Rapid Transfer of Plant Photosynthates to Soil Bacteria via Ectomycorrhizal Hyphae and Its Interaction With Nitrogen Availability

Abstract

Plant roots release recent photosynthates into the rhizosphere, accelerating decomposition of organic matter by saprotrophic soil microbes ("rhizosphere priming effect") which consequently increases nutrient availability for plants. However, about 90% of all higher plant species are mycorrhizal, transferring a significant fraction of their photosynthates directly to their fungal partners. Whether mycorrhizal fungi pass on plant-derived carbon (C) to bacteria in root-distant soil areas, i.e., incite a "hyphosphere priming effect," is not known. Experimental evidence for C transfer from mycorrhizal hyphae to soil bacteria is limited, especially for ectomycorrhizal systems. As ectomycorrhizal fungi possess enzymatic capabilities to degrade organic matter themselves, it remains unclear whether they cooperate with soil bacteria by providing photosynthates, or compete for available nutrients. To investigate a possible C transfer from ectomycorrhizal hyphae to soil bacteria, and its response to changing nutrient availability, we planted young beech trees (Fagus sylvatica) into "split-root" boxes, dividing their root systems into two disconnected soil compartments. Each of these compartments was separated from a litter compartment by a mesh penetrable for fungal hyphae, but not for roots. Plants were exposed to a ¹³C-CO₂-labeled atmosphere, while ¹⁵N-labeled ammonium and amino acids were added to one side of the split-root system. We found a rapid transfer of recent photosynthates via ectomycorrhizal hyphae to bacteria in root-distant soil areas. Fungal and bacterial phospholipid fatty acid (PLFA) biomarkers were significantly enriched in hyphae-exclusive compartments 24 h after ¹³C-CO₂-labeling. Isotope imaging with nanometer-scale secondary ion mass spectrometry (NanoSIMS) allowed for the first time in situ visualization of plant-derived C and N taken up by an extraradical fungal hypha, and in microbial cells thriving on hyphal surfaces. When N was added to the litter compartments, bacterial biomass, and the amount of incorporated ¹³C strongly declined. Interestingly, this effect was also observed in adjacent soil compartments where added N was only available for bacteria through hyphal transport, indicating that ectomycorrhizal fungi were acting on soil bacteria. Together, our results demonstrate that (i) ectomycorrhizal hyphae rapidly transfer plant-derived C to bacterial communities in root-distant areas, and (ii) this transfer promptly responds to changing soil nutrient conditions.

Keywords: NanoSIMS; PLFAs; ectomycorrhiza; hyphal carbon transfer; hyphosphere bacteria; hyphosphere priming; mycorrhizosphere.

Figure 1

Design of split–root boxes. Plants grew in two separated soil compartments where the root system was divided into two halves. Stems were stabilized in a cylinder filled with quartz sand. Each soil compartment was connected to a leaf litter compartment, separated by a mesh (35 μm) which is penetrable by fungal hyphae, but not by roots. Plant canopies were exclusively exposed to a ¹³C-CO₂ enriched atmosphere. A mixture of ¹⁵N-labeled ammonium and amino acids was added to only one litter compartment per box. (a) Schematic drawing of the setup. One of the two litter compartments is drawn without litter to illustrate the chamber design with the separating mesh. For the experiment both outer compartments were filled with beech litter. (b) Photo of a plant used in the labeling experiment.

Figure 2

Bar graph illustrating the proportion of ECM fungal orders in total ECM fungal communities in litter samples derived from planted (dark blue) and unplanted boxes (gray), as well as from the rhizosphere soil (light blue) as average proportion of the libraries ± standard error (SE). The insert graph depicts the proportion of ECM fungi in the total fungal communities (average proportion of the libraries ± SE).

Figure 3

Relative enrichment of (A) ¹³C and (B) ¹⁵N isotopes in plant, soil, and litter samples 24 h after ¹³C-CO₂ labeling of plants and 48 h after addition of ¹⁵N-labeled NH₄ and amino acids to litter compartments, indicating transfer of these supplies through the experimental system. Colors indicate the experimental system parts, i.e., plant (green), soil (red), and litter (blue). Dashed lines represent natural abundance values (mean of the respective system part). Symbols show results from dual-labeled (¹³C and ¹⁵N) plant boxes. Open circles, untreated side of the box; closed circles, N-treated side. This “side” designation refers only to roots, litter, and soil, as aboveground plant parts (i.e., leaves and stem) are connected to both sides. Error bars represent standard error. Asterisks indicate significant difference from natural abundance (^*p < 0.05, ^**p < 0.01, ^***p < 0.001; Mann-Whitney U-test; n = 7).

Figure 4

Relative enrichment of ¹³C in fungal-specific (18:1ω9 and 18:2ω6,9) and bacteria-specific PLFAs in rhizosphere, bulk soil, and litter pools shown as atom% excess ¹³C (calculated by subtracting the natural abundance atom% value of the respective PLFA biomarkers). Bars show the weighted mean of each pool from N-treated and untreated sides from dual-labeled (¹³C and ¹⁵N) plant boxes. This format was chosen to emphasize general trends of relative ¹³C enrichment in fungal and bacterial groups across the different pools irrespective of N addition. No significant differences in atom% excess ¹³C could be detected between the N-treated and the untreated sides (+N/noN; Mann-Whitney U-test for paired samples). For visualization of differences between +N/noN box sides see Figure S5. Litter compartment enrichment indicates allocation of ¹³C via ECM hyphae to root-distant bacteria. Error bars represent the standard error; n = 6. Differences in pools were analyzed with the Kruskal-Wallis rank sum test. Significant tests (p < 0.05) were followed by Dunn's post-hoc test of multiple comparisons with Bonferroni correction (adj. p < 0.05); significant differences are indicated by lowercase letters. All microbial groups were significantly enriched in ¹³C in all pools, except for 18:2ω6,9, Gram-negative bacteria and Actinobacteria, each in bulk soil (comparison of atom% ¹³C values via Mann-Whitney U-test, p < 0.05; not shown).

Figure 5

NanoSIMS and SEM images of a root-distant fungal hypha extracted from a litter compartment of a dual-labeled plant box, depicting ¹³C and ¹⁵N labeled regions within the hypha. The SEM image shows the morphology of the sample prior to NanoSIMS measurement. NanoSIMS images were acquired at an erosion depth below the hyphal surface, visualizing the presence of isotopically enriched compounds inside the hypha. Isotopic label contents are displayed as atom%. The white arrows at the color-scales indicate the natural isotopic abundance values, determined on the unlabeled control (1.08 atom% ¹³C and 0.37 atom% ¹⁵N). The upper limit of the atom% ¹³C scale is set to 2.2; however, maximum local values, extracted from individual cycles, range up to 3.6 atom% ¹³C. Secondary ion signal intensity thresholds were set to 62 and 60 counts/(sec^*pixel) for ¹²C¹³C⁻ and ¹²C¹⁵N⁻, respectively. SEM, scanning electron microscopy image (secondary electrons); C⁻, accumulated ¹²C⁻ and ¹³C⁻ secondary ion signal intensity distribution images, indicating the morphology of the sample during NanoSIMS analysis; at% ¹³C, carbon isotope composition image; at% ¹⁵N, nitrogen isotope composition image. Overlay images are composites of the C⁻ and isotope images. NanoSIMS images consist of accumulated z-stacks obtained from 19 consecutive scans (displayed in Animation S1). Scale bars, 5 μm.

Figure 6

Boxplots depicting the ¹³C (A) and ¹⁵N (B) isotopic content (in atom%) for a fungal hypha (displayed in Figure 5) and hyphosphere microorganisms (displayed in Figure 7) from dual-labeled plant-boxes, along with the values obtained from measurement on the natural abundance control. Each data point represents an individual region of interest (ROI) extracted from NanoSIMS images depicted in Figures 5, 7, error bars refer to the estimated analytical uncertainty (1σ) due to counting statistics (see section Materials and Methods). Significant levels of isotopic enrichment (p < 0.001) are indicated with an asterisk (^*). Data for the natural abundance control were obtained from microbial cells associated to the surface of a hypha extracted from a non-labeled control box. With respect to the values of the fungal hypha it should be noted that these refer to the averages over 19 consecutive scans. Owing to the observed variation of enrichment within the z-stack (see Animation S1), the displayed maxima rather represent conservative estimates.

Figure 7

NanoSIMS and SEM imaging of cells attached to root-distant fungal hyphae extracted from a litter compartment of a dual-labeled (¹³C and ¹⁵N) plant-box. SEM, scanning electron microscopy image (secondary electrons); C⁻, accumulated ¹²C⁻ and ¹³C⁻ secondary ion signal intensity distribution images, indicating the morphology of the sample during NanoSIMS analysis; at% ¹³C, carbon isotope composition image, showing ¹³C enrichment in microbial cells on the surface of the hyphae; at% ¹⁵N, nitrogen isotope composition image. Isotopic label contents are displayed as atom%. The white arrows at the color-scales indicate the natural isotopic abundance values, determined on the unlabeled control (1.08 atom% ¹³C and 0.37 atom% ¹⁵N). The upper limit of the atom% ¹³C scale is set to 1.50, maximum values range up to 1.56 atom%. Secondary ion signal intensity thresholds were set to 119, 51 and 58 counts/(sec^*pixel) for ¹²C¹³C⁻, ¹²C¹⁵N⁻, and ³¹P⁻ respectively. (P/C)_rel., relative phosphor-to-carbon elemental ratio image as inferred from C⁻ normalized ³¹P⁻ secondary ion signal intensities, indicating the presence of microbial cells on hyphal surfaces. Overlay images are composites of the SEM and NanoSIMS images. NanoSIMS images consist of accumulated z-stacks obtained from 24 consecutive scans. Scale bars, 5 μm.

Figure 8

Biomass of fungi and bacterial groups in rhizosphere, bulk soil and litter pools of N-treated and untreated sides of dual-labeled (¹³C and ¹⁵N) plant boxes, represented by PLFA biomarkers. Biomass of Actinobacteria and Gram-positive bacteria (excluding the Actinobacterial PLFA) declined significantly in both litter and soil compartments with N addition (i.e., to the litter compartment). Significant differences between N-treated (+N) and untreated side (no N) are indicated with asterisks (Mann-Whitney U-test for paired samples; p < 0.05; n = 6). Although not statistically significant, the adverse effect of N addition is also present in Gram-negative bacteria in litter, and the fungal PLFA 18:1ω9 in litter and bulk soil (p < 0.01). Error bars represent the standard error. Differences between soil and litter pools of the untreated side were analyzed with the Kruskal-Wallis rank sum test. Significant test results (p < 0.05) were followed by Dunn's post-hoc test of multiple comparisons with Bonferroni correction (adj. p < 0.05). Rhizosphere soil and litter, as well as bulk soil and litter, were significantly different in all groups except for Actinobacteria (not shown).

Figure 9

Correspondence analysis (CA) of C in PLFAs (μg C g⁻¹ dry weight) in N-treated and untreated compartments of dual-labeled (¹³C and ¹⁵N) plant boxes. Close distances between individual PLFAs (depicted as text) and soil/litter-pools of individual split-root boxes (depicted as symbols) indicate higher abundance of respective PLFAs in concerned pools. Axes notations give the proportion of variance explained on each coordinate in percent. Colors refer to box-sides with untreated (no N) and N-treated (+N) litter compartments. ANOSIM analysis shows significant difference between pools (R = 0.566, p = 0.001) as well as between N-treatments (R = 0.103, p = 0.011). This analysis indicates (i) a distinctly different microbial community between soil pools (rhizosphere soil, bulk soil) and litter, and (ii) that N addition affects not only community structure in hyphae-only litter compartments, but also in soil compartments (cf. Figure 8).

Figure 10

¹³C-excess in PLFAs per g dry weight in N-treated and untreated compartments of dual-labeled (¹³C and ¹⁵N) plant boxes. Enrichment of bacteria-specific PLFAs in the litter compartment indicates allocation of photoassimilated ¹³C via ECM hyphae to root-distant bacteria. Significant differences between untreated (no N) and N-treated side (+N) are indicated with asterisks (Mann-Whitney U-test for paired samples; p > 0.05; n = 6). Error bars represent the standard error.