Portal de Periódicos da CAPES

Are there geographic mosaics of mycorrhizal specificity and partial mycoheterotrophy? A case study in M oneses uniflora ( E ricaceae)

2015; Wiley; Volume: 208; Issue: 4 Linguagem: Inglês

10.1111/nph.13587

1469-8137

Nicole A. Hynson, Martin I. Bidartondo, D. J. Read,

Ecology and Vegetation Dynamics Studies

The majority of plants require symbiotic interactions with other organisms to complete at least a portion of their life cycles. However, the reliance of plants on these interactions varies, and the net benefit to plant individuals is dependent on the environmental context in which they occur (Thompson, 2005). One particularly interesting group of obligate symbiotic plants are mycoheterotrophs. Rather than deriving carbon from photosynthesis, mycoheterotrophic plants meet all or a portion of their carbon demands via symbiotic interactions with fungi that are often simultaneously engaged in mutualisms with surrounding autotrophic trees (Merckx, 2013). Via these tripartite networks, autotrophic trees are the ultimate carbon source for many mycoheterotrophic species. Mycoheterotrophic plants can be either fully mycoheterotrophic – where they have lost the ability to photosynthesize and are completely dependent on fungi to meet their carbon demands, or partially mycoheterotrophic – a kind of mixotrophy where some autotrophy is retained (Merckx et al., 2009). Similar to many host–parasite interactions, a hallmark of many fully mycoheterotrophic plants is extreme specificity to species (Bidartondo & Bruns, 2005), or even genotypes of fungal hosts (Barrett et al., 2010). Although some exceptions exist (Hynson & Bruns, 2009; Roy et al., 2009), the reason(s) for this frequent extreme specificity are not fully understood. However, two non-exclusive explanations prevail: (1) as exploiters of the mycorrhizal mutualism, mycoheterotrophs fine-tune their physiology to maximize their interactions with specific fungal hosts, thus preventing broad host switching (Hynson & Bruns, 2010); (2) neighborhood interactions such as partner filtering, prevent mycoheterotrophs from exploiting certain fungi (Egger & Hibbett, 2004). Because they retain some autotrophy (and thus the ability to reciprocate on the mycorrhizal mutualism), under each of these scenarios partial mycoheterotrophs would not be expected to form specific mycorrhizal associations unless the fitness benefits of fungal exploitation outweigh those of cooperation. Accordingly, among partial mycoheterotrophs studied thus far evidence of fungal partner specificity is somewhat limited; Pyrola japonica Klenze ex Alef. (Ericaceae) and Limodorum abortivum (L.) Sw. (Orchidaceae) associate with a range of ectomycorrhizal taxa, but predominantly partner with Russula Pers. spp., and Corallorhiza trifida Châtel (Orchidaceae) partners solely with ectomycorrhizal Tomentella Pers. ex Pat. spp., but appears to gain little carbon from photosynthesis (Girlanda et al., 2006; Zimmer et al., 2008; Cameron et al., 2009; Matsuda et al., 2012). Furthermore, the majority of studies on partial mycoheterotrophs examine the degree of partial mycoheterotrophy and fungal partner specificity within single or among a few plant populations (Bidartondo et al., 2004; Julou et al., 2005; Abadie et al., 2006; Tedersoo et al., 2007; Matsuda et al., 2012; Johansson et al., 2015). While these efforts provide good baseline data, a more complete test of mycorrhizal specificity among partial mycoheterotrophs would include wider geographic samplings of these species and their fungal symbionts. Assuming coevolution, across a species range its symbiotic partnerships will be shaped by geographic selection mosaics that depend upon population-specific biotic and abiotic conditions (sensu Thompson, 2005). Here we set out to examine the fungal partnerships of a putative partial mycoheterotroph, Moneses uniflora (L.) A. Gray (Ericaceae), across a large portion of its natural range. Analyzing the naturally abundant carbon and nitrogen stable isotope ratios of understory plants has been a useful technique to infer partial mycoheterotrophy in nature (Gebauer & Meyer, 2003; Julou et al., 2005; Hynson et al., 2013). The carbon stable isotope composition of partial mycoheterotrophs tends to be enriched in the heavy isotope of carbon (13C) compared to surrounding autotrophic species, but depleted in 13C relative to full mycoheterotrophs (Hynson et al., 2013). Within an ecological foodweb as substrates are processed and consumed there is a corresponding isotopic enrichment that can be detected in the consumer (Fry, 2006). Therefore, the enrichment in 13C found in mycoheterotrophs is owed to all or a portion of their carbon demands being met through the uptake of compounds that have been previously processed by fungi, rather than direct uptake of atmospheric CO2 through photosynthesis (Gebauer & Meyer, 2003; Hynson et al., 2013). Many partially and fully mycoheterotrophic taxa are also enriched in the heavy isotope of nitrogen (15N) relative to surrounding autotrophic species, and some leafy green ericaceous understory species are enriched only in 15N without any detectable differences in their 13C composition from surrounding autotrophs (Hynson et al., 2013). This latter group has been referred to as 'cryptic mycoheterotrophs' where their 15N enrichment may be indicative of the uptake of fungal-assimilated organic compounds that would inherently include carbon. However, whether this carbon is used by plants for their own growth is unclear (Hynson et al., 2013). Therefore, green plants that associate with ectomycorrhizal fungi and are enriched in both 13C and 15N relative to surrounding autotrophs are the clearest examples of partial mycoheterotrophy. In addition to examining fungal partner specificity, we determine the dependency of M. uniflora populations on fungal nutrition through the use of carbon and nitrogen stable isotope analyses. Over a 2-d period in July 2011 we sampled seven sites containing M. uniflora from the northwest corner of the contiguous United States on the Olympic Peninsula in Washington State (Supporting Information Table S1). The temperate rainforest vegetation of these sites is dominated by dense Tsuga heterophylla (Raf.) Sarg. and Picea sitchensis (Bong.) Carrière canopies. All sampling sites were old growth forests (> 100 yr old) with dark shaded understories located in one of the wettest regions on Earth with 343 cm of average precipitation annually. We selected our sampling sites of M. uniflora based on the following criteria: at least one cluster of M. uniflora was growing within 0.5 m of at least two other understory species, with at least two individuals per species. Due to the rhizomatous nature of M. uniflora, determining individuals in the field is challenging. To avoid sampling the same individual twice, we separated our sampling sites by at least 10 m. At each site we excavated one cluster of M. uniflora and separated the leaves from the roots for the isotopic analyses. A subset of our root samples were used for molecular identification of their mycorrhizal fungi. At some sites, additional M. uniflora roots were harvested along with neighboring spruce seedlings, which were analyzed to determine the identities of their mycorrhizal fungi as described later. As site specific references for the stable isotope composition of fully autotrophic and fully mycoheterotrophic plants, tissues of each group (when available for the latter) were also collected (Table S1). The green orchid Neottia cordata (L.) Rich. was also sampled from three of the seven sites (Table S1). To control for microsite variability that may influence the stable isotope composition of plant tissues, autotrophic and mycoheterotrophic leaf samples were taken from within a 0.5 m radius of their corresponding M. uniflora cluster and from similar heights as the leaves of M. uniflora. In July 2010, roots of five M. uniflora individuals from two additional populations were sampled in northeastern Scotland and we followed up the root sampling efforts with leaf samples for stable isotope analyses from the same sites in August 2013. One site is in Balblair Wood near Golspie in East Sutherland and another in the nearby woods around the Bonar Bridge, Ardgay Golf Club (Table S1). The plants from Scotland receive 98.5 cm average annual precipitation in open, 100-yr-old Scots pine forests (Pinus sylvestris L.) mixed with silver birch (Betula pendula Roth), beech (Fagus sylvatica L.) and larch (Larix decidua Mill.). Leaf samples were collected and processed in a similar manner to those from the Olympic Peninsula except that three M. uniflora rosettes were sampled in each forest, along with three replicates of three species of understory autotrophic reference plants collected from a slightly larger area than those on the Olympic Peninsula (1 m radius from M. uniflora, Table S1). In July 2010 and 2011 roots from 15 individuals of M. uniflora were sampled near the Gideåbergs bog south of Ramsele in northern Sweden and in August 2013 this site was re-sampled for stable isotope analyses of leaves. The forest is dominated by spruce (Picea abies (L.) H.Karst.) with a diversity of mosses and herbs in the understory. This area has an average of 49.87 cm of precipitation a year. At our sampling site, leaves of eight individuals of M. uniflora were collected as outlined earlier. Leaves from replicates of neighboring autotrophic species and flowering stalks of the fully mycoheterotrophic orchid Epipogium aphyllum Sw. were also collected as references (Table S1). Molecular identification of mycorrhizal fungi from M. uniflora roots collected at all sites, and P. stichensis seedlings from the Olympic Peninsula were carried out as in Bidartondo & Duckett (2010). All sequences were compared to GenBank sequences and unique representatives were accessioned in the same database. Sequences with ≥ 99% similarity to known fungal species were given the same name, those with < 99% similarity were given generic or familial names accordingly (Nilsson et al., 2008, accession numbers KP896155–KP896159, Table 1). Stable isotope analyses were done following the methods of Hynson et al. (2009). Because leaf stable isotope compositions are influenced by local environmental conditions, to make comparisons of these values across our study sites we used a data normalizing calculation known as an isotope enrichment factor (ε) approach (Preiss & Gebauer, 2008). Details on the calculation of enrichment factors and our statistical analyses can be found in Methods S1. We generated DNA sequence data from 27 M. uniflora plants, and independent of locality, all M. uniflora roots analyzed were associated with fungi in the family Atheliaceae (Table 1). On the Olympic Peninsula, independent of site, all M. uniflora roots were associated with the ectomycorrhizal species Tylospora fibrillosa (Burt) Donk, or another Atheliaceae species likely in the genus Tylospora (Table 1). These species were also detected on the ectomycorrhizal roots of neighboring P. sitchensis seedlings, confirming the potential for a common mycelial network between M. uniflora and surrounding autotrophs. All M. uniflora roots from the two Scottish sites and the single Swedish site were associated with another ectomycorrhizal athelioid species – Amphinema byssoides (Pers.) J. Erikss. or an unknown species of Atheliaceae likely in the genus Amphinema (Table 1). Except for plants from Sweden, M. uniflora ε13C values were significantly different from surrounding autotrophs (P = 0.003 for Scotland, and P = 0.004 for the Olympic Peninsula, Fig. 1). These differences however, were owed to a significant depletion in 13C relative to autotrophs found in Scottish M. uniflora plants and a significant enrichment in 13C for Olympic Peninsula plants. All M. uniflora plants were significantly enriched in 15N relative to their localities' autotrophic species (P = 0.001 for Sweden, and P < 0.001 for Scotland and the Olympic Peninsula, Fig. 1). In Sweden, the fully mycoheterotrophic orchid E. aphyllum was significantly enriched in both 13C and 15N relative to M. uniflora and autotrophic species (P < 0.001, Fig. 1). On the Olympic Peninsula the fully mycoheterotrophic orchid Corallorhiza striata Lindl. was also significantly enriched in both 13C and 15N relative to M. uniflora, autotrophic species, and the green orchid N. cordata (P < 0.001, Fig. 1). Neottia cordata was significantly enriched in 13C (P = 0.037), but not 15N compared to autotrophic species; while it was not significantly different from M. uniflora for 13C enrichment, it was significantly more depleted in 15N (P = 0.027, Fig. 1). The single sample of Hypopitys monotropa Crantz did not allow statistical comparisons to other species, although it is clearly enriched in both 13C and 15N compared to autotrophic species and M. uniflora (Fig. 1). Mean isotope enrichment factors for both 13C and 15N for all species and autotrophic reference plants from each locality are reported in Table S2. From sampling across a large portion of its geographic range, we found that M. uniflora is specialized on ectomycorrhizal Atheliaceae species. However, depending on locality, the fungal partners either belonged to the genus Tylospora (Olympic Peninsula) or Amphinema (Sweden and Scotland, Table 1). Furthermore, there was some variation in the species or genotypes of fungi associating with individual M. uniflora plants collected from relatively nearby populations on the Olympic Peninsula and in Sweden (Table 1). These results point to geographic selection mosaics for M. uniflora where fungal partner identity is dependent upon the local abiotic and biotic environments in which these interactions occur (Piculell et al., 2008). Tylospora fibrillosa is a common, and often-dominant ectomycorrhizal species associated with P. sitchensis plantations in Europe (Erland, 1995; Palfner et al., 2005). Conversely, A. byssoides is common and often dominant among conifer seedlings in nurseries, but rarely encountered in the wild (A. F. S. Taylor, pers. comm.). However, this genus is phylogenetically diverse (Larsson et al., 2004), thus more intensive sampling of conifer trees at our sites would be necessary to determine the overall abundance of the Amphinema spp. found associating with M. uniflora. We found that M. uniflora from the Olympic Peninsula behaves isotopically as a partial mycoheterotroph where it is significantly enriched in the heavy isotopes of carbon and nitrogen relative to surrounding autotrophic understory plants. This evidence was corroborated through finding the same ectomycorrhizal fungus colonizing both M. uniflora and P. sitchensis (Table 1). While M. uniflora plants from the Olympic Peninsula were only on average 1.33(SE 0.45)‰ enriched in 13C relative to autotrophs (Table S1), this enrichment is within the range of other Pyroleae species considered to be partially mycoheterotrophic (Tedersoo et al., 2007; Zimmer et al., 2007). Conversely, despite also associating with specific ectomycorrhizal fungi, M. uniflora populations from Sweden and Scotland show no indications of partial mycoheterotrophy based on their 13C enrichment (Fig. 1). This finding is similar to another recent study of M. uniflora in Sweden by Johansson et al. (2015). All M. uniflora plants were significantly enriched in 15N by a minimum average of 2.52(SE 0.29)‰ relative to surrounding autotrophic species, possibly indicating cryptic mycoheterotrophy in Swedish and Scottish populations. When interpreting our isotope data, factors other than reliance on fungal nutrition that can lead to 15N and 13C enrichment such as remobilization of stored carbon, changes in stomatal conductance and photosynthetic rate, should be considered, as discussed in detail in Hynson et al. (2012). However, because M. uniflora isotope enrichment factors from the Olympic Peninsula were significantly and consistently enriched in both 13C and 15N relative to six other understory species, one of which is from the same plant family (Vaccinium parvifolium Sm.), partial mycoheterotrophy at these sites is the most likely explanation. Differences in the carbon stable isotope compositions of M. uniflora from different localities could be related to differences in neighborhood interactions with different fungal communities. Relative to other ectomycorrhizal fungi, members of Atheliaceae may be more vulnerable to exploitation by M. uniflora, but based on individuals' local abundance and local interactions with autotrophic hosts, vary in their ability to support mycoheterotrophy. However, additional factors that could affect plant 13C abundance, such as environmental conditions, and differences in plant or fungal partners' physiology at the intraspecific and interspecific levels, deserve further attention. Mycorrhizal specificity and variations in the carbon isotope enrichment of M. uniflora across a significant portion of its geographic range are now revealed. However, how this specificity and phenotypic plasticity relates to abiotic and biotic interactions within and among plant populations needs to be further explored. While most fully mycoheterotrophic taxa are highly specialized on particular fungal partners, fungal partner specificity is not a requisite for plants to become fully mycoheterotrophic (Hynson & Bruns, 2009; Roy et al., 2009). Therefore, the role of fungal partner specificity in the evolution of mycoheterotrophy deserves further attention. Geographic selection mosaics have now emerged as a useful framework for examining the biotic and abiotic factors that may select for mycorrhizal specificity in conjunction with partial mycoheterotrophy. However, the question remains: are partial mycoheterotrophs or cryptic mycoheterotrophs coevolving with their fungal partners? To explicitly test this would involve a biogeographic approach that identifies coevolutionary hot and cold spots, selection mosaics and trait mixing (Gomulkiewicz et al., 2007). The first two can only be measured if the fitness of partnering genotypes can be accurately assessed (a persistent challenge for mycorrhizal fungi). Thus, tests of trait mixing (the production of locally mismatched phenotypes) may be relatively more approachable (Bidartondo & Bruns, 2005). In addition to biotic interactions with fungi, abiotic conditions such as light availability could select for the evolution of partial mycoheterotrophy. The understories of the temperate rainforests on the Olympic Peninsula are exceedingly light limited, and the plant species that occur there have evolved adaptations in order to do so (Kirk & Franklin, 1992). So, it seems intuitive that light limitation may exert a strong selective pressure for partial mycoheterotrophy in M. uniflora (but see Preiss et al. (2010), Hynson et al. (2012), and Matsuda et al. (2012) for diverging results on the effect of light availability on partial mycoheterotrophy). Until now, fungal partner specificity was not a trait clearly associated with other partial mycoheterotrophs or Pyroleae species in general. The current study adds to the growing body of knowledge on the evolution of mycoheterotrophy and fungal partner specificity, the path to which appears to have many origins. The authors would like to thank Steve Trudell and Anthony Amend for assistance with fieldwork on the Olympic Peninsula. The authors would also like to acknowledge the valuable feedback provided by Marc-André Selosse and anonymous reviewers. D.J.R. was funded by the Leverhulme Trust. 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. Table S1 Sampling site locations including latitude and longitude, species collected and replicates (n) from each site and group Table S2 Mean ε13C or ε15N isotope enrichment factors by group or species with replicates (n), and locality Methods S1 Enrichment factor calculation equation and statistical analyses. 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.