inputs, soil temperature humification with mean annual rainfall along a temperate grassland climose-
and moisture. Global Change Biology, 13, 2018–2035. quence. Geoderma, 81, 305–311.
Descamps-Julien B, Gonzalez A (2005) Stable coexistence in a fluctuating environment: Matthews DP, Gonzalez A (2007) The inflationary effects of environmental fluctuations
an experimental demonstration. Ecology, 86, 2815–2824. ensure the persistence of sink metapopulations. Ecology, 88, 2848–2856.
Drummond A, Ashton B, Cheung M et al. (2009) Geneious v4.8. Available at http:// McCune B, Grace JB (2002) Analysis of Ecological Communities. MjM Software Design,
www.geneious.com/ Gleneden Beach, OR.
Epstein HE, Burke IC, Lauenroth WK (2002) Regional patterns of decomposition and McCune B, Mefford MJ (2006) PC-Ord for Windows v. 5.15. Multivariate Analysis of
primary production rates in the U.S. great plains. Ecology, 83, 320–327. Ecological Data. MjM Software, Gleneden Beach OR, USA.
Fitter AH, Heinemeyer A, Staddon PL (2000) The impact of elevated CO2 and global McGonigle TP, Miller MH, Evans DG, Fairchid GL, Swan JA (1990) A new method
climate change on arbuscular mycorrhizas: a mycocentric approach. New Phytolo- which gives an objective measure of colonization of roots by vesicular-arbuscular
gist, 147, 179–187. mycorrhizal fungi. New Phytologist, 115, 495–501.
Fitzjohn RG, Dickie IA (2007) TRAMPR: an R package for analysis and matching of Mielke PW Jr, Berry KJ (2007) Permutation Methods: A Distance Function Approach.
terminal-restriction fragment length polymorphism (TRFLP) profiles. Molecular Springer-Verlag, New York.
Ecology Notes, 7, 583–587. Miller RM, Reinhardt DR, Jastrow JD (1995) External hyphal production of vesicular–
Frey SD, Elliott ET, Paustian K (1999) Bacterial and fungal abundance and biomass in arbuscular mycorrhizal fungi in pasture and tallgrass prairie communities.
conventional and no-tillage agroecosystems along two climatic gradients. Soil Oecologia, 103, 17–23.
Biology and Biochemistry, 31, 573–585. Nylund JE, Wallander H (1992) Ergosterol analysis as a means of quantifying
Frey SD, Six J, Elliott ET (2003) Reciprocal transfer of carbon and nitrogen by mycorrhizal biomass. Methods in Microbiology, 24, 77–88.
decomposer fungi at the soil-litter interface. Soil Biology and Biochemistry, 35, Pett-Ridge J, Firestone MK (2005) Redox fluctuation structures microbial
1001–1004. communities in a wet tropical soil. Applied and Environmental Microbiology, 71,
Gardes M, Bruns TD (1993) ITS primers with enhanced specificity for Basidiomycetes – 6998–7007.
application to the identification of mycorrhizae and rusts. Molecular Ecology, 2, 113–118. Pruesse E, Quast C, Knittel K, Fuchs B, Ludwig W, Peplies J, Glöckner FO
Godbold D, Hoosbeek M, Lukac M et al. (2006) Mycorrhizal Hyphal Turnover as a (2007) SILVA: a comprehensive online resource for quality checked and aligned
Dominant Process for Carbon Input into Soil Organic Matter. Plant and Soil, 281, 15–24. ribosomal RNA sequence data compatible with ARB. Nucleic Acids Research, 35,
Griffiths RI, Whiteley AS, O’Donnell AG, Bailey MJ (2000) Rapid method for coex- 7188–7196.
traction of DNA and RNA from natural environments for analysis of ribosomal Querejeta JI, Egerton-Warburton LM, Allen MF (2009) Topographic position modu-
DNA- and rRNA-based microbial community composition. Applied and Environ- lates the mycorrhizal response of oak trees to interannual rainfall variability.
mental Microbiology, 66, 5488–5491. Ecology, 90, 649–662.
Hausmann NT, Hawkes CV (2009) Plant neighborhood control of arbuscular mycor- Raich JW, Potter CS (1995) Global patterns of Carbon Dioxide emissions from soils.
rhizal community composition. New Phytologist, 183, 1188–1200. Global Biogeochemical Cycles, 9, 23–36.
Hausmann NT, Hawkes CV (2010) Order of plant host establishment alters the Rillig MC (2004) Arbuscular mycorrhizae, glomalin, and soil aggregation. Canadian
composition of arbuscular mycorrhizal communities. Ecology, 91, 2333–2343. Journal of Soil Science, 84, 355–363.
Hawkes CV, Wren IF, Herman DJ, Firestone MK (2005) Plant invasion alters nitrogen Rillig MC, Wright SF, Nichols KA, Schmidt WF, Torn MS (2001) Large contribution of
cycling by modifying the soil nitrifying community. Ecology Letters, 8, 976–985. arbuscular mycorrhizal fungi to soil carbon pools in tropical forest soils. Plant and
Heinemeyer A, Hartley IP, Evans SP, Carreira De La Fuente JA, Ineson P (2007) Forest Soil, 233, 167–177.
soil CO2 flux: uncovering the contribution and environmental responses of ecto- Ripa J, Ives AR (2003) Food web dynamics in correlated and autocorrelated environ-
mycorrhizas. Global Change Biology, 13, 1786–1797. ments. Theoretical Population Biology, 64, 369–384.

r 2010 Blackwell Publishing Ltd, Global Change Biology, 17, 1637–1645

 F U N G A L C L I M AT E R E S P O N S E 1645

Rosier CL, Hoye AT, Rillig MC (2006) Glomalin-related soil protein: assessment Vargas R, Baldocchi DD, Querejeta JI et al. (2010) Ecosystem CO2 fluxes of arbuscular
of current detection and quantification tools. Soil Biology & Biochemistry, 38, 2205– and ectomycorrhizal dominated vegetation types are differentially influenced by
2211. precipitation and temperature. New Phytologist, 185, 226–236.
Ruokolainen L, Lindén A, Kaitala V, Fowler MS (2009) Ecological and evolutionary Waldrop M, Firestone M (2006) Response of microbial community composition and
dynamics under coloured environmental variation. Trends in Ecology & Evolution, function to soil climate change. Microbial Ecology, 52, 716–724.
24, 555–563. Waldrop MP, Zak DR, Blackwood CB, Curtis CD, Tilman D (2006) Resource avail-
Sala OE, Parton WJ, Joyce LA, Lauenroth WK (1988) Primary production of the central ability controls fungal diversity across a plant diversity gradient. Ecology Letters, 9,
grassland region of the United States. Ecology, 69, 40–45. 1127–1135.
Skopp J, Jawson MD, Doran JW (1990) Steady-state aerobic microbial activity as a Wardle DA, Nicholson KS (1996) Synergistic effects of grassland plant species on soil
function of soil water content. Soil Sci Soc Am J, 54, 1619–1625. microbial biomass and activity: implications for ecosystem-level effects of enriched
Solomon S, Qin D, Manning M et al (eds) (2007) Climate Change 2007: The Physical plant diversity. Functional Ecology, 10, 410–416.
Science Basis. Cambridge University Press, Cambridge, UK. White TJ, Bruns T, Lee S, Taylor JW (1990) Amplification and direct sequencing of
Staddon PL, Thompson K, Jakobsen I, Grime JP, Askew AP, Fitter AH (2003) fungal ribosomal RNA genes for phylogenetics. In: PCR Protocols: A Guide
Mycorrhizal fungal abundance is affected by long-term climatic manipulations in to Methods and Applications (eds Innis MA, Gelfand DH, Sninsky JJ, White TJ),
the field. Global Change Biology, 9, 186–194. pp. 315–322. Academic Press Inc., New York, NY.
Steinberg PD, Rillig MC (2003) Differential decomposition of arbuscular Wilson GWT, Rice CW, Rillig MC, Springer A, Hartnett DC (2009) Soil aggregation and
mycorrhizal fungal hyphae and glomalin. Soil Biology & Biochemistry, 35, carbon sequestration are tightly correlated with the abundance of arbuscular
191–194. mycorrhizal fungi: results from long-term field experiments. Ecology Letters, 12,
Suttle KB, Thomsen M, Power ME (2007) Species interactions reverse grassland 452–461.
responses to changing climate. Science, 315, 640–642. Wright SF, FrankeSnyder M, Morton JB, Upadhyaya A (1996) Time-course study and
Thorn G (1997) The fungi in soil. In: Modern Soil Microbiology (eds Dirk Van partial characterization of a protein on hyphae of arbuscular mycorrhizal fungi
Elsas J, Trevors JT, Wellington EMH), pp. 63–127. Marcel Dekker Inc., New York, during active colonization of roots. Plant and Soil, 181, 193–203.
NY, USA. Wright SF, Upadhyaya A (1996) Extraction of an abundant and unusual protein from
Trent JD, Svejcar TJ, Blank RR (1994) Mycorrhizal colonization, hyphal lengths, and soil and comparison with hyphal protein of arbuscular mycorrhizal fungi. Soil
soil moisture associated with two Artemisia tridentata subspecies. Great Basin Science, 161, 575–586.
Naturalist, 54, 291–300. Xu L, Baldocchi DD (2004) Seasonal variation in carbon dioxide exchange over a
Treseder KK, Allen MF (2000) Mycorrhizal fungi have a potential role in soil Mediterranean annual grassland in California. Agricultural and Forest Meteorology,
carbon storage under elevated CO2 and nitrogen deposition. New Phytologist, 147, 123, 79–96.
189–200. Zhao XR, Lin Q, Brookes PC (2005) Does soil ergosterol concentration provide a
Vandenkoornhuyse P, Ridgway KP, Watson IJ, Fitter AH, Young JPW (2003) reliable estimate of soil fungal biomass? Soil Biology and Biochemistry, 37, 311–317.
Co-existing grass species have distinctive arbuscular mycorrhizal communities. Zhu Y-G, Miller RM (2003) Carbon cycling by arbuscular mycorrhizal fungi in soil-
Molecular Ecology, 12, 3085–3095. plant systems. Trends in Plant Science, 8, 407–409.

r 2010 Blackwell Publishing Ltd, Global Change Biology, 17, 1637–1645