The role of phosphorus, magnesium and potassium availability in soil ...

3 downloads 263 Views 1002KB Size Report
nutrients to their plant symbionts is responsive to plant demand. Nor is it ... demonstrate a small but significant increase in the weathering rate of apatite in the ...
Research

The role of phosphorus, magnesium and potassium availability in soil fungal exploration of mineral nutrient sources in Norway spruce forests Nicholas P. Rosenstock1*, Christoffer Berner2*, Mark M. Smits3, Pavel Kram4 and H akan Wallander5 Center for Environmental and Climate Research, Lund University, SE-22362 Lund, Sweden; 2Centre for Ecology and Evolution in Microbial model Systems – EEMiS, Linnaeus University,

1

SE-39182 Kalmar, Sweden; 3Center for Environmental Sciences, Hasselt University, Building D, Agoralaan, Diepenbeek, 3590 Limburg, Belgium; 4Czech Geological Survey, Klarov 3, 118 21, Prague 1, Czech Republic; 5MEMEG, Department of Biology, Lund University, SE-22362 Lund, Sweden

Summary Author for correspondence: Nicholas P. Rosenstock Tel: + 46 76 1488184 Email: [email protected] Received: 4 June 2015 Accepted: 2 February 2016

New Phytologist (2016) doi: 10.1111/nph.13928

Key words: ectomycorrhizal fungi (EMF), ergosterol, fungal community composition, hyphae, minerals, Norway spruce (Picea abies), soil nutrients.

 We investigated fungal growth and community composition in buried meshbags, amended

with apatite, biotite or hornblende, in Norway spruce (Picea abies) forests of varying nutrient status. Norway spruce needles and soil collected from forests overlying serpentinite had low levels of potassium and phosphorus, those from granite had low levels of magnesium, whereas those from amphibolite had comparably high levels of these nutrients.  We assayed the fungal colonization of meshbags by measuring ergosterol content and fungal community with 454 sequencing of the internal transcribed spacer region. In addition, we measured fine root density.  Fungal biomass was increased by apatite amendment across all plots and particularly on the K- and P-deficient serpentinite plots, whereas hornblende and biotite had no effect on fungal biomass on any plots. Fungal community (total fungal and ectomycorrhizal) composition was affected strongly by sampling location and soil depth, whereas mineral amendments had no effect on community composition. Fine root biomass was significantly correlated with fungal biomass.  Ectomycorrhizal communities may respond to increased host-tree phosphorus demand by increased colonization of phosphorus-containing minerals, but this does not appear to translate to a shift in ectomycorrhizal community composition. This growth response to nutrient demand does not appear to exist for potassium or magnesium limitation.

Introduction Ectomycorrhizal fungi (EMF) play a fundamental role in the nutrient uptake of forest trees (Aquino & Plassard, 2004; Cairney, 2005; Hobbie & Hobbie, 2008). In boreal and temperate forests EMF communities are species rich, with over 50 species commonly (Wallander et al., 2010; Pickles et al., 2012) and over 100 taxa occasionally reported from monodominant forest stands (Tedersoo et al., 2006; Parrent & Vilgalys, 2007). Although EMF have been found to take up and translocate nitrogen (N), potassium (K), magnesium (Mg) and phosphorus (P) to their hosts (Jentschke et al., 2001; Smith & Read, 2008), it is not known to what extent mycorrhizal provision of these nutrients to their plant symbionts is responsive to plant demand. Nor is it known to what extent the ability to exploit particular nutrient sources contributes to niche separation among EMF species, although it has been found that some EMF species have large capacities to release nutrients from *These authors contributed equally to this work. Ó 2016 The Authors New Phytologist Ó 2016 New Phytologist Trust

complex organic matter (Read & Perez-Moreno, 2003) and others are associated predominantly with mineral substrates in the soil (Landeweert et al., 2001). Ectomycorrhizal community composition has been observed to shift in response to liming (Kjøller & Clemmensen, 2009; Rineau et al., 2010), fertilization (Fransson et al., 2000; Wright et al., 2009) and anthropogenic N deposition (Lilleskov et al., 2002; Kjøller et al., 2012). In addition, EMF taxa have been observed to preferentially colonize different soil horizons (Dickie et al., 2002; Rosling et al., 2003). These coarse community responses to whole-forest nutrient amendment or soil horizons are not sufficient to explain the assembly processes of these diverse communities. Dickie & Koide (2014) suggest that mechanisms other than vertical niche partitioning may also contribute to the diversity of EMF found in soil at a very fine scale, including uptake of nutrients from specific minerals. Many EMF species have the potential to extract nutrients from minerals. (Hoffland et al., 2004; Finlay et al., 2009), but the actual role of EMF in mineral weathering has been widely debated (Finlay et al., 2009; Sverdrup, 2009). In pot/microcosm New Phytologist (2016) 1 www.newphytologist.com

New Phytologist

2 Research

settings, EMF have been demonstrated to enhance dissolution of and phosphorus uptake from apatite (Wallander et al., 1997; Wallander, 2000a; Smits et al., 2012), dissolution of and potassium uptake from phlogopite (Leyval & Berthelin, 1989), biotite and muscovite (Wallander & Wickman, 1999; Wallander, 2000b; van Sch€oll et al., 2006; Bonneville et al., 2011), and dissolution of magnesium from biotite (Bonneville et al., 2011). Because weathering of minerals is a slow process, it is difficult to determine the contribution of EMF in absolute terms, especially under field conditions, but Turpault et al. (2009) were able to demonstrate a small but significant increase in the weathering rate of apatite in the presence of mycorrhizal roots at 20-cm soil depth. Forests may increasingly become limited, or co-limited, by P, K, calcium (Ca) or Mg as a result of atmospheric N deposition (Fenn et al., 1998; Gradowski & Thomas, 2008), more intense harvesting (Akselsson et al., 2007) or N fertilization (Wallander & Hagerberg, 2004; Akselsson et al., 2008). Such transitions may be delayed if EMF increase the weathering of, or nutrient uptake from, apatite or other minerals under these conditions. When trees are subjected to P-deficiency, enhanced carbon allocation belowground stimulates root growth (Ericsson, 1995; Marschner et al., 1996) and EMF growth (Wallander & Nylund, 1992; Ekblad et al., 1995; Berner et al., 2012), particularly around P-rich mineral patches in the soil (Hagerberg et al., 2003; Wallander & Thelin, 2008). This elevated belowground carbon allocation may be a driver for EMF-induced weathering processes in the soil (Leake et al., 2008). In contrast to P-deficiency, Mgand K-deficiencies usually result in reduced carbon allocation belowground (Ericsson & K€ahr, 1993; Hermans et al., 2006), which may inhibit EMF growth and establishment of EMF symbiosis (Ericsson & K€ahr, 1993; Wikstr€om & Ericsson, 1995). It is unclear whether EMF may preferentially colonize mineral sources of K and Mg when these nutrients are in low supply. When using ingrowth meshbags, Hagerberg et al. (2003) found similar production of EMF mycelia in Norway spruce forests with different needle K status. It was, however, unclear to what extent K availability was limiting the growth rates of the trees in these forests. Certain EMF species, such as Paxillus involutus, have larger capacities to take up P at lower concentrations than other tested EMF species (Colpaert et al., 1999), and it might thus be expected that enhanced plant carbon allocation to P-rich microsites in the soil would favor such species, especially in Plimited forests. Berner et al. (2012) did not find a significant effect of apatite amendment in meshbags on the composition of the EMF community, although fungal biomass increased considerably in response to apatite in the P-deficient forest; however, their study was conducted with cloning and Sanger-sequencing and, as such, may have failed to achieve sufficient sequencing depth to observe a community shift, particularly among the less common species. Much less is known about the species-specific uptake and translocation abilities of different EMF in relation to K and Mg or the effect of K- and Mg-rich minerals on the composition of EMF communities in Norway spruce forests limited by these elements. New Phytologist (2016) www.newphytologist.com

In order to test if apatite (a P-rich mineral), biotite (a K-rich mineral) or hornblende (a Mg-rich mineral) influenced the biomass and the composition of EMF communities in conditions where these nutrients were poorly available, we incubated meshbags (Wallander et al., 2001) amended with these minerals at three different proximally located Norway spruce forests in the Czech Republic. Streamwater and minerological analysis indicate that these soils have exceptionally low availabilities of K, Mg or P (Kram et al., 2012), which offer unique possibilities to test the effect of mineral amendments to forests which are limited by these nutrients. We measured fungal biomass and fungal community composition in the meshbags. We hypothesized that in cases where the added mineral is a source of the nutrients deficient at that site (i.e. apatite and biotite on serpentinite, and hornblende on granite) we would observe greater fungal biomass and altered EMF species composition compared with control quartz-filled meshbags. We expected that these responses would be more pronounced at greater soil depth where fungi, confined to mineral substrates, can be expected to be more adapted to releasing nutrients from minerals than species inhabiting more organic-matterrich horizons. Two experiments were conducted; the first focused on examining fungal ingrowth and community composition across all three sites, and the second focused on depth effects on fungal growth and the effect of apatite and biotite on community composition in the P- and K-deficient serpentinite site.

Materials and Methods Study sites Nine plots (40 9 40 m) across three catchments (three in each catchment, 300–800 m apart) were established in the Slavkov Forest in western Bohemia, Czech Republic. The catchments, c. 5–7 km apart, differ in their underlying bedrock (parent material): Lysina (50°030 N, 12°400 E) is underlain by acidic (felsic) rocks, referred to as ‘granite’; Na Zelenem (50°020 N, 12°430 E) by basic (mafic) rocks, referred to as ‘amphibolite’; Pluhuv Bor (50°040 N, 12°460 E) by ultrabasic (ultramafic) rocks, referred to as ‘serpentinite’ (Kram et al., 2012). The catchments also differed significantly in soil type and nutrient ratios. The soil types (following the World Reference Base classification) were Podzols low in Mg on the granite plots, Stagnosols low in K and P on the serpentinite plots, and Cambisols comparatively replete in Mg, K and P on the amphibolite plots. On all plots the soils were formed from the underlying parent material. All nine plots are covered by mature Norway spruce (Picea abies (L) Karst.) stands. Mean annual air temperature is 5–6°C, and mean annual precipitation is between 800 and 1000 mm. All three catchments are part of the Czech Critical Zone Observatory (Regelink et al., 2015). Meshbag construction and harvest Meshbags were used to sample the growing fungal mycelium in the soil. These sand-filled meshbags are constructed from 50 lM nylon mesh which allows fungal mycelia to enter, but excludes plant roots (Wallander et al., 2001). For a discussion of this and Ó 2016 The Authors New Phytologist Ó 2016 New Phytologist Trust

New Phytologist similar methods to quantify EMF biomass, see Wallander et al. (2013). In this study, meshbags were filled with 20 g acid-washed quartz sand (0.5–1 mm) and amended with 0.2 g (1%) of one of three different minerals (0.05–0.63 mm); fluorapatite (P source; igneous pegmatite from Madagascar), hornblende (Mg source; scapolite from Krager€o, Norway) and biotite (K source; carbonatite from Moen, Norway) or left un-amended as controls. All minerals were purchased from Kranz GmbH, Bonn, Germany. These meshbags were incubated in the soil for 4–16 months (time period detailed later). Following harvest, the meshbags were frozen, freeze-dried, and divided into equal parts for ergosterol and molecular analysis. Experimental design: Study 1 In Study 1, eight replicate sets of meshbags (four bags in each set, one of each amendment) were buried along each of two transects in each plot in May, 2009 (eight replicate bags, buried at 3–5 m intervals 9 four minerals 9 two transects 9 three parent materials 9 three plots = 576 bags). The bags were buried at the interface between the organic and mineral layer. One transect (220 bags; 5–8 bags of each mineral amendments) was harvested after 4 months and the other transect (180 bags; five to eight bags of each mineral amendments) after 16 months (September, 2010). Some bags were lost due to disturbance. Both sets of bags were analysed for ergosterol content, and the set of meshbags harvested after 16 months was used for fungal community analysis. Fungal community analysis was conducted after pooling meshbags with the same mineral treatment in a plot (5–8 meshbags) into one composite sample, giving four samples per plot (one for each mineral amendment) for a total of 36 samples (four mineral amendments 9 three parent materials 9 three plots). Experimental design: study 2 In Study 2, meshbags identical to those from Study 1 were buried in each plot in May 2009. The bags were buried at three depths: in the middle of the organic horizon (c. 3 cm depth), 10 and 20 cm below the organic-mineral horizon boundary (c. 15 and 25 cm soil depth). The three parent materials 9 three plots 9 three depths 9 four mineral amendments 9 three replicates treatments gave 324 bags in total. Most of the bags (243 in total) were harvested after 16 months in September 2010. Some bags were lost due to disturbance. All recovered bags were analyzed for ergosterol content. Based on the mycelial ingrowth data from Study 1, which indicated a growth stimulation by apatite in the serpentinite plots, as well as the soil and needle elemental contents, which indicated both K and P deficiency at the serpentinite plots, and previous work (Kram et al., 2012), which also indicated both K and P deficiency at the serpentinite plots; we chose to perform community analysis in Study 2 on the control, K- and P-containing minerals in the three serpentinite plots. The quartz, biotite-amended and apatite-amended meshbags from the serpentinite plots were used for fungal community analysis, and individual bags were not pooled before DNA extraction (52 bags in total). Ó 2016 The Authors New Phytologist Ó 2016 New Phytologist Trust

Research 3

Soil and fine root sampling and analysis In September 2010 when meshbags were harvested, samples were collected to measure fine root density, pH and soil elemental contents with a 3-cm-diameter soil core (30 cm deep). Samples for soil analysis were separated into organic (humus) and mineral (B and E or A horizons combined) soil layers. Root and soil samples were each a composite of two cores and three samples were taken from each plot; three replicates 9 three plots 9 three parent materials yielded 27 root samples; two soil layers 9 three replicates 9 three plots 9 three parent materials yielded 54 samples for soil analysis. Soil samples were stored at 4°C for < 1 wk and freeze-dried. pH was measured on freeze-dried samples in 0.01 M CaCl2. Soil samples were ground, digested with aqua regia and elemental contents analyzed with inductively coupled plasma optical emission spectrometry (ICP-OES) (AcmeLabs, Vancouver, Canada). Reported elemental contents are > 20 times the detection limit. Root samples were stored at 4°C for < 2 wk. Root samples were cleaned over a sieve under running water, examined under a stereo microscope, and all live turgid spruce roots < 2 mm were collected, dried at 60°C and their dry mass was recorded. Root density was calculated as the mass of roots per volume of soil sampled. Spruce needle sampling and elemental analysis In June 2009 needles from the current year were collected from five branches on the lower part of the crown of three trees at each plot. The needles were pooled and dried at 40°C for 72 h. Nutrient contents were analyzed on 0.5 g of dried, ball-milled material after digestion in 7 ml HNO3 and 3 ml dH2O. Samples were refluxed at 180°C for 1 h and diluted to 50 ml. Element analysis was performed by atomic emission spectrometry (ICP PerkinElmer Optima 3000 DV; PerkinElmer, Waltham, MA, USA). Nitrogen was measured using a Vario MAX CN elemental analyser (Elementar Analysensysteme GmbH, Hanau, Germany). Ergosterol analysis Ergosterol analysis was used to estimate the fungal biomass in the meshbags. Ergosterol from a subsample of 5 g material from the meshbags was extracted and analyzed as per Berner et al. (2012). Ergosterol values are reported as lg ergosterol g1 DW of meshbag contents. DNA extraction, PCR and 454 sequencing DNA was extracted from the homogenized samples by adding CTAB-SDS buffer (2% cetyltrimetylammoniumbromid, 2% sodium dodecyl sulfate, 1.4 M NaCl, 20 mM EDTA, 100 mM Tris-HCl, pH 8), vortexing and then incubating at 65°C for 1.5 h, followed by chloroform addition, vortexing, supernatant removal and isopropanol/ethanol precipitation. The pellet was resuspended in 50 ll of MiliQ-water (Millipore) and further cleaned using Wizard DNA clean-up kit (Promega). New Phytologist (2016) www.newphytologist.com

New Phytologist

4 Research

PCR amplification of the ribosomal internal transcribed spacer (ITS) DNA was carried out for each sample in three triplicate 25-ll reactions using the fungal-specific primers ITS1F (Gardes & Bruns, 1993) and ITS4 (White et al., 1990). Each primer was elongated with adaptors required for 454 pyrosequencing (ITS1F-B adaptor and ITS4-A adaptor). The ITS4-A primer contained a sample specific 8-bp multiplex identifier (MID) tag consisting of 8 bases (50 -CCATCTCATCCCTGCGTGTC TCCGACTCAGXXXXXXXXTCCTCCGCTTATTGATATGC -30 ), which the ITS1F-B adaptor lacked (50 -CCTATCCCCTGT GTGCCTTGGCAGTCTCAGCTTGGTCATTTAGAGGAAG TAA-30 ). PCR products were purified with Agencourt AMPure kit (Agencourt Bioscence Corp., Beverly, MA, USA) to remove residual salts, primers and primer dimers. The concentration of the purified PCR products were measured with the PicoGreen ds DNA Quantification Kit (Molecular Probes, Eugene, OR, USA) on a FLUOstar Optima (BMG Labtech GmbH, Ortenberg, Germany). Equal amounts of DNA from each sample were pooled into one single pool and submitted for 454 pyrosequencing. Sequencing (starting from the ITS4 fragment end) was performed on a FLX 454 (Roche Applied Biosystems) using the Lib-L chemistry at the Pyrosequencing facility at Lund University, Lund, Sweden. The 36 samples from Study 1 comprised half of a pool submitted for sequencing on ¼ of a 454 sequencing plate, whereas the 52 samples from Study 2 were separately sequenced in a different 454 run on ¼ of a plate. Bioinformatic analysis After sequencing, sequences without template primers (allowing two mismatches) or MID sample tag (allowing one mismatch) were removed. After template primers were removed sequences < 175 bp long were excluded. Sequences were then trimmed to include only the ITS2 region with ITSx (Bengtsson-Palme et al., 2013), using the additional chimera removal tool. Sequences were clustered using USEARCH at 96% sequence similarity. Clusters containing < 10 reads or only found in one meshbag sample (one PCR reaction) were removed. Searches for sequence identities were performed by iteratively BLASTing against three different sequence databases: first a self-created database of sequences obtained by Sanger sequencing of mycorrhizal root tips gathered from the same study area; second the UNITE (Koljalg et al., 2005; http://unite.ut.ee/index.php) reference/representative sequence database (21 000 seqs, release date 9 February 2014); and third the full UNITE + INSD sequence database (377 000 seqs, release date 15 February 2014). The UNITE+INSD databases were purged of all sequences – nearly 25% of the total – that did not have any taxonomic information as they were primarily environmental samples from soils and roots; this was achieved using search terms ‘unidentified’, ‘endophyt’, ‘uncultured’, ‘fungal sp’ and ‘root associated’. Query sequences failing to match a sequence from the database were then queried against the next database, and sequences were assigned to operational taxonomic units (OTUs) when there was at least 97% similarity between query sequence and top hit, given at least 90% coverage of the query sequence length. Sequences that failed to match any New Phytologist (2016) www.newphytologist.com

of the three databases at this threshold were excluded. All sequences used for analysis were submitted to the International Nucleotide Sequence Database Sequence Read Archive (accession no. SRP050046). The total fungal community was divided by both phylum (Basidiomycota, Ascomycota, Zygomycota and Chytridiomycota) and function (known ectomycorrhizal fungi (EMF), unknown ectomycorrhizal status, saprotrophic fungi); OTUs were considered known EMF based on the most current knowledge of the ecology of known close relatives (genera or species) according to Tedersoo et al. (2010). After filtering, each sample was rarified to the median number of reads using the ‘rrarefy’ function in the VEGAN package (Oksanen et al., 2013) in R (R Core Team, 2013). For community comparison (total, or for ectomycorrhizal fungi), sequence read abundances were converted to fractional abundance, such that the read abundances for all OTUs for each sample totaled to 1. Statistical analysis All statistical analyses were performed using the VEGAN package (Oksanen et al., 2013) in R (R Core Team, 2013), except for indicator species analysis, which was conducted with the indicspecies package (De Caceres & Legendre, 2009). Two-way ANOVAs were performed to determine the effect of mineral amendment and parent material, and the interaction between them, on fungal growth. One-way ANOVA was performed for each specific parent material to test the effect of mineral addition on EMF biomass in the meshbags. Post-hoc Tukey tests were performed for both one- and two-way ANOVAs to examine the significance of differences between parent materials and mineral amendments. One-way ANOVA and Tukey tests were also employed to examine differences in soil chemistry and fine root abundance between parent materials. The correlation between meshbag ingrowth and fine root density was examined with linear regression and one-way ANOVA. Fungal communities were visualized with ordination using nonparametric multidimensional scaling (NMDS) using the metaMDS function. All ordinations were conducted in two dimensions. Differences in community structure were compared visually with centroids and the associated 95% confidence interval associated with a v2-distribution around the SE of the centroid. Differences between communities were compiled into distance matrices (calculated using the VEGDIST function with Bray–Curtis dissimilarity) and compared between plots, depths and mineral-amendments with permutational multivariate analysis of variance (PERMANOVA; Anderson, 2001). PERMANOVA analysis of the effect of mineral amendment or biomass was conducted using the parent material, the depth class or the plot as the ‘strata’, allowing us to isolate the differences arising from mineral amendment from differences arising from sampling location. Single factors that were stratified were additionally analyzed with the ‘nested.npmanova’ function in the BiodiversityR package (Kindt & Coe, 2005) to obtain corrected F-values. Multifactor PERMANOVA was conducted after singlefactor analysis, and factors found significant in the single-factor Ó 2016 The Authors New Phytologist Ó 2016 New Phytologist Trust

New Phytologist

Research 5

analysis were analyzed together in a multifactor analysis; the order of factors was randomized, and only factors found significant under all factor orders are reported with the exception of plot and parent material (parent material always preceding plot). Ordination and PERMANOVA analyses were conducted for both the total community (all fungal OTUs) and the EMF community across all sample locations (parent materials in Study 1, depth classes in Study 2). Additional analyses were performed for communities within a given parent material or depth class. One-way ANOVA followed by the Tukey’s HSD test was performed to determine whether the different fungal phyla or ecological groups or any of the more abundant species (> 0.5% of total sequence read abundance) were significantly associated with a specific soil depth and whether there was an effect of mineral addition on their abundance. In addition, indicator species analysis was conducted with the ‘multipatt’ function to determine whether any particular mineral amendment was associated with a significantly greater or lower abundance of any of the more abundant species.

ingrowth in the meshbags after 16 months (P < 0.002, F = 5.54, df = 3), but not after 4 months. Amongst the four mineral amendments (none, hornblende, biotite, apatite) apatite amendment was associated with significantly greater fungal ingrowth (P < 0.01) (Fig. 1). There was also a significant interaction between the effect of mineral amendment and parent material (P < 0.01, F = 3.05, df = 6). When the relationship between fungal ingrowth and mineral amendments was examined for each parent material separately, mineral amendment was associated with a significant effect on fungal ingrowth only in the serpentinite plots (P < 0.002, F = 5.78, df = 3); on the serpentinite plots apatite was associated with significantly greater fungal ingrowth (P < 0.01) (Fig. 1).

Results

Table 2 Chemical measurements of nutrient concentrations in needles collected from each catchment (SE in parentheses) with Norway spruce deficiency levels according to Thelin et al. (2002)

Fungal ingrowth: Study 2 Across all samples, parent material (P < 0.0001, F = 29.0, df = 2), depth (P < 0.02, F = 4.1, df = 2) and mineral amendment

Soil and needle chemistry Across all soil samples pH was significantly lower in granite soils and higher in serpentinite soils (Table 1). The serpentinite soil had significantly less K (P < 0.01) and P (P < 0.0001) by mass, and the granite soils had significantly less Mg (P < 0.0001) (Table 1). Needle chemistry analysis showed a deficiency of Mg at the Lysina site (granite) and deficiency of both P and K at the Pluhuv Bor site (serpentinite) (Table 2). Spruce needles at the amphibolite site had sufficient amounts of all nutrients according to Thelin et al. (2002).

Nutrients in needles (mg g1)

Lysina (granite)

Pluhuv Bor (serpentinite)

Na Zelenem (amphibolite)

Deficiency level (mg g1)

Magnesium Phosphorus Potassium Nitrogen

0.5 (0.1) 1.2 (0.2) 4.8 (0.7) 12.9 (0.9)

1.8 (0.2) 0.8 (0.1) 2.9 (0.1) 11.7 (0.5)

0.9 (0.1) 1.3 (0.1) 4.8 (0.2) 14.3 (1.0)

0.6 1.3 4.5 12

Fungal ingrowth: Study 1 Fungal ingrowth (as measured by ergosterol content) in the meshbags differed significantly between the three parent materials, both after 4 (P < 0.001, F = 32.3, df = 2) and 16 months (P < 0.0001, F = 22.9, df = 2) (Fig. 1). Granite had the lowest and the serpentinite the highest fungal ingrowth, whereas the amphibolite was intermediate (all differences significant at P < 0.005). The fungal biomass in the meshbags was greater after 16 months compared with 4 months (P < 0.001, F = 30.4, df = 1). Mineral amendment significantly affected fungal

Fig. 1 Ergosterol concentrations in meshbags from Study 1, amended with different minerals or unamended (quartz), incubated in plots from three different parent materials for 16 months. Error bars show  1 SE.

Table 1 Soil analysis results from mineral and organic soil from the three parent materials used in this study; SE in parentheses and values followed by different letters are significantly different at P < 0.05 Parent material

Soil horizon

pH (CaCl2)

%P (g g1 dry mass)

%Mg (g g1 dry mass)

%K (g g1 dry mass)

Granite Granite Serpentinite Serpentinite Amphibolite Amphibolite

Organic Mineral Organic Mineral Organic Mineral

2.61 (0.038)a 2.94 (0.009)b 3.13 (0.023)b 4.29 (0.044)d 3.02 (0.006)b 3.47 (0.007)c

0.108 (0.006)c 0.108 (0.004)c 0.062 (0.0004)b 0.032 (0.001)a 0.100 (0.001)c 0.089 (0.003)c

0.064 (0.004)a 0.141 (0.007)a 1.426 (0.095)b 7.584 (0.088)c 0.707 (0.041)a 2.069 (0.053)b

0.89 (0.162)c 2.73 (0.057)d 0.23 (0.007)a 0.72 (0.011)bc 0.41 (0.005)ab 1.05 (0.033)c

P, phosphorus; Mg, magnesium; K, potassium. Ó 2016 The Authors New Phytologist Ó 2016 New Phytologist Trust

New Phytologist (2016) www.newphytologist.com

New Phytologist

6 Research

(P < 0.01, F = 3.89, df = 3) had a significant effect on fungal ingrowth, and there was no significant interaction between the three factors. The meshbags placed on granite had significantly lower fungal ingrowth than those on amphibolite (P < 0.0001) or serpentinite (P < 0.0001); meshbags from serpentinite had the highest ingrowth (serpentinite > amphibolite; P < 0.037) (Fig. 2a,b). Fungal ingrowth decreased with depth; the uppermost bags had significantly greater fungal ingrowth than the deepest bags (P < 0.02) (Fig. 2a). Apatite amendment was associated with higher meshbag fungal ingrowth across all parent materials (P < 0.05) (Fig. 2b). When examining the meshbag ingrowth on each parent material individually, no single parent material had a significant association between fungal ingrowth and mineral amendment, although on each individual parent material apatite was associated with higher ingrowth (not significant) and this increase was greatest on serpentinite (Fig. 2b). When only the 52 bags from serpentinite that were used for community composition analysis (amended with biotite or apatite, or control) were analyzed, a significant effect of mineral amendment was observable (P < 0.02, F = 4.27, df = 2) with apatite associated with greater ingrowth (apatite > biotite, P < 0.024; apatite > quartz, P < 0.077); the lack of a significant effect of mineral amendment when all the serpentinite samples in Study 2 were analyzed appears to be a function of the high variability in ergosterol contents from the hornblende-amended bags (Fig. 2b). Fine root biomass Fine root density in the top 30 cm of soil was significantly greater on serpentinite than on granite (P < 0.014), whereas that on amphibolite was intermediate. Ergosterol content of the meshbags from Study 1 (incubated at the organic–mineral horizon interface) and fine root density in the top 30 cm (n = 9, three plots 9 three parent materials) were strongly and significantly correlated (r2 = 0.78, F = 18.84, P < 0.005). Fungal community composition: study 1 After quality filtering and rarefaction to 1200 reads per sample the sequence data contained 37 540 reads representing 241 OTUs. Of the 241 OTUs, 130 were Ascomycetes, 90 were Basidiomycetes, 19 were Zygomycetes and two were Chytridiomycetes, representing an average (per sample, averaged across (a)

all 35 samples) of 19.7% ( 2.2% SE), 69.1% ( 3.0% SE), 8.9% ( 0.88% SE) and 2.8% ( 0.54% SE) of the total reads, respectively. EMF accounted for an average of 56.9% of ( 6.3% SE) of the reads across all samples; 38% ( 4.9% SE) in granite samples, 51.8% of ( 6.3% SE) in serpentinite samples and 80.9% ( 1.9% SE) in amphibolite samples; amphibolite samples were associated with a significantly higher fraction of reads corresponding to known EMF (P < 0.0001). Across all samples 2–3% of the reads were from taxa with unclear mycorrhizal status (primarily certain taxa from the Heliotiales, Oidodendron maius and Phialocephala fortinii) and are not included in our examination of the EMF. There was no significant relationship between mineral amendment, either across all parent materials, or within samples from a single parent material, and the fraction of reads from EMF, basidiomycetes, ascomycetes or zygomycetes. Examination of the total fungal community (241 OTUs) across all samples (n = 36) showed that parent material was strongly and significantly associated with the fungal community composition (R2 = 0.314, P < 0.001; Table 3; Figs 3a, 4). When only the EMF community was examined (50 OTUs), a strong influence of parent material (R2 = 0.321, P < 0.001) and plot (R2 = 0.283, P < 0.001) was observed (Table 3; Fig. 3b). No association between ergosterol content or mineral amendment and EMF community composition was observed, either across all parent materials or amongst samples from a single parent material (Table 3). The 21 most abundant EMF OTUs each contributed > 0.5% of total EMF reads and collectively comprised 87–99% of all EMF reads. No fungal species, EMF or non-EMF, exhibited a significant response in relative abundance to mineral amendment. Fungal community composition: study 2 After quality filtering and rarefaction to 1400 reads per sample, the sequence data contained 63 787 reads representing 231 OTUs. Of the 231 OTUs, 142 were Ascomycetes, 67 were Basidiomycetes, 20 were Zygomycetes and two were Chytridiomycetes, representing an average (per sample, averaged across all 52 samples) of 34.0% ( 2.7% SE), 55.3% ( 3.3% SE), 10.7% ( 1.5% SE) and 0.034% ( 0.01% SE) of the total reads, respectively. Across all samples 41.7% ( 0.01% SE) of the reads were from known EMF; 2.4% of the reads were from taxa with unknown mycorrhizal status. Meshbags incubated in the upper (c. 3 cm depth), middle (c. 15 cm) and lower (c. 25 cm)

(b)

Fig. 2 Ergosterol concentrations in meshbags from Study 2 incubated at different soil depths across plots from three parent materials: (a) by incubation depth (all mineral amendments combined; n = 26–30); (b) by mineral amendment (all incubation depths combined; n = 18–21). Error bars show  1 SE. New Phytologist (2016) www.newphytologist.com

Ó 2016 The Authors New Phytologist Ó 2016 New Phytologist Trust

New Phytologist

Research 7

Table 3 Test results from permutational multivariate analysis of variance (PERMOVA) in Study 1 Total fungal community All plots combined Single factor Parent material Ergosterol content (strat by par. mat.) Mineral amendment (strat by par. mat.) Plot (strat by par. mat.) Multi factor Parent material Plot (strat by par. mat.) Ectomycorrhizal fungal community All plots combined Single factor Parent material Ergosterol content (strat by par. mat.) Mineral amendment (strat by par. mat.) Plot (strat by par. mat.) Multi factor Parent material Plot (strat by par. mat.)

df

2 1 3 8 2 8

2 1 3 8 2 8

F

R2

9.119 2.687 0.769^ 3.92^ 12.673 3.920

7.816 0.675 0.92^ 3.216^ 10.965 3.216

P

Ectomycorrhizal fungal community

df

F

R2

P

0.356 0.073 0.045 0.656

0.001*** 0.083 0.874 0.001***

Amphibolite parent material Ectomycorrhizal fungal community Single factor Ergosterol content Mineral amendment Plot

1 3 2

1.047 0.385 2.925

0.095 0.126 0.394

0.322 0.934 0.019*

0.314 0.291

0.001*** 0.001***

Granite parent material Ectomycorrhizal fungal community Single factor Ergosterol content Mineral amendment Plot

1 3 2

2.270 0.603 2.588

0.185 0.184 0.365

0.284 0.696 0.005**

Serpeninite parent material Ectomycorrhizal fungal community Single factor Ergosterol content Mineral amendment Plot

1 3 2

2.270 0.603 3.880

0.185 0.184 0.463

0.258 0.718 0.005**

0.321 0.019 0.044 0.604

0.001*** 0.822 0.821 0.001***

0.321 0.283

0.001*** 0.001***

Total fungal community and ectomycorrhizal community were examined, both across all parent materials (n = 35) and, for the ectomycorrhizal community, individually on each parent material (par. mat.) (n = 11,12). Only factors found significantly associated with community in single-factor analysis were examined in multifactor analysis. Factors with significant associations have P-values in bold (, P = 0.1–0.05; *, P = 0.05–0.01; **, P = 0.01–0.001; ***, P < 0.001). F-values with a ‘^’ were corrected by nested analysis of variance. ‘strat’ indicates groups within which ordination was constrained.

mineral soil depths had 42.9% ( 7% SE), 51.7% ( 6.8% SE) and 30.4% ( 5.0% SE) of the their reads, respectively, corresponding to known EMF (15 cm > 25 cm, P < 0.058). There was no significant relationship between mineral amendment, either across all depths, or within samples from a single depth, and the fraction of reads arising from EMF. Examination of the total fungal community (231 OTUs) across all samples (n = 52, serpentinite plots only) showed that depth was strongly and significantly associated with the community composition (R2 = 0.095, P < 0.001; Table 4), whereas mineral amendment was not. When only the EMF community was examined (28 OTUs), a strong effect of depth (R2 = 0.076, P < 0.001) and plot (R2 = 0.245, P < 0.007) on the community composition were also observed (Table 4; Figs 4, 5). The effect of depth was only observable between the deepest (25 cm mineral soil depth) depth class and the rest of the community (R2 = 0.086, P < 0.001). Neither ergosterol content nor mineral amendment was significantly associated with EMF community composition overall, or when the communities within a single plot or depth class were analyzed individually (Table 4). ANOVA and Tukey tests indicate that Tylospora asterophora (P < 0.005) was significantly more abundant in the deepest (c. 25 cm) bags, whereas Tylospora fibrillosa (P < 0.075) and Xerocomus badius (P < 0.1) were significantly less abundant at a significance level of P < 0.1 (Fig. 4). Indicator species analysis indicated that Laccaria laccata (P < 0.001; IndVal 0.641), Russula integra (P < 0.01; IndVal 0.617) and Inocybe assimilata (P < 0.015; IndVal 0.484) had significantly stronger associations with the deepest soil incubation depth. No species was significantly more or less abundant in only the middle or uppermost bags, and no species were significantly more or less abundant in relation to mineral amendment. Ó 2016 The Authors New Phytologist Ó 2016 New Phytologist Trust

Discussion The positive effect of apatite amendment on fungal biomass under phosphorus (P) limitation supports earlier studies (Hagerberg et al., 2003; Wallander & Thelin, 2008; Potila et al., 2009; Berner et al., 2012) and suggests that apatite may be an important source of P for fungi. By contrast, neither biotite nor hornblende stimulated fungal growth when spruces were deficient in potassium (K) and magnesium (Mg), as indicated by the concentrations of these elements in the needles. This suggests that allocation to fungi does not respond in accordance with plant demand for these elements, these minerals are not important sources of K or Mg for fungi, or the fungi were not subjected to K- and Mg-deficiency under these conditions. Recent studies have found enhanced weathering of specific minerals to be strongly correlated with ectomycorrhizal fungi (EMF) colonization of mineral surfaces and preferential secretion of oxalic acid by EMF in laboratory settings (Schmalenberger et al., 2015). It is possible that similar effects occurred in our meshbags, but our experimental design did not allow any quantitative measurements of mineral weathering or assessment of the contribution of EMF to this process. As neither the community composition nor the proportion of EMF sequences varied between mineral amendments within a parent material, we believe our observations of elevated growth in response to apatite on P-deficient sites, and a lack of such a response for K and Mg, is equally applicable to EMF and to saprotrophic fungi. The presence of specific minerals does not seem to be a strong driver for shaping the EMF communities because we found little effect of mineral addition on the EMF community structure in any of the three parent materials (Table 3; Fig. 3b) in Study 1 or New Phytologist (2016) www.newphytologist.com

8 Research (a)

(b)

Fig. 3 Nonparametric multidimensional scaling (NMDS) ordination of (a) the total fungal community and (b) the ectomycorrhizal fungal community across all parent materials and mineral amendments (n = 36) in Study 1. Points represent pooled samples from 5 to 8 individual meshbags. Symbols represent mineral amendments to the meshbags (quartz is the nonamended control) and colors represent the parent material. Ellipses are drawn around the 95% confidence interval for the centroid (the standard error multiplied by the v2 distribution) of all samples from a particular parent material (n = 11,12). Gran, granite; Serp, serpentinite; Amph, amphibolite.

in the more exhaustive investigation on the serpentinite plots in Study 2 (Table 4). Similar results were obtained by Koele et al. (2014) who studied the effects of apatite addition on EMF communities in P-limited sites in New Zealand. EMF may obtain P released from apatite, but if some species exhibit greater capacity to obtain such nutrients then our results suggest that they are not rewarded with more carbon from the host. There does not seem to be a selection pressure to favor EMF taxa that are adept in New Phytologist (2016) www.newphytologist.com

New Phytologist taking up mineral P in P-rich niches when host plant P demand increases. As expected, our finding of comparatively low hyphal ingrowth and fine root density observed on granite suggests lower carbon allocation belowground by the Mg-deficient spruce compared with the other parent materials. Serpentinite had low availability of both K and P, and it seemed that P- rather than K-deficiency has been the driver for the carbon allocation pattern, as we found the highest hyphal ingrowth and fine root density in the serpentinite plots. In contrast to the belowground allocation, the forests overlying the granite and amphibolite plots had more than fivefold greater annual aboveground tree biomass increment than the  y et al., 1996; Pavlo serpentinite plots (Cern nova et al., 2008), indicating that the greater allocation to EMF and fine roots on serpentinite, and lower allocation on granite, represent differences in proportional allocation as well as total allocation. One factor that could explain some of the differences in EMF community composition between parent materials or depths is pH, which differed between the three parent materials (highest in serpentinite, lowest in granite; Table 1), and can be expected to increase with depth in the mineral soil. However, the pH was highest on the serpentinite plots and the EMF community on serpentinite was intermediate and not significantly different from the other parent materials (Fig. 3b). Another factor that may affect the EMF community composition is the amount of carbon allocated belowground, which the meshbag ingrowth and fine root density data indicate is lowest on granite and highest on serpentinite, although again the EMF community on serpentinite was intermediate and not significantly different from the other parent materials. Additionally, within each parent material, the total fungal biomass in each meshbag was not significantly associated with EMF community composition. A number of studies have observed that if factors such as overstorey species, aspect, slope, climate, stand history and management regime are constant, then the scale of spatial autocorrelation for fungal communities is on the order of meters and tens of meters, not hundreds of meters (Lilleskov et al., 2004; O’Hanlon, 2012; Pickles et al., 2012; Bahram et al., 2013). In their review of biogeographic and spatiotemporal influences on mycorrhizal communities, Bahram et al. (2015) concluded that the variation related to vertical distribution in EMF communities was significantly higher than variation related to spatial variation. In our study, plots on each parent material were c. 40 9 40 m and 300–800 m apart, and the community composition differences between plots on each parent material (Tables 3, 4) were comparable in magnitude to the difference between soil depths (Table 4; Fig. 5). This indicates that larger distances (300–800 m between plots vs 10– 40 m distance between sampling locations within each plot) between sampling points have a significant effect on fungal community composition. These findings are analogous to those of Talbot et al. (2014) who found, in a continental-scale study of the relationship between sampling distance and fungal community composition, a steady increase in community dissimilarity over distances of tens to hundreds of meters to kilometers. Additionally, they concluded that distance between Ó 2016 The Authors New Phytologist Ó 2016 New Phytologist Trust

New Phytologist

Research 9

Fig. 4 Relative abundance of the 25 most abundant fungi (by read abundance in meshbags) across the different parent materials in Study 1 and depths in Study 2. Horizontal axis labels represent granite (Gran), serpentinite (Serp) and amphibolite (Amph) parent materials for Study 1, and the lowermost (25 cm), middle (15 cm) and uppermost (3 cm) incubation depths for Study 2. Operation taxonomic units (OTUs) are sorted by overall abundance from bottom to top. Proportions represent the average of all samples on that parent material (n = 11,12) or depth class (n = 17,18). Taxa with asterisks (*) in the legend are known ectomycorrhizal fungi. Table 4 Test results from permutational multivariate analysis of variance in study 2 Total fungal community Single factor Depth Depth (strat by plot) Ergosterol content (strat by depth) Ergosterol content (strat by plot) Mineral amendment (strat by depth) Mineral amendment (strat by plot) Plot Plot (strat by depth) Multi factor Depth (strat by plot) Ergosterol content (strat by depth) Plot (strat by depth)

df

F

R2

P

2 2 1 1 2 2 2 2

3.820 4.762^ 2.058 2.058 0.411^ 0.429^ 6.518 4.323^

0.071 0.104 0.040 0.040 0.022 0.022 0.210 0.210

0.001*** 0.001*** 0.017* 0.184 0.943 0.961 0.001*** 0.001***

2 1 2

3.315 2.228 7.297

0.095 0.032 0.210

0.001*** 0.004** 0.001***

Ectomycorrhizal fungal community Single factor Depth Depth (strat by plot) Ergosterol content (strat by depth) Ergosterol content (strat by plot) Mineral amendment (strat by depth) Mineral amendment (strat by plot) Plot Plot (strat by depth) Multi factor Depth (strat by plot) Plot (strat by depth)

df

F

R2

P

2 2 1 1 2 2 2 2

3.406 4.667^ 2.431 2.431 0.582^ 0.634^ 8.569 2.17^

0.064 0.087 0.046 0.046 0.029 0.029 0.259 0.259

0.015* 0.003** 0.056 0.277 0.667 0.555 0.001*** 0.001***

2 2

2.761 8.887

0.076 0.245

0.007** 0.001***

Total fungal community and ectomycorrhizal community were examined. Only factors found significantly associated with community in single-factor analysis were examined in multifactor analysis. Factors with significant associations have P-values in bold (, P = 0.1–0.05; *, P = 0.05–0.01; **, P = 0.01–0.001; ***, P < 0.001). F-values with a ‘^’ were corrected by nested analysis of variance. ‘strat’ indicates groups within which ordination was constrained.

sampling points was a greater factor in fungal community composition than soil horizon or climate effects. As our research questions involved comparisons between meshbags within each plot, this effect of distance on fungal community composition does not affect our ability to draw conclusions regarding the effect of mineral composition and forest nutrient status on fungal growth and mineral colonization. Our results Ó 2016 The Authors New Phytologist Ó 2016 New Phytologist Trust

do indicate, however, that future studies, involving plot-scale treatments should be designed to account for the potential for a steady increase in community dissimilarity over greater distances than previously reported. The proportion of sequence reads in the meshbags of EMF origin in the amphibolite plots, c. 80%, is in line with previously observed results (Parrent & Vilgalys, 2007; Hedh et al., 2008; New Phytologist (2016) www.newphytologist.com

New Phytologist

10 Research

1

3 cm 15 cm 25 cm

Serp_2

0

NMDS2

top mid Serp_3

Serp_1

–2

–1

bottom

plot 1 plot 2 plot 3

Stress = 0.15

–1

0

1

2

NMDS1 Fig. 5 Nonparametric multidimensional scaling (NMDS) ordination of the ectomycorrhizal fungal community from Study 2. Points represent individual meshbags. Point colors represent the plot; symbols represent incubation depths. Ellipses are drawn around the 95% confidence interval for the centroid of all samples (SE 9 v2 distribution) from a particular plot (Serp (serpentinite)–1–3) (in green; n = 17,18) or incubation depth (top, middle, bottom) (in black; n = 17,18).

Wallander et al., 2010). However, of these studies only Wallander et al. (2010) employed next-generation metagenomic methods (pyrosequencing of meshbags in Swedish Norway spruce forests), and they found that in young (< 10 yr) and old (> 100 yr) stands the fraction of sequence reads belonging to EMF was much lower (< 30%). We did not expect to find such a low proportion of sequence reads of EMF origin as we observed on granite and serpentinite. Also unexpected was our observation that the deepest meshbags in Study 2 had a lower, rather than higher, fraction of sequence reads of EMF origin. This finding contradicts both our hypothesis and a major assumption behind the utilization of meshbags to assay EMF growth; that EMF will be selected for in low organic carbon substrates. It is possible that the granite catchment with its very low belowground carbon allocation due to magnesium deficiency, and the serpentinite catchment, with the unique chemistry of a serpentine soil (Kram et al., 2009), in conjunction with the high historic acid deposition levels this area sustained during the second half of the 20th century, have combined to make these study sites anomalous in their relative EMF : saprotrophic fungal compositions. It is also important to note, that read abundance of a particular taxa, or group of taxa, may not accurately reflect active fungal biomass (Amend et al., 2010), and thus the fraction of sequence reads from EMF should probably be at best taken as an approximation for the proportion of fungal hyphal ingrowth of mycorrhizal origin. ITS copy number can vary significantly, as can the DNA : fungal biomass ratio across different fungal strains, species, and higher phylogenetic groups of fungi (Baldrian et al., 2013, and references therein). Nonetheless, using read abundance New Phytologist (2016) www.newphytologist.com

as part of the fingerprint of a fungal community’s composition for the purposes of investigating those factors influencing community composition is a well-established technique, and has been shown to yield repeatable results (Kauserud et al., 2012), independent of sequencing technology (Smith & Peay, 2014). The variability of, and on two of the three parent materials the low number of, EMF reads obtained from our meshbags suggest that caution should be employed when making the assumption that ergosterol assays in meshbags are an accurate reflection of EMF mycelial growth, particularly when comparing across different soils or soil depths. However, our finding of a strong correlation between meshbag fungal biomass and fine root density in the soil do suggest that meshbag ingrowth may be a good indication of belowground carbon allocation. In conclusion, although soils in different catchments, developed from different parent materials (granite, serpentinite and amphibolite), had distinctly different EMF communities, no EMF species exhibited a tendency to preferentially colonize certain minerals in the soil. The presence of specific minerals does not seem to be a strong driver for shaping EMF communities. We suggest that EMF may have the potential to compensate P deficiencies in forests by preferential colonization of P-containing minerals, but this mechanism does not exist for K and Mg deficiencies. The physiology behind this differential response to nutrient deficiency may stem from the fact that plants respond strongly to P deficiency by allocating more carbon belowground and to areas of P availability, whereas this response is nonexistent, weak or negative for K and Mg. One reason for these different responses by plants could be that K and Mg are relatively abundant in most ecosystems and rarely limiting to forest growth, and thus specific mechanisms to increase the availability of these nutrients have not evolved. However, P deficiency is common in many ecosystems and plants have evolved a number of traits to cope with this (Vance et al., 2003). For these reasons, directions for sustainable forest management should make sure that the removal of nutrients through harvesting and leaching does not result in K and Mg deficiencies unless these nutrients are added as fertilizers.

Acknowledgements We would like to thank the Swedish Research Council and the BECC (Biodiversity and Ecosystem Services in a Changing Climate) consortium for funding this research, and the European Commission SoilTrEC project (244118) for maintaining these research sites and permission to use them. We would also like to thank four anonymous reviewers and the editor for their input during the review process.

Author contributions N.P.R., C.B., M.M.S, P.K. and H.W. contributed to design of research and conducted field work. N.P.R. and C.B. conducted sample analysis and data analysis. N.P.R., C.B., M.M.S, P.K. and H.W. contributed to interpretation of results and writing of manuscript. Ó 2016 The Authors New Phytologist Ó 2016 New Phytologist Trust

New Phytologist References Akselsson C, Westling O, Alveteg M, Thelin G, Fransson AM, Hellsten S. 2008. The influence of N load and harvest intensity on the risk of P limitation in Swedish forest soils. Science of the Total Environment 404: 284–289. Akselsson C, Westling O, Sverdrup H, Gundersen P. 2007. Nutrient and carbon budgets in forest soils as decision support in sustainable forest management. Forest Ecology and Management 238: 167–174. Amend AS, Seifert KA, Bruns TD. 2010. Quantifying microbial communities with 454 pyrosequencing: does read abundance count? Molecular Ecology 19: 5555–5565. Anderson MJ. 2001. A new method for non-parametric multivariate analysis of variance. Austral Ecology 26: 32–46. Aquino MT, Plassard C. 2004. Dynamics of ectomycorrhizal mycelial growth and P transfer to the host plant in response to low and high soil P availability. FEMS Microbial Ecology 48: 149–156. Bahram M, Koljalg U, Courty P, Diedhiou AG, Kjøller R, Polme S, Ryberg M, Veldre V, Tedersoo L. 2013. The distance decay of similarity in communities of ectomycorrhizal fungi in different ecosystems and scales. Journal of Ecology 101: 1335–1344. Bahram M, Peay KG, Tedersoo L. 2015. Local-scale biogeography and spatiotemporal variability in communities of mycorrhizal fungi. New Phytologist 205: 1454–1463. Baldrian P, Vetrovsk y T, Cajthaml T, Dobiasova P, Petra nkova M, Snajdr J, Eichlerova I. 2013. Estimation of fungal biomass in forest litter and soil. Fungal Ecology 6: 1–11. Bengtsson-Palme J, Veldre V, Ryberg M, Hartmann M, Branco S, Wang Z, Godhe A, Bertrand Y, De Wit P, Sanchez M et al. 2013. ITSx: improved software detection and extraction of ITS1 and ITS2 from ribosomal ITS sequences of fungi and other eukaryotes for use in environmental sequencing. Methods in Ecology and Evolution 4: 914–919. Berner C, Johansson T, Wallander H. 2012. Long-term effect of apatite on ectomycorrhizal growth and community structure. Mycorrhiza 22: 615–621. Bonneville S, Morgan DJ, Schmalenberger A, Bray A, Brown A, Banwart SA, Benning LG. 2011. Tree-mycorrhiza symbiosis accelerate mineral weathering: evidences from nanometer-scale elemental fluxes at the hypha-mineral interface. Geochimica et Cosmochimica Acta 75: 6988–7005. Cairney JWG. 2005. Basidiomycete mycelia in forest soils: dimensions, dynamics and roles in nutrient distribution. Mycological Research 109: 7–20.  y M, Parez J, Malık Z. 1996. Growth and mensuration tables of the main tree Cern species of the Czech Republic (spruce, pine, beech, oak). (In Czech, English summary). Jılove u Prahy, Czech Republic: IFER, Institute for Forest Ecosystems Research. Colpaert JV, van Tichelen KK, van Assche JA, van Laere A. 1999. Short-term phosphorus uptake rates in mycorrhizal and non-mycorrhizal roots of intact Pinus sylvestris seedlings. New Phytologist 143: 589–597. De Caceres M, Legendre P. 2009. Associations between species and groups of sites: indices and statistical inference. Ecology 90: 3566–3574. Dickie IA, Koide RT. 2014. Deep thoughts on ectomycorrhizal fungal communities. New Phytologist 201: 1083–1085. Dickie IA, Xu B, Koide RT. 2002. Vertical niche differentiation of ectomycorrhizal hyphae in soil as shown by T-RFLP analysis. New Phytologist 156: 527–535. Ekblad A, Wallander H, Carlsson R, Huss-Danell K. 1995. Fungal biomass in roots and extramatrical mycelium in relation to macronutrients and plant biomass of ectomycorrhizal Pinus sylvestris and Alnus incana. New Phytologist 131: 443–451. Ericsson T. 1995. Growth and shoot: root ratio of seedlings in relation to nutrient availability. Plant and Soil 168–169: 205–214. Ericsson T, K€a hr M. 1993. Growth and nutrition of birch seedlings in relation to potassium supply rate. Trees–Structure and Function 7: 78–85. Fenn EM, Poth MA, Aber JD, Baron JS, Bormann BT, Johnson DW, Lemly AD, McNulty SG, Ryan DF, Stottlemeyer R. 1998. Nitrogen excess in North American ecosystems: predisposing factors, ecosystem responses and management strategies. Ecological Applications 8: 706–733. Ó 2016 The Authors New Phytologist Ó 2016 New Phytologist Trust

Research 11 Finlay R, Wallander H, Smits M, Holmstrom S, Van Hees P, Lian B, Rosling A. 2009. The role of fungi in biogenic weathering in boreal forest soils. Fungal Biology Reviews 23: 101–106. Fransson PMA, Taylor AFS, Finlay RD. 2000. Effects of continuous optimal fertilization on belowground ectomycorrhizal community structure in a Norway spruce forest. Tree Physiology 20: 599–606. Gardes M, Bruns TD. 1993. ITS Primers with enhanced specificity for Basidiomycetes – application to the identification of mycorrhizae and rusts. Molecular Ecology 2: 113–118. Gradowski T, Thomas SC. 2008. Responses of Acer saccharum canopy trees and samplings to P, K and lime additions under high N deposition. Tree Physiology 28: 173–185. Hagerberg D, Thelin G, Wallander H. 2003. The production of ectomycorrhizal mycelium in forests: relation between forest nutrient status and local mineral sources. Plant and Soil 252: 279–290. Hedh J, Wallander H, Erland S. 2008. Ectomycorrhizal mycelial species composition in apatite amended and non-amended meshbags buried in a phosphorus-poor spruce forest. Mycological Research 112: 681–688. Hermans C, Hammond JP, White PJ, Verbruggen N. 2006. How do plants respond to nutrient shortage by biomass allocation? Trends in Plant Science 11: 610–617. Hobbie EA, Hobbie JE. 2008. Natural abundance of 15N in nitrogen-limited forests and tundra can estimate nitrogen cycling through mycorrhizal fungi: a review. Ecosystems 11: 815–830. Hoffland E, Kuyper TW, Wallander H, Plassard C, Gorbushina AA, Haselwandter K, Holmstr€om S, Landeweert R, Lundstr€om US, Rosling A et al. 2004. The role of fungi in weathering. Frontiers in Ecology and the Environment 2: 258–264. Jentschke G, Brandes B, Kuhn AJ, Schroder WH, Godbold DL. 2001. Interdependence of phosphorus, nitrogen, potassium and magnesium translocation by the ectomycorrhizal fungus Paxillus involutus. New Phytologist 149: 327–337. Kauserud H, Kumar S, Brysting AK, Norden J, Carlsen T. 2012. High consistency between replicate 454 pyrosequencing analyses of ectomycorrhizal plant root samples. Mycorrhiza 22: 309–315. Kindt R, Coe R. 2005. Tree diversity analysis. A manual and software for common statistical methods for ecological and biodiversity studies. Nairobi: World Agroforestry Centre (ICRAF). Kjøller R, Clemmensen KE. 2009. Belowground ectomycorrhizal fungal communities respond to liming in three southern Swedish coniferous forest stands. Forest Ecology and Management 257: 2217–2225. Kjøller R, Nilsson LO, Hansen K, Schmidt IK, Vesterdal L, Gundersen P. 2012. Dramatic changes in ectomycorrhizal community composition, root tip abundance and mycelial production along a stand-scale nitrogen deposition gradient. New Phytologist 194: 278–286. Koele N, Dickie IA, Blum JD, Gleason JD, de Graaf L. 2014. Ecological significance of mineral weathering in ectomycorrhizal and arbuscular mycorrhizal ecosystems from a field-based comparison. Soil Biology and Biochemistry 69: 63–70. Koljalg U, Larsson KH, Abarenkov K, Nilsson RH, Alexander IJ, Eberhardt U, Erland S, Hoiland K, Kjoller R, Larsson E et al. 2005. UNITE: a database providing web-based methods for the molecular identification of ectomycorrhizal fungi. New Phytologist 166: 1063–1068. Kra m P, Hruska J, Shanley JB. 2012. Streamwater chemistry in three contrasting monolithologic Czech catchments. Applied Geochemistry 27: 1854–1863. Kra m P, Oulehle F, Stedra V, Hruska J, Shanley JB, Minocha R, Traister E. 2009. Geoecology of a forest watershed underlain by serpentine in central Europe. Northeastern Naturalist 16(Spec. 5): 309–328. Landeweert R, Hoffland E, Finlay RD, Kuyper TW, van Breemen N. 2001. Linking plants to rocks: ectomycorrhizal fungi mobilize nutrients from minerals. Trends in Ecology and Evolution 16: 248–254. Leake JR, Duran AL, Hardy KE, Johnson I, Beerling DJ, Banwart SA, Smits MM. 2008. Biological weathering in soil: the role of symbiotic root-associated fungi biosensing minerals and directing photosynthate-energy into grain-scale mineral weathering. Mineralogical Magazine 72: 85–89. New Phytologist (2016) www.newphytologist.com

12 Research Leyval C, Berthelin J. 1989. Experimental weathering of mica by mycorrhizal and non-mycorrhizal beech and pine. Annals of Forest Science 46: 762s–764s. Lilleskov EA, Bruns TD, Horton TR, Taylor DL, Grogan P. 2004. Detection of forest stand-level spatial structure in ectomycorrhizal fungal communities. FEMS Microbiol Ecology 49: 319–332. Lilleskov EA, Fahey TJ, Horton TR, Lovett GM. 2002. Belowground ectomycorrhizal fungal community change over a nitrogen deposition gradient in Alaska. Ecology 83: 104–115. Marschner H, Kirkby EA, Cakmak I. 1996. Effect of mineral nutritional status on shoot–root partitioning of photoassimilates and cycling of mineral nutrients. Journal of Experimental Botany 47: 1255–1263. O’Hanlon R. 2012. Below-ground ectomycorrhizal communities: the effect of small scale spatial and short term temporal variation. Symbiosis 57: 57–71. Oksanen J, Blanchet G, Kindt R, Legendre P, Minchin PR, O’Hara RB, Simpson GL, Solymos PM, Stevens MHH, Wagner H. 2013. vegan: Community Ecology Package. R package version 2.0-9. URL http://CRAN.Rproject.org/package=vegan. Parrent JL, Vilgalys R. 2007. Biomass and compositional responses of ectomycorrhizal fungal hyphae to elevated CO2 and nitrogen fertilization. New Phytologist 176: 164–174. Pavlo nova G, Cichra D, Hanychova M, Nikl M. 2008. Biomass balance for the fourteen GEOMON catchments. Research Report (in Czech). Brandy s nad Labem, Czech Republic: Forest Management Institute UHUL. Pickles BJ, Genney DR, Anderson IC, Alexander IJ. 2012. Spatial analysis of ectomycorrhizal fungi reveals that root tip communities are structured by competitive interactions. Molecular Ecology 21: 5110–5123. Potila H, Wallander H, Sarjala T. 2009. Growth of ectomycorrhizal fungi in drained peatland forests with variable P and K availability. Plant and Soil 316: 139–150. R Core Team. 2013. R: a language and environment for statistical computing. v.3.1.2. Vienna, Austria: R Foundation for Statistical Computing. URL http:// www.R-project.org/. Read DJ, Perez-Moreno J. 2003. Mycorrhizas and nutrient cycling in ecosystems – a journey towards relevance? New Phytologist 157: 475–492. Regelink IC, Stoof CR, Rousseva S, Weng L, Lair GJ, Kram P, Nikolaidis NP, Kercheva M, Banwart S, Comans RNJ. 2015. Linkages between aggregate formation, porosity and soil chemical properties. Geoderma 247–248: 24–37. Rineau F, Maurice JP, Nys C, Voiry H, Garbaye J. 2010. Forest liming durably impact the communities of ectomycorrhizas and fungal epigeous fruiting bodies. Annals of Forest Science 67: 1–12. Rosling A, Landeweert R, Lindahl BD, Larsson KH, Kuyper TW, Taylor AFS, Finlay RD. 2003. Vertical distribution of ectomycorrhizal fungal taxa in a podzol soil profile. New Phytologist 159: 775–783. Schmalenberger A, Duran AL, Bray AW, Bridge J, Bonneville S, Benning LG, Romero-Gonzalez ME, Leake JR, Banwart SA. 2015. Oxalate secretion by ectomycorrhizal Paxillus involutus is mineral-specific and controls calcium weathering from minerals. Scientific Reports 5: 12 187. van Sch€oll L, Smits MM, Hoffland E. 2006. Ectomycorrhizal weathering of the soil minerals muscovite and hornblende. New Phytologist 171: 805–814. Smith DP, Peay KG. 2014. Sequence depth, not PCR replication, improves ecological inference from next generation DNA sequencing. PLoS ONE 9: 1–12. Smith SE, Read DJ. 2008. Mycorrhizal symbiosis. Cambridge, UK: Academic Press. Smits MM, Bonneville S, Benning LG, Banwart SA, Leake JR. 2012. Plantdriven weathering of apatite – the role of an ectomycorrhizal fungus. Geobiology 10: 445–456. Sverdrup H. 2009. Chemical weathering of soil minerals and the role of biological processes. Fungal Biology Reviews 23: 94–100.

New Phytologist (2016) www.newphytologist.com

New Phytologist Talbot JM, Bruns TD, Taylor JW, Smith DP, Branco S, Glassman SI, Erlandson S, Vilgalys R, Liao HL, Smith ME et al. 2014. Endemism and functional convergence across the North American soil mycobiome. Proceedings of the National Academy of Sciences, USA 111: 6341–6346. Tedersoo L, May TW, Smith ME. 2010. Ectomycorrhizal lifestyle in fungi: global diversity, distribution, and evolution of phylogenetic lineages. Mycorrhiza 20: 217–263. Tedersoo L, Suvi T, Larsson E, K~oljalg U. 2006. Diversity and community structure of ectomycorrhizal fungi in a wooded meadow. Mycological Research 110: 734–748. Thelin G, Rosengren U, Nihlg ard B. 2002. Barrkemi p a Sk anska Gran- och Tallprovytor. Rapport 20. Scania, Sweden: County administrative board of Sk ane. Turpault M-P, Nys C, Calvaruso C. 2009. Rhizosphere impact on the dissolution of test minerals in a forest ecosystem. Geoderma 153: 147–154. Vance CP, Uhde-Stone C, Allan DL. 2003. Phosphorus acquisition and use: critical adaptations by plants for securing a nonrenewable resource. New Phytologist 157: 423–447. Wallander H. 2000a. Uptake of P from apatite by Pinus sylvestris seedlings colonised by different ectomycorrhizal fungi. Plant and Soil 218: 249–256. Wallander H. 2000b. Use of strontium isotopes and foliar K content to estimate weathering of biotite induced by pine seedlings colonised by ectomycorrhizal fungi from two different soils. Plant and Soil 222: 215–229. Wallander H, Ekblad A, Godbold DL, Johnson D, Bahr A, Baldrian P, Bj€ork RG, Kieliszewska-Rokicka B, Kjøller R, Kraigher H et al. 2013. Evaluation of methods to estimate production, biomass and turnover of ectomycorrhizal mycelium in forests soils – a review. Soil Biology and Biochemistry 57: 1034– 1047. Wallander H, Hagerberg D. 2004. Do ectomycorrhizal fungi have a significant role in weathering of minerals in forest soil? Symbiosis 37: 249–257. Wallander H, Johansson U, Sterkenburg E, Brandstr€om Durling M, Lindahl BD. 2010. Production of ectomycorrhizal mycelium peaks during canopy closure in Norway spruce forests. New Phytologist 187: 1124–1134. Wallander H, Nilsson LO, Hagerberg D, B a ath E. 2001. Estimation of the biomass and seasonal growth of external mycelium of ectomycorrhizal fungi in the field. New Phytologist 151: 753–760. Wallander H, Nylund JE. 1992. Effects of excess nitrogen and phosphorus starvation on the extramatrical mycelium of ectomycorrhizas of Pinus sylvestris L. New Phytologist 120: 495–503. Wallander H, Thelin G. 2008. The stimulating effect of apatite on ectomycorrhizal growth diminishes after PK fertilization. Soil Biology and Biochemistry 40: 2517–2522. Wallander H, Wickman T. 1999. Biotite and microcline as a K source in mycorrhizal and non-mycorrhizal Pinus sylvestris seedlings. Mycorrhiza 9: 25– 32. Wallander H, Wickman T, Jacks G. 1997. Apatite as a P source in mycorrhizal and non-mycorrhizal Pinus sylvestris seedlings. Plant and Soil 196: 123–131. White TJ, Bruns T, Lee S, Taylor J. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ, eds. Pcr protocols: a guide to methods and applications. San Diego, CA, USA & London, UK: Academic Press, 315–322. Wikstr€om F, Ericsson T. 1995. Allocation of mass in trees subject to nitrogen and magnesium limitation. Tree Physiology 15: 339–344. Wright SHA, Berch SM, Berbee ML. 2009. The effect of fertilization on the below-ground diversity and community composition of ectomycorrhizal fungi associated with western hemlock (Tsuga heterophylla). Mycorrhiza 19: 267–276.

Ó 2016 The Authors New Phytologist Ó 2016 New Phytologist Trust