Dual plant host effects on two arbuscular mycorrhizal fungi

Pedobiologia 54 (2011) 209–216

Contents lists available at ScienceDirect

Pedobiologia - International Journal of Soil Biology journal homepage: www.elsevier.de/pedobi

Dual plant host effects on two arbuscular mycorrhizal fungi Antonio J. Golubski ∗ University of Illinois at Chicago, Department of Biological Sciences, 845 West Taylor, Chicago, IL 60607, USA

a r t i c l e

i n f o

Article history: Received 14 October 2010 Received in revised form 1 March 2011 Accepted 2 March 2011 Keywords: Arbuscular mycorrhizal fungi Multiple partners Hyphae Feedback Spatial structure

a b s t r a c t Mycorrhizal fungi may simultaneously associate with multiple plant hosts, and the implications of this for the fungi involved are not well understood. To address this question, two arbuscular mycorrhizal fungi (AMF), Glomus clairoideum (a treatment referred to as “Glo”) and Scutellospora fulgida (a treatment referred to as “Scut”), were grown separately in pots that each consisted of two plant compartments separated by a root-free-compartment (RFC). Fungi within each two-plant-compartment pot were exposed to either two individuals of indiangrass (Sorghastrum nutans), two individuals of big bluestem (Andropogon gerardii), or one of each. A non-inoculated treatment (“Non”) was included to help gauge the potential influence of greenhouse contaminant fungi, cross-contamination, or any misidentification of non-AMF hyphae. The two host species had additive effects on the growth of AM hyphae in plant compartments of Scut, Glo, and Non pots, and in the RFCs of Scut pots. In Glo RFCs, however, they were antagonistic in their effects. Synergism between hosts in Non RFCs suggested that any potential contaminants or misidentification could not explain this result. Underyielding was not seen in shoot weight, root weight, or root length in dual host pots, and also therefore could not explain the result. Hyphal growth in the Scut treatment was evenly distributed between the RFC and plant compartments (or marginally skewed toward the RFC), while hyphal growth in the Glo treatment was skewed toward plant compartments (nearer roots). However, hyphal lengths were more highly correlated across plant compartments within a common pot in the Glo treatment, suggesting that this AMF bridged the RFC to experience the entire two-host pot as a single environment to a greater extent than Scut did. These AMF differed in how they responded to both the species composition of the two-host environment and its spatial structure; potential implications for mycorrhizal community dynamics are discussed. © 2011 Elsevier GmbH. All rights reserved.

1. Introduction In natural communities, mycorrhizal fungi are often affected by multiple plant species simultaneously. Effects that plants may have on one another via shared mycorrhizal fungi have received a great deal of attention, particularly in relation to the potential redistribution of resources among plants that may occur due to hyphal linkages between plants (e.g., reviewed by Jakobsen 2004; Simard and Durall 2004). The fact that the fungi maintain these linkages and that the carbon transported via the linkages may remain within the fungal symbiote (e.g., Pfeffer et al. 2004) suggests that they are important for the fungi as well, but the effects of them on the fungi responsible have received less attention, as have interactions between plant species in their effects on mycorrhizal fungi that may occur in the absence of hyphal linkages. It is possible for example that arbuscular mycorrhizal fungi (AMF) colonizing a focal

∗ Present address: University of Michigan, 2041 Kraus Natural Science Building, 830 North University Avenue, Ann Arbor, MI 48109, USA. Tel.: +1 7346159805. E-mail address: [email protected] 0031-4056/$ – see front matter © 2011 Elsevier GmbH. All rights reserved. doi:10.1016/j.pedobi.2011.03.003

plant may be affected by inoculum from neighboring plants, or by direct or indirect influences of neighbors on the focal plant’s allocation to its AMF. It has also been suggested that seasonality in plant–fungal interactions has the potential to cause fungi to benefit from occurring on multiple plant species, through extension of the fungi’s growing season (Allen and Allen 1992, p. 463). Plant community composition has been repeatedly shown to affect both AM and ectomycorrhizal fungi. The composition of the mycorrhizal fungal community on a plant’s roots may be affected by the presence of neighboring plants (e.g., Johnson et al. 2003; Dickie et al. 2004; Mummey et al. 2005; Hausmann and Hawkes 2009), and this has been found even when those neighbors do not form the same type of mycorrhizas as the focal plant (Haskins and Gehring 2004). Neighbors may also influence the extent of a plant’s mycorrhizal colonization (e.g., Jastrow and Miller 1993; Urcelay et al. 2003; Dickie et al. 2004; McHugh and Gehring 2006). AMF hyphal length density has been shown to increase with plant community richness (Bingham and Biondini 2009). The infectivity of mycorrhizal inoculum may be affected by the identities of plants which have previously occupied a location (e.g., Johnson et al. 1991; Dickie et al. 2006).

210

A.J. Golubski / Pedobiologia 54 (2011) 209–216

Most studies that have examined the effects of host plant neighborhood on AMF have focused on the composition or aggregate growth/biomass response of an entire community of co-occurring (and presumably often competing) AMF. It is also important to examine the effects of the host plant community’s makeup on separate fungi, in order to more mechanistically understand the dynamics behind AMF responses at the community level. One question of particular interest is whether multiple plant hosts interact additively, synergistically, or antagonistically in their effects on a given AM fungus. Theory suggests that non-additive effects of multiple mutualists could facilitate coexistence and diversity (Golubski 2002) and/or affect the relative fitnesses of generous vs. exploitative mutualistic partners (Golubski 2007). Synergistic effects of multiple AMF on a shared host plant have been observed (Jansa et al. 2008), and the importance of synergistic effects of multiple mutualists over long timescales has recently been shown for ant-acacia mutualisms (Palmer et al. 2010). Here, I examine the growth response of arbuscular mycorrhizal fungi (AMF) to pairs of conspecific or heterospecific hosts separated by small root-free compartments (RFCs), in order to determine whether the two host species are additive, synergistic, or antagonistic in their effects on the fungi. Two AMF of differing genera, Glomus clairoideum Schenck & Smith (which I will refer to as “Glo” for conciseness) and Scutellospora fulgida Koske & Walker (referred to as “Scut”), were used, as well as a non-inoculated treatment (referred to as “Non”). Two C4 prairie grasses, indiangrass (Sorghastrum nutans (L.) Nash) and big bluestem (Andropogon gerardii Vitman), were used as the host plant species. Hyphal length per gram dry soil was measured as the indicator of AMF growth. I also examine whether any of three plant parameters (shoot weight, root weight, or root length) predict fungal growth in each combination of host and fungal species, and/or are able to explain the response of a fungus to the combination of the two plant hosts. The effects of the fungi on these plant parameters as well as three parameters that are constructs of them: whole plant dry weight, root:shoot ratio, and specific root length (SRL; m root per g dry weight), are also shown.

2. Methods 2.1. Study species Big bluestem and indiangrass were chosen as hosts in large part because of their similarities (both being large C4 prairie grasses). Synergism or antagonism would be indicated by an increase or depression of fungal growth when both plants are present relative to what would be expected based on fungal growth with each plant alone. If a fungus’ growth on one species differed greatly from its growth on another, synergism or antagonism between hosts might be more difficult to detect and the results would be more sensitive to discrepancies between what the fungus perceives the relative abundances of the two hosts to be vs. what relative abundances are assigned in the analyses. Acting in opposition to that concern is the reasoning that greater differences between plant species should provide greater opportunities for those plants to interact synergistically or antagonistically (i.e., not be functionally redundant) in their effects on fungi. Fungal species were chosen to represent two families (Gigasporaceae (S. fulgida) and Glomeraceae (G. clairoideum)) that tend, in general, to differ in their relative amounts of internal vs. external mycelium (e.g., Hart and Reader 2002b) as well as relative dependence on different structures as inoculum (Klironomos and Hart 2002). This was done in order to attempt to examine the responses of two AMF species that are as different as possible.

2.2. Setup Pots were constructed from 4 in. schedule 40 PVC. They consisted of two plant compartments 33 cm tall, each of which held approximately 800 ml of medium, located on opposite sides of a root-free-compartment (RFC). Each plant compartment was separated from the RFC by 20 ␮m nylon mesh, through which hyphae, water, and nutrients could pass but roots could not. The RFC ranged from 0.5 cm wide at their narrowest point to 2 cm at their widest. Pots were assigned to one of three inoculum treatments: Glo, Scut, or Non (as outlined above). In each of these treatments, there were six replicate pots in which an individual of big bluestem would be grown in each of the two plant compartments, six pots in which an individual of indiangrass would be grown in each plant compartment, and 6 pots in which a big bluestem individual would be grown in one plant compartment and an indiangrass individual in the other. The former two treatments are referred to as “single host” treatments while the latter is referred to as the “dual host” treatment. Hosts were quantified by number of individuals, so the ratio of hosts in the dual host treatment was 1:1. Supplementary Fig. S1 illustrates the pots used. Four pots were excluded from analyses because one plant had died by the end of the experiment; these represented one replicate from each of the following treatments: single host big bluestem with Scut inoculum, single host indiangrass with Non inoculum, dual host with Non inoculum, and dual host with Glo inoculum. Plant seeds were purchased from the Prairie Moon Seed Nursery (Winona, MN; www.prairiemoon.com). They were surface sterilized for 5 min in 5% bleach, then rinsed and stratified in wet sand in a refrigerator for 6 weeks. Fungal inocula were obtained from James Bever, and had been grown from isolates originally collected from an Indiana prairie. Each inoculum was bulked up on sudangrass (Sorghum bicolor (L.) Moench subsp. drummondii (Steud.) de Wet) for 4 months in a growth chamber, with a 2-week period at the end during which pots (with plants still present) a were allowed to dry to encourage sporulation. A homogenized mixture of plant roots and medium (with roots cut to approximately 1 cm segments) was used as experimental inoculum. The resulting material was diluted 1:10 with pasteurized medium to produce the inoculum for the experiment. Pasteurized medium consisted of 9 parts autoclaved sand to 1 part prairie soil which had been collected from Fermi National Accelerator Laboratory, passed through a 1 mm sieve, and autoclaved twice at 100 ◦ C for 1 h with 24 h between the two autoclaving steps. All RFCs, as well as plant compartments of Non pots were filled with only pasteurized medium; plant compartments in Glo or Scut pots were inoculated with 250 ml of the appropriate inoculum, located as a band approximately 3/4 of the way to the top of the compartment. This corresponded to approximately 1/6 of the volume of each compartment after taking into account the amount by which the medium settled when watered for the first time. Plant seeds were added to the top of the medium in each plant compartment, and covered with autoclaved gravel. The experiment was planted on June 25, 2004. Plants were thinned to one per pot on July 16, and grown in a greenhouse through December 6. The greenhouse was not air-conditioned, and windows were whitewashed to control temperatures, which reduced the amount of light available to the plants. Plants were provided with some supplemental (fluorescent) lighting to partially compensate for this. Each plant compartment was given 100 ml of deionized water daily, except for twice weekly when they were instead given 100 ml of nutrient solution (one teaspoon of 2020-20 fertilizer (Peters Professional, Scotts Miracle-Gro Company, Marysville, OH) diluted in 4 gallons of deionized water). Fertilization began in the 3rd week of plant growth after thinning; it was originally hoped that the changing conditions that might be generated through the course of the experiment by depletion of a

A.J. Golubski / Pedobiologia 54 (2011) 209–216

Pots with each inoculum were harvested on three sequential days; medium from each compartment was homogenized and frozen for future extraction of hyphae. Shoots and roots were separated and refrigerated until their fresh weights could be taken. Shoots were then dried and re-weighed. Subsamples of roots were taken, and root lengths in these were measured using WinRHIZO (Regent Instruments, Canada). The remaining roots were dried and re-weighed. Root dry weights and root lengths of the two subsamples, along with the fraction of fresh weight represented by each subsample, were used to estimate root dry weights and root lengths for the original sample. Hyphae were extracted and stained in trypan blue by a modification of the method of Miller and Jastrow, 1992 (as described in Brundrett et al. 1994 pp.24–28). Because the experimental media was very sandy and had relatively little organic debris and little soil aggregation, the centrifugation steps were omitted and soil samples were only soaked and stirred in sodium hexametaphosphate for 30 and 10 min, respectively. AMF hyphal lengths per g dry soil were estimated microscopically at 100× magnification by the grid line intersect method of Tennant, 1975 (as described in Brundrett et al. 1994 pp. 29–33). AMF hyphae were identified on the basis of being aseptate, having irregular (unsmooth) walls, and branching at approximately right angles.

m hyphae per g dry soil

2.3. Harvest and measurements

a

30

Mean plant compartments AM hyphae

24

p=0.5833 18

12

p=0.9779 6

p=0.7104 0

IG & IG

b m hyphae per g dry soil

limited nutrient pool might maximize the opportunities for plants to interact in their effects on the fungi. There were indications of nutrient limitation as the experiment progressed, however, and it was decided that keeping the plants healthy via fertilization was necessary.

211

30

IG & BB

BB & BB

Root-free compartments AM hyphae

p=0.9705

24

18

12

p=0.0347 6

p=0.0293 0

IG & IG

IG & BB

BB & BB

2.4. Statistical analysis For each combination of response variable and inoculum type, the additivity of the two plant hosts was determined by a contrast testing the null hypothesis that the sum of the mean hyphal length density in each of the two single host treatments equaled twice the mean in the dual host treatment. In cases where the contrast showed a significant difference, hyphal length density in the dual host treatment was also compared to each of the corresponding single host treatments in order to look for transgressive over/underyielding (growth in the dual-host treatment that is either higher or lower than in both single-host treatments). Paired t-tests were also used to investigate the effects of compartment type (plant vs. root-free) on hyphal length density within each inoculum. Pearson correlation tests were used to test for correlation between hyphal lengths in the two plant compartments within a pot; note that plant compartments within a pot were considered separately for this test but pooled for all others. Because the contrasts were pre-planned comparisons while paired t-tests for effect of compartment type and correlation tests were not, and because t-tests comparing single-host treatments to dual-host treatments were contingent on significance of the contrast (since the dual host treatment could not be significantly higher or lower than both single-host treatments if it was not significantly different from the mean of the two single-host treatments), the critical p value for the contrasts are not adjusted for multiple comparisons. The critical values for all tests after these involving hyphal data were adjusted ˇ for five multiple comparisons by the Dunn-Sidák method (Sokal and Rohlf 1981 p. 242), resulting in a critical p value of 0.0102. Each data point for each of the directly measured plant parameters (shoot dry weight, root dry weight, and root length) was involved in four comparisons: a contrast similar to that done for hyphal length density to test for over/underyielding in the plant parameter, and comparisons of the plant parameter’s value within a pot type between each inoculum treatment (Glo vs. Non, Scut vs. Non, and Glo vs. Scut).

Fig. 1. Results of and p values for contrasts testing additivity of host species in their effects on AMF hyphal length density in Glo (gray diamonds and gray p values), Scut (black squares and bold p values), and Non (white circles and thin black p values) treatments, in plant compartments (top panel) and root free compartments (bottom panel). Pot types are: indiangrass in both plant compartments (IG & IG), big bluestem in both (BB & BB), or one of each (IG & BB). Hosts were always additive for plant compartments, but had antagonistic effects on hyphae in RFCs with Glo inoculum and synergistic effects on hyphae in non-inoculated RFCs. In the RFCs, the mean for Non in IG & BB pots differed significantly from that in IG & IG pots but not from BB & BB pots. The IG & BB mean for Glo did not differ from either that in IG & IG or BB & BB.

Critical values for these tests were adjusted for four multiple comparisons, resulting in a critical p value of 0.0127. Data are also shown for whole plant weight, root:shoot ratio, and specific root length, but analyses are not carried out on these parameters because of their non-independence from the parameters from which they are calculated. Correlations between hyphal lengths and plant parameters in each pot type were also examined, but no tests were done on those associations. Most analyses were carried out in SAS version 9.1 (SAS institute, Cary, IN). Correlation tests were run in R. Paired t-tests were carried out in SYSTAT version 10 (SPSS, Chicago, IL). 3. Results Hyphal growth was extensive in the experiment. Although not quantified, spores of G. clairoideum were observed throughout the Glo treatment, and spores and auxiliary cells of S. fulgida throughout the Scut treatment, in plant compartments as well as RFCs (which had not received inoculum). Presence of these structures essentially only in the appropriate treatments (single Glomus spores were seen on a small number of Scut slides) suggests that cross-contamination was minimal. Along with the high hyphal length densities and generally good quality of hyphae observed, it also demonstrates that inoculation was successful and AMF did colonize plants in the

212

A.J. Golubski / Pedobiologia 54 (2011) 209–216

experiment. Correlation in hyphal length between plant compartments within pots observed in the Glo treatment (discussed later) also strongly indicates that those hyphae grew during the course of the experiment. Indiangrass and big bluestem were additive in their effects on hyphal length density in the plant compartments for all three inocula (F1,14 = 0.00 and p = 0.9779, F1,14 = 0.32 and p = 0.5833, and F1,13 = 0.14 and p = 0.7104 for Glo, Scut, and Non, respectively). They were also additive in their effects on hyphal length density for the Scut inoculum in the RFCs (F1,14 = 0.00, p = 0.9705). The two plants were antagonistic in their effects on hyphal length density in the RFCs for the Glo inoculum (F1,14 = 5.47, p = 0.0347), and synergistic in their effects in the RFCs for the Non inoculum (F1,13 = 6.00, p = 0.0293; Fig. 1). No significant over/underyielding was observed in shoot weight, root weight, or root length (Fig. 2). The lowest p value for the contrasts testing for such effects was 0.2574, and in fact for each plant parameter other than specific root length, the mean of the dual-host treatment was higher than the average of the two single-host treatment means. This means that the antagonistic effect on Glomus RFC hyphal length cannot be explained by decreased availability of host as measured by any of these host plant parameters. In terms of effects of the inocula on plants, Glo significantly increased root dry weight compared to the Non treatment in dual host (IG & BB) pots (F1,8 = 11.14, p = 0.0103) and Scut significantly decreased root length in BB & BB pots (F1,9 = 10.40, p = 0.0104). Root dry weight was also lower in BB & BB pots with Scut than with Glo inoculum (F1,9 = 15.87, p = 0.0032; Fig. 2). Fig. 2 highlights additional comparisons that either were significant at the within-comparison level but not significant when controlling for multiple comparisons, or were marginally significant at the within-comparison level. It should be noted, however, that any effects of inocula in this experiment may include effects from associated microbiota as well as AMF, because filtrate re-additions to control for non-AMF differences in microbial communities were not performed. Correlations between plant parameters and hyphal length densities were generally weak (data not shown). Furthermore, hyphal length densities were often negatively correlated with one or more of these parameters, positively correlated with a plant parameter in one host but negatively correlated with the same plant parameter in the other host, or positively correlated with a plant parameter in one compartment type but negatively correlated or uncorrelated in the opposite compartment type. Thus it could not be determined which if any of the plant parameters measured indicated the quality of the host environment to the fungi. Hyphal length density in one plant compartment of a pot was better correlated to the matching plant compartment within the pot in Glo than in Scut (Fig. 3a–c; Scut t15 = 0.697, p = 0.4964, R = 0.1772; Glo t15 = 3.4793, p = 0.0034, R = 0.4466; Non t14 = 0.1023, p = 0.9200, R = 0.0273). In the Scut and Non treatments, hyphal lengths per gram dry soil were not significantly different in the RFCs than in the plant compartments (t16 = −0.94 and p = 0.362, and t15 = −1.65 and p = 0.120, respectively), while Glo pots had lower hyphal lengths in their RFCs than in their plant compartments (t16 = 3.59, p = 0.002; Fig. 3d–f).

4. Discussion Previous studies have shown that plant neighbors’ effects on one another can depend on the presence of AMF (e.g., Hartnett et al. 1993; Marler et al. 1999; Kytöviita et al. 2003), and that the effects of AMF on a plant can further depend on the origin (Ronsheim and Anderson 2001) and/or density (reviewed in Koide and Dickie 2002) of neighboring plants present. This study complements those listed previously by focusing on the effects of co-occurring plants

on separate species of mycorrhizal fungi themselves, rather than on plants’ effects on one another via mycorrhizal fungi or plants’ effects on each other’s associations with mycorrhizal fungi. It complements experiments that have explored in greater detail the separate effects of competing plant species on mycorrhizal fungi and other members of the soil microbial community (e.g., Hetrick and Bloom 1986 and several reviewed in Bever 2003), by comparing such single-host effects with the effects of the hosts when mixed. It also complements work highlighting the effects of plant neighborhood on the growth or composition of entire communities of AMF (Johnson et al. 2003; Dickie et al. 2004; Mummey et al. 2005; Bingham and Biondini 2009; Hausmann and Hawkes 2009) by focusing on the plant neighborhood’s effects on one AMF species at a time. Indiangrass and big bluestem were additive in their effects on hyphal length density in the Scut treatment. They were antagonistic in the Glo treatment, but only in RFCs (Fig. 1). Synergism between hosts was observed in the Non RFCs, suggesting that the antagonism observed with Glo was not likely a result of either the response of greenhouse contaminant fungi or any potential misidentification of non-mycorrhizal fungi as AMF. Hyphal length densities in the two plant compartments within a pot were more correlated with one another in the Glo treatment than in the Scut treatment (Fig. 3), which may mean that Glo experienced the entire pot as one environmental patch to a greater extent than Scut did. This would be consistent with dual hosts being additive in their effects on Scut but interacting in their effects on Glo. It is also interesting that even though Glo seemed more likely than Scut to be bridging the RFC and experiencing the entire pot as one patch, within pots Glo hyphae grew preferentially in the plant compartments relative to the RFCs, while Scut hyphae did not (Fig. 3). These patterns suggest that traits related to how AMF interact with the spatial structure of their host environment may not fall out into straightforward categories such as expansive vs. limited dispersers: the AMF species here that seemed to more readily span multiple hosts across a distance actually had the more heterogeneously distributed hyphae, concentrated near plant roots. It is plausible that the optimal locations for AMF to forage for new hosts to infect may not correspond to the optimal locations for AMF to forage for nutrients to supply hosts, and the possibility that AMF growth in different locations relative to plant roots may have different implications for the symbiosis may be an interesting one for future studies to explore. Mycorrhizal fungi are affected by the spatial structure of host communities. The composition of mycorrhizal fungal communities has been found to change across co-occurring host species within a site (e.g., Vandenkoornhuyse et al. 2002), across sites (e.g., Brundrett and Abbott 1995; Pringle and Bever 2002; Husband et al. 2002), with distance from the edge of a habitat type (Dickie and Reich 2005), or in canopy gaps vs. under vegetation (Lovelock and Miller 2002). AMF community variability has also been shown to be affected by plant community heterogeneity and assembly history (van de Voorde et al. 2010). The results of this experiment suggest another mechanism by which the spatial arrangement of hosts might affect mycorrhizal fungi: if multiple hosts are synergistic or antagonistic for some fungi but not others in zones where the hosts overlap in their influence on the fungi, competitive hierarchies among the fungi in those zones of overlap could differ from those found in zones dominated by the effects of one host or another. The spatial specificity of the non-additive effects seen here (in the RFCs but not in the plant compartments) further implies that the importance of such effects for a given fungus may depend on the density of hosts in the environment. This also extends previous findings of differences in the spatial patterns of external hyphal growth between members of these two genera (Smith et al. 2000), by showing that the makeup of the host community can affect those spatial patterns.

A.J. Golubski / Pedobiologia 54 (2011) 209–216

a

Sum plant compartments shoot dry weight

Sum plant compartments whole plant dry weight

d

4

213

6

3

g

g

4 2 1

*

*

2

**

+ 0

IG & IG

IG & BB

0

BB & BB

Sum plant compartments root dry weight

2

0.8

0

BB & BB

g⋅g-1

e 1.2

1

IG & BB

Mean plant compartments root:shoot ratio

3

g

b

IG & IG

* *

***

**

++

+++

IG & IG

IG & BB

0.4

0

BB & BB

Sum plant compartments root length

c 40

IG & IG

IG & BB

BB & BB

Mean plant compartments specific root length

f

30

m⋅g-1

m

30 20

20

10 10

++ 0

*** ++

0

IG & IG

IG & BB

BB & BB

IG & IG

IG & BB

BB & BB

Fig. 2. Results for plant parameters. Data are either sums or means across the two plant compartments within a pot; pot abbreviations are the same as in Fig. 1. Statistical analyses were not conducted on plant parameters which are constructs of others (whole plant dry weight, root:shoot ratio, or specific root length; d–f), to minimize multiple comparison concerns. Contrasts equivalent to those used to test for additivity of hosts in their effects on hyphal length density were non-significant for each combination of inoculum treatment and the remaining plant parameters. Symbols below groups of plant parameter means indicate significance of differences between inoculum treatments Glo and Non (gray asterisk), Scut and Non (black asterisk), and Glo and Scut (plus). One symbol indicates marginal significance: 0.1 < p < 0.05, two symbols indicates withincomparison significance 0.05 < p < 0.0127, and three symbols indicates significance when adjusted for four non-independent comparisons using the same data p < 0.0127 (the latter being the cases reported in the main text). Other symbols are as in Fig. 1.

The non-additive effect of multiple hosts seen in this experiment has implications for plant-environment feedback. If habitat quality for plants depends in part on the amount of some AMF present, then antagonistic interactions between plant species on that AMF could cause habitat quality to be particularly low at the boundaries between monospecific aggregations of plants. Theory suggests that under some circumstances such a scenario could promote founder effects among competing plants, reinforce intraspecific clumping of plants, and favor plants that increase habitat quality, even if they pay a cost in terms of short-term reproductive output to do so (Golubski 2007). One clear limitation of the current study is the use of hyphal length as the sole indicator of fungal growth. The responses seen in hyphal length density may not reflect the responses of all fungal structures in this experiment, and spores or structures colonizing roots, which were not quantified, could have provided complementary metrics of fungal performance. The hyphal length responses seen here also may not generalize across hosts: Ehinger et al. (2009) for example recently found relative abundance of hyphae to spores to vary across three host plant species, in ways that differed among three isolates of the same AMF species tested. Using a mixed-AMF inoculum and one of the host species in this study (big

bluestem), Schultz et al. (2001) found that the combination of soil origin and plant ecotype leading to the highest spore volume led to the lowest hyphal length density, and vice versa (although not all of the relevant interaction terms were statistically significant). In an experiment comparing the responses of four AMF isolates to elevated CO2 , Klironomos et al. (1998) found that two isolates responded in their hyphal length density while two responded in their spore numbers (interestingly, the former were the two with relatively higher hyphal length density and the latter were the two with higher spore numbers under ambient CO2 as well). While spores are generally thought to be of greater relative importance as inoculum to the Gigasporaceae than the Glomeraceae (Hart and Reader 2002a and references therein), it was two Glomus species that responded by spore numbers and a Scutellospora species that was among the two responding by hyphal length density to elevated CO2 in Klironomos et al. (1998). It is therefore possible that hyphal length density was not the parameter that would have best predicted the future inoculum potential of one or both of the AMF in this study. A clearer understanding of the relative importance of hyphae in each AMF species for generating new infections vs. foraging for nutrients to supply to hosts would aid the interpretation of the patterns observed here.

214

A.J. Golubski / Pedobiologia 54 (2011) 209–216

60

Scut :

60

r=0.177

Scut:

p=0.362

p=0.496 45

45

30

30

15

15

a

0 15

30

45

0

60

20

15

30

45

60

20

Glo:

r=0.668

m hyphae per g dry soil in RFC

m hyphae per g dry soil in plant compartment "A"

0

d

0

p=0.003 15

10

5

Glo:

p=0.002

15

10

5

b

e

0

0

0

5

10

15

20

10

0

5

10

15

20

10

Non:

r=0.027

Non:

p=0.920

8

p=0.120

8

6

6

4

4

2

2

c

f

0

0 0

2

4

6

8

10

m hyphae per g dry soil in plant compartment "B"

0

2

4

6

8

10

m hyphae per g dry soil in mean of plant compartments

Fig. 3. Relationships between hyphal length densities among the two plant compartments (a–c) or between the mean of the plant compartments and the RFC (d–f) within each pot. A critical p value of 0.0102 controls for five comparisons. In (a–c) Pearson r values reflect the degree of correlation and p values are for t-tests on the significance of that correlation. In (d–f), p values are for paired t-tests on effect of compartment type. The 1:1 line is drawn in each panel for reference. The scales of the axes differ between panels. Removing the data point to the far right in (d) changed the p value to 0.072, with the trend toward higher growth in the RFC. The correlation tests in (a) changed only slightly if the data point to the far right was removed.

No plant parameter seemed capable of explaining the antagonistic effect seen in this experiment (Fig. 2), and in general plant parameters were not well correlated with the hyphal length densities observed (data not shown). In fact, some of the strongest correlations between plant parameters and hyphal length densities were negative. These results mean that it cannot be determined which, if any, of the plant parameters measured indicated the amount or quality of partner available to the fungi. Reproductive structures (seeds in the case of big bluestem and rhizomes, often

with buds or tillers, in the case of indiangrass) were obvious on most plants by the end of the experiment, and it is possible that allocation to these structures may have contributed to masking the relationship between one or more of the plant parameters measured and hyphal length densities of the fungi. It is also interesting to note that the two inocula affected different plant parameters. Both inocula reduced specific root length in big bluestem for example (Fig. 2f), but for different reasons: Glo increased root weight (Fig. 2b) while Scut reduced root length

A.J. Golubski / Pedobiologia 54 (2011) 209–216

(Fig. 2c). However, it should be noted that because microbial rinses were not applied, differences between inocula in microbes other than the focal AMF species could have affected the inoculum treatments’ effects on plants. Potential microbial community differences between inocula could not be responsible for the effects of plant treatments within an inoculum type, and so do not affect conclusions regarding interactive effects of plants on each AMF separately. However, it cannot be ruled out that the fungal growth responses to host environment seen here might depend on soil microbial community composition (so that, for example, plants might have had antagonistic effects on S. fulgida hyphae in the presence of the Glo inoculum’s microbial community). In this aspect the two AMF species’ responses were not measured in a common environment, so that their individual responses to particular host environments might not entirely reflect what those responses would be when in competition with one another. As more is learned about AMF responses to various host environments, the potential impact of the wider soil microbial community (including competing AMF, which were absent from all inocula here) on those responses will be a valuable facet to explore. These results show that plant species can have non-additive effects on at least one aspect of AMF growth. This work further suggests that the response of AMF to zones between hosts as well as zones under specific plants, even at short distances, might be important in mycorrhizal communities. AMF species may differ in their interaction with a spatially structured host environment in complex ways, with the degree to which a fungus is affected by neighboring hosts at a distance not necessarily being reflected by the expansiveness of that fungus’ hyphal network. It is important to identify such differences between AMF species if we are to understand how those species compete for plant hosts and partition the resource represented by the host plant community so as to allow a diverse AMF community to coexist. This may also help elucidate the mechanisms behind the small-scale spatial heterogeneity that has been observed in AMF communities (Wolfe et al. 2007; Mummey and Rillig 2008). Acknowledgments A. Golubski was supported during part of this work by a University Fellowship from the University of Illinois at Chicago. Thanks to co-advisors John Lussenhop and Joel Brown, and committee members Hormoz BassiriRad, Jim Bever, and Mike Miller for feedback, materials and use of lab space and resources, Rod Walton and Fermilab for use of their site, Wittaya Kaoongbua for help with inoculum, Louise Egerton-Warburton for help with microscopy methods, and Larry Sikora, Jim Scios, Erin Haase, Kathy Craft, Barbara Zorn-Arnold, Sarah O’Brien, Erin O’Brien, Ellen Marquez, Jason Moll, Kara Borden, George Konstantinopolous, and John Conners for their help setting up, maintaining, and/or harvesting the experiment. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.pedobi.2011.03.003. References Allen, M.F., Allen, E.B., 1992. Mycorrhizae and plant community development: mechanisms and patterns. In: Carroll, G.C., Wicklow, D.T. (Eds.), The Fungal Community; Its Organization and Role in the Ecosystem. , second edition. Marcel Decker, New York, pp. 455–479. Bever, J.D., 2003. Soil community feedback and the coexistence of competitors: conceptual frameworks and empirical tests. New Phytol. 157, 465–473. Bingham, M.A., Biondini, M., 2009. Mycorrhizal hyphal length as a function of plant community richness and composition in restored northern tallgrass prairies (USA). Rangeland Ecol. Manage. 62, 60–67.

215

Brundrett, M.C., Abbott, L.K., 1995. Mycorrhizal fungal propagules in the jarrah forest II spatial variability in inoculum levels. New Phytol. 131, 461–469. Brundrett, M., Melville, L., Peterson, L. (Eds.), 1994. Practical Methods in Mycorrhiza Research. Mycologue Publications. Dickie, I.A., Guza, R.C., Krazewski, S.E., Reich, P.B., 2004. Shared ectomycorrhizal fungi between a herbaceous perennial (Helianthemum bicknellii) and oak (Quercus) seedlings. New Phytol. 164, 375–382. Dickie, I.A., Oleksyn, J., Reich, P.B., Karolewski, P., Zythowiak, R., Jagodzinski, A.M., Turzanska, E., 2006. Soil modification by different tree species influences the extent of seedling ectomycorrhizal infection. Mycorrhiza 16, 73–79. Dickie, I.A., Reich, P.B., 2005. Ectomycorrhizal fungal communities at forest edges. J. Ecol. 93, 244–255. Ehinger, M., Koch, A.M., Sanders, I.R., 2009. Changes in arbuscular mycorrhizal fungal phenotypes and genotypes in response to plant species identity and phosphorous concentration. New Phytol. 184, 412–423. Golubski, A.J., 2002. Potential impacts of multiple partners on mycorrhizal community dynamics. Theor. Popul. Biol. 62, 47–62. Golubski, A.J., 2007. Interactive effects of builders and exploiters on environmental quality and the outcome of competition between the two. J. Theor. Biol. 249, 46–57. Hart, M.M., Reader, R.J., 2002a. Host plant benefit from association with arbuscular mycorrhizal fungi: variation due to differences in size of mycelium. Biol. Fertil. Soils 36, 357–366. Hart, M.M., Reader, R.J., 2002b. Taxonomic basis for variation in the colonization strategy of arbuscualr mycorrhizal fungi. New Phytol. 153, 335–344. Hartnett, D.C., Hetrick, B.A.D., Wilson, G.W.T., Gibson, D.J., 1993. Mycorrhizal influence on intra- and interspecific neighbour interactions among co-occurring prairie grasses. J. Ecol. 81, 787–795. Haskins, K.E., Gehring, C.A., 2004. Interactions with juniper alter pinyon pine ectomycorrhizal fungal communities. Ecology 85, 2687–2692. Hausmann, N.T., Hawkes, C.V., 2009. Plant neighborhood control of arbuscular mycorrhizal community composition. New Phytol. 183, 1188–1200. Hetrick, B.A.D., Bloom, J., 1986. The influence of host plant on production and colonization ability of vesicular–arbuscular mycorrhizal spores. Mycologia 78, 32–36. Husband, R., Herre, E.A., Turner, S.L., Gallery, R., Young, J.P.W., 2002. Molecular diversity of arbuscular mycorrhizal fungi and patterns of host association over time and space in a tropical forest. Mol. Ecol. 11, 2669–2678. Jakobsen, I., 2004. Hyphal fusion to plant species connections – giant mycelia and community nutrient flow. New Phytol. 164, 4–7. Jansa, J., Smith, F.A., Smith, S.E., 2008. Are there benefits of simultaneous root colonization by different arbuscular mycorrhizal fungi? New Phytol. 177, 779–789. Jastrow, J.D., Miller, R.M., 1993. Neighbor influences on root morphology and mycorrhizal fungus colonization in tallgrass prairie plants. Ecology 74, 561–569. Johnson, N.C., Pfleger, F.L., Crookston, R.K., Simmons, S.R., Copeland, P.J., 1991. Vesicular–arbuscular mycorrhizas respond to corn and soybean cropping history. New Phytol. 117, 657–663. Johnson, D., Vandenkoornhuyse, P.J., Leake, J.R., Gilbert, L., Booth, R.E., Grime, J.P., Young, J.P.W., Read, D.J., 2003. Plant communities affect arbuscular mycorrhizal fungal diversity and community composition in grassland microcosms. New Phytol. 161, 503–515. Klironomos, J.N., Hart, M.M., 2002. Colonization of roots by arbuscular mycorrhizal fungi using different sources of inoculum. Mycorrhiza 12, 181–184. Klironomos, J.N., Ursic, M., Rillig, M., Allen, M.F., 1998. Interspecific differences in the response of arbuscular mycorrhizal fungi to Artemisia tridentata grown under elevated atmospheric CO2 . New Phytol. 138, 599–605. Koide, R.T., Dickie, I.A., 2002. Effects of mycorrhizal fungi on plant populations. Plant Soil 244, 307–317. Kytöviita, M.M., Vestberg, M., Toumi, J., 2003. A test of mutual aid in common mycorrhizal networks: established vegetation negates benefit in seedlings. Ecology 84, 898–906. Lovelock, C.E., Miller, R., 2002. Heterogeneity in inoculum potential and effectiveness of arbuscular mycorrhizal fungi. Ecology 83, 823–832. Marler, M.J., Zabinski, C.A., Callaway, R.M., 1999. Mycorrhizae indirectly enhance competitive effects of an invasive forb on a native bunchgrass. Ecology 80, 1180–1186. McHugh, T.A., Gehring, C.A., 2006. Below-ground interactions with arbuscular mycorrhizal shrubs decrease the performance of pinyon pine and the abundance of its ectomycorrhizas. New Phytol. 171, 171–178. Mummey, D.L., Rillig, M.C., 2008. Spatial characterization of arbuscular mycorrhizal fungal molecular diversity at the submetre scale in a temperate grassland. FEMS Microbial. Ecol. 64, 260–270. Mummey, D.L., Rillig, M.C., Holben, W.E., 2005. Neighboring plant influences on arbuscular mycorrhizal fungal community composition as assessed by T-RFLP analysis. Plant Soil 271, 83–90. Palmer, T.M., Doak, D.F., Stanton, M.L., Bronstein, J.L., Kiers, E.T., Young, T.P., Goheen, J.R., Pringle, R.M., 2010. Synergy of multiple partners, including freeloaders, increases host fitness in a multispecies mutualism. P. Natl. Acad. Sci. U.S.A. 107, 17234–17239. Pfeffer, P.E., Douds Jr., D.D., Bucking, H., Schwartz, D.P., Shachar-Hill, Y., 2004. The fungus does not transfer carbon to or between roots in an arbuscular mycorrhizal symbiosis. New Phytol. 163, 617–627. Pringle, A., Bever, J.D., 2002. Divergent phenologies may facilitate the coexistence of arbuscular mycorrhizal fungi in a North Carolina grassland. Am. J. Bot. 89, 1439–1446.

216

A.J. Golubski / Pedobiologia 54 (2011) 209–216

Ronsheim, M.L., Anderson, S.E., 2001. Population-level specificity in the plantmycorrhizae association alters intraspecific interactions among neighboring plants. Oecologia 128, 77–84. Schultz, P.A., Miller, R.M., Jastrow, J.D., Rivetta, C.V., Bever, J.D., 2001. Evidence of a mycorrhizal mechanism for the adaptation of Andropogon gerardii (Poaceae) to high- and low-nutrient prairies. Am. J. Bot. 88, 1650–1656. Simard, S.W., Durall, D.M., 2004. Mycorrhizal networks: a review of their extent, function, and importance. Can. J. Bot. 82, 1140–1165. Smith, F.A., Jakobsen, I., Smith, S.E., 2000. Spatial differences in acquisition of soil phosphate between two arbuscular mycorrhizal fungi in symbiosis with Medicago truncatula. New Phytol. 147, 357–366. Sokal, R.R., Rohlf, F.J., 1981. Biometry, second edition. W. H. Freeman and Company, San Fransisco.

Urcelay, C., Bret-Harte, M.S., Diaz, S., Chapin, F.S., 2003. Mycorrhizal colonization mediated by species interactions in arctic tundra. Oecologia 137, 399–404. Vandenkoornhuyse, P., Husband, R., Daniel, T.J., Watson, I.J., Duck, J.M., Fitter, A.H., Young, J.P.W., 2002. Arbuscular mycorrhizal community composition associated with two plant species in a grassland ecosystem. Mol. Ecol. 11, 1555–1564. van de Voorde, T.F.J., van der Putten, W.H., Gamper, H.A., Hol, W.H.G., Bezemer, T.M., 2010. Comparing arbuscular mycorrhizal communities of individual plants in a grassland biodiversity experiment. New Phytol. 186, 746–754. Wolfe, B.E., Mummey, D.L., Rillig, M.C., Klironomos, J.N., 2007. Small-scale spatial heterogeneity of arbuscular mycorrhizal fungal abundance and community composition in a wetland plant community. Mycorrhiza 17, 175–183.