Small-scale heterogeneity of abiotic and biotic factors is expected to play a crucial role in species coexistence. It is known that plants are able to concentrate their root biomass into areas with high nutrient content and also acquire nutrients via symbiotic microorganisms such as arbuscular mycorrhizal (AM) fungi. At the same time, little is known about the small-scale distribution of soil nutrients, microbes and plant biomass occurring in the same area. We examined small-scale temporal and spatial variation as well as covariation of soil nutrients, microbial biomass (using soil fatty acid biomarker content) and above- and belowground biomass of herbaceous plants in a natural herb-rich boreonemoral spruce forest. The abundance of AM fungi and bacteria decreased during the plant growing season while soil nutrient content rather increased. The abundance of all microbes studied also varied in space and was affected by soil nutrient content. In particular, the abundance of AM fungi was negatively related to soil phosphorus and positively influenced by soil nitrogen content. Neither shoot nor root biomass of herbaceous plants showed any significant relationship with variation in soil nutrient content or the abundance of soil microbes. Our study suggests that plants can compensate for low soil phosphorus concentration via interactions with soil microbes, most probably due to a more efficient symbiosis with AM fungi. This compensation results in relatively constant plant biomass despite variation in soil phosphorous content and in the abundance of AM fungi. Hence, it is crucial to consider both soil nutrient content and the abundance of soil microbes when exploring the mechanisms driving vegetation patterns.
Citation: Koorem K, Gazol A, Öpik M, Moora M, Saks Ü, Uibopuu A, et al. (2014) Soil Nutrient Content Influences the Abundance of Soil Microbes but Not Plant Biomass at the Small-Scale. PLoS ONE 9(3): e91998. https://doi.org/10.1371/journal.pone.0091998
Editor: Kurt O. Reinhart, USDA-ARS, United States of America
Received: November 27, 2013; Accepted: February 15, 2014; Published: March 17, 2014
Copyright: © 2014 Koorem et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This research was funded by grants from the Estonian Science Foundation (9050, 9157), targeted financing (IUT 20-28), European Union's Seventh Framework Programme for research, technological development and demonstration under grant agreement no GA-2010-267243 – PLANT FELLOWS, ERMOS programme grant 14 (co-funded by Marie Curie Actions), and the European Regional Development Fund (Centre of Excellence FIBIR). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: MM and MÖ (co-authors of this study) are PLOS ONE Editorial Board members. This does not alter our adherence to PLOS ONE Editorial policies and criteria.
Resource heterogeneity in space and time is a ubiquitous phenomenon in natural ecosystems and plays a fundamental role in determining plant community dynamics , . Small-scale resource heterogeneity can be especially important in determining plant productivity since plants can concentrate their root biomass in areas of high nutrient content –. However, plant growth can also be strongly influenced by soil microorganisms . Symbiotic microorganisms, such as mycorrhizal fungi, contribute to plant nutrient uptake, thereby affecting plant growth and biomass allocation . Similarly, bacteria play an important role in increasing nutrient availability for plants . Despite increasing recognition of the roles played by soil nutrients and especially soil microbes in determining vegetation patterns, information about how soil nutrients, microbes and plant biomass covary in natural ecosystems remains limited.
The spatial distribution of soil microbes is heterogeneous at various scales ,  and can be closely related to soil fertility , –. Less is known about how temporal variation in soil nutrient content  influences soil microbial communities, though the abundance of fungi and bacteria has been shown to display significant temporal variation, with maximum abundance in spring and autumn . Indeed, the abundance of soil microbes is expected to be high in spring and autumn when nutrient availability is high and decline as the plant growth season proceeds and nutrient availability subsequently decreases . However more evidence is needed to better understand these dynamics and the mechanisms driving them.
Studies have found significant small-scale patchiness in the abundance in AM fungi, ectomycorrhizal fungi, other fungi and bacteria – but knowledge about the simultaneous dynamics of different microbial groups remains limited. Both positive and negative associations in the abundance of co-occuring microbes have been reported: in saprotrophic fungi and bacteria  and AM fungi and bacteria , . Positive interactions may be the result of one group providing substrate for the other , while negative interactions may indicate competition for nutrients . A recent study demonstrated that high N:P ratio in soil can provide an advantage, in the form of plant assimilated carbon, to AM fungi, in competition with non-mycorrhizal fungi that also inhabit plant roots .
Mycorrhizal fungi contribute to plant nutrient acquisition from the soil  and their abundance has been found to decrease with increasing nutrient availability , , . Hence, mycorrhizal fungi and other soil microbes are expected to play an important role in determining plant productivity in ecosystems with low soil nutrients . However, it is unknown whether the effect of soil microbes on plant productivity remains dependent on soil nutrient content at small scales. A recent experimental study did not find any change in the effect of AM fungi on plant productivity in conditions where local soil fertility varied , but that study did not consider the effect of microbes besides AM fungi.
We simultaneously measured the temporal and spatial variation of soil nutrient content, soil microbial biomass and plant biomass in a natural herb-rich boreonemoral spruce forest. We focused on a scale that corresponds to the approximate size of adult plant individuals in the understory (i.e. small scale, 15 cm2). The study site is characterized by heterogeneous light and soil conditions, patchy understory vegetation , and remarkably taxon rich AM fungal communities in the roots of herbaceous species , . We ask:
i) Does soil nutrient content and microbial abundance vary during the first half of the growing season when plant growth is rapid?
ii) Are there negative associations in the small-scale spatial distribution of soil microbial groups?
iii) Is the spatial distribution of soil microbes and plant biomass related to soil nutrient content?
Materials and Methods
Study site and sampling design
The study was carried out in a boreonemoral forest near Koeru, Estonia (58°58′N; 26°03′E). The climate of the study site is transitional between maritime and continental. The mean annual precipitation in the area is 665 mm and the mean annual air temperature is 5.0 °C (−5.4 °C in January and 16.6 °C in July on average; Estonian Meteorological and Hydrological Institute, 2013). The study site is a 130 ha patch of Hepatica type forest on a calcareous cambisol (cf. ). The soil of the area is relatively uniform in terms of pH, ranging from 5.2–5.7, but exhibits some variation in nutrient content , . The tree layer is dominated by Picea abies (L.) Karst. with individuals of Fraxinus excelsior L. and Acer platanoides L. also present. The shrub layer is dominated by Corylus avellana L. Common species in the herb layer are Oxalis acetosella L., Hepatica nobilis Schreb., Paris quadrifolia L. and Viola mirabilis L. The forest can be classified as relatively undisturbed old-growth forest, but clear cutting of approximately 1–2 ha patches has taken place repeatedly in parts of the study site. Detailed descriptions of the area are provided by Moora et al.  and Zobel et al. . This study complies with the laws of Estonia in which no permits are required to carry out research on public land, unless specified otherwise in legislation. No regulation applies to the Koeru site and no protected species were sampled during the study.
Three 105×105 cm plots (A, B, C) were positioned at 5 m intervals in an area of homogenous forest in terms of vegetation cover and light conditions as estimated by eye. All plots were divided into 49 (15×15 cm) quadrats. During preliminary measurements conducted in summer 2007, five plant individuals of the five most common understory species (Oxalis acetosella, Hepatica nobilis, Fragaria vesca L., Galeobdolon luteum Huds., Viola mirabilis) were excavated to a depth of 15 cm, and the root systems of those plants measured with a ruler to determine the size of an average adult plant root system. Soil cores studied during the preliminary study harboured most of the roots in the upper 10 cm of the soil and the average size of plant roots systems was 15×15 cm (data not presented). The size of quadrats used in the main experiment therefore corresponds to the average size of the root system of an adult plant (the average of the species listed above), and is expected to harbour the majority of the biomass of herbaceous plant roots and microbes present in the rhizosphere. Plot A was used to address temporal variation, and plots A, B and C were used to address spatial variation in soil nutrient content and soil microbial activity at small scales. Plot A was sampled twice, at the beginning (26.05.2008) and in the middle (21.07.2008) of the growing season. The small spatial scale of the study meant that it was not possible to address temporal variation in the abundance of soil nutrients and microbes over a longer time period, as additional sampling would have necessitated considerable disturbance. Soil samples (50 g and 10 g) were collected from the soil surface (depth 10 cm) at both sampling times for soil chemical, and fatty acid analysis respectively. Soil samples collected at beginning of the plant growing season (in plot A only) were excavated with a soil corer (2 cm diameter and 10 cm deep) at the corner of each quadrat (n = 49) in order to minimize disturbance to the plots. The hole was filled with soil taken from nearby, to minimize sampling effects on soil conditions. Collected soil was mixed (to avoid any effect of vertical structure on measured parameters) and subjected to soil chemical and fatty acid analysis (see below). Soil samples collected in the middle of the growing season (July) were collected during destructive harvesting (see below). One core was taken from each quadrat; in plot A, the locations where the first samples were collected were avoided. The soil corer used during destructive harvesting had a diameter of 15 cm and a depth of 10 cm. All soil cores were transported to the laboratory where, after the removal of soil samples (50 and 10 grams as before) from all of the soil cores (n = 147), the shoots and roots of herbaceous plant species were separated carefully by hand from the samples from plots B and C (n = 98). Plant biomass was dried to a constant weight at 70 °C, and weighed. As the vegetation in plot A was disturbed by the sampling at the beginning of the growing season, biomass was not collected from that plot.
Soil chemical and fatty acid analysis
Soil samples for chemical analyses were kept at −80°C prior to analysis, in order to prevent the loss of soil nutrients. Analysis of soil nutrient content followed International Organization for Standardization (ISO) protocols. Specifically, determination of soil ammonium (NH4) and nitrate (NO3) content (mg/kg) was based on ISO 14256–2: 2005. Organic carbon (C) and total nitrogen (N) content (%) were determined based on ISO 10694: 1995 and ISO 13878: 1998, respectively. As soil NH4, NO3 and C content were highly correlated with total N content, only the latter was used in further analysis. The Mehlich III procedure was used to determine soil phosphorus (P) and potassium (K) content (mg/kg): P was extracted by reaction with acetic acid and fluoride compounds; exchangeable K was extracted by the action of ammonium nitrate and nitric acid . Chemical analyses were performed at the laboratory of the Agricultural Research Centre in Saku, Harjumaa, Estonia.
Ester-linked fatty acid (ELFA) content in soil samples was used as a proxy for the biomass of major microbial groups: AM fungi, other fungi and bacteria. Fatty acid profiling is widely used to measure microbial biomass and coarse microbial community composition , –. The abundance of soil microbes can also be estimated by quantifying their DNA by qPCR  but the advantage of fatty acid profiling is that it enables the biomass of different microbial groups to be determined simultaneously. Three grams of fresh soil (randomly taken from collected 10 g) was lyophilized immediately after separation and stored at room temperature. Phospholipids were converted to ester-linked fatty acids and extracted from the soil following the protocol of Schutter & Dick , which methylates only ester-linked and not free fatty acids. Briefly, 15 ml of 0.2 M KOH in methanol was added to 2 g of dry soil, mixed and incubated with shaking at 200 rpm at 37°C for 1 h. The sample was neutralized with 3 ml of 1 M acetic acid. Fatty acid methyl esters were extracted by adding 10 ml of hexane, the sample was mixed by vortexing and left to settle into two layers. The upper hexane layer was transferred to a clean tube and evaporated under a vacuum in a Jouan RC1022 Vacuum Concentrator. Fatty acids were re-suspended in 100 μl hexane and transferred to a GC vial for analysis using gas chromatography.
The ester-linked fatty acid peak areas were normalized against the internal standard and expressed as μg per g of dry soil. Fatty acid nomenclature follows Ranneklev & Bååth . Ester-linked fatty acid 16:1ω5c was used as a biomarker for AM fungi. This marker is widely used as an indicator for AM fungi , , though the different fatty acid fractions of cells may be of different indicative value for AM fungi. Namely, NLFA (neutral lipid fatty acids, storage lipids) 16:1ω5c occurs only in AM fungi, while PLFA (phospholipid fatty acids, structural lipids in cell membranes) 16:1ω5c can also occur in some Gram-negative bacteria , . NLFA is therefore proposed as a more suitable indicator of AM fungi in samples with potentially high-bacterial content, such as soils , . However, ELFA 16:1ω5c has been shown to correlate very well with AM fungal inoculation in an experiment containing both AM fungi and microbes from the AM fungal inocula, where ELFA 16:1ω5c was absent in microbial wash-treated, but AM fungal-free soils . Therefore, ELFA measurements used in this paper represent whole-cell fatty acids, i.e. they contain both the PLFA and NLFA fraction. Fatty acid 18:2ω6,9 was used as a biomarker for fungi other than AM fungi, as it is a dominant fatty acid in most fungi  but is only found at very low levels in AM fungi . We calculated the ratio between the levels of fatty acids 16:1ω5c and 18:2ω6,9 (hereafter AM fungi: other fungi ratio) to estimate the relative abundance of AM fungi in the soil fungal community. Fatty acids i15:0, a15:0, 15:0, i16:0, 16:1ω9c, br17:0, 10Me16:0, i17:0, a17:0, 17:0, cy17:0, br18:0, 10Me17:0, 18:1ω7c, 10Me18:0, cy19:0 ω 9, cy19:0 ω 7 were used as bacterial biomarkers , .
Linear Mixed-Effect Models (LMM)  were used to address temporal differences in soil nutrient content (N, P, K) and soil microbial biomass (fatty acid biomarkers for AM fungi, other fungi, AM fungi: other fungi ratio, bacteria). The abundance of other fungi and bacteria as well as the AM fungi: other fungi ratio was log transformed prior to analysis to meet the assumptions of parametric tests. For each variable, a model using time (sampling time: May or July) as a fixed factor and quadrat identity as a random factor were fitted. The homogeneity assumption for model residuals was evaluated graphically . When any sign of heterogeneity was detected in the residuals, it was tested if keeping the variance fixed between the two time periods resulted in homogeneity for the residuals (i.e. including a fixed variance structure in the model; ). Models were fitted using Restricted Maximum Likelihood (REML) in order to compare models with different variance structures .
Moran's I was used to determine the presence of spatial autocorrelation in the distribution of soil microbes. Moran's I calculates the degree of correlation between observations as a function of the spatial distance separating them . A correlogram is a graphical representation of the Moran's I statistic against distance classes, and provides information about the spatial autocorrelation of a variable and the size of patches. We constructed a spatial correlogram for each group of soil microbes (AM fungi, other fungi and bacteria) in each plot (A, B and C). Moran's I was evaluated at different distance classes (0 to 50 cm) and its significance was assessed by performing 9999 unrestricted permutations. Since the quadrats were distributed regularly, all correlograms were constructed using distance classes with increments of 0.075 m up to a maximum distance of 0.75 m. The first distance class (0.000–0.075) was discarded since it did not contain any pairs of quadrats.
The Modified t-Test for correlation  was used to study pair-wise relationships between the abundance of soil microbes: AM fungi, other fungi and bacteria in each plot (A, B and C). This method evaluates the correlation between two spatial patterns using the Pearson's product-moment correlation, but corrects the significance of the correlation for the presence of spatial autocorrelation.
Generalized least square models (GLS)  were used to determine which factors best explain variations in soil microbial abundance (in plots A, B, C) and both above- and belowground biomass of herbaceous plants (in plots B, C). The set of predictor variables were different depending on the response variable considered. Specifically, in the models of soil microbial abundance (AM fungi, other fungi, bacteria and AM fungi: other fungi ratio), the influence of plot identity (A, B and C) and soil nutrients (N, P, K) were considered. In the models of above- and belowground biomass of herbaceous plants, the influence of plot identity, soil nutrients and microbial abundance were considered. In order to select the most parsimonious model for each variable, the protocol proposed by Zuur et al.  was followed. First, a model which included the full set of covariates as potential predictors was fitted for each variable. After that, the homogeneity of the residuals of the model was evaluated graphically. In addition, due to the nested structure of the sampling design, within plot spatial autocorrelation of residuals was studied. When spatial autocorrelation was present, the inclusion of correlation structures of different form were evaluated . Finally, models were fitted with Maximum Likelihood (ML) and the best set of predictor variables was selected using Akaike's Information Criterion (AIC). If several models showed similar AIC values (difference lower than two), the most parsimonious one, i.e. the one with fewer fixed components, was chosen . The final model was then refitted using REML to obtain estimates of factor effects and their significance.
LMM, GLS and modified t-Tests and correlograms were performed in the R environment , using the packages nlme , SpatialPack  and ncf . Maps of the spatial patterns of AM fungal abundance and soil P content were created using ArcGIS 9.3 software (ESRI Redlands, NY, USA). For each variable, the “geostatistical analyst” extension of ArcGIS 9.3 was used to generate semivariograms of density and create kriging maps.
After the removal of outliers, data from 46 quadrats from plot A was subjected to analysis of temporal variation in soil nutrient content and soil microbial abundance. The linear mixed-effect models (LMMs) revealed significant temporal variation in the abundance of arbuscular mycorrhizal (AM) fungi, bacteria and soil K content, but not in the other variables (Table 1, Figure 1). Means of the measured variables are given in Table S1. The abundance of AM fungal and bacterial fatty acid biomarkers in the soil was higher in May than in July. Soil K content was lower at the beginning of the growing season. The model fitted to explain temporal variation in the abundance of AM fungi included a fixed variance structure as it substantially improved the model fit.
The abundance of arbuscular mycorrhizal (AM) fungi, other fungi and bacteria was estimated as the content of respective ester-linked fatty acid (y-axis on graph A, see Methods for details). Soil N concentration was measured in % (left y-axis of graph B) whereas P and K content were measured as mg/kg (right y-axis of graph B). Significant differences (Linear Mixed-Effect Models; p<0.05) in the measured parameters over time are marked with *.
After the outliers were removed 47, 46 and 44 quadrats from plots A, B and C respectively were subjected to analysis to test the spatial variation of soil nutrient content and the abundance of soil microbes. Limited evidence was found for the presence of spatial autocorrelation in the abundance of soil microbes studied in each plot, since only the abundance of bacteria showed significant autocorrelation (Figure S1). The abundance of AM fungi and soil bacteria tended to show positive spatial autocorrelation up to a distance of 20 cm, but this trend was significant only in the case of bacteria for plots A and C (Figure S1). The Modified t-Test for correlation revealed significant correlations between the abundance of AM fungi and other fungi, AM fungi and bacteria and between bacteria and other fungi (Table 2).
Generalized least square models revealed that plot identity was an important predictor of the abundance of AM fungi (Table 3); the mean abundance was significantly lower in plots A and B compared to plot C (Table 4). In addition, the abundance of AM fungi was significantly positively related to the content of soil N and negatively affected by soil P content (Table 3, Figure 2). Plot identity also played a significant role in predicting the abundance of other fungi (Table 3); the mean abundance was considerably higher in plot C compared to plots A and B (Table 4). In addition, the abundance of other fungi was significantly positively influenced by soil K (Table 3). The AM fungi: other fungi ratio was also predicted by plot identity (Table 3); the mean AM fungi: other fungi ratio differed significantly between all of the plots, being lowest in plot A and highest in plot C (Table 4). In addition, the AM fungi: other fungi ratio was negatively associated with soil P content (Table 3). The abundance of bacteria was the only variable that was best explained by a model where a spatial correlation structure was included, indicating small-scale autocorrelation in bacterial biomass. The abundance of bacteria was explained by plot identity (Table 3), as the mean differed significantly between all plots, being lowest in plot A and highest in plot C (Table 4). In addition, the abundance of bacteria was significantly positively influenced by soil K content (Table 3).
Maps were created by kriging, using data from soil samples collected from 15×15 cm quadrats in each 105×105 cm plot.
Of all the variables considered, the AM fungi: other fungi ratio was the only one with a significant and positive effect on the shoot biomass of herbaceous plants (F = 5.29; P = 0.02). However, the mean shoot biomass of herbaceous plants did not differ between plots (Table 4). None of the studied variables explained the root biomass of herbaceous plants, which also did not vary significantly between plots (Table 4).
The small-scale abundance of soil microbes varied through time in the studied boreonemoral herb-rich spruce forest. We recorded a lower abundance of soil microbes in the middle compared to the beginning of the growing season, whereas no decrease in soil nutrient content occurred. Contrary to our expectations, the small-scale spatial abundance of different soil microbes, including bacteria, arbuscular mycorrhizal (AM) fungi and other fungi, was positively related, suggesting little or no importance of resource competition among these groups in our study system. However, the abundance of soil microbes was significantly influenced by soil nutrient content. In particular, the abundance of AM fungi was strongly negatively related to soil P content. Plant biomass on the other hand showed little variation. Based on these results, we suggest that plants potentially compensate for low soil P content with alternative resource acquisition mechanisms, for example via symbiosis with AM fungi, and therefore achieve equal biomass production despite variations in soil nutrient content and the abundance of soil microbes. Soil microbes (especially plant symbionts such as AM fungi) thus play a fundamental role in the relationship between soil nutrients and plant productivity.
The abundance of soil microbes (especially AM fungi and bacteria) was lower in the middle of the plant growing season, which matches the findings of previous studies that have addressed the temporal variation of soil microbes at larger and smaller spatial scales , . Active plant growth during the first half of the growing season did not bring about a decrease in soil nutrient content. Thus, the importance of soil nutrients in determining temporal changes in the abundance of soil microbes seem to be minor . Some previous studies have reported a contrasting pattern of AM fungal abundance during the plant growing season: an initial increase with a peak in the summer, followed by a decrease in the autumn , . A potential explanation for this inconsistent variation of AM fungal abundance in soil can be the particular weather conditions of the sampling year, as soil AM fungal hyphal density is considerably affected by temperature and precipitation , . Thus, data from long-term measurements with multiple sampling times during the growing season is needed in order to draw solid conclusions about the dynamics of soil microbes.
We found a strong positive association between the abundance of bacteria, AM fungi and other fungi, suggesting a limited role of resource competition among these microbial groups in a natural ecosystem. Several previous experimental studies have reported negative associations between the abundance of ectomycorrhizal fungi, AM fungi and bacteria , , as well as between the abundance of ectomycorrhizal and saprotrophic fungi , and AM fungi and other fungi , . On the other hand, studies conducted in natural conditions have demonstrated similar biomass dynamics of fungi other than AM fungi and bacteria during both the growing season  and over decades . These contrasting results highlight the importance of studies addressing the pattern of soil microbes in natural environments. The strong positive relationship between the abundance of bacteria, AM fungi and other fungi found in this study suggests that the use of slightly different resources by soil microbes promotes their coexistence. Indeed, AM fungi, which are expected not to be highly saprotrophic themselves, have been shown to promote the activity of saprotrophic microbes . In addition, fungal hyphae and spores can function as substrates for bacteria , , and carbon released from dead bacteria can promote the growth of saprotrophic fungi . Also, living roots excrete carbon compounds , which can be consumed by saprotrophic soil microorganisms . In addition, AM fungi as well as ectomycorrhizal fungi have been shown to increase their growth in the presence of root exudates , .
The abundance of all groups of soil microorganisms tested depended on the sampling location (plot identity), which confirms the small-scale spatial heterogeneity of soil microbes previously demonstrated . In addition, soil nutrients played an important role in determining the abundance of soil microorganisms. The negative effect of soil P content on the abundance of AM fungi has also been reported in earlier field fertilization experiments in grasslands and forests , . As AM fungi can improve the acquisition of soil nutrients, especially P, by plants , , they can play an important role when soil nutrient content is low. When soil nutrient levels increase, the cost to a plant of supporting AM fungi may outweigh the benefits, and can result in a decrease of AM fungal colonization of plant roots . However, soil AM fungal fatty acid content has also been reported to be positively or insignificantly related to soil P content , . We recorded a positive effect of soil N on the abundance of AM fungi, a pattern that has also been registered at a large scale  and can be explained by the high N demand of AM fungi . In addition, AM fungal communities have been shown to be highly influenced by soil pH . However, Dumbrell et al.  focused on sites with high variation in soil pH (ranging from 4.5–8), whereas our study area is characterised by fairly uniform pH values , . Therefore, the role of soil nutrients in determining the abundance of AM fungi in current study site could be expected to be more pronounced, while the effect of small-scale variation in soil pH remains to be fully considered in future studies.
Saprotrophic and ectomycorrhizal fungi as well as bacteria were positively influenced by soil K content. A recent study demonstrated an increase in bacterial biomass, and growth of extramatrical mycelia of ectomycorrhizal fungi in response to K addition in peatland soils with severe K deficiency . Our results suggest that K also has an important role to play in determining the abundance of soil microbes in soils where K content is not extremely low. On the other hand, the observed positive relationship can also reflect increased weathering and K influx resulting from active growth of bacteria and saprotrophic fungi , .
Plants have been shown to concentrate their biomass into patches with high nutrient content –, but in the current study, conducted in an old and stable ecosystem, we found no direct effect of nutrients on plant biomass. One explanation for the absence of the expected pattern is that plants are continuously foraging, while the pattern of biomass measured in studies such as this one reflects dynamics that have operated over numerous years. Greenhouse experiments have shown that nutrient poor patches contain higher plant biomass when next to nutrient rich patches ,  but those experiments were short-term. In the longer term, there will be natural turnover of nutrient rich patches (both temporal and spatial), and active foraging can therefore result in relatively uniform plant biomass. Secondly, the natural variation in nutrient content could be too low to generate distinct patterns. Indeed, understory plant biomass in this study area has not shown changes in response to induced variation in soil nutrient content either . Thirdly, the effect of soil nutrients can be mitigated by the influence of soil microbes such as AM fungi. AM fungi seemed to be the only group of soil microbes to affect understory plant growth, as the relative increase of AM fungi compared to other fungi resulted in increased plant shoot biomass. As AM fungi can greatly improve plant P acquisition , , the presence of AM fungi can be more important for plant growth when soil P content is low and less beneficial for plants when soil nutrient content is high. In accordance with this, a decrease of AM fungal hyphal density in soil and AM colonization in plant roots has been reported in response to increased P availability , ; whereas the nutrient content of plant tissues and plant biomass did not exhibit any change , . Our study, based on data from a natural environment, demonstrates stable plant biomass in conditions of low abundance of AM fungi and high P content and vice versa. We support the suggestion of Beauregard et al.  that plants can compensate for low available nutrient concentration with more efficient symbiosis with soil microbes. Hence, it is crucial to consider both the soil nutrient content and the abundance of soil microbes when exploring the mechanisms driving vegetation patterns.
Spatial correlograms of microbes. Spatial correlograms of arbuscular mycorrhizal fungi (a–c), other fungi (d–f), and bacteria (g–i) in plot A (a, d, g), B (b, e, h) and C (c, f, i). Distances with significant spatial autocorrelation are marked with solid circles.
Temporal variation in the abundance of soil nutrients and microbes. Summary statistics of the soil nutrient content and ester-linked fatty acid (ELFA) biomarkers of arbuscular mycorrhizal (AM) fungi, other fungi and bacteria measured from plot A at the beginning (May) and in the middle (July) of the growing season. The mean, standard deviation (SD) and the range are presented for each variable.
We thank Sandra Caul at The James Hutton Institute, Dundee, UK, for help with fatty acid analyses. We are grateful to Jodi Price and John Davison for their comments on earlier drafts of this manuscript.
Conceived and designed the experiments: MZ MM MÖ ÜS AU KK. Performed the experiments: KK MÖ MM ÜS AU VS MZ. Analyzed the data: AG MZ MÖ MM KK. Contributed reagents/materials/analysis tools: AG MM MZ MÖ. Wrote the paper: KK AG MÖ MM ÜS AU VS MZ.
- 1. Farley RA, Fitter AH (1999) Temporal and spatial variation in soil resources in a deciduous woodland. J. Ecol. 87: 688–696.
- 2. Pausas JG, Carreras J, Ferre A, Font X (2003) Coarse-scale plant species richness in relation to environmental heterogeneity. J. Veg. Sci. 14: 661–668.
- 3. Hutchings MJ, Wijesinghe DK (1997) Patchy habitats, division of labour and growth dividends in clonal plants. Trends Ecol. Evol. 12: 390–394.
- 4. Day KJ, Hutchings MJ, John EA (2003) The effects of spatial pattern of nutrient supply on yield, structure and mortality in plant populations. J. Ecol. 91: 541–553.
- 5. Hodge A (2004) The plastic plant: root responses to heterogeneous supplies of nutrients. New Phytol. 162: 9–24.
- 6. van der Heijden MGA, Bardgett RD, van Straalen NM (2008) The unseen majority: soil microbes as drivers of plant diversity and productivity in terrestrial ecosystems. Ecol. Lett. 11: 296–310.
- 7. Smith SE, Read DJ (2008) Mycorrhizal Symbiosis. Third edition. Academic Press, Amsterdam
- 8. Ettema CH, Wardle DA (2002) Spatial soil ecology. Trends Ecol. Evol. 17: 177–183.
- 9. Baldrian P, Vetrovsky T (2012) Scaling down the analysis of environmental processes: monitoring enzyme activity in natural substrates on a millimeter resolution scale. Appl. Environ. Microbiol. 78: 3473–3475.
- 10. Olsson PA, Bååth E, Jakobsen I (1997) Phosphorus effects on the mycelium and storage structures of an arbuscular mycorrhizal fungus as studied in the soil and roots by analysis of fatty acid signatures. Appl. Environ. Microbiol. 63: 3531–3538.
- 11. Nilsson LO, Bååth E, Falkengren-Grerup U, Wallander H (2007) Growth of ectomycorrhizal mycelia and composition of soil microbial communities in oak forest soils along a nitrogen deposition gradient. Oecologia 153: 375–384.
- 12. Franklin RB, Mills AL (2009) Importance of spatially structured environmental heterogeneity in controlling microbial community composition at small spatial scales in an agricultural field. Soil Biol. Biochem. 41: 1833–1840.
- 13. Kaiser C, Koranda M, Kitzler B, Fuchslueger L, Schnecker J, et al. (2010) Belowground carbon allocation by trees drives seasonal patterns of extracellular enzyme activities by altering microbial community composition in a beech forest soil. New Phytol. 187: 843–858.
- 14. Paul EA, Clark FE (1996) Soil Microbiology and Biochemistry. Second edition. Academic Press, San Diego.
- 15. Klironomos JN, Rillig MC, Allen MF (1999) Designing belowground field experiments with the help of semi-variance and power analyses. Appl. Soil Ecol. 12: 227–238.
- 16. Mummey DL, Rillig MC (2008) Spatial characterization of arbuscular mycorrhizal fungal molecular diversity at the submetre scale in a temperate grassland. FEMS Microbiol. Ecol. 64: 260–270.
- 17. Baldrian P, Merhautova V, Cajthaml T, Petrankova M, Snajdr J (2010) Small-scale distribution of extracellular enzymes, fungal, and bacterial biomass in Quercus petraea forest topsoil. Biol. Fertil. Soils 46: 717–726.
- 18. Rousk J, Bååth E (2011) Growth of saprotrophic fungi and bacteria in soil. FEMS Microbiol. Ecol. 78: 17–30.
- 19. Lecomte J, St-Arnaud M, Hijri M (2011) Isolation and identification of soil bacteria growing at the expense of arbuscular mycorrhizal fungi. FEMS Microbiol. Lett. 317: 43–51.
- 20. Leigh J, Fitter AH, Hodge A (2011) Growth and symbiotic effectiveness of an arbuscular mycorrhizal fungus in organic matter in competition with soil bacteria. FEMS Microbiol. Ecol. 76: 428–438.
- 21. Chagnon P-L, Bradley RL (2013) Evidence that soil nutrient stoichiometry controls the competitive abilities of arbuscular mycorrhizal vs. root-borne non-mycorrhizal fungi. Fungal Ecol. 6: 557–660.
- 22. McNaughton SJ, Oesterheld M (1990) Extramatrical mycorrhizal abundance and grass nutrition in a tropical grazing ecosystem, the Serengeti National Park, Tanzania. Oikos 59: 92–96.
- 23. Koorem K, Saks Ü, Sõber V, Uibopuu A, Öpik M, et al. (2012) Effects of arbuscular mycorrhiza on community composition and seedling recruitment in temperate forest understory. Basic Appl. Ecol. 13: 663–672.
- 24. Moora M, Daniell TJ, Kalle H, Liira J, Püssa K, et al. (2007) Spatial pattern and species richness of boreonemoral forest understorey and its determinants – a comparison of differently managed forests. For. Ecol. Manage. 250: 64–70.
- 25. Öpik M, Metsis M, Daniell TJ, Zobel M, Moora M (2009) Large-scale parallel 454 sequencing reveals host ecological group specificity of arbuscular mycorrhizal fungi in a boreonemoral forest. New Phytol. 184: 424–437.
- 26. Davison J, Öpik M, Daniell TJ, Moora M, Zobel M (2011) Arbuscular mycorrhizal fungal communities in plant roots are not random assemblages. FEMS Microbiol. Ecol. 78: 103–115.
- 27. Zobel M, Kalamees R, Püssa K, Roosaluste E, Moora M (2007) Soil seed bank and vegetation in mixed coniferous forest stands with different disturbance regimes. For. Ecol. Manage. 250: 71–76.
- 28. Koorem K, Moora M (2010) Positive association between understory species richness and a dominant shrub species (Corylus avellana) in a boreonemoral spruce forest. For. Ecol. Manage. 260: 1407–1413.
- 29. Sen Tran T, Simard RR (1993) Mehlich III-extractable elements. In: Carter MR, editor. Soil sampling and methods of analysis. pp. 43–48.
- 30. Graham JH, Hodge NC, Morton JB (1995) Fatty acid methyl ester profiles for characterization of Glomalean fungi and their endomycorrhizae. Appl. Environ. Microbiol. 61: 58–64.
- 31. Frostegård A, Bååth E (1996) The use of phospholipid fatty acid analysis to estimate bacterial and fungal biomass in soil. Biol. Fertil. Soils 22: 59–65.
- 32. Hinojosa MB, Carreira JA, Garcia-Ruiz R, Dick RP (2005) Microbial response to heavy metal-polluted soils: community analysis from phospholipid-linked fatty acids and ester-linked fatty acids extracts. J. Environ. Qual. 34: 1789–1800.
- 33. De Deyn GB, Quirk H, Bardgett RD (2011) Plant species richness, identity and productivity differentially influence key groups of microbes in grassland soils of contrasting fertility. Biol. Lett. 23: 75–78.
- 34. Schnoor TK, Martensson LM, Olsson PA (2011) Soil disturbance alters plant community composition and decreases mycorrhizal carbon allocation in a sandy grassland. Oecologia 167: 809–819.
- 35. Birkhofer K, Schöning I, Alt F, Herold N, Klarner B, et al. (2012) General relationships between abiotic soil properties and soil biota across spatial scales and different land-use types. PLoS ONE 7: e43292.
- 36. Schutter ME, Dick RP (2000) Comparison of fatty acid methyl ester (FAME) methods for characterizing microbial communities. Soil Sci. Soc. Am. J. 64: 1659–1668.
- 37. Ranneklev SB, Bååth E (2003) Use of phospholipid fatty acids to detect previous self-heating events in stored peat. Appl. Environ. Microbiol. 69: 3532–3539.
- 38. Nichols P, Stulp BK, Jones JG, White DC (1986) Comparison of fatty acid content and DNA homology of the filamentous gliding bacteria Vitreoscilla, Flexibacter, Filibacter. Arch. Microbiol. 146: 1–6.
- 39. Olsson PA (1999) Signature fatty acids provide tools for determination of the distribution and interactions of mycorrhizal fungi in soil. FEMS Microbiol. Ecol. 29: 303–310.
- 40. Frostegård A, Tunlid A, Bååth E (2011) Use and misuse of PLFA measurements in soils. Soil Biol. Biochem. 43: 1621–1625.
- 41. Walder F, Niemann H, Lehmann MF, Boller T, Wiemken A, et al. (2013) Tracking the carbon source of arbuscular mycorrhizal fungi colonizing C3 and C4 plants using carbon isotope ratios (delta 13C). Soil Biol. Biochem. 58: 341–344.
- 42. Zelles L (1999) Fatty acid patterns of phospholipids and lipopolysaccharides in the characterisation of microbial communities in soil: a review. Biol. Fertil. Soils 29: 111–129.
- 43. Pinheiro JC, Bates DM (2000) Mixed-Effect Models in S and S-PLUS. Springer.
- 44. Zuur AF, Ieno EN, Walker NJ, Saveliev AA, Smith GM (2009) Mixed effects models and extensions in ecology with R. Springer, New York.
- 45. Legendre P, Legendre L (1998) Numerical Ecology. Second English edition. Elsevir Science, Amsterdam.
- 46. Dutilleul P, Clifford P, Richardson S, Hemon D (1993) Modifying t test for assessing the correlation between two spatial processes. Biometrics 49: 305–314.
- 47. Dormann CF, McPherson JM, Araujo MB, Bivand R, Bolliger J, et al. (2007) Methods to account for spatial autocorrelation in the analysis of species distributional data: a review. Ecography 30: 609–628.
- 48. R Development Core Team (2011) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org/ Accessed 2013 February 27
- 49. Pinheiro J, Bates D, DebRoy S, Sarkar D (2011) the R Development Core Team (2011) Nlme: linear and nonlinear mixed effects models. R package version 3: 1–98.
- 50. Osorio F, Vallejos R, Cuevas F (2012) SpatialPack: package for analysis of spatial data.
- 51. Bjornstad ON (2012) Ncf: spatial nonparametric covariance functions. R package version 1.1–4.
- 52. Hamel C, Hanson K, Selles F, Cruz AF, Lemke R, et al. (2006) Seasonal and long-term resource-related variations in soil microbial communities in wheat-based rotations of the Canadian prairie. Soil Biol. Biochem. 38: 2104–2116.
- 53. Kabir Z, O'Halloran IP, Fyles JW, Hamel C (1998) Dynamics of the mycorrhizal symbiosis of corn (Zea mays L.): effects of host physiology, tillage practice and fertilization on spatial distribution of extra-radical mycorrhizal hyphae in the field. Agric. Ecosyst. Environ. 68: 151–163.
- 54. Staddon PL, Thompson K, Jakobsen I, Grime JP, Askew AP, et al. (2003) Mycorrhizal fungal abundance is affected by long-term climatic manipulations in the field. Glob. Chang. Biol. 9: 186–194.
- 55. Hu Y, Rillig MC, Xiang D, Hao Z, Chen B (2013) Changes of AM fungal abundance along environmental gradients in the arid and semi-arid grasslands of Northen China. PLoS ONE 8: e57593.
- 56. Olsson PA, Chalot M, Bååth E, Finlay RD, Söderström B (1996) Ectomycorrhizal mycelia reduce bacterial activity in a sandy soil. FEMS Microbiol. Ecol. 21: 77–86.
- 57. Pereira E, Coelho V, Tavares RM, Lino-Neto T, Baptista P (2012) Effect of competitive interactions between ectomycorrhizal and saprotrophic fungi on Castanea sativa performance. Mycorrhiza 22: 41–49.
- 58. Olsson PA, Francis R, Read DJ, Söderström B (1998) Growth of arbuscular mycorrhizal mycelium in calcareous dune sand and its interaction with other soil microorganisms as estimated by measurement of specific fatty acids. Plant Soil 201: 9–16.
- 59. Campbell CD, Cameron CM, Bastias BA, Chen CR, Cairney JWG (2008) Long term repeated burning in a wet sclerophyll forest reduces fungal and bacterial biomass and responses to carbon substrates. Soil Biol. Biochem. 40: 2246–2252.
- 60. Cheng L, Booker FL, Tu C, Burkey KO, Zhou L, et al. (2012) Arbuscular mycorrhizal fungi increase organic carbon decomposition under elevated CO2. Science 337: 1084–1087.
- 61. Roesti D, Ineichen K, Braissant O, Redecker D, Wiemken A, et al. (2005) Bacteria associated with spores of the arbuscular mycorrhizal fungi Glomus geosporum and Glomus constrictum. Appl. Environ. Microbiol. 71: 6673–6679.
- 62. Jones DL, Hodge A, Kuzyakov Y (2004) Plant and mycorrhizal regulation of rhizodeposition. New Phytol. 163: 459–480.
- 63. Buee M, Rossignol M, Jauneau A, Ranjeva R, Becard G (2000) The pre-symbiotic growth of arbuscular mycorrhizal fungi is induced by a branching factor partially purified from plant root exudates. Mol. Plant Microbe Interact. 13: 693–698.
- 64. Kikuchi K, Matsushita N, Suzuki K, Hogetsu T (2007) Flavonoids induce germination of basidiospores of the ectomycorrhizal fungus Suillus bovinus. Mycorrhiza 17: 563–570.
- 65. Groffman PM, Fisk MC (2011) Phosphate additions have no effect on microbial biomass and activity in a northern hardwood forest. Soil Biol. Biochem. 43: 2441–2449.
- 66. Liu YJ, Shi GX, Mao L, Cheng G, Jiang SJ, et al. (2012) Direct and indirect influences of 8 yr of nitrogen and phosphorus fertilization on Glomeromycota in an alpine meadow ecosystem. New Phytol. 194: 523–535.
- 67. Smith SE, Smith FA, Jakobsen I (2003) Mycorrhizal fungi can dominate phosphate supply to plants irrespective of growth responses. Plant Physiol. 133: 16–20.
- 68. Beauregard MS, Hamel C, Nayyar A, St-Arnaud M (2010) Long-term phosphorus fertilization impacts soil fungal and bacterial diversity but not AM fungal community in alfalfa. Microb. Ecol. 59: 379–389.
- 69. Thoms C, Gattinger A, Jacob M, Thomas FM, Gleixner G (2010) Direct and indirect effects of tree diversity drive soil microbial diversity in temperate deciduous forest. Soil Biol. Biochem. 42: 1558–1565.
- 70. Hodge A, Fitter AH (2010) Substantial nitrogen acquisition by arbuscular mycorrhizal fungi from organic material has implications for N cycling. Proc. Natl. Acad. Sci. U.S.A. 107: 13754–13759.
- 71. Dumbrell AJ, Nelson M, Helgason T, Dytham C, Fitter AH (2010) Relative roles of niche and neutral processes in structuring a soil microbial community. ISME J. 4: 337–345.
- 72. Potila H, Wallander H, Sarjala T (2009) Growth of ectomycorrhizal fungi in drained peatland forests with variable P and K availability. Plant Soil 316: 139–150.
- 73. Hoffland E, Giesler R, Jongmans AG, van Breemen N (2003) Feldspar tunneling by fungi along natural productivity gradients. Ecosystems 6: 739–746.
- 74. Calvaruso C, Turpault MP, Frey-Klett P (2006) Root-associated bacteria contribute to mineral weathering and to mineral nutrition in trees: a budgeting analysis. Appl. Environ. Microbiol. 72: 1258–1266.
- 75. Gazol A, Tamme R, Price JN, Hiiesalu I, Laanisto L, et al. (2013) A negative heterogeneity–diversity relationship found in experimental grassland communities. Oecologia 173: 545–555.
- 76. Nayyar A, Hamel C, Forge T, Selles F, Jefferson PG, et al. (2008) Arbuscular mycorrhizal fungi and nematodes are involved in negative feedback on a dual culture of alfalfa and Russian wildrye. Appl. Soil Ecol. 40: 30–60.