Does fungal competitive ability explain host specificity or rarity in ectomycorrhizal symbioses?

Two common ecological assumptions are that host generalist and rare species are poorer competitors relative to host specialist and more abundant counterparts. While these assumptions have received considerable study in both plant and animals, how they apply to ectomycorrhizal fungi remains largely unknown. To investigate how interspecific competition may influence the anomalous host associations of the rare ectomycorrhizal generalist fungus, Suillus subaureus, we conducted a seedling bioassay. Pinus strobus seedlings were inoculated in single- or two-species treatments of three Suillus species: S. subaureus, S. americanus, and S. spraguei. After 4 and 8 months of growth, seedlings were harvested and scored for mycorrhizal colonization as well as dry biomass. At both time points, we found a clear competitive hierarchy among the three ectomycorrhizal fungal species: S. americanus > S. subaureus > S. spraguei, with the competitive inferior, S. spraguei, having significantly delayed colonization relative to S. americanus and S. subaureus. In the single-species treatments, we found no significant differences in the dry biomasses of P. strobus seedlings colonized by each Suillus species, suggesting none was a more effective plant symbiont. Taken together, these results indicate that the rarity and anomalous host associations exhibited by S. subaureus in natural settings are not driven by inherently poor competitive ability or host growth promotion, but that the timing of colonization is a key factor determining the outcome of ectomycorrhizal fungal competitive interactions.


Introduction
A fundamental axiom in biology is the existence of tradeoffs, which are commonly defined as an increase in performance in one area being correlated with a decrease in performance in another area [1]. In ecology and evolution, tradeoffs generally focus on specific traits that shape the life history strategy of a given species [2]. For example, many organisms exhibit tradeoffs among growth, storage, and reproduction based on energetic constraints [3]. These tradeoffs are frequently conceptualized as bifurcations in allocation, commonly referred to as Y models (e.g. allocation to survival versus allocation to fecundity, James [4]). While tradeoffs typically function at the level of the individual organism, their consequences can have important effects on the structure of ecological communities. Specifically, the presence of tradeoffs underpins much of the basic ecological theory explaining species coexistence [5,6], as species generally have a particular set of traits that make them well suited to persist in certain environments but not in others. Like many other organisms, ectomycorrhizal (ECM) fungi, which form symbiotic associations with the roots of many plants [7], have been shown to exhibit ecological tradeoffs that can be linked to their life history strategies. For example, with the rise of molecular-based analyses in fungal ecology [8,9], it was found that many ECM fungal species that produce sporocarps abundantly aboveground were not the dominant species colonizing host root tips belowground [10]. This discordance suggested that allocation to sporocarps, which favor greater dispersal capacity and colonization of new areas, may come at the cost of proliferation of belowground mycelium [11]. Mycelial spread is essential for nutrient and water acquisition by all fungi, but for ECM fungi it is also important for colonizing additional root tips (and thereby gaining carbon from plant hosts). As such, a lower investment in mycelium would be likely to reduce root tip competitive ability for ECM fungi [12]. Subsequent empirical and theoretical studies of ECM fungal communities have provided clear support for this tradeoff between competition and colonization ability [13][14][15], mirroring similar tradeoffs observed in plant and microbial communities [16,17].
A second possible competition-related trade-off for ECM fungi involves host specificity. In parasite/pathogen systems, competition for hosts is thought to select for greater parasite/pathogen specificity [18,19], with specialists being stronger competitors than generalists on their preferred hosts (i.e. 'the jack of all trades is the master of none', 5). While specialization on a single host may counter the negative effects experienced by symbionts on hosts colonized by multiple species, other studies have shown that interspecific competition among symbionts favors the evolution of host generalism as a means of promoting local coexistence [20]. To date, there has only been one test of a putative tradeoff between competitive ability and host specificity for ECM fungi. Parlade and Alvarez [21] examined interspecific competition among four ECM fungal species, two generalists (Laccaria bicolor and Pisolithus arrhizus) and two specialists (Rhizopogon roseolus and R. subareolatus) on the common host Pseudotsuga menziesii. Comparing all two-species pairings, the authors found that the outcome of competition generally favored host generalists (in 3 of the 4 pairings, the generalist won). Those results suggest that competitive ability and host specificity among ECM fungi may not be tightly linked, which is consistent with many of the dominant species in ECM fungal communities being classified as host generalists [22][23][24].
Species abundances have also been considered in the broader ecological literature about competition-related tradeoffs. It is often assumed that rare species have low abundances because of lowered competitive abilities [25]. Despite this intuition, there are many explanations for species rarity that are not related to competitive ability [26], such as the occupation of uncommon niches [27]. A direct test of rarity and competitive ability in plants found that multiple locally rare grass species were actually strong competitors against more abundant grass species, which may facilitate their persistence despite their low abundances [28]. Similarly, Lloyd et al. [29] found that the relationship between plant range size and competitive ability was inconsistent, depending on both the lineage tested and soil fertility level. While the broader topic of rarity has received less attention in the study of ECM fungi, it has been demonstrated that less abundant species do not appear to have consistently lower nutrient acquisition potential (based on extracellular enzyme activity of ECM root tips sampled from mature trees) relative to more abundant species [30].
We recently became interested in the aforementioned tradeoffs to help understand the enigmatic life history strategy of the ECM fungus, Suillus subaureus. Unlike other Suillus species, which are largely host-specific to one of three different host genera in the family Pinaceae [31], S. subaureus is able to colonize both Pinus and Quercus host species [32]. However, this expansion of host range is complicated by the fact that the spores of S. subaureus will only germinate in the presence of Pinaceae host [32]. As such, the establishment on Quercus individuals requires spread of S. subaureus mycelium growing from colonized Pinaceae root tips. Curiously, in the three sites where we have encountered S. subaureus sporocarps, Pinus hosts are always locally absent, although they can be found as isolated individuals or in small patches less than one km away. Under those nearby Pinus individuals, we have encountered other Suillus species but never encountered S. subaureus despite extensive searching. Thus, it appears that the presence of S. subaureus on Quercus hosts represents a legacy effect of Pinus individuals that are no longer present [32]. Further, in our extensive collecting of Suillus throughout Minnesota, USA, we have regularly encountered many other Suillus species in Pinus forests, but only S. subaureus in two locations. The latter suggests that in addition to its anomalous host associations, S. subaureus is also a much rarer species than other Suillus.
To investigate how interspecific competition may contribute to the curious life history of Suillus subaureus, we conducted a seedling bioassay. In addition to S. subaureus, we collected spores of two other common Suillus species, S. americanus and S. spraguei, which regularly produce abundant sporocarps in Pinus strobus forests in eastern North America. We then inoculated Pinus strobus seedlings with all combinations of single-and two-species treatments of these Suillus species. We hypothesized that S. subaureus would be a competitive inferior to both S. americanus and S. spraguei based on its absence in extant Pinus strobus forests. From analyses of host performance in other Suillus bioassays [32,33], we did not expect to observe major differences in P. strobus seedling biomass when colonized individually by the three different species. Finally, to better qualify the potential rarity of S. subaureus relative to other Suillus species, we compared the herbarium records and citizen science observations of these three Suillus species (in terms of both numbers of collections and geographic range) in Minnesota and throughout eastern North America.

Mushroom collection
In fall 2018, multiple sporocarps of each of the three Suillus species were collected at two locations in Minnesota, USA. S. subaureus mushrooms were collected from Interstate State Park (45.3947˚N, 92.6678˚W), while mushrooms of S. americanus and S. spraguei were collected at Cedar Creek Ecosystem Science Reserve (45.4020˚N, 93.1994˚W). Whenever possible, individual sporocarps were collected >5 m apart to try to maximize genetic diversity [34]. In the laboratory, the stipe of each sporocarp was removed and the pileus was placed onto an individual piece of aluminum foil and then covered with a glass jar. After 12 hours, the jar and pileus were both removed and the resulting spore prints were folded, placed in plastic ziptop bags, and stored at -20˚C.

Spore inoculum
To generate spore slurries for seedling inoculations, spores of each species were washed off the aluminum foil prints using deionized water and collected into separate 50 ml tubes by species. Alcohol was used to wipe down surfaces and tools were flame-sterilized between species to eliminate cross-contamination. Under 40x magnification on a Nikon light microscope, the number of spores per slurry was counted in eight separate 10 ul aliquots using a haemocytometer, after which the number of spores per ml was calculated. Slurries were stored at 4˚C for 24 h ahead of inoculation.

Seedling bioassays
Pinus strobus seeds were obtained from Sheffield's Seed Company (Locke, NY, USA) in winter 2019. The seeds were soaked in deionized water for 24 hours and then stratified in a plastic bag at 4˚C for 30 days to facilitate germination. In April 2019, seeds were bulk planted into trays containing sterilized Sunshine grow mix soil #4 (Sun Gro, Agawam, MA, USA) and placed in a growth chamber set with a 14:10 light:dark cycle at 22˚C. For the competition bioassay, a soil mixture consisting of sand, peat moss, and forest loam soil (2:2:1 by volume) was created. The sand came from the Cedar Creek Ecosystem Science Reserve, the peat moss from a commercial source (Professional Sphagnum Peat Moss, Berger Co., Saint-Modeste, Quebec, Canada), and the forest soil from a woodland on the campus of the University of Minnesota. Both the peat moss and forest soils were sieved using a 5 mm mesh to remove sticks, roots, and other debris. The soil mixture was composited in a large heavy-duty plastic bag, shaken to homogenize, and then placed into multiple autoclave trays, spread evenly no more than 5 cm thick. To eliminate any existing fungal inoculum, soils were autoclaved at 121˚C for 90 minutes, cooled for >24 hours, and autoclaved a second time. The soil was then stored in sterile 20 L plastic buckets at 4˚C until use.
Surface-sterilized 100 mL cone-tainers (Steuwe and Sons, Tangent, OR, USA) (pots) were filled with a small amount of polyester pillow filling to cover the pot drainage holes. Each pot was then filled with 90 mL of sterilized soil. The soil was then moistened with water; some manual mixing was required due to the hydroscopic nature of the soil following autoclaving. In June 2019, one seedling was transplanted into each pot (which was harvested on the same day from the germination trays, with the root systems rinsed of any adhering media) and immediately watered to ensure root-water contact. Spores were then injected into the top layer of soil via pipetting to each treatment at a concentration of 5 x 10 5 spores/species per ml of soil/pot. The inoculated seedlings were grown in the same growth chamber as the germination trays with the same light:dark cycle and watered 2-3 times per week throughout the duration of the experiment. The temperature in the chamber, which was set to 22˚C, fluctuated somewhat during the experiment due to multiple short-term chamber malfunctions during the final month of growth, but the seedlings did not appear to suffer from this variation.
At four (September 2019) and eight (January 2020) months after spore inoculation, seedlings in each treatment were harvested. After being removed from the pots, the root systems of each seedling were gently washed to remove all soil. The entire root system of each harvested seedling was examined under 10-20x magnification on a Nikon dissecting scope for mycorrhizal colonization. Root tips colonized by the three Suillus species were distinguished from uncolonized root tips by the presence of a white mantle (S1 Fig). For the scoring of percent colonization, all root tips less than 1 mm long were not counted. Similarly, root tips with poorly developed mantles were not counted as colonized. Any tips with coralloid clusters at their apex were counted as a single root tip. After scoring colonization, roots and shoots of each seedling were dried at 60˚C for 48 hours and then weighed.

Molecular identification
Because the color and shape of the mycorrhizal root tips of the three Suillus species was very similar (S1 Fig), we used molecular techniques to identify which species were present in the two-species treatments. Ten colonized tips per seedling in each two-species treatment were randomly selected and placed into 10 μl of tissue Extraction Solution (Sigma Aldrich, St. Louis, MO, USA) (if less than 10 tips were colonized, then all tips were removed). The tubes containing the tips and Extraction Solution were placed onto a thermocycler and incubated at 65˚C for 10 minutes and 95˚C for 10 minutes. Afterwards, 30 μl of Neutralization Solution were added to each tube and all extractions were stored at 4˚C ahead of PCR. The Internal Transcribed Spaced (ITS) region of the rRNA gene was amplified using the fungal-specific primer pair ITS1F [35] and ITS4 [36]. 15 μl PCR reactions were conducted using 7.5 μl of REDE PCR mix, 0.75 μl of each primer (at [10 μM]), 1 μl of DNA template, and 5 μl of sterile water. PCR cycling conditions were as follows: 95˚C for 1 minute, then repeating 95˚C 30 seconds, 55˚C for 20 seconds, and 72˚C for 50 seconds, 34 times, then 72˚C for 5 minutes and cooled to 12˚C. Success of PCR was confirmed by running 5 μl of reaction product on a 1% agarose gel. Restriction Fragment Length Polymorphism (RFLP) reactions were then performed on the products from successful PCRs using the restriction enzyme Alul and Cutsmart buffer (New England Biolabs, Waltham, MA, USA). Gel electrophoresis was run on the products using 2%/1% (regular/low melt temperature) agarose gel. RFLP band patterns were compared to patterns from known samples of each species (S2 Fig).

Statistical analysis
All statistical analyses were conducted in R (v3.6.2 [37]). The percent mycorrhizal colonization of each Suillus species in the two-species treatments was determined by taking the total percent mycorrhizal colonization and dividing it by the ratio of the counts for each species on the 10 tips analyzed (i.e. if 9 of 10 tips belonged to S. subaureus and 1 belonged to S. spraguei, 90% of the total mycorrhizal colonization would be assigned to S. subaureus and 10% of the total mycorrhizal colonization to S. spraguei). For each species, the percent mycorrhizal colonization was compared using Type-II sum of squares two-way analyses of variance (ANOVAs) in the 'car' package (https://cran.r-project.org/web/packages/car/car.pdf), with harvest date and treatment (single-versus two-species) as the predictor variables. At each harvest date, total seedling biomass was compared using a one-way ANOVA, with inoculation treatment as the predictor variable. Before running all ANOVAs, variance and normality assumptions were tested and found to be met, except for the percent mycorrhizal colonization by S. spraguei, which was dominated by zero values. Differences among means were assessed using a posthoc Tukey HSD Test run in the 'emmeans' package (https://cran.r-project.org/web/packages/ emmeans/emmeans.pdf).

Herbarium collections
On March 15, 2020, the Mycology Collections Portal (www.mycoportal.org) was queried for records of S. subaureus, S. americanus, and S. spraguei. Nguyen et al. [31] determined that S. spraguei is the correct name for North American collections labeled as S. pictus, so here records of S. spraguei and S. pictus were combined. We made two separate queries: 1) the number of On the same date, we also searched for observations of all three Suillus species on iNaturalist (inaturalist.org). In both datasets (i.e. collections and observations), we limited our final records to those within the native range of P. strobus.

Results
After 4 months, the number of seedlings colonized varied depending on treatment (Table 1). In the single-species treatments, 100% of the seedlings inoculated with S. americanus were colonized, 85% of the seedlings inoculated with S. subaureus were colonized, and none of the seedlings inoculated with S. spraguei were colonized. In the two-species treatments, all seedlings except for one seedling in the S. subaureus/S. spraguei treatment were colonized. In terms of mycorrhizal colonization, all three species showed no significant differences among the single-and two-species treatments (Fig 1A-1C), despite trends indicating that S. subaureus was the competitive inferior to S. americanus and that S. spraguei was competitively inferior to both other Suillus species due to its lack of colonization in all treatments.
After 8 months, the colonization frequency across treatments remained similar for S. americanus and S. subaureus, but differed for S. spraguei ( Table 1). As opposed to zero colonization after 4 months, S. spraguei colonized the majority of the seedlings in the single-species treatment (5 of 7), although at a consistently low percent mycorrhizal colonization (mean = 4%, Fig  1F). S. spraguei also colonized a single tip on one of the seedlings in the S. subaureus/S. spraguei treatment, but was not detected on any of the other seedlings in that treatment and was completely absent from the S. americanus/S. spraguei treatment. The amount of mycorrhizal colonization of S. americanus remained similar across treatments (indicating no significant negative effect of competition) (Fig 1D), while the percent mycorrhizal colonization of S. subaureus was significantly lower in the S. americanus/S. subaureus treatment than the single-species treatment and the S. subaureus/S. spraguei treatments (Fig 1E). Additionally, the percent mycorrhizal colonization of S. spraguei was significantly higher in the single-species treatment than in either of the two-species treatments (Fig 1F). The total biomass of seedlings after 4 months was not significantly different across treatments, including the non-inoculated control seedlings (Table 1). After 8 months, seedlings in the S. americanus single-species treatment had significantly higher total biomass than the seedlings in the S. americanus/S. spraguei treatment, with the mean total biomass of seedlings in all other treatments being intermediate.
There was a total of 85 records for the three Suillus species in the Bell Museum herbarium (Minnesota, USA), 324 records in the combined herbaria of eastern North America, and 1689 records in the iNaturalist database. While the geographical range of all three species was generally similar in both east-west and north-south extent (Fig 2), S. subaureus was the least abundant species in all three queries. Specifically, S. subaureus represented only~20% (60/324) of the records in the combined herbaria in eastern North America,~6% (5/85) of records in the Bell Museum herbarium, and~1% (22/1789) of the iNaturalist observations. S. americanus was more than twice as common in both the Bell and other herbaria collections as S. sprageui (Bell: S. americanus = 56, S. spraguei = 24; Other herbaria: S. americanus = 179, S. spraguei = 80),  americanus (a,d), S. subaureus (b,e), and S. spraguei (c,f) on Pinus strobus seedlings after 4 and 8 months of growth. Boxes represent the 2nd and 3rd interquartile ranges; the horizontal lines in the boxes represent the median; the upper and lower bars outside the boxes represent the 1st and 4th quartiles, respectively. Letters indicate significant statistical differences (p < 0.05) between means based on Tukey HSD tests. N = 7 for each treatment at both time points. https://doi.org/10.1371/journal.pone.0234099.g001

Discussion
We found a clear competitive hierarchy among the three ECM fungal species studied, with S. americanus being competitively dominant, S. subaureus competitively intermediate, and S. spraguei competitively inferior. This ordering is not consistent with our hypothesis that S. subaureus would be the competitive inferior due to its absence from extant P. strobus forests. While our results only pertain to a limited set of congeneric ECM fungal species, they do suggest that the local absence of S. subaureus in P. strobus forests is not likely due to inherently low competitive ability. This finding is consistent with other studies of rare species, which indicate that species abundance and competitive ability are not necessarily tightly coupled [28,29]. In our study system, the unusual capacity of S. subaureus to colonize Quercus individuals creates a competitive refuge, at least from other Pinaceae-specific ECM fungal species, which are among the dominant colonists of young pine seedlings in Minnesota and elsewhere [38]. We suggest these two factors, i.e. sufficient competitive ability against some other ECM fungal species and the unique ability to colonize novel hosts, combine to support the overall persistence of S. subaureus mixed conifer-angiosperms forests in spite of small population sizes.
The competitive dynamics observed among these Suillus species reveal important parallels with other studies of ECM fungal competition. For example, it appears that the competitively dominant species, here S. americanus, possesses spores that germinate both rapidly and consistently. This was most apparent at the first harvest (i.e. four months), where S. americanus colonized every seedling onto which it was inoculated (including in the two-species treatments), while S. subaureus colonized many but not all of the seedlings onto which it was inoculated, and S. spraguei, which colonized none of the seedlings onto which it was inoculated. Because early colonization of a single root tip facilitates additional root tip colonization via mycelium on the same seedling [39], this early and consistent colonization likely explains the higher levels of percent mycorrhizal colonization of S. americanus across treatments. The results from the second harvest confirm that all three species had viable spores, as S. spraguei was found in the majority of the single-species seedling replicates and on a single root tip of one S. subaureus/S. spraguei seedling. However, the delay in colonization by S. spraguei that resulted in notably lower percent mycorrhizal colonization, particularly relative to the other two species, is likely a significant factor explaining why it was outcompeted in all two-species treatments. The slower spore germination by S. spraguei may also be linked to its preference for older P. strobus forests, in contrast to S. americanus which is more common in younger P. strobus forests and where rapid spore germination is particularly important for initial establishment (N. Nguyen, pers. obs.). The presence of preemptive colonization driving competitive outcomes, a.k.a. priority effects, appears to be a common pattern observed in ECM fungal competition, both in lab and field settings (see Kennedy [12] for a review). Yet other studies have found that competitive outcomes can be reversed depending on order or arrival [40] or when based on mycelial rather than spore colonization [13]. We imagine that these latter possibilities in combination with the highly patchy nature of ECM fungal assemblages [41,42] allow for sufficient opportunities for S. spraguei to colonize P. strobus, thereby preventing competitive exclusion in natural settings.
A second possible explanation for the absence of S. subaureus in extant P. strobus forests is that it is not a beneficial symbiont. Recent work on plant discrimination in ECM fungal symbioses suggests that plants can reward portions of the root system that provide more nitrogen and that there is some amount of pre-colonization chemical screening that controls ECM fungal colonization [33]. However, in the case of S. subaureus, it does not appear that there is active host inhibition, as it colonized many of the seedlings just as quickly as S. americanus. Similarly, the lack of differences in seedling dry biomass across treatments is also consistent with S. subaureus being a functionally equivalent symbiont to the other two Suillus species. The soil mixture used is not particularly nutrient rich, but it is possible that the experimental conditions used masked potential growth effects across treatments, especially since none of the inoculated seedlings were significantly larger than the non-mycorrhizal control seedlings (Table 1). We did, however, find that ECM fungal competition lowered seedling growth, which is similar to previous results in other studies of ECM symbioses [14,43]. Moving forward, more detailed studies using isotopic tracers to track carbon and nitrogen exchange ([e.g. 44]) will be key to determining whether S. subaureus has different physiological interactions with its host relative to other species. Based on our results here and those presented in Lofgren et al. [32], however, we suspect that symbiont benefit is not a primary driver of the anomalous host associations of S. subaureus.
Both the herbarium records and citizen science observations of these three Suillus species correspond well with our own field observations. Specifically, we have found S. subaureus fruiting only 3 times in 5 years of regular mushroom collecting in the midwestern USA, while we have found both S. americanus and S. spraguei fruiting every year during the same time period in a wide variety of locations in the midwestern USA. Members of the genus Suillus produce relatively large sporocarps are generally considered prolific fruiters [34], so we doubt that the patterns observed for S. subarueus are due to cryptic missed collections. Similarly, the distinctive coloring of S. subauerus, which has a more orange pore surface and different pileus characteristics than other co-occurring Suillus species (S3 Fig), makes it unlikely that the differences in collections and observations reflect misidentification. Instead, we believe that S. subaureus is truly a rare Suillus species. While fungal conservation efforts lag behind those of plants and animals, there is growing global interest in documenting fungal species abundances for assigning protected status [45]. Although neither host of S. subaureus appears to be in need of conservation, given the limited abundance of S. subaureus and its particularly unique life history, it may be a good candidate for being designated at greater conservation status.

Conclusions
This study provides new insight into the curious ecology of the rare ectomycorrhizal fungus, S. subaureus. Our results suggest that two apparent hypotheses are unlikely to explain its anomalous host associations: 1) a weak competitive ability forcing S. subaureus to take refuge on alternative hosts and 2) a poor symbiont that is actively discriminated against by its Pinus host. Future competitive tests against other species common in ECM fungal spore banks [38] and mature Pinus forests [46] will help in determining the extent to which the results observed here hold against more distantly related species. Additionally, comparisons of the genome content of S. subaureus [47] as well as its gene expression during symbiotic establishment [48] relative to other Suillus and ECM fungal species will help shed light on its unique capacity to associate with both Pinus and Quercus hosts.