Responses of arbuscular mycorrhizal fungi to multiple coinciding global change drivers

Responses of arbuscular mycorrhizal fungi to multiple coinciding global change drivers

Fungal Ecology 40 (2019) 62e71 Contents lists available at ScienceDirect Fungal Ecology journal homepage: www.elsevier.com/locate/funeco Responses ...

443KB Sizes 0 Downloads 86 Views

Fungal Ecology 40 (2019) 62e71

Contents lists available at ScienceDirect

Fungal Ecology journal homepage: www.elsevier.com/locate/funeco

Responses of arbuscular mycorrhizal fungi to multiple coinciding global change drivers € ren Eliot Weber a, *, Jeffrey M. Diez a, Lela V. Andrews b, Michael L. Goulden c, So Emma L. Aronson d, Michael F. Allen e a

Department of Botany and Plant Sciences, University of California Riverside, USA Department of Biological Sciences, Northern Arizona University, USA Department of Earth System Science, University of California Irvine, USA d Department of Microbiology and Plant Pathology, University of California Riverside, USA e Center for Conservation Biology, University of California Riverside, USA b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 12 November 2017 Received in revised form 24 April 2018 Accepted 20 November 2018 Available online 23 January 2019

A significant challenge for understanding how fungal communities may change in the Anthropocene are the multiple aspects of simultaneous environmental change. To address this challenge, we used a sevenyear multi-factorial field experiment in southern California to examine how root-associated fungi respond to aridity, nitrogen deposition, and plant invasions. We hypothesized that all three global change drivers reduce the abundance of arbuscular mycorrhizal fungi responsible for nutrient uptake (edaphophilic AMF), while increasing the abundance of AMF that colonize roots at high rates (rhizophilic AMF). We found that invasive grasses hosted lower abundances of edaphophilic AMF, and higher abundances of rhizophilic AMF and opportunistically parasitic fungi. Aridity reduced overall AMF abundance while N addition altered the allocation of AMF biomass, increasing root colonization while reducing the density of extraradical hyphae. Overall, these results imply that ongoing global change will alter both the composition of AMF and how these fungi interact with plants. © 2018 Elsevier Ltd and British Mycological Society. All rights reserved.

Corresponding Editor: Petr Kohout and Lynne Boddy Keywords: AMF Fungi Global change Mycorrhizae Community ecology Fungal ecology Nitrogen deposition Biotic invasion Coastal sage scrub

1. Introduction Plant and animal responses to global environmental changes have been extensively studied, but little is known about the sensitivity or direction of fungal responses to these shifts. Multiple important dimensions of the environment are changing simultaneously, and one of the major challenges is to understand how fungi will respond to the diversity of, and interactions between, these drivers. In the US southwest, precipitation frequency is expected to decline over the next century (Seager and Vecchi, 2010;

* Corresponding author. Current Address: Institute of Evolutionary Biology and Environmental Sciences, University of Zurich, Winterthurerstrasse 190, 8057, Zürich. E-mail address: [email protected] (S.E. Weber). https://doi.org/10.1016/j.funeco.2018.11.008 1754-5048/© 2018 Elsevier Ltd and British Mycological Society. All rights reserved.

Allen and Luptowitz, 2017), which coupled with warming is expected to increase aridity. In this same region, anthropogenic atmospheric nitrogen deposition has increased N availability, leading to changes in the composition of arbuscular mycorrhizal fungi (AMF) and other groups of fungi (Fig. 1A; Egerton-Warburton and Allen, 2000; Amend et al., 2016). Meanwhile, exotic annual grasses from the Mediterranean are invading plant communities previously dominated by native perennial shrubs and grasses in the US southwest. AMF are plant mutualists currently grouped in the subphylum Glomeromycotina of Mucoromycota (Spatafora et al., 2016). These fungi associate with plant roots and provide a variety of services in exchange for photosynthetically derived carbon, including uptaking nutrients (especially P, but also N, see (Hodge and Storer, 2014), and micronutrients), altering plant drought response (Allen and Allen, 1986; Auge, 2001; Worchel et al., 2013), and reducing root

S.E. Weber et al. / Fungal Ecology 40 (2019) 62e71

A

63

Host Plant Life History

Grasses

Shrubs

Nitrogen Deposition

Rhizophilic AMF

Parasitic Fungi

Drought

x

Edaphophilic AMF

Ancestral AMF AMF Abundance

Non-AMF root coln.

AMF root coln.

AMF hyphal density

Average spore diameter

AMF Biomass

B

Parasitic Fungi

Host Plant Life History

Grasses

*

Shrubs

Nitrogen Deposition

Rhizophilic AMF

Drought

x

Edaphophilic AMF

Ancestral AMF AMF Abundance

Non-AMF root coln.

AMF root coln.

*

AMF hyphal density

Average spore diameter

AMF Biomass

Fig. 1. Panel A: Conceptual diagram of hypotheses. Solid lines denote positive correlations while dashed lines denote negative associations. Panel B: Summary diagram of results from generalized linear models. Supported hypotheses are depicted with black lines, unsupported with grey lines. Solid lines denote positive correlations while dashed lines denote negative associations. Asterisks denote significant interactions that were the opposite direction than hypothesized.

pathogen infection (Sikes et al., 2009, 2010). The current view of AMF community responses to shifting environmental conditions is limited to coarse patterns estimated from spore composition and fungal morphological observations. However, recent developments in molecular sequencing allow a more detailed understanding of fungal responses to changing environmental conditions. New challenges arise with the use of these molecular data, including interpreting the ecological meaning of shifts in the richness and

abundance of fungal taxa (Peay, 2014). While a functional trait approach could improve our understanding of responses of fungal communities to environmental drivers (Crowther et al., 2014), functional traits are sparsely described across AMF taxa. AMF can be broadly classified into guilds by their patterns of biomass allocation to extraradical hyphae, intraradical hyphae and spores (Maherali and Klironomos, 2007). Based on published descriptions of these patterns at the family level, we refer to these

64

S.E. Weber et al. / Fungal Ecology 40 (2019) 62e71

guilds as ‘edaphophilic’ (high allocation to extraradical hyphae, low allocation to intraradical hyphae), ‘rhizophilic’ (high allocation to intraradical hyphae, with limited extraradical hyphae) and ‘ancestral’ (lower biomass and lack allocation preference, hypothesized to be the ancestral condition for AMF by Powell et al., 2009). These AMF guilds differ in how they interact with, and affect, plant hosts. Rhizophilic AMF (e.g. Glomeraceae, Table 1) appear to play a role in reducing root pathogen infection, primarily benefiting plants with fine-roots prone to pathogen infection (Sikes et al., 2010). Edaphophilic AMF (e.g. Gigasporaceae, Table 1) increase plant nutrient uptake via their extensive extraradical mycelium, primarily benefiting plants with coarse-roots poorly suited to uptake nutrients directly (Sikes et al., 2010) and are thus more reliant on AMF for nutrient acquisition than plants with finer roots (Eissenstat et al., 2015). While the effect of ancestral AMF on plant performance remains unclear, Maherali and Klironomos (2007) showed that plant biomass responded positively to increasing overall AMF species diversity, suggesting functional complementarity between ancestral and other AMF guilds. These guilds of AMF also have different affinities for soil moisture and N, and likely differ in host plant association (Fig. 1A). Estimates from spore abundances suggest that edaphophilic AMF are less diverse and abundant in communities in arid environments, whereas rhizophilic and ancestral AMF appear to be tolerant of low soil moisture (Allen et al., 1995; Stutz and Morton, 1996; Stutz et al., 2000; Chaudhary et al., 2014). This apparent sensitivity of edaphophilic AMF could be due to increased exposure to the surrounding soil matrix because of long extraradical hyphae (Fig. 1A). Additional N increases the abundance of rhizophilic AMF spore density, while decreasing the abundance of edaphophilic AMF in southern California (Egerton-Warburton and Allen, 2000), though theory and field experiments suggest that this pattern is dependent on initial soil N:P (Treseder and Allen, 2002; Egerton-Warburton et al., 2007). The abundance of edaphophilic AMF may decline with increasing N because of a decreased dependence of host plants on nutrient uptake from edaphophilic AMF (Sikes et al., 2010) as nutrient availability increases (Johnson, 1993). Additionally, plants  pezappear to prefer different taxa of AMF (Davison et al., 2011; Lo García et al., 2017), possibly to maximize those benefits contingent on host plant root architecture (Sikes et al., 2010). EgertonWarburton and Allen (2000) used spore counts to find that invasive annual grass species in southern California host a lower abundance of Gigasporaceae (edaphophilic AMF) than native shrubs and suggest that invasion by these fine-rooted invasive grasses may be facilitated by an increased abundance of taxa in that we describe as rhizophilic AMF as result of nitrogen deposition (Fig. 1A). In this study, we tested the effects of these multiple global changes on AMF fungal communities. We hypothesized that Table 1 Description of AMF guilds. Guild

Intraradical hyphae

Extraradical hyphae

Families

Citationa

Rhizophilic

High

Low

1,2,3,4,5 2

Edaphophilic Low

High

Ancestral

Low

Glomeraceae Claroideoglomeraceae Paraglomeraceaeb Gigasporaceae Diversisporaceae Archaeosporaceaeb Ambisporaceaeb Pacisporaceae Acaulosporaceae

Low

1,2,5 2,5

5 1,2

a 1. Hart and Reader (2002), 2. Powell et al., (2009), 3. Varela-Cervero et al., (2015), 4. Varela-Cervero et al., (2016a), 5. Varela-Cervero et al., (2016b). b Families assigned based on past classifications.

increasing aridity, atmospheric nitrogen deposition, and species invasion would all reduce relative abundance of edaphophilic AMF, while increasing the relative abundance of rhizophilic AMF in roots (Fig. 1A). Furthermore, we hypothesized that these responses of edaphophilic and rhizophilic AMF would alter the abundance of parasitic fungi in roots as well as change patterns of fungal biomass allocation (Fig. 1A). Because we know little about the ecology of ancestral AMF we also asked how aridity, nitrogen deposition and host plant affect ancestral AMF abundance? We tested these hypotheses and questions using a precipitation and atmospheric nitrogen deposition experiment established in 2007, and by coupling molecular-based estimates of fungal community composition with traditional measurements of AMF biomass within a synthesized guild approach. 2. Methods 2.1. Site description The study site is a coastal sage scrub plant community in the foothills of the Santa Ana Mountains at Loma Ridge, near Irvine, CA (33.7428 N, 117.7048 W; Kimball et al., 2014; Amend et al., 2016; Martiny et al., 2017). The plant community is a mix of native crownsprouting shrubs (e.g. Salvia mellifera, Artemisia californica, and Eriogonum fasciculatum) with interspaces now predominantly filled by invasive grasses (e.g. Bromus rubens, Schismus barbatus and Avena sp.). The climate is Mediterranean, with a wet winter season lasting from November to April and a dry summer from May through October. Soils are Myford Sandy Loam (Kimball et al., 2014), and a characterization of soil chemistry from our samples is provided in Table 2. This site burned in 1914, 1948, 1967, 1998 and 2007. In 2007, we initiated an experiment that manipulated precipitation (3 levels) and nitrogen deposition (2 levels) in a split-plot design. Three plots (18.29  12.19 m each) within each of eight blocks received:(1) reduced precipitation (40% reduction, using ‘rain-out’ shelters that were deployed during storms), (2) added precipitation (40% addition, from captured runoff from shelters, and (3) control plot with unaltered precipitation. We added N to one side of each plot as 2 g m2 of quick-release CaNO3 prior to growing season (November), and 4 g m2 as slow-release CaNO3 a month into the growing season (January, Supplemental Figure 1), resulting in two N treatment levels: added (ambient plus supplemental 10.2 kg N ha1 yr1) and ambient (estimated at 2.56 kg N ha1yr1). The variation in nitrogen deposition spans the typical range of high and low nitrogen deposition sites in southern California. 2.2. Sampling and storage We sampled roots and soils for six plant species across the precipitation and N treatments in September 30, 2015 (dry season) and February 25, 2016 (wet season; Supplemental Figure 1). While our nominal goal was to sample three individual plants from each species in each treatment, we weighted our sampling effort towards A. californica, S. mellifera and B. rubens, which are abundant at this site (Kimball et al., 2014, Table 3). Individual plants were randomly selected at a minimum of 1 m apart to reduce spatial autocorrelation (Hart et al., 2015). We collected approximately 500 mL of bulk roots and soils from the base of each plant. Samples collected during the dry season originated from two blocks and were stored in paper bags at ambient temperatures. Samples collected during the wet season were collected from a single block, stored in Ziplock bags on ice in the field and transported to a 20  C freezer until processed. We sieved each sample with 500 mm mesh, which was bleach-sterilized between samples,

S.E. Weber et al. / Fungal Ecology 40 (2019) 62e71

65

Table 2 Characterization of soil chemistry at Loma Ridge by nitrogen and precipitation treatments (Means ± standard deviation). Nitrogen treatment

Precipitation Treatment

P (ppm)

N (ppm)

C (ppm)

NO 3 as N (ppm)

NH4þ as N (ppm)

pH

Added

Added Ambient Reduced Added Ambient Reduced

20.8 ± 8.4 19.3 ± 11 17.2 ± 8.5 20.9 ± 8.6 18.1 ± 6.5 17.9 ± 6.5

1082 ± 234 1280 ± 348 1109 ± 199 1088 ± 225 1229 ± 133 1157 ± 228

12496 ± 3060 14772 ± 3443 12900 ± 2903 12367 ± 2447 14257 ± 1525 13055 ± 2286

8.8 ± 7.8 26.7 ± 82.2 45.9 ± 48.8 13.4 ± 37.7 6.6 ± 5.6 11.8 ± 18.5

18.5 ± 14.3 15.6 ± 13.2 19.5 ± 13.8 16.9 ± 12.9 20 ± 16.9 16.9 ± 13.8

6.39 ± 0.37 6.41 ± 0.3 6.43 ± 0.26 6.42 ± 0.33 6.26 ± 0.38 6.36 ± 0.34

Ambient

Table 3 Description of host plant species and sampling coverage. ‘Total samples’ refers to the number of soil and root samples collected from beneath a plant species while ‘(DNA)’ refers to the number of root samples collected from each species that were analyzed for their fungal composition using molecular methods. Species

Growth Form

Origin

Total Samples (DNA)

Salvia mellifera Artemisia californica Eriogonum fasciculatum Bromus rubens Schismus barbatus Avena sp.

Shrub Shrub Shrub Grass Grass Grass

California native California native California native Europe Europe/North Africa Europe/North Africa

54 (17) 54 (16) 2 (2) 53 (17) 8 (4) 3 (3)

and collected roots with sterilized forceps. Samples collected in both wet and dry seasons were analyzed for fungal biomass and soil chemistry, but only wet season root samples could be used for molecular analyses. 2.3. Environmental covariates We measured soil pH, available N (NH4 & NO3), total C, N and P of the soil samples (Table 2). We estimated available N by pooling N as NO3 and NH4 from KCl extractable fractions (Keeney and Nelson, 1982; Hofer, 2003). Using a pH meter, we measured pH from a soil paste (Salinity Laboratory, 1954). We extracted P with a bicarbonate solution and quantified concentrations using molybdenum-blue chemistry and 880 nm absorbance on a spectrophotometer (Olsen, 1954). We measured total C and N with Dumas combustion on a Leco Truspec CN Analyzer. For all samples, we measured total C, N and P at the USDA-ARS Soils Lab in Reno, NV. For samples collected in September, we measured pH and KCl extractable N at the lab USDA-ARS, while for winter samples we measured pH and KCl extractable N at the University of California, Davis Analytical laboratory. 2.4. DNA extraction, library construction and sequencing DNA was extracted from homogenized frozen root samples using a Qiagen Plant Minikit (Qiagen, Germantown, MD). Frozen DNA samples were transported on dry ice to the Northern Arizona University Environmental Genetics and Genomics Laboratory (EnGGen). Samples were further purified using PEG-8000 and carboxylated magnetic beads (Rohland and Reich, 2012) and DNA concentrations were determined with Quant-iT PicoGreen dsDNA Reagent (Molecular Probes Inc., Eugene OR, USA) and standardized to 10 ng/mL. Ribosomal DNA PCR amplicons were produced for sequencing using a two-step protocol as suggested by Berry et al. (2011). Samples were amplified for ITS2 using the fungal-specific primers 5.8SFun and ITS4Fun (Taylor et al., 2016), or for the 18S region using the Glomeromycotina-specific primers AML2 (Lee et al., 2008; Dumbrell et al., 2011) and WANDA (Lee et al., 2008; Dumbrell et al., 2011). First-round PCR primers were modified with 50 tails to

facilitate indexing and Illumina sequencing (Alvarado et al., 2018; Supplemental Table 1). First round PCR reactions were performed at three separate dilutions (1/10, 1/100, 1/1000) in 384-well plates. The 10 mL reactions contained the following: 200 nM each primer (Eurofins Genomics, Louisville, KY), 200 mM each dNTP (Phenix Research, Candler, NC), 0.01 U/mL Phusion HotStart II DNA Polymerase (Thermo Fisher Scientific, Waltham, MA), 1X HF Phusion Buffer (Thermo Fisher Scientific), 3.0 mM MgCl2, 6% glycerol, and 1 mL template DNA normalized to c. 1 ng/mL. Cycling conditions were: 2 min at 95  C, followed by 35 cycles of 30 s at 95  C, 30 s at 55  C, 4 min at 60  C. Reactions were checked on an agarose gel, and successful amplifications were pooled by sample. Reaction pools were purified and resuspended in 20 mL Tris-HCl pH 8.0 prior to the second round of PCR. Indexing and flowcell sequences were added during the second round of PCR using primers with sequences matching the universal tails at the 3’ end (Supplemental Table 1). Reaction conditions were the same as the first round of PCR but only one reaction was conducted per sample, reactions contained only 100 nM of each primer, and only 15 total cycles were performed. Indexed PCR products were checked on an agarose gel, purified, quantified by PicoGreen fluorescence, and an equal mass of each sample was combined into a final sample pool. The pool was purified and quantified by qPCR against Illumina DNA Standards (Kapa Biosystems, Wilmington, DE). Sequencing was carried out on a MiSeq Desktop Sequencer (Illumina Inc, San Diego, CA) running in paired end 2  300 mode. Raw sequencing data is available at the NCBI Sequence Read Archive under accession SRP141740. 2.5. Bioinformatics Contaminating PhiX sequence was removed using the akutils phix_filtering command in akutils v1.2 (Andrews, 2016) and locusspecific primer sequences were trimmed with the akutils strip_primers command. ITS2 paired-end reads were merged with the akutils join_paired_reads command. Merged ITS2 sequences and forward read 18S sequences were inspected for quality with FastQC (Andrews, 2010) and trimmed with FastX Toolkit (http:// hannonlab.cshl.edu/fastx toolkit) to remove 3’ low-quality base calls (200 bases 18S, 309 bases ITS2). Demultiplexing and quality filtering was carried out with the split_libraries_fastq.py command

66

S.E. Weber et al. / Fungal Ecology 40 (2019) 62e71

in QIIME 1.9.1 (Caporaso et al., 2010) using a minimum quality threshold of q20, 0 bad characters allowed, and retaining only reads which satisfied these requirements for at least 95% of their length (command options, -q 19 -r 0 -p 0.95). Chimeras were removed with VSEARCH 1.1.1 (Rognes et al., 2016) in using the -uchime_denovo option for 18S or using the -uchime_ref option against the UNITE fungal chimera reference for ITS2 (Nilsson et al., 2015). ITS2 sequences were screened for fungal-appropriate hidden Markov models with ITSx (Bengtsson-Palme et al., 2013). For both loci, sequences were dereplicated on the first 100 bases with prefix_suffix OTU picker in QIIME. OTUs were clustered de novo with a resolution  et al., 2014), corresponding to approximately of d4 in Swarm (Mahe 98% sequence similarity for 18S and 98.7% sequence similarity for ITS2. Taxonomy was assigned to these OTUs with BLAST (Altschul et al., 1990) against the UNITE database for ITS2 (Koljalg et al., € 2014) or the MaarJAM database for 18S (Opik et al., 2010). Reference database sequences were trimmed to the primer sets used with the format_database command in akutils. OTUs constituting less than 0.005% of the total data set were removed (Bokulich et al., 2013). OTU tables were rarified to 2500 for ITS2 data or 4900 reads for 18S data for richness estimates. For tests of differential abundance, OTU counts were transformed by cumulative sum scaling (CSS) normalization (Paulson et al., 2013).

2.6. Guild assignment We assigned families of Glomeromycotina to AMF guilds based on studies detailing functionally relevant morphological differences between AMF families, generally from culturable taxa (Hart and Reader, 2002; Maherali and Klironomos, 2007; Powell et al., 2009; Sikes et al., 2009, 2010; Varela-Cervero et al., 2015, 2016a,b). Because of recent changes in the taxonomy of Glomeromycotina, not all currently described families have been studied for their patterns of biomass allocation. Families that did not contain taxa examined for their allocation to intra- and extraradical hyphae were assigned to the guilds of phylogenetically related families, assuming phylogenetic trait conservatism, supported in part by (Powell et al., 2009). To examine responses of pathogenic and saprotrophic fungal taxa, we assigned OTUs to functional groups based on taxonomy through the online FUNGuild application (“http://www.stbates.org/guilds/app.php”, Nguyen et al., 2015). We filtered away taxa that were not assigned as ‘probable’ or ‘highly probable’, and further curated these assignments for interpretability by considering ‘pathotrophs’, ‘pathotroph-saprotrophs’, ‘pathotroph-symbiotrophs’, ‘saprotrophs’ and ‘saprotrophpathotrophs’ as ‘parasitic fungi’. Additionally, we removed OTUs classified as symbionts belonging to Glomeromycotina to reduce overlap between 18S and ITS datasets.

2.7. Fungal root colonization We estimated percent root length colonized (PRLC) by AM and non-AM fungi from a subset of our fine roots following Vierheilig et al. (1998). Prior to staining AMF structures, we first cleared root tissues by boiling roots in 10% KOH for 3e5 min until blonde, rinsed these roots with de-ionized H2O, acidified and stained AMF structures in boiling 5% India ink:acetic acid for 3 min and destained root tissue in mildly acidified de-ionized H2O until roots were unstained and fungal structures were discernible. Roots were mounted on microscope slides with PVLG. We measured PRLC under a Zeiss Axioskop 2 compound microscope at 200x using a modified version of the random intersect method (Giovannetti and Mosse, 1980).

2.8. AMF extraradical hyphae We estimated AMF hyphal density in soils using a modified version of the membrane filter protocol described in (Allen and MacMahon, 1985). We added approximately 10 g of soil to 500 mL of 0.4% w/v sodium hexametaphosphate:DI H2O detergent solution, and stirred for 5 min (Holden et al., 2013). We then transferred 20 mL of this soil solution into an additional 180 mL of detergent solution and stirred for 2 min. We then pipetted 5 mL of this solution onto gridded 25 mm diameter, 0.2 mm EMD Millipore Nylon Hydrophilic Membrane Filters, vacuum filtered through Fisher Millipore Fritted Glass, and stained hyphae on the filters with acid fuchsin (14:1:1 lactic acid:glycerol:DI H2O with 0.01 acid fuchsin stain). We repeated this last step, resulting in two filters per sample. We then placed these filters onto microscope slides and dried overnight. Once dry, we mounted filters with poly-vinyl alcohol lactic-acid glycerol (PVLG) and dried overnight at 60  C. We counted hyphal length per gram soil from these slides under a Zeiss Axioskop 2 compound microscope at 200x using the gridline length estimate method (Marsh, 1971). 2.9. AMF spores We extracted spores from 2.5 to 5 g soil using a sucrosedetergent centrifugation method (Allen et al., 1979), and then counted by diameter class (25e50, 50e100, 100e200, and 200e300 mm; no spores larger than 300 mm were observed in our samples) under an Olympus SZ40 dissecting scope at 40 magnification. We did not separate AMF spores taxonomically as the morphological distinctions were minimal from these field conditions, but these spore diameter classes were chosen to roughly correspond to rhizophilic AMF (25e100 mm), and ancestral and edaphophilic AMF (100e300 mm). From these binned spore counts we then calculated average spore diameter (Ianson and Allen, 1986). 2.10. Statistical modeling We used generalized linear models (GLMs) to test our three primary hypotheses (Fig. 1A): (i) invasive grasses, experimental drought and nitrogen deposition would alter the richness and abundance of rhizophilic and edaphophilic AMF, (ii) these changes in AMF would affect the relative abundance of parasitic fungi, and (iii) these changes in relative abundances of fungal guilds would alter patterns of fungal biomass allocation. Because we estimated fungal composition from samples collected from one treatment block of our experiment, we did not include treatment blocks as random effects. We considered water and nitrogen treatments numerically, corresponding to their input values (e.g. 1.4, 1.0, 0.6 for 40% added, ambient and 40% reduced water, 12.76 and 2.56 kg N ha1yr1 for added and ambient nitrogen treatments). We standardized environmental covariates prior to inclusion in models. For count data we used Poisson distributions (with log link functions) and Gaussian distributions (with identity link function) when responses were normally distributed (Table 4). Zero-inflated Poisson models were used to model count data when there were many zeros, a common feature of sequencing data (Table 4; zeroinflated models were fitted with ‘zeroinfl’ of the ‘pscl’ package in R; (Zeileis et al., 2008; Jackman et al., 2015). We selected models for parsimony using stepwise AIC model selection (‘stepAIC’ from MASS package in R, Venables and Ripley, 2002). We evaluated contrasts between grass and shrub roots with ‘lsmeans’ of the ‘lsmeans’ package in R (Lenth, 2016). Table 4 gives a complete list of variables used in the full and final models, Table 5 presents posthoc contrasts, Fig. 1B presents hypotheses supported by final

S.E. Weber et al. / Fungal Ecology 40 (2019) 62e71

67

Table 4 Summary of generalized linear models. Variables in bold are significant at p < 0.05, italicized marginally significant at p < 0.10. Variables in parentheses were evaluated, but not retained in final models after model selection. Poisson models use a log link and Gaussian models use an identity link function. Final model output is provided in Supplemental Tables 2-14. Model Probability # Distribution

Response

Predictors

1

Water, Host Plant, N, Water X N, pH

4

Zero-Inflated Poisson Poisson

5

Poisson

Edaphophilic AMF Abundance Rhizophilic AMF Abundance Ancestral AMF Abundance Parasitic Fungi Abundance AMF PRLC

6

Poisson

Non-AMF PRLC

7

Poisson

8

Gaussian

AMF Soil Hyphal Density Average AMF Spore Diameter

2 3

Zero-Inflated Poisson Poisson

Water, Water X N, Host Plant, pH, N Water, N, Host Plant, Water X N, pH Water, N, Host Plant, Host Plant X N, Rhizophilic AMF Abundance, Ancestral AMF Abundance (Water X N, pH, Edaphophilic AMF Abundandance) Water, N, Water X N, Host Plant, pH, Rhizophilic AMF Abundance (Edaphophilic AMF Abundance, Ancestral AMF Abundance) N, Host Plant, Host Plant X N, pH, AMF PRLC (Water, Water X N, Edaphophilic AMF Abundance, Ancestral AMF Abundance, Parasitic Fungi Abundance) Water, N, AMF PRLC, Edaphophilic AMF Abundance, Rhizophilic AMF Abundance, pH (Water X N, Ancestral AMF Abundance) Rhizophilic AMF Abundance, Ancestral AMF Abundance, Edaphophilic AMF Abundance, pH, (Water, N, Water X N, Host Plant, AMF PRLC)

models, and full output of final models can be found in Supplemental Tables 2-9. 3. Results 3.1. 18S (AMF - Glomeromycotina) We observed 37 recognized taxa (121 OTUs) with an average of 330 ± 105 (SD) reads per sample. The OTUs belonged to 4 orders, 8 families and 8 genera. On average 245 ± 77 reads belonged to taxa in Glomeromycotina, while 85 ± 40 reads were not assigned to taxonomy. The most abundant families in our samples were the Glomeraceae (21 recognized taxa, 84 OTUs) with 51 ± 6% of reads per sample, followed by the Paraglomeraceae (3 recognized taxa, 8 OTUs) with 29 ± 25% of reads per sample and the Claroideoglomeraceae (4 recognized taxa, 10 OTUs) with 10 ± 4% of reads per sample. We placed these OTUs into three guilds based on our earlier descriptions (Table 1), with rhizophilic AMF (65 ± 8% of reads per sample), edaphophilic AMF (7 ± 4% of reads per sample), ancestral AMF (2 ± 2% of reads per sample). Percent abundances of these guilds across treatments are presented in Supplemental Table 10, richness for 18S at the genus level is presented in Supplemental Table 11, and a full list of observed taxa and their guild placement is provided in Supplemental Table 13.

recognized taxa, 160 OTUs, 43 ± 6% of reads per sample), Sordariomycetes (50 recognized taxa, 97 OTUs, 25 ± 7% of reads per sample) and Eurotiomycetes (22 recognized taxa, 29 OTUs, 6 ± 3% of reads per sample). After using FUNGuild, these reads and OTUs were assigned to trophic based functional groups, saprotrophs (69 recognized taxa, 111 OTUs, 38 ± 14% of reads per sample), pathogens (49 recognized taxa & 80 OTUs total, with 41 ± 15% of reads per sample), and non-glomeromycotinan symbionts (12 recognized taxa and 15 OTUs total, 2 ± 2% of reads per sample). Family-level fungal richness observed with ITS2 is provided in Supplemental Table 12, while a full list of taxa assigned to functional group by FUNGuild is provided in Supplemental Table 14. 3.3. Fungal activity We found an average AMF root colonization of 42 ± 19% (SD) across samples. Most AMF root colonization was hyphal (29 ± 20%). Non-AMF root colonization averaged 36.5 ± 20% across samples. We also found that root spaces were frequently co-colonized by both AM and non-AM fungi (17 ± 12% across samples). Soil hyphal length of AM fungi averaged 21 ± 11 m/g soil across samples. Across our samples the most abundant AMF spores were those from 50 to 100 mm in diameter (16 ± 11 spores per sample), resulting in an average spore diameter of 82 ± 12.5 mm across samples. 3.4. Effect of host plant

3.2. Internal transcribed spacer 2 (broader fungi) We found 203 recognized taxa (451 OTUs) with an average of 622 ± 163 (SD) reads per sample belonging to 5 phyla, 14 classes, 43 orders, 73 families and 119 genera. Our most abundant phylum was Ascomycota (159 recognized taxa, 371 OTUs), with an average of 564 ± 144 reads per sample, followed by Basidiomycota (36 recognized taxa, 53 OTUs) with 38 ± 23 reads per sample. Of these groups, the most abundant classes were the Dothideomycetes (57 Table 5 Summary of life-history contrasts. Contrasts were evaluated on models of same name as responses. All contrasts are presented as between grasses and shrubs. Grass - Shrub contrasts

Estimate

SE

Z ratio

P value

Edaphophilic AMF Abundance Rhizophilic AMF Abundance Parasitic Fungi Abundance AMF PRLC NAMF PRLC

2.919 0.193 0.103 0.268 0.241

1.455 0.019 0.021 0.046 0.030

2.005 10.042 4.915 5.803 7.922

0.045 <0.0001 <0.0001 <0.0001 <0.0001

Edaphophilic AMF were more abundant in shrub roots than in grasses (Table 5), whereas rhizophilic AMF were more abundant in grass roots than in shrubs (Table 5). Ancestral AMF abundance did not differ significantly between grass and shrub roots. Parasitic fungi were more abundant in grass roots than in shrub roots (Table 5). AMF PRLC was higher in shrub than in grass roots while non-AMF PRLC was higher in grass roots than in shrubs (Table 5). AMF hyphal density and average spore diameter did not significantly differ between grass and shrub roots (models 8 and 9). 3.5. Effect of altered precipitation The abundance of edaphophilic and ancestral AMF responded positively to increasing amounts of water (Z ¼ 3.6, p < 0.001; Z ¼ 2.2, p ¼ 0.025; models 1 and 3). Rhizophilic AMF abundance responded positively to increased water (Z ¼ 4.2, p < 0.0001, model 2) and this positive response was mediated to some degree by a

68

S.E. Weber et al. / Fungal Ecology 40 (2019) 62e71

positive interaction between water and nitrogen treatments (Z ¼ 6.5, p < 0.0001, model 2). The abundance of parasitic fungi responded positively to increasing water (Z ¼ 4.8, p < 0.0001, model 4). Additional water increased the percent root-length colonized (PRLC) by AMF (Z ¼ 5.6, p < 0.0001, model 5) however there was a negative interaction between water and nitrogen altering AMF PRLC responses to water (Z ¼ 10.01, p < 0.0001, model 5). NonAMF PRLC was not correlated with water availability (model 6). AMF hyphal density responded positively to water availability (Z ¼ 0.07, p ¼ 0.015, model 7). Average AMF spore diameter did not respond significantly to water (model 9). 3.6. Effect of altered nitrogen deposition While edaphophilic AMF abundance did not significantly respond to altered nitrogen, there was a marginal positive correlation (Z ¼ 1.9, p ¼ 0.055, model 1). While rhizophilic AMF abundance did not respond directly to altered nitrogen (Z ¼ 1.9, p ¼ 0.052, model 2), rhizophilic AMF abundance did positively respond to the interaction between water and nitrogen (Z ¼ 6.5, p < 0.0001, model 2). Ancestral AMF abundance responded positively to additional nitrogen (Z ¼ 3.5, p < 0.001, model 3). The abundance of parasitic fungi had a negative correlation with additional nitrogen (Z ¼ 5.8, p¼<0.0001, model 4), mediated in part by the interaction between nitrogen and host plant (Z ¼ 6.3, p¼<0.0001, model 4). AMF PRLC responded positively to increasing nitrogen (Z ¼ 8.2, p < 0.0001, model 5), but negatively to the interaction between nitrogen and water treatments (Z ¼ 10.7, p < 0.0001, model 5). Non-AMF PRLC responded negatively to increasing nitrogen (Z ¼ 12.5, p < 0.0001, model 6), and to the interaction between host plant and nitrogen. AMF hyphal density negatively correlated to increasing nitrogen (Z ¼ 3.6, p < 0.0001, model 7). AMF average spore diameter did not significantly respond to nitrogen (model 9). 3.7. Biotic interactions between AMF and parasitic fungi The abundance of parasitic fungi positively covaried with rhizophilic AMF abundance (Z ¼ 8.7, p < 0.0001, model 4) and negatively with the abundance of ancestral AMF (Z ¼ 4.2, p < 0.0001, model 4). Non-AM fungal PRLC positively covaried with AMF PRLC (Z ¼ 7.9, p < 0.0001, model 6). 3.8. Biomass allocation responses to fungal composition Percent root length colonized by AM fungi was positively correlated with the relative abundance of rhizophilic AMF (Z ¼ 4.4, p < 0.0001, model 5). The density of AMF hyphae in soils surrounding plant roots was positively correlated with the relative abundance of edaphophilic AMF in plant roots (Z ¼ 2.4, p ¼ 0.018, model 7), negatively correlated with the relative abundance of rhizophilic AMF in plant roots (Z ¼ 3.5, p ¼ 0.001, model 7) and negatively correlated with overall AMF root colonization (Z ¼ 4.2, p < 0.0001, model 7). Non-AMF PRLC did not respond to fungal composition. The average diameter of AMF spores in soil surrounding plant roots was negatively correlated with the relative abundance of rhizophilic AMF in plant roots (t ¼ 2.6, p ¼ 0.011, model 8). 4. Discussion We evaluated how precipitation, nitrogen and host plant life histories affect AMF and parasitic fungi abundance in plant roots during the winter growing season within coastal sage scrub in

southern California. Our results suggest that prolonged drought, nitrogen deposition and annual grass invasion interact to cause significant changes in the composition and allocation of biomass in soil fungal communities (Fig. 1B). Drought reduced AMF root colonization and extraradical hyphal density, as well as the relative abundances of rhizophilic, edaphophilic, ancestral AMF and parasitic fungi. Nitrogen addition altered the allocation of biomass by AMF, increasing root colonization while decreasing the density of extraradical hyphae. Additionally, we found that invasive grasses host higher abundances of rhizophilic AMF than native shrubs, while native shrubs host more edaphophilic AMF. We also found that invasive grasses host higher abundances of parasitic fungi, and higher rates of root colonization by non-AM fungi. Rather than a negative correlation between rhizophilic AMF and pathogenic and saprotrophic fungi, we found that these fungi positively covaried. We also found that the abundance of edaphophilic AMF was positively correlated to the density of extraradical hyphae, while the abundance of rhizophilic AMF was positively correlated with root colonization and negatively correlated with the average diameter of AMF spores in rhizosphere soils. 4.1. Effect of host plant We hypothesized that shrubs would host a higher abundance of edaphophilic AMF because of their greater need for nutrient acquisition from AMF, and we found that shrubs hosted a greater abundance of edaphophilic AMF than grasses. Additionally, we found that edaphophilic AMF abundance was positively correlated with AMF extraradical hyphae. This suggests that shrubs do indeed host more edaphophilic AMF than grasses because of the importance of these fungi for nutrient acquisition, corroborating past observations of spore abundance response to host plant (EgertonWarburton and Allen, 2000) and implications from greenhouse work (Maherali and Klironomos, 2007; Powell et al., 2009; Sikes et al., 2010). We also hypothesized that grasses would host more rhizophilic AMF to protect against infection from parasitic fungi, and we found that grasses did host a higher abundance of rhizophilic AMF, and the abundance of these AMF was positively correlated with AMF root colonization. However, grasses also hosted higher abundances of parasitic fungi, higher rates of root colonization by non-AMF and had lower rates of root colonization by AMF than shrubs. Additionally, parasitic fungal abundance and rhizophilic AMF abundance positively covaried, as did root colonization by non-AMF and AMF (17 ± 12% of root space were cocolonized by AMF and non-AMF). This suggests that grass roots are indeed more susceptible to infection from parasitic fungi, but whether grasses associate with rhizophilic AMF to ameliorate this susceptibility is not immediately clear. One possibility is that these rhizophilic AMF are indeed competing with these parasitic fungi, and if rhizophilic AMF were not present parasitic fungi would be even more abundant in these grass roots. Alternatively, the fine roots of grasses may simply be more easily colonized by both groups of fungi, or colonization by one group may facilitate colonization by another. While distinguishing between these three possibilities was beyond the scope of the current study, past research has examined interactions in a controlled experiment and found that Glomus spp. (a rhizophilic AMF genus) reduced colonization by parasitic Fusarium oxysporum and Pythium sp. in Plantago lanceolata roots (Sikes et al., 2010). Future field and greenhouse experiments could help to clarify the importance of these proposed mechanisms for driving patterns of AMF abundance and biomass allocation in local ecosystems. These host effects on fungi are also worth considering within the broader context of invasive species dynamics in this ecosystem. In particular, the greater abundance of rhizophilic AMF and

S.E. Weber et al. / Fungal Ecology 40 (2019) 62e71

parasitic fungi in the roots of invasive grasses suggests that grass invasion has altered the composition of fungi. These grasses are very abundant in shrub interspaces, and it is likely that the different fungi that they host influence fungal population dynamics. However, a more rigorous test of the influence of grass invasion on this community could evaluate abundances of rhizophilic AMF and parasitic fungi in native shrub and forb roots in sites where invasive grasses are present and where they are absent. 4.2. Altered precipitation We hypothesized that edaphophilic AMF were more sensitive to drought than other AMF because of their relatively high allocation of biomass to extraradical hyphae and that rhizophilic AMF were more tolerant to drought because of their high allocation of biomass to root colonization. While we did find differences among these groups in their sensitivities to drought, it is important to note that all fungi analyzed in this study were impacted to some degree by water availability, highlighting the primary role of water in shaping fungal communities in semi-arid ecosystems. The relative abundances of edaphophilic, ancestral and rhizophilic AMF in plant roots all decreased with reduced water availability. Additionally, the relative abundances of parasitic fungi, AMF root colonization and AMF extraradical hyphal density were also lower under reduced water availability. Therefore, we conclude that edaphophilic AMF are not necessarily more susceptible to drought than other AMF, nor are rhizophilic AMF more tolerant to drought than other AMF. 4.3. Atmospheric nitrogen deposition We hypothesized that additional nitrogen from atmospheric nitrogen deposition would reduce edaphophilic AMF abundance, but instead found that these fungi had a marginally significant positive response to nitrogen addition. This increase in edaphophilic AMF with additional nitrogen contrasts with previous work in this system examining the responses of AMF using spore abundances, but the positive response of rhizophilic AMF in plant roots is consistent with previous findings (Egerton-Warburton and Allen, 2000). While additional nitrogen did not reduce the relative abundance of edaphophilic AMF in plant roots, it did reduce the density of AMF extraradical hyphae. Additionally, the density of AMF extraradical hyphae was positively correlated with the relative abundance of edaphophilic AMF. This suggests that atmospheric nitrogen deposition does not reduce the abundance of edaphophilic AMF but may reduce the biomass that edaphophilic AMF allocate to extraradical hyphae, and likely the extent to which these fungi transfer nutrients, specifically phosphorous, to host plants (Treseder et al., 2018). Our related hypothesis that rhizophilic AMF would respond positively to additional nitrogen was supported by these data but may also depend on interactions with water availability. For example, AMF PRLC was higher under added nitrogen, but there was a significant negative interaction between nitrogen and water. Additionally, AMF PRLC was positively correlated with rhizophilic AMF abundance, and rhizophilic AMF abundance showed a significant positive interaction between nitrogen and water treatments (with a marginally positively main effect of nitrogen addition). This suggests that while the overall extent of root colonization by AMF declines when nitrogen is added under higher levels of soil moisture, a higher proportion of the AMF within roots are rhizophilic. Because adding nitrogen and water simultaneously reduces two limiting factors to plant growth, this likely reduces plant dependence on AMF for nutrient acquisition, resulting in the observed greater colonization by those AMF species providing the least

69

amount of nutrient transfer (Johnson et al., 2003; Sikes et al., 2010). The abundance of parasitic fungi in plant roots decreased with additional nitrogen, as did root colonization by non-AM fungi. Moreover, this negative response of parasitic fungal abundance and non-AMF root colonization to nitrogen was mediated by the interaction between host plant and nitrogen: additional nitrogen in grass roots was associated with a lower abundance of parasitic fungi and non-AMF root colonization, compared to additional nitrogen in shrub roots. This suggests that the differences in root structure between grasses and shrubs may mediate the responses of the fungal community to added nitrogen. In particular, grasses with fine roots may better defend against pathogens when more nitrogen is available, while shrubs may grow more fine roots to uptake the additional nitrogen, increasing their susceptibility to these fungi. 4.4. Caveats The results of this study relied on the ability to assign fungi to functional guilds, based either on a combination of morphological differences and existing databases (FUNGuild application). The characterization of AMF guilds relied heavily on a set of papers that described differences in biomass allocation patterns between culturable AMF taxa (Table 1). Because most AMF taxa have not been cultured, or if cultured not examined for differences in biomass allocation, these results must be interpreted with caution. Additionally, the larger number of uncultured and undescribed fungal taxa globally, and their often context-dependent functional variation, functional classifications from FUNGuild are only a coarse description of fungal ecological diversity. This uncertainty regarding fungal functional classifications likely introduced a degree of ambiguity to our results. Although we have used the best currently available methods, our understanding of fungal responses to environmental change will increase substantially with improved knowledge of fungal life histories. 4.5. Implications Our findings imply that increasing aridity, continued atmospheric nitrogen deposition and invasion by annual grasses will reduce the abundance of edaphophilic AMF in coastal sage scrub, leading to potential feedbacks between host plants and soil biota. Given predictions of increased aridity and nitrogen deposition in the southwestern US, our study suggests that rhizophilic AMF will become more abundant in plant roots, and AMF biomass will increase in plant roots relative to surrounding soil. The predicted reduction in overall interactions between plants and AMF in coastal sage scrub, and associated AMF nutrient foraging through extraradical hyphae, may further impact ecosystem processes. However, this alteration may be most prominent during dry years and the AMF community may be resilient to drought and rebound in wet years, as implied by the resilience of litter decomposing fungi to drought at this site (Martiny et al., 2017). Our findings also suggest that atmospheric nitrogen deposition will reduce the abundance of edaphophilic AMF and the density of nutrient foraging hyphae, while increasing root colonization and potentially increasing the abundance of rhizophilic AMF, corroborating past work in coastal sage scrub and in other systems (Egerton-Warburton and Allen, 2000; Treseder and Allen, 2002; Johnson et al., 2003; Egerton-Warburton et al., 2007; Johnson, 2010). Persistence of invasive annual grasses in coastal sage scrub may increase the abundance of rhizophilic AMF at the expense of edaphophilic AMF taxa. While we found parasitic fungi positively covaried with rhizophilic AMF at the plant level, the scope of this project could not evaluate the direction of interactions between

70

S.E. Weber et al. / Fungal Ecology 40 (2019) 62e71

these fungi within roots. While we describe broad patterns, the finer ecological implications of these responses within the fungal community to multiple, coinciding global change drivers is still unclear. Currently there is little information regarding the natural history of most arbuscular mycorrhizal fungi specifically, and fungi in general. Progress towards understanding AMF community ecology in an increasingly changing world will require an increased effort to connect molecular data to basic biological information within a unified framework (Peay, 2014). Acknowledgements The authors greatly appreciate revisions from two anonymous reviewers. SW thanks Michala Phillips, Tye Morgan, Eufrocina Palaganas, Minjin Yim, Courtney G. Collins, Teresa Bohner, Linh Anh Cat, Edith Allen, Mia Maltz and all members of the Allen labs for laboratory help and constructive criticism. SW thanks Chris Karounos and Jon Botthoff for assistance in collecting samples. This study was financially supported by the Agricultural Experiment Station Hatch Projects (CA-R-PPA-6689-H), the Center for Conservation Biology at UCR and the California Native Plant Society (G.Ledyard Stebbins Award) for financial support. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.funeco.2018.11.008. References Allen, E.B., Allen, M.F., 1986. Water relations of xeric grasses in the field: interactions of mycorrhizas and competition. New Phytol. 104, 559e571. Allen, E.B., Allen, M.F., Helm, D.J., Trappe, J.M., Molina, R., Rincon, E., 1995. Patterns and regulation of mycorrhizal plant and fungal diversity. Plant Soil 170, 47e62. Allen, M.F., MacMahon, J.A., 1985. Impact of disturbance on cold desert fungi: comparative microscale dispersion patterns. Pedobiologia 28, 215e224. Allen, M.F., Moore, T.S., Christensen, M., Stanton, N., 1979. Growth of vesiculararbuscular-mycorrhizal and nonmycorrhizal Bouteloua gracilis in a defined medium. Mycologia 71, 666e669. ~ o-like teleconnection increases California Allen, R.J., Luptowitz, R., 2017. El Nin precipitation in response to warming. Nat. Commun. 8, 16055. Altschul, S.F., Gish, W., Miller, W., Myers, E.W., Lipman, D.J., 1990. Basic local alignment search tool. J. Mol. Biol. 215, 403e410. Alvarado, P., Teixeira, M. de M., Andrews, L., Fernandez, A., Santander, G., Doyle, A., Perez, M., Yegres, F., Barker, B.M., 2018. Detection of Coccidioides posadasii from xerophytic environments in Venezuela reveals risk of naturally acquired coccidioidomycosis infections. Emerg. Microb. Infect. 7, 46. Amend, A.S., Martiny, A.C., Allison, S.D., Berlemont, R., Goulden, M.L., Lu, Y., Treseder, K.K., Weihe, C., Martiny, J.B.H., 2016. Microbial response to simulated global change is phylogenetically conserved and linked with functional potential. ISME J. 10, 109e118. Andrews, L.V., 2016. Akutils-v1.2: akutils-v1.2: Facilitating Analyses of Microbial Communities through QIIME. https://doi.org/10.5281/zenodo.61581. Andrews, S., 2010. FastQC: a Quality Control Tool for High Throughput Sequence Data. Auge, R.M., 2001. Water relations, drought and vesicular-arbuscular mycorrhizal symbiosis. Mycorrhiza 11, 3e42. Bengtsson-Palme, J., Ryberg, M., Hartmann, M., Branco, S., Wang, Z., Godhe, A., De nchez-García, M., Ebersberger, I., de Sousa, F., Amend, A., Wit, P., Sa Jumpponen, A., Unterseher, M., Kristiansson, E., Abarenkov, K., Bertrand, Y.J.K., Sanli, K., Eriksson, K.M., Vik, U., Veldre, V., Nilsson, R.H., 2013. Improved software detection and extraction of ITS1 and ITS2 from ribosomal ITS sequences of fungi and other eukaryotes for analysis of environmental sequencing data. Methods Ecol. Evol. 4, 914e919. Bokulich, N.A., Subramanian, S., Faith, J.J., Gevers, D., Gordon, J.I., Knight, R., Mills, D.A., Caporaso, J.G., 2013. Quality-filtering vastly improves diversity estimates from Illumina amplicon sequencing. Nat. Methods 10, 57e59. Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K., ~ a, A.G., Goodrich, J.K., Gordon, J.I., Huttley, G.A., Kelley, S.T., Fierer, N., Pen Knights, D., Koenig, J.E., Ley, R.E., Lozupone, C. a., Mcdonald, D., Muegge, B.D., Pirrung, M., Reeder, J., Sevinsky, J.R., Turnbaugh, P.J., Walters, W. a., Widmann, J., Yatsunenko, T., Zaneveld, J., Knight, R., 2010. QIIME Allows Analysis of HighThroughput Community Sequencing Data, vol. 7. Nature Publishing Group, pp. 335e336. Chaudhary, V.B., O'Dell, T.E., Rillig, M.C., Johnson, N.C., 2014. Multiscale patterns of

arbuscular mycorrhizal fungal abundance and diversity in semiarid shrublands. Fungal Ecol. 12, 32e43. Crowther, T.W., Maynard, D.S., Crowther, T.R., Peccia, J., Smith, J.R., Bradford, M.A., 2014. Untangling the fungal niche: the trait-based approach. Front. Microbiol. 5, 579. € Davison, J., Opik, M., Daniell, T.J., Moora, M., Zobel, M., 2011. Arbuscular mycorrhizal fungal communities in plant roots are not random assemblages. FEMS Microbiol. Ecol. 78, 103e115. Dumbrell, A.J., Ashton, P.D., Aziz, N., Feng, G., Nelson, M., Dytham, C., Fitter, A.H., Helgason, T., 2011. Distinct seasonal assemblages of arbuscular mycorrhizal fungi revealed by massively parallel pyrosequencing -Supplement. New Phytol. 190, 794e804. Egerton-Warburton, L.M., Allen, E.B., 2000. Shifts in arbuscular mycorrhizal communities along an anthropogenic nitrogen deposition gradient. Ecol. Appl. 10, 484e496. Egerton-Warburton, L.M., Johnson, N.C., Allen, E.B., 2007. Mycorrhizal community dynamics following nitrogen fertilization: a cross-site test in five grasslands. Ecol. Monogr. 77, 527e544. Eissenstat, D.M., Kucharski, J.M., Zadworny, M., Adams, T.S., Koide, R.T., 2015. Linking root traits to nutrient foraging in arbuscular mycorrhizal trees in a temperate forest. New Phytol. 208, 114e124. Giovannetti, M., Mosse, B., 1980. An evaluation of techniques for measuring vesicular arbuscular mycorrhizal infection in roots. New Phytol. 84, 489e500. Hart, M.M., Aleklett, K., Chagnon, P.-L., Egan, C., Ghignone, S., Pickles, B.J., Waller, L., € Lekberg, Y., Opik, M., 2015. Navigating the labyrinth: a guide to sequence-based, community ecology of arbuscular mycorrhizal fungi. New Phytol. https:// doi.org/10.1111/nph.13340. Hart, M.M., Reader, R.J., 2002. Does percent root length colonization and soil hyphal length reflect the extent of colonization for all AMF? Mycorrhiza 12, 297e301. Hodge, A., Storer, K., 2014. Arbuscular mycorrhiza and nitrogen: implications for individual plants through to ecosystems. Plant Soil 386, 1e19. Hofer, S., 2003. QuikChem Method 12-107-06-2-A: Determination of Ammonia (Salicylate) in 2 M KCl Soil Extracts by Flow Injection Analysis. Lachat Instrum., Loveland, CO. Holden, S.R., Gutierrez, A., Treseder, K.K., 2013. Changes in soil fungal communities, extracellular enzyme activities, and litter decomposition across a fire chronosequence in alaskan boreal forests. Ecosystems 16, 34e46. Ianson, D.C., Allen, M.F., 1986. The effects of soil texture on extraction of vesiculararbuscular mycorrhizal fungal spores from arid sites. Mycologia 78, 164e168. Jackman, S., Tahk, A., Zeileis, A., Maimone, C., Fearon, J., Jackman, M.S., 2015. Package “pscl. https://cran.r-project.org/web/packages/pscl/pscl.pdf. Johnson, N.C., 2010. Resource stoichiometry elucidates the structure and function of arbuscular mycorrhizas across scales. New Phytol. 185, 631e647. Johnson, N.C., 1993. Can fertilization of soil select less mutualistic mycorrhizae? Ecol. Appl. 3, 749e757. Johnson, N.C., Rowland, D.L., Corkidi, L., Egerton-Warburton, L.M., Allen, E.B., 2003. Nitrogen enrichment alters mycorrhizal allocation at five mesic to semiarid grasslands. Ecology 84, 1895e1908. Keeney, D., Nelson, D.W., 1982. NitrogendInorganic Forms. Methods of Soil Analysis. Part 2. Chemical and Microbiological Properties, pp. 643e698. Kimball, S., Goulden, M.L., Suding, K.N., Parker, S., 2014. Altered water and nitrogen input shifts succession in a southern California coastal sage community. Ecol. Appl. 24, 1390e1404. Koljalg, U., Nilsson, R.H., Abarenkov, K., Tedersoo, L., Taylor, A.F.S., Bahram, M., et al., 2014. Towards a unified paradigm for sequence-based identification of fungi. Mol. Ecol. 22, 5271e5277. Lee, J., Lee, S., Young, J.P.W., 2008. Improved PCR primers for the detection and identification of arbuscular mycorrhizal fungi. FEMS Microbiol. Ecol. 65, 339e349. Lenth, R., 2016. Least-squares means: the R package lsmeans. J. Stat. Software 69, 1e33.  Varela-Cervero, S., Vasar, M., Opik, €  pez-García, A.,  nLo M., Barea, J.M., Azco Aguilar, C., 2017. Plant traits determine the phylogenetic structure of arbuscular mycorrhizal fungal communities. Mol. Ecol. 26, 6948e6959. , F., Rognes, T., Quince, C., de Vargas, C., Dunthorn, M., 2014. Swarm: robust Mahe and fast clustering method for amplicon-based studies. Peer J. 2, e593. Maherali, H., Klironomos, J.N., 2007. Influence of phylogeny on fungal community assembly and ecosystem functioning. Science 316, 1746e1748. Marsh, B.A., 1971. Measurement of length in random arrangements of lines. J. Appl. Ecol. 8, 265e267. Martiny, J.B.H., Martiny, A.C., Weihe, C., Lu, Y., Berlemont, R., Brodie, E.L., Goulden, M.L., Treseder, K.K., Allison, S.D., 2017. Microbial legacies alter decomposition in response to simulated global change. ISME J. 11, 1e10. Nguyen, N.H., Song, Z., Bates, S.T., Branco, S., Tedersoo, L., Menke, J., Schilling, J.S., Kennedy, P.G., 2015. FUNGuild: an open annotation tool for parsing fungal community datasets by ecological guild. Fungal Ecol. 20, 241e248. Nilsson, R.H., Tedersoo, L., Ryberg, M., Kristiansson, E., Hartmann, M., Unterseher, M., Porter, T.M., Bengtsson-Palme, J., Walker, D.M., de Sousa, F., ~ljalg, U., Edgar, R.C., Abarenkov, K., Gamper, H.A., Larsson, E., Larsson, K.-H., Ko 2015. A comprehensive, automatically updated fungal ITS sequence dataset for reference-based chimera control in environmental sequencing efforts. Microb. Environ. 30, 145e150. Olsen, S.R., 1954. Estimation of Available Phosphorus in Soils by Extraction with Sodium Bicarbonate. United States Department of Agriculture, Washington. € Opik, M., Vanatoa, A., Vanatoa, E., Moora, M., Davison, J., Kalwij, J.M., Reier, Ü.,

S.E. Weber et al. / Fungal Ecology 40 (2019) 62e71 Zobel, M., 2010. The online database MaarjAM reveals global and ecosystemic distribution patterns in arbuscular mycorrhizal fungi (Glomeromycota). New Phytol. 188, 223e241. Paulson, J.N., Stine, O.C., Bravo, H.C., Pop, M., 2013. Differential abundance analysis for microbial marker-gene surveys. Nat. Methods 10, 1200e1202. Peay, K.G., 2014. Back to the future: natural history and the way forward in modern fungal ecology. Fungal Ecol. 12, 4e9. Powell, J.R., Parrent, J.L., Hart, M.M., Klironomos, J.N., Rillig, M.C., Maherali, H., 2009. Phylogenetic trait conservatism and the evolution of functional trade-offs in arbuscular mycorrhizal fungi. Proc. Biol. Sci. 276, 4237e4245. , F., 2016. VSEARCH: a versatile Rognes, T., Flouri, T., Nichols, B., Quince, C., Mahe open source tool for metagenomics. Peer J. PrePrints 1e30. Rohland, N., Reich, D., 2012. Cost-effective, high-throughput DNA sequencing libraries for multiplexed target capture. Genome Res. 22, 939e946. Salinity Laboratory, U., 1954. pH Reading of Saturated Soil Paste. Seager, R., Vecchi, G.A., 2010. Greenhouse warming and the 21st century hydroclimate of southwestern North America. Proc. Natl. Acad. Sci. U. S. A 107, 21277e21282. Sikes, B.A., Cottenie, K., Klironomos, J.N., 2009. Plant and fungal identity determines pathogen protection of plant roots by arbuscular mycorrhizas. J. Ecol. 97, 1274e1280. Sikes, B.A., Powell, J.R., Rillig, M.C., 2010. Deciphering the relative contributions of multiple functions within plant-microbe symbioses. Ecology 91, 1591e1597. Spatafora, J.W., Chang, Y., Benny, G.L., Lazarus, K., Smith, M.E., Berbee, M.L., Bonito, G., Corradi, N., Grigoriev, I., Gryganskyi, A., James, T.Y., O'Donnell, K., Roberson, R.W., Taylor, T.N., Uehling, J., Vilgalys, R., White, M.M., Stajich, J.E., 2016. A phylum-level phylogenetic classification of zygomycete fungi based on genome-scale data. Mycologia 108, 1028e1046. Stutz, J.C., Copeman, R., Martin, C.A., Morton, J.B., 2000. Patterns of species composition and distribution of arbuscular mycorrhizal fungi in arid regions of southwestern North America and Namibia, Africa. Can. J. Bot. 78, 237e245. Stutz, J.C., Morton, J.B., 1996. Successive pot cultures reveal high species richness of arbuscular endomycorrhizal fungi in arid ecosystems. Can. J. Bot. 74,

71

1883e1889. Taylor, D.L., Walters, W.A., Lennon, N.J., Bochicchio, J., Krohn, A., Caporaso, J.G., Pennanen, T., 2016. Accurate estimation of fungal diversity and abundance through improved lineage-specific primers optimized for Illumina amplicon sequencing. Appl. Environ. Microbiol. 82, 7217e7226. Treseder, K.K., Allen, E.B., Egerton-Warburton, L.M., Hart, M.M., Klironomos, J.N., Maherali, H., Tedersoo, L., 2018. Arbuscular mycorrhizal fungi as mediators of ecosystem responses to nitrogen deposition: a trait-based predictive framework. J. Ecol. 106, 480e489. Treseder, K.K., Allen, M.F., 2002. Direct nitrogen and phosphorus limitation of arbuscular mycorrhizal fungi: a model and field test. New Phytol. 155, 507e515.  Barea, J.M., Azco  pez-García, A., n-Aguilar, C., 2016a. Differences Varela-Cervero, S., Lo in the composition of arbuscular mycorrhizal fungal communities promoted by different propagule forms from a Mediterranean shrubland. Mycorrhiza 26, 489e496.  Barea, J.M., Azco pez-García, A., n-Aguilar, C., 2016b. Spring to Varela-Cervero, S., Lo autumn changes in the arbuscular mycorrhizal fungal community composition in the different propagule types associated to a Mediterranean shrubland. Plant Soil 408, 1e14. €  n-Aguilar, C., Varela-Cervero, S., Vasar, M., Davison, J., Barea, J.M., Opik, M., Azco 2015. The composition of arbuscular mycorrhizal fungal communities differs among the roots, spores and extraradical mycelia associated with five Mediterranean plant species. Environ. Microbiol. https://doi.org/10.1111/ 1462e2920.12810. Venables, W.N., Ripley, B.D., 2002. Modern Applied Statistics with S. , Y., 1998. Ink and vinegar, a simple Vierheilig, H., Coughlan, A.P., Wyss, U., Piche staining technique for arbuscular-mycorrhizal fungi. Appl. Environ. Microbiol. 64, 5004e5007. Worchel, E.R., Giauque, H.E., Kivlin, S.N., 2013. Fungal symbionts alter plant drought response. Microb. Ecol. 65, 671e678. Zeileis, A., Kleiber, C., Jackman, S., 2008. Regression models for count data in R. J. Stat. Software 27, 1e25.