Portal de Periódicos da CAPES

Partial mycoheterotrophy in green plants forming Paris ‐type arbuscular mycorrhiza requires a thorough investigation

2022; Wiley; Volume: 234; Issue: 4 Linguagem: Inglês

10.1111/nph.18049

1469-8137

Satoe Murata‐Kato, Risa Sato, Shigeki Abe, Yasushi Hashimoto, Hiroki Yamagishi, Jun Yokoyama, Hiroshi Tomimatsu,

Legume Nitrogen Fixing Symbiosis

Mycorrhizas are usually considered as mutualistic symbioses, in which green plants provide photosynthetic carbon (C) to mycorrhizal fungi in exchange for mineral resources, such as nitrogen (N) and phosphorus, which are acquired by fungi from the soil (Smith & Read, 2008; van der Heijden et al., 2015). However, some green plants have evolved the ability to reverse this C flow and rely on fungal C in addition to their own photosynthesis, a phenomenon known as 'partial mycoheterotrophy' (or 'mixotrophy'). While evidence for partial mycoheterotrophy has been well established in many orchids and pyroloids associating with ectomycorrhizal fungi (Hynson et al., 2013; Selosse et al., 2017), the question remains whether this intermediate nutritional strategy is also common among plants forming arbuscular mycorrhiza (AM). Arbuscular mycorrhiza is the widespread symbiosis formed by c. 71% of vascular plants with Glomeromycotina fungi (Brundrett & Tedersoo, 2018). Recently, Giesemann et al. (2020b, 2021) suggested that partial mycoheterotrophy is common among green plants forming Paris-type AM. The fungal structures in AM are usually described as Paris- or Arum-type, with roughly equal distribution among plant species (Smith & Smith, 1997; Dickson et al., 2007). Paris-type AM is frequent among forest plants, and all achlorophyllous fully mycoheterotrophic AM plants possess the Paris-type (Dickson et al., 2007). Thus, the functional link between Paris-type AM and mycoheterotrophy has been suspected (Imhof, 1999). The evidence of C gain from fungi may be deduced from the natural abundance of stable isotopes, because achlorophyllous fully mycoheterotrophic plants are enriched in heavy isotopes (typically 13C and 15N) compared to surrounding autotrophic plants in the same vegetation layer, thanks to a higher content of these isotopes in their fungi (Gebauer & Meyer, 2003; Tedersoo et al., 2007; Gomes et al., 2020). Giesemann et al. (2020b) revealed that Paris quadrifolia (after which Paris-type was named) was significantly enriched in heavy isotopes, not only 13C and 15N but also 2H compared to neighbouring understorey plants with Arum-type AM. The enrichment of 13C and 2H could also occur due to reduced transpiration (Farquhar et al., 1989; Cernusak et al., 2016), but the abundance of 18O (a proxy of transpiration rate) did not significantly differ from that of the neighbouring plants. Therefore, P. quadrifolia was presumed to be partially mycoheterotrophic. The subsequent analysis of published isotopic abundance for 135 plant species (mostly in Europe) further showed that approximately half of the green Paris-type AM plants were significantly enriched in 13C (Giesemann et al., 2021), by which they claimed that the capacity to parasitize AM fungi is widespread. The suggestion by Giesemann et al. is appealing, because, if this is true, many understorey plants supplement C from dominant AM trees via common mycorrhizal networks, allowing resource sharing among linked plants. However, as detailed below, we argue that the partial mycoheterotrophy among green plants with Paris-type AM requires a thorough investigation. The Giesemann et al.'s work (2020b, 2021) brought up two issues. First, although they showed that Paris-type AM plant species were generally more enriched in heavy isotopes than Arum-type species, the isotopic abundance varied considerably among species. Plant species differ greatly in the extent of AM fungal colonization, which may also reflect the benefits provided by the fungi, including C gain. Second, Giesemann et al. (2020b, 2021) estimated the proportion of C gained from fungi using the two-source linear mixing model (Gebauer & Meyer, 2003), which requires information on two endpoints represented by the 13C abundance of autotrophic reference (i.e. lower endpoint) and achlorophyllous fully mycoheterotrophic plants or their fungi (i.e. upper endpoint). For example, P. quadrifolia was estimated to receive nearly half (48%) of its C from AM fungi (Giesemann et al., 2020b). The mean 13C abundance of fully mycoheterotrophic AM plants was used as an upper endpoint in their models, and this value was almost exclusively derived from data collected in the tropics because of the difficulty in finding such plants in temperate forests. However, previous research suggests that the degree of 13C enrichment in canopy trees (predominant C source of AM fungi in forests) relative to autotrophic understorey plants may differ between temperate and tropical forests (Quay et al., 1989; Hanba et al., 1997) and thus inferred a latitudinal difference in 13C enrichment of AM fungi. Here, we analysed the natural abundance of stable isotopes (13C and 15N) in plants and AM fungal spores in two broad-leaved deciduous forest stands in Hokkaido, northern Japan. As the 15N enrichment might also be attributable to the colonization by dark septate endophytes (Giesemann et al., 2020a), we place a greater emphasis of our argument on the 13C abundance. In this study, we particularly aimed to (1) examine whether the variability in isotopic abundance among understorey plant species is explained by the extent of AM fungal colonization in addition to the AM morphology and to (2) obtain more reliable estimates of the proportion of C gained from fungi by analysing the 13C abundance of AM fungal spores in the soil. Although the isotopic data on AM fungal hyphae are scarce and potentially subject to contamination from nonmycorrhizal materials (Walder et al., 2013; Klink et al., 2020), AM fungal spores showed similar 13C signatures to both host plants and fully mycoheterotrophic plants exploiting AM fungi (Courty et al., 2011; Walder et al., 2013). Because our results indicated that the 13C enrichment in understorey plants may also involve processes other than its mode of nutrition, we further aimed to (3) explore whether the 13C-enriched Paris-type AM plants are truly partially mycoheterotrophic using Trillium camschatcense as a model system. In some partially mycoheterotrophic plant species exploiting ectomycorrhizal fungi, the degree of 13C enrichment relative to autotrophic plants decreases with increasing light availability, suggesting that higher irradiance drives individual plants towards autotrophy (Preiss et al., 2010; Lallemand et al., 2017). In a similar vein, partially mycoheterotrophic individuals under higher-light conditions may associate less closely with fungi (Matsuda et al., 2012). We tested the relationships of 13C enrichment and AM fungal colonization with light availability to see whether there is any evidence that supports partial mycoheterotrophic nutrition. The forest in which our study was conducted is dominated by AM trees, Fraxinus mandshurica, Ulmus davidiana var. japonica and Acer pictum (Supporting Information Methods S1). In May–July 2017, we collected leaves from the three canopy tree species (n = 47) and 24 understorey plant species (n = 226) within 12 circular plots of 25 m2 (Table S1). The leaves of canopy trees were collected from the crowns (c. 30 m high) using Big Shot Throw Weight Launching System (SherrillTree Inc., Greensboro, NC, USA; Youngentob et al., 2016). The leaves were oven-dried at 60°C for 72 h and ground to a fine powder with a ball mill, and stored in desiccators with silica gel until isotopic analysis. Also, four soil cores (10 cm deep) from each plot were collected and pooled to isolate AM fungal spores by density-gradient centrifugation (Courty et al., 2011; Walder et al., 2013). The spores (> 700 for each plot) were put into tin capsules, oven-dried for 24 h and sealed. As harvesting spores requires a substantial effort, this was done only in half of the plots (n = 6). The natural abundance of stable isotopes was analysed by elemental analyser–isotope ratio mass spectroscopy. The resulting isotopic abundance was denoted by δ-values: δ13C or δ15N = (Rsample/Rstandard − 1) × 1000 (‰), whereby R is the molar ratio of the heavy to the corresponding light isotope. The reproducibility of isotope measurements was < 0.2‰ for both δ13C and δ15N. Following Preiss & Gebauer (2008), the δ-values were normalized to account for plot-specific environmental variability by calculating enrichment factors (ε) as the difference between δ-values of individual target samples and the mean values of Arum-type AM and nonmycorrhizal understorey plants (i.e. reference plants) in the respective plots. However, the variability among plots was small (SD = 0.07‰ for δ13C and 0.17‰ for δ15N), and the use of δ-values instead of ε-values did not qualitatively affect the results. Between 2017 and 2019, root samples were also collected from two to seven individuals of each plant species (total n = 101) and immediately stored in 70% ethanol to assess AM fungal colonization. The roots were stained with trypan blue, and the percentage of root length colonized was assessed by the magnified intersection method (McGonigle et al., 1990). Based on the observations of stained roots, plant species were categorized into either Paris-type AM, Arum-type AM or nonmycorrhizal (Table S1). Paris- and Arum-types are characterized by intracellular hyphal coils and arbuscules, respectively (Fig. S1). The resulting classification was consistent with previous information (Dickson et al., 2007) when compared at the family level. We first compared enrichment factors of 13C and 15N between Paris-type AM and reference understorey plants and then analysed how the enrichment factors were explained by both the percentage of AM colonization and AM morphology. The Arum-type AM plus nonmycorrhizal understorey plants exhibited mean enrichment factors (ε) of zero by definition, with SD of 2.4‰ for 13C and 1.7‰ for 15N (n = 112; Fig. 1). When only the effect of AM morphology was considered, Paris-type AM species were almost significantly enriched in 13C (ε13C = 0.9 ± 0.1‰; mean ± SE; n = 114) compared with the reference plants (Mann–Whitney U test, P = 0.05; Table S2), but their ranges considerably overlapped (Fig. S2). Among Paris-type AM species, the 13C abundance in Veratrum oxysepalum (2.7 ± 0.2‰), T. camschatcense (2.6 ± 0.2‰) and Paris verticillata (2.3 ± 0.2‰) was significantly greater than that in the reference plants (Fig. 1; Table S2). However, Allium victorialis subsp. platyphyllum (3.3 ± 0.4‰) and Pachysandra terminalis (2.8 ± 0.3‰) showed comparable or even greater 13C abundance despite forming Arum-type AM (Fig. S1a,b). Similarly, 15N abundance was exceptionally high in Symplocarpus renifolius (3.6 ± 0.3‰) and Cardiocrinum cordatum var. glehnii (2.0 ± 0.1‰), wherein only the latter formed Paris-type AM. Paris-type AM species were generally enriched in 15N (0.4 ± 0.1 ‰) compared with the reference plants (n = 236, P < 0.01), but again, their ranges overlapped considerably (Fig. S2). Rather, the linear model analysis showed that the percentage of AM colonization explained greater variability in isotopic abundance across species than AM morphology (Table S3). Both 13C abundance and 15N abundance were positively related to the percentage of AM colonization (n = 24; t = 3.27, P < 0.01 for ε13C and t = 2.33, P < 0.05 for ε15N; Fig. 2), whereas the effect of AM morphology was not significant. Thus, the heavy-isotope enriched understorey species were not necessarily those forming Paris-type AM, but rather those associating closely with AM fungi. The 13C abundance of AM fungal spores (3.4 ± 0.2‰, n = 6; P < 0.01) was significantly greater than that of the reference plants but did not differ from canopy leaves (3.5 ± 0.2‰, n = 47; P = 1.00; Fig. 1; Table S2). This was an expected result, because the 13C enrichment of AM fungi is mainly due to the supply of 13C-enriched carbohydrates from canopy trees (Courty et al., 2011). However, the degree of 13C enrichment in our fungi was lower than the average in achlorophyllous fully mycoheterotrophic plants across the globe (5.0‰), which was used as the upper endpoint in the two-source linear mixing model of Giesemann et al. (2021). Whereas the 13C enrichment in canopy trees (and thus AM fungi) relative to understorey plants is largely due to the effect of irradiance on 13C discrimination in leaves (e.g. the discrimination changes by 2.7‰ from understorey to upper canopy in a temperate forest; Hanba et al., 1997), another reason is that understorey plants absorb 13C-depleted CO2 originating from soil respiration. Compared with temperate forest floors, the ambient CO2 in tropical forest floors, where most fully mycoheterotrophic AM species occur, is likely to be more 13C-depleted due to higher rates of respiratory CO2 released from the soil. This is supported by the difference in vertical profiles of air δ13C between the temperate and tropical forests (Quay et al., 1989; Hanba et al., 1997). Also, in a comparative study of isotopic abundance among fully mycoheterotrophic AM plants, the enrichment of 13C was considerably lower in species from temperate climates (Gomes et al., 2020). Accordingly, using the mean 13C enrichment of fully mycoheterotrophic plants occurring mostly in the tropics as an upper endpoint probably underestimates the relative amount of fungal C that putative partially mycoheterotrophic plants may receive in temperate forests. We instead used the on-site mean 13C enrichment of AM fungal spores to estimate the relative amount of fungal C. The enrichment factor ε13C of our spores was similar to that of fully mycoheterotrophic plants in temperate climates (3.4‰; derived from the data of Gomes et al., 2020). The y-intercept of the fitted line in Fig. 2(a) (i.e. ε13C in nonmycorrhizal plants) was employed as the lower endpoint, because several Arum-type AM species exhibited large 13C abundance, as previously described. Our calculation showed that the proportional C gain from fungi (mean ± SD) was 86.0 ± 13.9% for V. oxysepalum, 82.9 ± 12.7% for T. camschatcense, and 75.9 ± 15.6% for P. verticillata. If A. victorialis subsp. platyphyllum, forming Arum-type AM, was assumed to obtain fungal C, it would receive 97.2 ± 15.5% of its C from fungi. The obtained estimates approach or exceed 100% for many individuals and seem unrealistically large considering the photosynthetic abilities of these plants. Our results suggest that the 13C enrichment in these AM plants may also involve other processes. Next, we employed T. camschatcense as a model (Fig. 3a) to further explore whether the 13C-enriched Paris-type AM plant species is partially mycoheterotrophic. This species, which is related to P. quadrifolia (both Melanthiaceae), forms Paris-type AM with dense hyphal coils (Fig. S1c) and was significantly enriched in both 13C and 15N (Fig. 1; Table S2). Additionally, the 13C enrichment in T. camschatcense was not likely due to reduced transpiration, because 18O abundance did not differ significantly between T. camschatcense (δ18O = 25.3 ± 0.3‰; mean ± SE; n = 6) and neighbouring nonenriched plants (24.7 ± 0.3‰, n = 12; Mann–Whitney U test, P = 0.12; Table S4). We examined the relationship between leaf δ13C of 30 flowering individuals and light availability (quantified as relative photosynthetic photon flux density under an overcast sky) in 2018. The leaves of nonmycorrhizal plants (n = 55) were also collected within 1 m from each T. camschatcense individual and subjected to isotopic analysis for reference. We then performed analysis of covariance of the leaf δ13C with the species group (T. camschatcense or nonmycorrhizal) as a fixed effect and light as a covariate. Light availability in late May, when canopy trees were expanding leaves, was used in the analysis, while the light levels in May and late June (after canopy closure) were positively correlated with one another (r = 0.72, P < 0.001). Leaf δ13C of both T. camschatcense and nonmycorrhizal plants increased with light availability at the same slope (Fig. 3b; light: t = 1.59, P < 0.05; species group: t = 4.87, P < 0.001; and light × species group: t = −0.67, P = 0.51), indicating that the degree of 13C enrichment in T. camschatcense relative to the reference plants did not change with light. Although the response of partially mycoheterotrophic plants to light varies among species (Lallemand et al., 2017; Selosse et al., 2017), our data did not support C acquisition from fungi. The positive response of δ13C to increasing light is expected in autotrophic plants, as low photosynthetic rates and enhanced stomatal conduction (due to low water stress) in shaded conditions facilitate efficient equilibration of 13CO2 concentrations between the atmospheric air and stomatal chamber (Farquhar et al., 1989), producing 13C-depleted photosynthates. The variability in AM colonization among individuals was also not in support of partial mycoheterotrophy in T. camschatcense. First, we assessed AM colonization of 30 flowering individuals used in the above analysis. Although individuals under lower-light conditions may associate more closely with fungi in partially mycoheterotrophic plants (Matsuda et al., 2012), the percentage of AM colonization was positively, not negatively, related to light availability (linear regression: t = 4.09, P < 0.001; Fig. 3c). Second, as T. camschatcense is slow-growing and requires an average of c. 10 yr to reach the flowering stage (Tomimatsu & Ohara, 2010), we also collected individuals of three other distinct growth stages (first-year seedling, one-leaf juvenile and three-leaf juvenile) within 1 m from flowering individuals and assessed their mycorrhizal colonization. Seedlings and one-leaf juveniles were so small (c. 5–10 cm height) that they grew in the shade of other understorey plants. A minimum age was estimated for each growth stage by counting scape scars on the rhizomes (Broyles et al., 2013). While the flowering individuals were closely associated with AM fungi (85.9 ± 2.0%), mycorrhizal association appeared to proceed quite slowly during vegetative growth (Fig. 3d). The colonization was low not only in first-year seedlings (1.1 ± 0.6%) but also in one-leaf juveniles (17.3 ± 3.1%) even > 1 yr after germination. Therefore, these results were opposite to the expectation based on partial mycoheterotrophy – AM colonization appears to be regulated by the plants' requirement for mineral nutrients rather than that for fungal carbohydrates. We do not suggest that partial mycoheterotrophy has not evolved in Paris-type AM plants. Because achlorophyllous fully mycoheterotrophic plants feeding on these fungi are generally enriched in 13C and 15N (and 2H when data are available) in forests (Gomes et al., 2020), understorey plants feeding partly on fungi are also expected to be enriched in these isotopes. The recent analysis of the published isotopic abundance (Giesemann et al., 2021) has revealed that the degree of 13C enrichment was remarkably large (> 4‰) for several ferns and horsetails, and such species with greater degrees of enrichment may be more likely to be partially mycoheterotrophic. However, our results, including the presence of 13C-enriched Arum-type AM plants (Fig. 1), were only partially consistent with the suggestion by Giesemann et al. (2020b, 2021). One explanation may be that the acquisition of fungal C also occurs in Arum-type AM species. Alternatively, as 13C abundance also reflects other processes, the reduced enrichment of AM fungi in temperate forests may limit the utility of stable isotopes for identifying partially mycoheterotrophic plants. For example, some spring-emerging species assimilate most C early in spring before canopy closure of overstorey trees (Heberling et al., 2019). At this time, the soil respiration rate is low and the ambient CO2 should therefore be less 13C-depleted in the forest floor. Moreover, high light levels in spring enhance photosynthesis and limit 13C discrimination in leaves. Thus, these plants are more likely to be 13C-enriched than those that continue photosynthesis well into the summer. In our study, the large 13C abundance in some spring plants such as T. camschatcense and V. oxysepalum, which wither early in summer, and low 13C abundance in Impatiens noli-tangere and Laportea bulbifera (Fig. 1), which only become conspicuous in summer, appear consistent with this argument. In addition, P. terminalis showed large 13C abundance despite forming Arum-type AM, possibly because this evergreen species takes advantage of both early spring and late autumn as photosynthetically active seasons. The positive relationship between AM colonization and isotopic abundance may also have multiple interpretations. One may think that the large enrichment of 13C in species associating closely with AM fungi (Fig. 2a) is an additional evidence of acquiring fungal C, although others may interpret it as evidence of high efficiency of water use resulting from AM symbiosis (Cernusak et al., 2004; Mariotte et al., 2017). With regard to N, the positive relationship between AM colonization and 15N abundance (Fig. 2b) may reflect a greater degree of dependence on mycorrhizas for N acquisition in species more closely associated with AM fungi, although the mechanisms warrant further investigation. In conclusion, our case study suggests that the enrichment of heavy isotopes may not always be the result of mycoheterotrophy in temperate forests, because the 13C enrichment of AM fungi is likely to be reduced compared with the tropics. As the natural abundance of stable isotopes can reflect processes other than the mode of C nutrition, additional studies are necessary to determine the functional link between Paris-type AM and mycoheterotrophy, and to evaluate the extent to which the intermediate nutritional strategy prevails among chlorophyllous, shade-tolerant plants. Laboratory or in situ CO2 labelling experiments, which can directly demonstrate the C transfers through common mycorrhizal networks, require a substantial effort but are needed. Experimental shading and the application of fungicides to target species could be useful in such studies to increase the detectability of transferred C and to use fungicide-applied individuals as reference controls in the field, respectively. We thank Shun Sasaki, Ryotaro Taniguchi and Saki Inomata for assistance in the field and laboratory work; Masanori Saito for technical support in handling fungal spores; Kenji Suetsugu and Jun Matsubayashi for helpful comments; and the landowners for kindly allowing us to conduct the research. This research was funded by the Japan Science Society (Sasakawa Scientific Research Grant no. 2018-5003 to SM-K) and the Japan Society for the Promotion of Science (KAKENHI JP18K19356 to JY and HT). SM-K, RS and SA developed the idea and designed the research under the supervision of JY and HT. SM-K, RS, SA, YH, HY, JY and HT collected the data. SM-K and HT wrote the manuscript with inputs from all authors. All data are available on Dryad (10.5061/dryad.bzkh18999). Fig. S1 Examples of morphological structures observed in arbuscular mycorrhizal roots. Fig. S2 Stable isotope enrichment factors of 13C and 15N for Paris-type arbuscular mycorrhizal (AM) plants vs Arum-type AM plus non-mycorrhizal plants. Methods S1 Additional description of the study sites. Table S1 Arbuscular mycorrhizal-type, stable isotope enrichment factors (ε13C and ε15N), and percentage of root length colonized ± SD for 27 plant species sampled in two deciduous forest stands of northern Japan. Table S2 Mann–Whitney U-tests for differences in stable isotope enrichment factors (ε13C and ε15N). Table S3 Effects of arbuscular mycorrhizal morphology (Paris-type vs reference plants) and the percentage of root length colonized by AM fungi on the enrichment factors of understory plants, analysed with linear models. Table S4 Natural abundance of leaf 18O (mean ± SE) in Trillium camschatcense and neighboring reference plants. Please note: Wiley Blackwell are not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.