Dead organic matter (OM) is a major source of nitrogen (N) for plants. The majority of plants support N uptake by symbiosis with arbuscular mycorrhizal (AM) fungi. Mineralization of N is regulated by microfauna, in particular, protozoa grazing on bacteria. We hypothesized that AM fungi and protozoa interactively facilitate plant N nutrition from OM.
In soil systems consisting of an OM patch and a root compartment, plant N uptake and consequences for plant carbon (C) allocation were investigated using stable isotopes.
Protozoa mobilized N by consuming bacteria, and the mobilized N was translocated via AM fungi to the host plant. The presence of protozoa in both the OM and root compartment stimulated photosynthesis and the translocation of C from the host plant via AM fungi into the OM patch. This stimulated microbial activity in the OM patch, plant N uptake from OM and doubled plant growth.
The results indicate that protozoa increase plant growth by both mobilization of N from OM and by protozoa–root interactions, resulting in increased C allocation to roots and into the rhizosphere, thereby increasing plant nutrient exploitation. Hence, mycorrhizal plants need to interact with protozoa to fully exploit N resources from OM.
Nitrogen (N) is the primary limiting element for plant growth in most terrestrial ecosystems (Vitousek & Howarth, 1991). In soil, most N is unavailable for plant uptake as it is bound in dead organic matter (OM) and patchily distributed. The majority of plants form associations with arbuscular mycorrhizal (AM) fungi to support the capture and uptake of N in exchange for carbon (C)-rich photosynthates (Hodge et al., 2001; Heinemeyer et al., 2006; Smith & Read, 2008). Although AM fungi are not able to mineralize N from OM on their own, their extraradical hyphae are highly efficient in the acquisition and translocation of inorganic N to host plants (Govindarajulu et al., 2005). Protozoan grazing stimulates microbial decomposition of patchily distributed OM (Hodge et al., 1998; Bonkowski et al., 2000) by excreting N (as ) from consumed bacterial biomass (Kuikman & Van Veen, 1989; Bonkowski, 2004). The excreted N is easily accessible to roots and promotes plant growth (Clarholm, 1985; Bonkowski, 2004). As has been shown for ectomycorrhizal fungi (Bonkowski et al., 2001), protozoa increase the availability of N from OM patches for mycorrhizal uptake, and may foster transportation of the mobilized N via fungal hyphae to plant roots. In addition, selective grazing on bacteria by protozoa may affect root growth indirectly and increase allocation of photosynthates to roots (Bonkowski & Brandt, 2002; Krome et al., 2009), thereby strengthening the AM fungi–plant symbiosis.
Despite the ubiquity of both AM fungi and protozoa, and their importance for plant N nutrition and growth, only a few studies have investigated their interactions (Rønn et al., 2002; Wamberg et al., 2003; Herdler et al., 2008; Vestergard et al., 2008). Studies examining the interplay between AM fungi, protozoa and roots exploiting nutrients from OM patches are lacking entirely.
We tested the following hypotheses: the grazing of protozoa on bacteria mobilizes N from OM, and this N is captured by AM fungi and transported to plant roots; the presence of protozoa in the root vicinity, in addition to their presence in the OM patch, increases plant N capture from OM via the enhancement of plant C fixation and C investment into rhizosphere interactions. To test these hypotheses, we set up soil microcosms with plants (Plantago lanceolata L.) consisting of one compartment with roots and a second with an OM patch with restricted access to AM fungal hyphae. To evaluate the role of protozoa in plant N capture and plant growth, the presence of protozoa in the two compartments was manipulated. The capture of N from OM by plants was investigated using OM labeled with 15N. To investigate plant investment into belowground interactions, we followed the translocation of plant C into roots, AM fungi and OM after labeling of the plants with 13CO2.
Materials and Methods
Experimental system and set-up
Experimental systems consisted of two 250-ml cell culture flasks (NUNC, Langenselbold, Germany), each containing a moist soil–sand mixture (ratio 1 : 1) equivalent to 320 g dry weight (Supporting Information Fig. S1). The two compartments were connected by openings (5.5 cm in diameter) facing each other. One compartment was planted (root compartment) and the other contained an OM patch (OM patch compartment). The opening was covered by a double layer of 0.45-μm blotting membrane (Sartorius, Göttingen, Germany) to prevent access by both roots and AM fungi (control treatment; n =6) or by a double layer of 20-μm mesh (Gebr. Stallmann; Rellingen, Hamburg, Germany) allowing access by AM fungi, but not by roots (treatment with only mycorrhiza in OM patch; n =6). To investigate the role of protozoa for N capture by mycorrhiza, protozoa were added to the OM patch compartment (treatment with mycorrhiza plus protozoa in OM patch; n =6). To investigate whether protozoa further affect plant growth if also present in the vicinity of roots, protozoa were added to both the OM patch and root compartment (treatment with mycorrhiza plus protozoa in OM patch and protozoa in root compartment; n =7). Seven instead of six replicates were set up for the latter treatment to account for the increased probability of contamination in this most intensively handled treatment.
The soil was taken from the uppermost 20 cm of an extensively managed grassland (Hedlund et al., 2003) and stored in plastic bags at 4°C for 4 wk. Before use, the soil was mixed with sand at a ratio of 1 : 1 and autoclaved (20 min, 121°C). Subsequently, the soil–sand mixture was washed with tap water (2 l water per 1 kg soil) to reduce nutrients and toxic compounds. Then, the soil was dried to c. 50% of the water holding capacity, autoclaved again and filled into sterile cell culture flasks. The soil contained 5.02 g kg−1 organic C, 0.33 g kg−1 total N and 0.07 g kg−1 P, and had a pH of 6.66.
To establish a protozoa-free bacterial community in autoclaved soil of the microcosms, fresh collected rhizosphere soil (300 g dry weight), taken from a grassland of the Biology Campus of Darmstadt University of Technology (Germany), was suspended in 300 ml Neff's modified Amoebae Saline (NMAS; Page, 1976). The soil slurry was filtered through paper filters (595½ Schleicher & Schuell, Dassel, Germany) to remove larger soil particles. Protozoa and fungi were excluded by subsequent filtering through 5-μm and then through 1.2-μm Isopore filters (Millipore, Schwalbach, Germany; Bonkowski & Brandt, 2002). To check for protozoan and fungal contaminations, the filtrate was cultured for 3 d in sterile nutrient broth (NB; Merck, Darmstadt, Germany) with NMAS at 1 : 9 v/v (NB-NMAS; Page, 1976). Five milliliters of the bacterial filtrate were homogeneously mixed into the soil of each compartment.
Protozoa treatments were inoculated with 200 μl (c. 400 000 individuals) of axenic Acanthamoeba castellanii (Rosenberg et al., 2009) in mineral water (Volvic, Frankfurt, Germany). Acanthamoeba castellanii was chosen as model protozoan because it is one of the most common soil protozoa (Page, 1988). Treatments without protozoa received 200 μl of sterile mineral water. All inoculation steps of soil microorganisms and the transfer of soil into the microcosm were performed under sterile conditions to avoid contamination by airborne cysts of protozoa and fungi. The microcosms were incubated in darkness in a climate chamber at 18°C and 70% humidity for 5 d before the addition of plants.
A mesh bag (20 μm) enclosing the OM patch of 15N-labeled grass powder was placed into the centre of the OM patch compartment (Fig. S1). The OM patch compartment was closed with a cellulose plug to avoid contamination from airborne cysts of protozoa and fungi, but enabling gas exchange. The 15N-labeled grass powder was obtained by mixing labeled (40.8% C, 2.7% N, 45.2 atom% 15N) and nonlabeled Lolium perenne L. (Appels Wilde Samen GmbH, Darmstadt, Germany) shoots (39.8% C, 2.4% N). For labeling, L. perenne was grown in a soil–vermiculite mixture in a glasshouse; 20 d after sowing, plants were watered for 21 d with a total of 3 g of 99 atom% 15NH415NO3 dissolved in distilled H2O. Then, plants were regularly watered for another 2 months and harvested by cutting shoots. The shoot material was dried at 80°C for 48 h, ground and analyzed for C, N and 15N concentrations. Nonlabeled plants were treated as labeled plants but received unlabeled NH4NO3. Labeled and nonlabeled shoots were milled to a fine powder and mixed to obtain c. 10 atom% 15N litter. The OM patch material was obtained by mixing 1.0 g of this litter powder with 9.0 g of soil. Before placement in the OM patch compartment, the mesh bags with the OM patch material were autoclaved at 121°C for 20 min.
The root compartment of the microcosm was planted with P. lanceolata (Conrad Appel, Darmstadt, Germany). Surface-sterilized (Hensel et al., 1990) seeds of P. lanceolata were germinated in Petri dishes with NMAS agar (1%) at room temperature. After 5 d, similar sized seedlings were aseptically planted into sterile polypropylene tubes (diameter, 1.2 cm; height, 3 cm) filled with sterile quartz sand. Plantago lanceolata seedlings were mycorrhized by placing next to the roots an agar piece of c. 27 mm3 containing spores and mycelium of an axenic culture of the AM fungus Glomus intraradices Schenk (Bago et al., 1996; Mycovitro S.L. Biotechnología Ecológica, Granada, Spain). The planted tubes were then placed aseptically on top of the root compartment of the microcosms (Fig. S1). The planted microcosms were incubated in a climate chamber at 18°C : 22°C night : day temperature, 70% humidity, 16 h photoperiod and 460 ± 80 μmol m−2 s−1 light photon flux density. Every second day, microcosms were randomized and soil moisture was checked gravimetrically and kept at 75% field capacity by adding sterile distilled water through a sterile filter (pore size, 0.2 μm; Sartorius) into the root compartment.
13C labeling of plants
Plants were labeled with 13CO2 38 d after transfer of the plants into the microcosms. Briefly, microcosms were transferred to an assimilation chamber, kept at the same climatic conditions as the microcosm chamber and incubated for 24 h. Before labeling, the shoot base was sealed with liquid silicon (Rhodorsil® RTV 3428; Rhone Poulenc, Lyon, France) preventing labeled CO2 from entering the root compartment. The CO2 concentration in the chamber was reduced by 50% to 180 μl l−1 by passing the incoming air through a soda lime cartridge. Then, the CO2 partial pressure was re-adjusted to 360 μl l−1 by injecting 13CO2 liberated from NaH13CO3 (99 atom%) by adding 1 M lactic acid. Thereafter, the concentration of CO2 was kept at c. 360 μl l−1 for the following 5 h by liberating 13CO2 from NaH13CO3 (50 atom%). Liberation of CO2 was monitored by an infrared gas analyzer (ADC 225 MK3, Hoddesdon, Hertfordshire, UK) and adjusted at 5-min intervals by a control unit. The labeling was stopped by flushing the assimilation chamber with ambient air.
Plants were harvested at the end of the vegetative growth stage, that is 42 d after transplantation of the plants into the microcosms. Four days after labeling and before destructively sampling the microcosms, the net C assimilation rate of the plants was measured on the apical part of a mature single leaf at mid-photoperiod with an infrared gas analyzer and photosynthetic leaf cuvette (PP system model CIRAS-1, Hitchin, Hertfordshire, UK). Thereafter, plant shoots were cut at soil surface level. Plant biomass was determined from freeze–dried shoot and root samples. Roots were separated from soil by hand sorting. Washed roots were scanned and total root length was determined using WinRhizo (Régent Instruments, Ottawa, ON, Canada). Subsamples of the OM patch were dried (80°C, 48 h), frozen (−20°C) or kept at 5°C for further analyses.
Microbial biomass (Cmic) in the OM patch was measured by substrate-induced respiration (Anderson & Domsch, 1978) using an automated respirometer based on electrolytic O2 microcompensation (Scheu, 1992), as outlined in Beck et al. (1997). For basal respiration, the average O2 consumption rate of samples not amended with glucose was measured during 15–20 h after attachment of samples to the respirometer. Microbial specific respiration (qO2, μl O2 μg−1 Cmic h−1) was calculated as the quotient between basal respiration and microbial biomass.
The incorporation of 13C from plant photosynthates via AM fungi into dissolved OM in the OM patch was determined from 1.5 g of OM patch material suspended in 6 ml of 0.05 M K2SO4 with agitation (130 rev min−1) for 1 h and subsequent filtering. 13C in K2SO4-extractable C was determined after precipitation of carbonates in saturated SrCl2 (Harris et al., 1997). The procedure extracts C in dissolved OM rather than liberating C from cells (Vance et al., 1987). Stable isotope signatures from plant and soil materials were determined from milled samples using an elemental analyzer (Na 1500 type II; Carlo Erba, Milan, Italy) coupled with an isotope ratio mass spectrometer (Finnigan Delta S, Bremen, Germany). Data are presented as 15N and 13C in excess of natural abundance. For background, 15N and 13C natural abundance data of control plants grown under the same conditions as labeled plants were used.
The amounts of total N (Ntot) and total 15N (15Ntot) of a given plant organ or soil compartment were calculated as
To calculate the amount of 13C, we used delta values defined as
Atom% 13C is defined as atom% 13C = 100 × F, where F is the fraction of the heavy isotope
The percentage root length colonized (RLC) by AM fungi was determined from 60 ± 20 intercepts of subsamples of roots before drying (Phillips & Hayman, 1970). Total root length colonization (TRLC) was calculated as
Differences between treatments were analyzed by the General Linear Model (GLM) procedure in SAS 9.1 (Cary, NC, USA) using four a priori contrasts:
Treatment with mycorrhiza having access to the OM patch (M) vs treatment without mycorrhiza and protozoa in the OM patch (control; CTRL), analyzing whether the access of AM fungi to the OM patch affects the transfer of N from the OM patch to the host plant via AM fungi.
Treatment with mycorrhiza plus protozoa in the OM patch (MP) vs treatment with mycorrhiza having access to the OM patch (M), analyzing whether protozoa in the OM patch affect the mineralization of N from OM and subsequent transfer of this N to the host plant via AM fungi.
Treatment with mycorrhiza plus protozoa in the OM patch and additional presence of protozoa in the root compartment (MPP) vs treatment with mycorrhiza plus protozoa in the OM patch (MP), analyzing whether the additional presence of protozoa in the vicinity of roots affects plant N capture from the OM patch via AM fungi.
Treatment with mycorrhiza plus protozoa in the OM patch and protozoa also in the root compartment (MPP) vs treatment with mycorrhiza having access to the OM patch (M), analyzing the overall effect of protozoa on plant N capture from the OM patch via AM fungi. Percentage data were arcsine square-root transformed before statistical analysis.
During sample preparation for stable isotope analysis, one of the six replicates of the CTRL treatment was lost and therefore only five replicates were measured.
Access of AM fungi to the OM patch compartment and the additional presence of protozoa in the OM patch did not increase shoot or root biomass (M vs CTRL and MP vs M, respectively; Table 1; Fig. 1). Only if protozoa were present in both the patch and root compartment did plant biomass increase (MPP vs M). Compared with the treatment with only AM fungi in the OM patch, the additional presence of protozoa in both the OM patch and root compartment increased plant shoot and root biomass by factors of 1.7 and 2.5, respectively (MPP vs M). The increase caused by the additional presence of protozoa in the root compartment was similar, even when compared with the treatment with protozoa and mycorrhiza in the OM patch compartment (MPP vs MP; factors of 1.9 and 3.7, respectively).
Table 1. Generalized Linear Model (GLM) table of F and P values of a priori contrasts between the four experimental treatments for plant (Plantago lanceolata) shoot and root biomass (g), relative root length colonized by arbuscular mycorrhizal fungi (AMF; Glomus intraradices) (%) and colonization of roots by AMF (AMF infections per root system): control without mycorrhiza and protozoa (Acanthamoeba castellanii) in the organic matter (OM) patch (CTRL); with mycorrhiza in the OM patch (M); with mycorrhiza plus protozoa in the OM patch (MP); and with mycorrhiza plus protozoa in the OM patch and protozoa also in the root compartment (MPP) (CTRL, M and MP, n =6; MPP, n =7)
Relative root length colonization by AMF
Total root length colonization by AMF
Significant F and P values are given in bold.
M vs CTRL
MP vs M
MPP vs MP
MPP vs M
Plant uptake of OM 15N
Access to the OM patch compartment by AM fungi did not affect the concentration or amount of excess atom% 15N in shoots and roots (M vs CTRL; Table 2; Fig. 2a,b). By contrast, additional colonization of the OM patch compartment by protozoa enhanced the concentration of excess atom% 15N in shoots and roots by factors of 1.6 and 1.7, respectively (MP vs M). However, because of reduced shoot and root biomass in the treatment with mycorrhiza plus protozoa in the OM patch compartment, when compared with the treatment with mycorrhiza only (MP vs M; Fig. 1), the amount of excess 15N in shoots and roots did not differ significantly between these two treatments. By contrast, if protozoa were also present in the root compartment, both the concentration and amount of excess atom% 15N in shoots and roots increased significantly by factors of 1.6 and 1.3 for the concentration and 3.2 and 3.1 for the amount, respectively, when compared with the treatment with mycorrhiza in the OM patch compartment (MPP vs MP). Notably, compared with the treatment with mycorrhiza only, the presence of protozoa in both the OM patch and root compartment increased the concentration and amount of excess atom% 15N in shoots and roots by factors of 2.5 and 2.3 for the concentration and 5.2 and 4.4 for the amount, respectively (MPP vs M).
Table 2. Generalized Linear Model (GLM) table of F and P values of a priori contrasts between the four experimental treatments for the incorporation of 15N from litter in the organic matter (OM) patch and 13C from atmospheric 13CO2 into shoots and roots of Plantago lanceolata, given as both concentrations (atom%) and amounts (mg) of excess 15N and 13C: control without mycorrhiza (Glomus intraradices) and protozoa (Acanthamoeba castellanii) in the OM patch (CTRL); with mycorrhiza in the OM patch (M); with mycorrhiza plus protozoa in the OM patch (MP); and with mycorrhiza plus protozoa in the OM patch and protozoa also in the root compartment (MPP) (CTRL, n =5; M and MP, n =6; MPP, n =7)
Concentration of 15N
Amount of 15N
Concentration of 13C
Amount of 13C
Significant F and P values are given in bold.
M vs CTRL
MP vs M
MPP vs MP
MPP vs M
Plant photosynthesis, 13C fixation, root colonization by AM fungi and transfer of 13C into the OM patch
Plant photosynthesis (measured as the net C assimilation rate) was not affected by access of AM fungi to the OM patch (M vs CTRL; F1,21 = 0.33, P =0.57) or by the presence of both mycorrhiza and protozoa in the OM patch compartment (MP vs M; F1,21 < 0.01, P =0.98). However, if protozoa were also present in the root compartment, photosynthesis more than doubled (5.9 ± 0.28 μmol m−2 s−1) compared with the treatment with mycorrhiza only in the OM patch compartment (2.8 ± 1.69 μmol m−2 s−1) or the treatment with both mycorrhiza and protozoa in the OM patch compartment (2.7 ± 1.45 μmol m−2 s−1; F1,21 = 19.14, P =0.0004 and F1,21 = 21.40, P =0.0002 for MPP vs M and MPP vs MP, respectively).
Similar to photosynthesis, neither the concentration nor the amount of excess atom% 13C in shoots and roots of P. lanceolata was affected by access of AM fungi to the OM patch (M vs CTRL) or by the presence of both mycorrhiza and protozoa in the OM patch compartment (MP vs M; Table 2; Fig. 2c,d). By contrast, if protozoa were also present in the root compartment, the concentration and amount of excess atom% 13C in shoots and roots were increased by factors of 1.4 and 1.7 for the concentration and 2.3 and 4.2 for the amount, respectively, when compared with the treatment with mycorrhiza in the OM patch compartment (MPP vs MP). Notably, compared with the treatment with mycorrhiza only, the presence of protozoa in both the OM patch and root compartment increased the concentration and amount of excess atom% 13C in shoots and roots by factors of 1.6 and 2.1 for the concentration and 2.5 and 4.3 for the amount, respectively (MPP vs M; Table 2; Fig. 2c,d).
Relative root colonization by AM fungi was reduced by 11% and 21% in treatments with protozoa in the OM patch compartment and protozoa in both the OM patch and root compartment, respectively, when compared with the treatment with mycorrhiza only (MP vs M and MPP vs M; Table 1; Fig. 3a). Total root length colonization by AM fungi was not affected by the presence of mycorrhiza only or by the presence of both mycorrhiza and protozoa in the OM patch compartment (Table 1; Fig. 3b). However, parallel to the enhanced root biomass, total root colonization by AM fungi increased by factors of 3.2 and 6.0 when protozoa were present in both the OM patch and root compartment, when compared with the treatment with mycorrhiza only in the OM patch and the treatment with both mycorrhiza and protozoa in the OM patch compartment, respectively (MPP vs M and MPP vs MP). Further, the translocation of plant-derived C into the OM patch was enhanced in the treatment with protozoa present in both the OM patch and root compartment, when compared with the treatment with mycorrhiza only in the patch compartment, as indicated by a significant shift in the δ13C value of dissolved OM in the OM patch from −27.72 ± 0.22 to −26.95 ± 1.03 (MPP vs M; F1,21 = 5.07, P =0.037; Fig. S2).
Parallel to the increased transfer of C into the OM patch, microbial specific respiration in the OM patch increased from 0.024 ± 0.004 to 0.033 ± 0.011 μl O2 μg−1 Cmic h−1 in the treatment with protozoa present in both the OM patch and root compartment, when compared with the treatment with mycorrhiza only in the patch compartment (MPP vs M; F1,21 = 4.17, P =0.05, Fig. S3).
We manipulated the presence of AM fungi and protozoa in both a patch of OM and in the rhizosphere of plants, thereby investigating the role of some of the most abundant and functionally important microorganisms interacting with plants below the ground. The access of AM fungi to the OM patch in the absence of protozoa did not increase plant N uptake from OM. Mineralization of N from OM depends on the C : N ratio of OM (Hodge et al., 2000); at the C : N ratio of OM in our experiment (c. 15), mineral N typically is immobilized by bacteria (Kaye & Hart, 1997; Hodge et al., 2000), suggesting that most available N in OM was incorporated into bacterial tissue (Leigh et al., 2009). By contrast, the concentration of 15N from OM increased in plant biomass if both mycorrhiza and protozoa were present in the OM patch. In agreement with our hypothesis, this suggests that the grazing of protozoa on bacteria mobilized N from OM, and this N was translocated by mycorrhizal hyphae to roots and allocated to shoots. The results indicate that N fixed in bacterial biomass is effectively mobilized by protists, as suggested by the microbial loop in soil concept (Clarholm, 1985; Bonkowski & Clarholm, 2012). Assuming that 14N and 15N from the OM patch were taken up in the same proportions, the amount of N transferred from the OM patch via AM fungi to P. lanceolata would be equivalent to 6% of total plant N if only mycorrhiza accessed the OM patch, but would be increased to 10% if both mycorrhiza and protozoa were present in the OM patch. By excluding mycorrhiza from the OM patch, only mass flow from the OM patch across the 0.45-μm membrane barrier by soluble products in water contributed to the capture of 15N by plants from the OM patch, as found previously (Leigh et al., 2009). Obviously, the addition of water only to the root compartment did not prevent mass flow from the OM patch compartment to the root compartment.
Compared with the treatment with protozoa in the OM patch compartment only, virtually all plant parameters analyzed responded in a more pronounced way if protozoa were also present in the root compartment. In agreement with our second hypothesis, the presence of protozoa in the root vicinity resulted in increased shoot and root growth, photosynthesis, concentration and amount of 13C assimilated from 13CO2 in shoots and roots. Notably, AM fungi, being a strong C sink (Lerat et al., 2002; Heinemeyer et al., 2006), benefited from the almost four-fold increase in root biomass at harvest, as indicated by the six-fold increase in the colonization of roots by AM fungi if protozoa were present in the root vicinity. In parallel, however, the relative root length colonization by AM fungi decreased, indicating that the relative C sink strength of AM fungi was reduced, potentially as a result of the protozoa-mediated increase in plant N supply from the OM patch.
Plant-derived C is critical for N uptake by AM fungi and N transfer to the host plant (Fellbaum et al., 2012), but plants allocate less C to AM fungi if nutrient limitation is alleviated (Johnson, 2010). Although data on 13C concentrations in AM hyphae are missing, the significant shift in δ13C values in the extract from the OM patch in the presence of protozoa in the root vicinity suggests that protozoa not only increased plant C allocation to roots, but also the transport of C from plant photosynthates into the OM patch via AM hyphae. Presumably, this C formed part of the dissolved OM in the OM patch, indicating that, in the presence of protozoa in the root vicinity, plants enhance investment into mycorrhizal symbionts, thereby strengthening nutrient foraging of AM fungi in OM hotspots and increasing the capture of N mobilized by protozoa to the benefit of plants. Increased activity of microorganisms (i.e. microbial specific respiration) in the OM patch suggests that the leaking of plant C from AM hyphae into the OM patch triggered increased microbial metabolism, potentially enhancing the decomposition of OM. Notably, the presence of protozoa in both the OM patch compartment and in the root vicinity enhanced the capture of litter N by the plant by more than a factor of five, when compared with the case in which only mycorrhiza was present, thereby increasing the contribution of litter N to total plant N to 15% (relative to 6% when only mycorrhiza was present). Remarkably, this protozoa-mediated increase in the capture of litter N by the plants resulted in a two-fold increase in plant biomass.
Although the design of our experiment did not allow the separation of the contribution of the root effect of protozoa for plant N uptake from the contribution of the patch effect of protozoa, for the first time, the results highlight that plants, AM fungi and protozoa intimately interact to the benefit of plants. Most importantly, the results indicate that two different effects of protozoa combine to stimulate plant growth (Fig. 4). Protozoa in the vicinity of roots triggered increased plant photosynthesis and plant C allocation to roots via hormonal effects (Fig. 4(1); Bonkowski & Brandt, 2002; Krome et al., 2009). Increased root C was channeled via mycorrhiza into hotspots of OM (Fig. 4(2)), stimulating nutrient acquisition by bacteria (Fig. 4(3)); by grazing on bacteria, protozoa in OM mobilized nutrients fixed in bacterial tissue, that is, by the microbial loop in soil (Fig. 4(4); Clarholm, 1985; Bonkowski & Clarholm, 2012), which were subsequently channeled to plants via AM fungi (Fig. 4(5)), resulting in increased plant photosynthesis and growth (Fig. 4(6)). The results suggest that AM fungi predominantly function as a bidirectional pipeline, transporting root-derived C into decomposing OM and nutrients mobilized from bacterial biomass back to roots without increasing nutrient mobilization themselves.
The beneficial effects of protozoa and AM fungi on plant growth are well documented, but their interactions and feedbacks to plants remain largely unknown. Although an acceleration of OM decomposition in the presence of AM fungi has been shown previously (Hodge et al., 2001; Atul-Nayyar et al., 2009), the results of the present study suggest that plant N acquisition from OM in soil crucially depends on interactions between AM fungi and protozoa. The results imply that protozoa form an essential control point for the capture of N by AM fungi and plant nutrition. Both protozoa-mediated hormonal effects in the vicinity of roots and mobilization of nutrients locked up in bacterial biomass by grazing of protozoa on bacteria complement each other, presenting protozoa as key biota for plant nutrient acquisition and growth.
R.K. and A.R. were funded by the European Union in the BIORHIZ project. M.B. was supported by the German program ‘Biodiversity and Global Change (BIOLOG)’, subproject ‘Microbial Terrestrial Biodiversity’, funded by the Bundesministerium für Bildung und Forschung (BMBF). We thank A. Bago (Estación Experimental del Zaidin, Granada, Spain) for supplying the axenic mycorrhizal cultures, C. Brechet (INRA, Champenoux, France) for isotope analyses, N. Amougou for help in harvesting the experiment, P. Baumhoff (TU Darmstadt, Darmstadt, Germany) for drawing Fig. S1, and H. Hillers (University of Cologne, Cologne, Germany) and two anonymous reviewers for helpful comments on drafts of the article.