Danthonia spicata (Poaceae) and Atkinsonella hypoxylon (Balansiae): Environmental Dependence of a Symbiosis

by Melissa K. McCormick, Robin A. Smith, Katherine L. Gross
Danthonia spicata (Poaceae) and Atkinsonella hypoxylon (Balansiae): Environmental Dependence of a Symbiosis
Melissa K. McCormick, Robin A. Smith, Katherine L. Gross
American Journal of Botany
Start Page: 
End Page: 
Select license: 
Select License






?W. K. Kellogg Biological Station, Michigan State University, 3700 East Gull Lake Drive, Hickory Cor~iers, Michigan 49060 USA;
'Department of Zoology and Program in Ecology, Evolutionary Biology and Behavior, Michigan State University, East Lansing,
Michigan 48823 USA; and 'Department of Botany and Plant Pathology and Program in Ecology,
Evolutionary Biology and Behavior, A4ichigan State University, East Lansing, ~Michigan 48823 USA

Epiphytic and endophytic fungal infections often enhance plant growth. Howevel; supporting active fungal tissue may be costly to plants in low-nutrient conditions and may affect the spatial distuibution of host plants in lieterogeneous environments. We examined the field distribution of Uaiztizorrin spicaftr infected and uni~ifected by the epiphytic fu~igus Arkirnronella Irvpo.r-);lorz relative to soil resource levels. We also conducted a greenhouse experiment to determine how L). spicntn growth and performance responded to soil fertility and moisture. In two of three field populations, locations where A. Izypo.vylon occurred had higher ammonia, but lower soil moisture. than locations where D. spicclfi~were uninfected. Infected and uliinfected plants had similar growth rates across greenhouse treatments, but infected plants bad a perforlnance (size X survival) disadvantage relative to uninfected plants in high-nutrient, high- moisture and low-nutrient, low-moisture conditions. Field locations with D. spicofn had low soil moisture, thus the performance disadvantage of infected plants in low-nutrient. low-moisture conditions correspo~ids to tield observations that infected plants are rare in habitats with low ammonia. In a field common garden, infected plants had higher nitrogen concentrations than uninfected plants, suggesting that high nitrogen demand by .4. hvpoxyloi~may exclude infected plants from low-fertility field locations.

Keg words: Dnnthorzin spicurrt; environmental variability; plant-fungal interactions; spatial heterogeneity; symbiosis.

Much attention has been focused recently on the role of fungal symbiotes in structuring interactions within and among plant populations and communities (Clay, 1990b; Dobson and Crawley, 1997; Thrall and Burdon, 1997; Clay and Holah, 1999). Many plant species, particularly grasses, commonly form symbiotic associations with epiphytic or endophytic fun- gi that have the potential to strongly alter a host plant's morphology, physiology. and response to environmental condi- tions (e.g., Bacon et al., 1986; Read and Camp. 1986; Belesky et al., 1987; Carroll, 1988; Marks and Clay, 1990). These fungi can be beneficial to their host plant in some environments, but may be parasitic in others (Clay, 1990a).

Nutrient requirements of active, sometimes toxin-producing, fungal biomass may outstrip nutrient availability when nutri- ents available to a host are very low, causing otherwise ad- vantageous fungal infections to be disadvantageous (Bacon, 1993). Soil fertility-dependent growth advantages of fungal infection have been demonstrated for endophyte-infected Loliun? perenlie and Festuca nrundi1tncea (Cheplick, Clay, and Marks, 1989; Marks and Clay, 1990; Cheplick, 1997). In both of these species, the growth advantage of infected plants was

' Nlanuscript received 9 March 2000: revision accepted 5 July 2000.

The authors thank Adrienne Purdy. Carrie Stolt, and Andrea Corbett for help with collecting, processing, and analyzing samples, and Keith Clay. Jeff Conner. Tom Getty, Sam Scheiner, and Alan Tessier for comments that im- proved the manuscript. Support for this project \vas provided by the NSF grants to the RTG (NSF grant DIR-9602252) and REU (DB1 9605 168) pro- grams at the W. K. Kellogg Biological Station and by an NSF dissertation improvement grant to K. L. Gross and M. K. McCormick (DEB-9623772). This is KBS contribution no. 914.

Author for correspondence, current address: Smithsonian Environmental Research Center, PO. Box 28, Edgewater, Maryland 21037 USA. Turrent address: Department of Botany, Duke University, Durham, North Carolina 27708 USA.

either small or negative at the lowest fertility levels and in- creased with fertility in the greenhouse.

In most plant species that have epiphytic or endophytic fun- gal associations, both infected and uninfected individuals oc- cur within and among populations (Bradshaw, 1959; Clay, 1990a). In mixed populations, infected and uninfected plants may be patchily distributed, with some areas having predom- inantly infected and other areas predominantly uninfected plants. The distribution of infected plants may have substantial impacts on population and community composition and inter- actions (Dobson and Crawley, 1994; Clay and Holah, 1999).

Soil resources also can vary spatially within and among plant communities over a range of scales (Jackson and Cald- well. 1993; Robertson and Gross, 1994; Gross, Pregitzer, and Burton, 1995). If infected and uninfected plants respond dif- ferently to soil conditions that are patchily distributed, then the spatial distribution of infected individuals in a population might correspond to patches of environmental conditions that are favorable to infected plants.

To determine whether patchy distributions of soil nitrogen and percentage water were related to observed spatial variation in a plant-fungal symbiosis, we exanlined the correlation be- tween Atki~zsonella hypoxylon infection and soil resource (ni- trogen and moisture) in three populations of Danthonia spicato in southwestern Michigan. We conducted a parallel greenhouse experiment to test whether growth and survival of D. .spicata plants infected by A. hypoxylon differed from that of unin- fected plants under high and low levels of moisture and fer- tility. A common garden planting of infected and uninfected culms was used to determine how the growth and survival of infected and uninfected plants differed in the field.


llonthornin spicnfn (Poaceae) is a native perennial C, bunch grass that com- monly occurs in dry, nutrient-poor oak savannas, old fields, and grassland openings throughout the eastern and northern United States and southern Can- ada (Clay, 1982). Throughout much of its range, D. spicata is infected by the epiphytic fungus Atkinsonella hypoxylon (Balansiae), which is specific to the genus Danthonin (Clay, 1994). Spread of A. hypoxylon in D. spicnta populations occurs asexually through seedlings from infected plants and sexually through dissemination of ascospores, but infection of previously healthy plants is very rarely observed in the field (Kover, Thomas, and Clay, 1997). Unin- fected D. .spicnta produce two flower types. Potentially outcrossed, wind- pollinated chasmogamous flowers are produced at the tip of the reproductive stalk. Obligately self-fertilized cleistogamous flowers are produced in the axils of the leaf sheaths along the reproductive stalk. In infected plants, a fungal sclerotium, or "choke," is produced at the initiation of host plant flowering and causes abortion of all but a few infected cleistogamous seeds at the base of the much-reduced reproductive stalk. Plants infected by A. hypoxylon often have higher growth rates than uninfected plants and so this symbiosis is gen- erally considered mutualistic (Diehl, 1950; Clay, 1984, 1990a). The proportion of plants infected by A. hypoxylon varies among populations in Michigan. Some populations have a high proportion of plants infected by A. hypoxylon (>75%; Leuchtmann and Clay, 1989), but others are entirely free of infection (Scheiner, 1989; McCormick, 1999).

The three populations of D. sl~icatnwe examined in this study were all partially infected by A. hypo.qloiz and growing in low-productivity old fields that had been abandoned at least 45 yr ago. Plant communities at these sites were dominated by herbaceous, perennial species, especially Andropogon vir- ginicus, Schiznchiriiirn scoparium, and Carex spp. Two populations (L, and L,) were located near the W. K. Kellogg Biological Station (KBS) of Michigan State University in Hickory Corners, Michigan, USA. The third population

(R) was located in the Rose Lake Wildlife Research Area near East Lansing, Michigan, USA.

In each population, we measured plant density (number of D. spicnta plants per square meter) and percentage infection along two randomly located belt transects (12 X 0.25 m). We also established a 4 X 6 m permanent plot where we quantified soil moisture and nitrogen supply. Seventy soil samples were taken from each plot, using a 10 cm deep, 2.5 cm diameter soil core, in a stratified, nested sampling design. At each location where a soil core was removed we inserted an ion exchange resin bag (Dowex MR-3, Sigma, St. Louis, Missouri, USA) and buried it under 5 cm of soil to measure the relative supply of ammonia-nitrogen in the soil. Soil moisture and relative supply of ammonia were chosen to characterize environmental quality because these resources often limit plant growth in the dry, nutrient-poor environments where D. spicata occurs. We also sampled nitrate-nitrogen in these locations, but nitrate levels were low and were not consistent among years in these sites (McCormick, 1999). Because relative nitrate-nitrogen supply was not spatially or temporally consistent, we felt that it would not be a reliable determinant of microhabitat conditions related to infection distribution.

We measured soil moisture gravimetrically in all three populations in mid- May 1998, corresponding to the time of maximum growth by D. spicata. The relative ammonia supply rate was estimated in all three populations for 4 mo after sampling for infection (June 1997-October 1997). Soil moisture and ammonia supply were also measured at other times using identical methods (see McCormick, 1999), but these other measurements were used only to assess temporal consistency in infected and uninfected quadrats. We used 1998 measures of soil moisture and 1997 measures of ammonia supply to characterize the spatial pattern of soil resources in each sites. These sampling periods were closest in time to when we measured the distribution of infected and uninfected plants in these populations.

To determine whether the pattern of infection in each population was cor- related with soil moisture or ammonia supply, we surveyed the incidence of

A. hypoxylon infection in each population after the plants had bolted in June 1997. Infected plants are easily distinguished from uninfected plants at this stage by the presence of a gray fungal sclerotium ("choke") on each aborted reproductive stalk. We randomly selected 75 quadrats (25 X 25 cm) from within each of our three permanent plots (4 X 6 m; 384 quadrats) to survey for incidence of infection. This small quadrat size was chosen to allow as- sessment of plant density without integrating across substantial variation in spatially heterogeneous environmental conditions. In each quadrat, we noted the number and location of infected and uninfected D. spicata plants. We used semivariance analysis and kriging interpolation (GS' version 3.1 1.6. 1999, Gamma Design Software, Plainwell, Michigan, USA) to estimate percentage water and ammonia supply for the center of each quadrat that we sampled for infection incidence. Kriging interpolation uses the variance-distance relationship, summarized in a semivariogram, to assign weights to sample points as a function of their distance from a point for which an estimate is desired (Robertson and Gross, 1994).

We used logistic regression (Systat 8.0 for Windows, Systat, 1998, Evans- ton, Illinois, USA) to analyze the distribution of infected quadrats over the range of soil moisture and ammonia conditions across the three populations. We designated a quadrat as infected if it contained at least one infected plant. Because of the possible confounding of current spatial structure of infection with point of infection introduction or other processes within a population, we can only draw conclusions about the association between environmental factors and infection incidence across all three populations. A significant pop- ulation effect in the regression would indicate that patterns were not the same across the three populations

We also examined whether there were indirect environmental effects on the distribution of infection acting through plant density. Favorable environments might promote higher plant densities and plant density could affect contagious spread of the fungus. To evaluate this effect, we estimated plant density across all quadrats and used a t test to compare the average plant density in infected and uninfected quadrats.

We conducted a 2 X 2 factorial greenhouse experiment in which watering regime and fertilizer were manipulated to determine whether variation in soil fertility and moisture levels could affect performance and thus influence the differential distribution of infected and uninfected plants in the field. We col- lected cleistogamous seeds from 21 infected and 28 uninfected individuals selected randomly from two D. spicntn populations. We collected seeds from Population L,, where we had conducted the field pattern survey, and a second, unsurveyed, population located at the Lux Arbor Reserve of KBS. The her- baceous vegetation at this site was similar to L, and L?, except for the presence of more woody species. Seedlings from these two populations had similar growth and survival in the greenhouse and thus were combined in all analyses.

We surface sterilized infected and uninfected seeds with bleach and ethanol according to the methods of Leuchtmann and Clay (1988) and nicked each with a sterile razor blade to stimulate germination. Seeds were then placed in moist, sterile sand in a growth chamber with 14-h days at 29°C and 10-h nights at 24°C. Of the 21 infected and 28 uninfected plants from which we collected seeds, 12 infected and 17 uninfected plants had sufficient seed ger- mination for experimental replication. Nine days after being sown in the ster- ile sand, we randomly assigned the 109 seedlings from infected and uninfected families to four treatment groups and planted them into random positions in 70-hole conetainers (each hole was 2.5 cm diameter X 15 cm deep) filled with sterile silica sand. Each infected family was represented by nine or ten seedlings, two in each treatment with the additional one or two seedlings randomly assigned to treatments. Each uninfected family was represented by six or seven seedlings, one per treatment with the other two or three randomly assigned to treatments. The plants were placed in the greenhouse under am- bient light. After a 4-d stabilization period with daily watering to saturation, we established the four treatments (2 X 2 factorial) in which fertility and watering regime were varied. Seedlings from each infected and uninfected family were grown under all combinations of high and low fertilizer and high and low watering.

The moisture and fertility levels used in the experiment were chosen to represent the range of conditions observed in the field, without imposing ex- tremely high mortality. Fertilizer levels consisted of 0.015 g/L (low fertility) or 0.500 gL (high fertility) of Peter's Peat-lite Special Fertilizer (20-10-20) applied at a rate of 4.5 mL per planting location 2-3 times per week. The nitrogen levels in these fertility treatments corresponded to nitrogen miner- alization rates in field incubations of 0.01 and 0.45 mg N.g dry soil-l.d-l, the approximate range of fertility found in field populations of D. spicatci (McCormick, 1999). Plants in high-moisture treatments were watered daily, while plants in the low-moisture treatments were watered every second or third day, when approximately half of the plants showed leaf rolling, symptomatic of


water stress. Average greenhouse temperatures were -30°C from June to September and -24°C from October to April with ambient light. The posi- tions of the 70-hole conetainers on the greenhouse bench were randomly ro- tated each week to minimize location effects.

We monitored plants monthly over 9 mo for size and survival, using the number of living leaves as our measure of plant size. After 6 mo, plants in high-fertility treatments were sufficiently large that they began to shade plauts in adjacent locations. Therefore, we transplanted plants in these treatments into staggered locations in 70-hole conetainers, with one empty cell on all sides to prevent shading.

Plant growth data were log transformed to minimize heterogeneous vari- ances produced by substantial growth differences between high- and low- fertility treatments. Transformed data were analyzed using a repeated measures MANOVA (Systat 8.0 for Windows, Systat, 1998, Evanston, Illinois, USA). We also calculated performance by using the percentage survival as a weight on the size of each plant in each treat~nent (log number of leaves). We compared seedling performance among treatments over the experimental period using a repeated measures ANOVA.

To assess differences in growth of infected and uninfected plants under field conditions, we established a common garden adjacent to our permanent plot in Population L,. We collected nine uninfected and three infected adult

D. spicntn plants from field Populations L, and L, in April 1995. The plants were divided into cul~ns and each culm was planted into a 10 X 10 crn pot filled with sterile sand in the greenhouse. Culms that produced vegetative shoots were again divided into individual culms and planted into new 10 X 10 cm pots in May 1995. The culms were grown in the greenhouse without fertilizer until June 1995, when we used them to establish the common garden experiment.

We established a common garden by tilling a 2 X 2 In area near Population L,. We planted individual culnls of each genotype (individual parent) into four randomly selected locations on a 10 X 10 cm grid established in the central 1 X I m of the tilled area. Remaining planting locations in the grid were occupied by D. spicntn culms from another experiment, We planted additional cullns around the perimeter of the colnmon garden to avoid edge effects on the study plants.

Culms were grown in the colnmon garden for 2 yr (June 1995-1997) and then harvested for determinat~on of aboveground biomass. At the time of harvest. plants were still separated by at least 5 cm of open ground. We measured final plant size by counting the number of tillers, leaves, and inflo- rescences produced. We also measured initial size, but its effect on final size was not significant (Pearson correlation. P > 0.3), so we only considered final size in these analyses. All harvested material was dried at 45'C for 48 h and then weighed to the nearest 0.01 g. We ground subsamples of leaf material from each culm to determine tissue nitrogen concentration. We also ground reproductive stalks, with seeds removed, from all infected culms and from three randomly selected uninfected culms from different parents. For each infected reproductive stalk we included the fungal sclerotium, which was in- timately associated with the stalk, in tissue to be ground so we could assess the overall nitrogen content of the reproductive stalk. Percentage nitrogen content of this tissue was analyzed using an elemental analyzer (Nitrogen Analyser 1500 Series 2, 1990, Carlo-Erba Instruments, Milan, Italy).


Plant density and ammonia nitrogen were not statistically different among the three sites, but Population R had signifi- cantly lower soil moisture than Populations L, and L2. The three sites also had similar percentages of infected plants (Ta- ble 1). Over all three sites, plant density in infected quadrats was not significantly different from that in uninfected quadrats

(40.8 t-0.79 plants/m2 vs. 35.6 i3.65 plants/m2, respectively [mean i 1 SE]; P > 0.2 from t test). Within each site, soil moisture estimates for 1996 or 1997 were significantly predic- tive of 1998 soil moisture (P < 0.04). Soil ammonia levels estimated in 1997 were significantly predictive of 1998 am- monia in two of the three populations (L, and R, P < 0.004;

TABLE1. Plant density (number of plants per square meter), percentage of plants infected, soil moisture (1998), atnmonia nitrogen, and nitrate-nitrogen (1997 pprn nittogen, i 1 SE) in three populations. Plant density measures in 10 X 0.25 m parallel transects; soil mois- ture and nitrogen measured in randomly located points within 4 X 6 m study plot. Values given are means t SE and range (maximum to minimum soil moisture, ammonia, and nitrate).


Vatiahle    L,    L2    R    SLZC
Density    37.8 i 3.8            
% infected    23            
5% HH,O    18.5 t 0.18            
NH,-N    1.25 i 0.09            
NO,-N    2.16 i 0.17            
L,,P = 0.540), suggesting that heterogeneity in soil resources at these sites was generally temporally consistent among years.

Logistic regression analysis indicated there was a significant interaction between soil moisture and ammonia supply in de- termining infection prevalence (P < 0.001), due largely to the absence of infected plants in locations with both high moisture and low ammonia. Populations L, and L, had substantial over- lap in soil nloisture and ammonia conditions. In both these populations, infected plants occurred primarily in areas with low moisture and high ammonia conditions (Fig. I). In con- trast, Population R had somewhat higher ammonia and sub- stantially lower soil moisture than Populations L, or L,, and infected plants were distributed throughout Population R without regard to soil moisture or ammonia conditions (Fig. 1).

In the greenhouse experiment, growth and survival differ- ences between infected and uninfected plants depended on both the fertility and moisture treatments. All plants were smaller in low-fertility than in high-fertility treatments (Fig. 2), resulting in a significant Fertility effect (P < 0.001). How- evel; no other between-subject effect was significant. Within subjects, Time X Water X Fertility and Time X Infection X Fertility effects were significant (P < 0.01), indicating that infected and uninfected plants grew differently in response to fertility treatment. The Time X Water X Fertility X Infection was not significant (Fig. 2).

Plant performance, calculated as the product of average group survival and plant size (log number of leaves) was sig- nificantly different among the experimental treatments (Fig. 3). Using performance as an index of relative group success also revealed differences within fertility groups: uninfected plants in high-moisture, high-fertility conditions outperformed all other high-fertility plants (Fig. 3). Uninfected plants were somewhat smaller than infected plants in the high-fertility, high-moisture treatment, but had higher survival than other treatment groups (Table 2). Infected plants grown in low-fer- tility, low-moisture conditions performed significantly less well than all other lower fertility treatment groups (Fig. 3). Infected plants in the low-fertility treatment were smaller (Fig. 2) and also had the lowest survival (Table 2) of all the treat- ment groups. This suggests that infection was a disadvantage for stressed plants and that A. hjyoxylon may be an important source of mortality in times of severe stress.

In the common garden experiment, infected and uninfected plants did not differ significarltly in total biomass (P = 0.310). Because reproductive stalks of infected plants were aborted by



Soil moisture (g H20/g dry soil)

Fig. 1. Field distribution of Danthoniu spicata plants uninfected (open symbols) and infected (solid symbols) by Atkiizsonella hypoxylon relative to soil moisture and ammonia in three populations (L,, L,, and R). The axis break in the soil-moisture axis runs from 0.10 to 0.15 g H20/g dry soil.

the fungal sclerotium at an early stage of growth, infected DISCUSSION plants had significantly less reproductive and more vegetative biomass than uninfected plants (Fig. 4A). Infected plants had In this study, we found that the effect of A. Izjyo~ylonon significantly higher nitrogen conccntrations in their vegetative D. spicntn plants varied across environments. Infected plants tissues than uninfected plants (1.80 vs. 1.64% nitrogen, re- performed less well than uninfected plants in both low-mois- spectively, t < 0.01; Fig. 4B). Reproductive stalks of infected ture. low-fertility conditions and high-moisture, high-fertility plants (including fungal sclerotia) had almost twice the nitro- conditions. These performance differences may partially ex- gcn co~lcentration of those from uninfected plants (1.36 vs. plain the variation In the distribution of infected and unin-

0.768 nitrogen, respectively, t < 0.001; Fig. 4B). fected plants we observed within and among the three field

0 4 8 I2 16 20 25 29 3 3 37 Weeks since plantlng

Fig. 2. Growth of Danthonia spicatn plants uninfected (open symbols) and infected (solid symbols) by Atkinsonelin hjpoxylon in the high (diamonds) and low (circles) fertility treatments at high (solid lines) and low (dashed lines) moisture. iV for the high-fertility treatment varies from 27-28 at week 0 to 16-22 at week 37. h7for the low-fertility treatment varies from 26-27 at week 0 to 11-14 at week 37. Values are mean log number of leaves i 1 SF,; P values indicate significant effects from a multivariate repeated-measures ANOVA. I = infection class.

0 4 8 12 16 20 25 29 33 37

Weeks since planting

Fig. 3. Performance (survival X log number of leaves) of Danthonia spicata plants uninfected (open symbols) and infected (solid symbols) by ANiinsonella hjpo.xyion in the high- (diamonds) and low- (circles) fertility treatments at high (solid lines) and low (dashed lines) moisture. N for the high-fertility treatment varies from 27-28 at week 0 to 16-22 at week 37. N for the low-fertility treatment varies from 26-27 at week 0 to 11-14 at week 37. Values are mean performance 2 I SE; P values indicate significant between-subject effects from a repeated-measures ANOVA. I = infection class.

populations we surveyed. Soil resources were spatially hetero- geneous in all three sites, and in two sites (L, and L2) the spatial heterogeneity corresponded to disease incidence. These results suggest that the nature of the interaction between D. spicata and A. hjpo.xylon may be dependent on environmental conditions.

Several studies that have addressed the role of fungal epi- and endophytic fungi in altering growth patterns and resource requirements of infected host plants have also found that the effect of fungal infection on a host plant varies with environ- mental conditions (e.g., Cheplick, Clay, and Marks, 1989; Ba- con, 1993). Epi- and endophytic fungal symbionts are often advantageous to plant growth in high-nutrient environments where they may increase plant growth, but can become dis- advantageous in low-nutrient environments where increased nutrient demands of infected plants preclude increased host growth rates (e.g., Bacon, 1993; Latch, 1993). While acknowl- edging the impact of environmental conditions on host plant performance, these studies have rarely assessed the importance of heterogeneous environmental conditions in structuring plant-fungal interactions (but see Thrall and Burdon, 1997). The uniqueness of our study is that we linked growth differ-

TABLE2. Percentage survival of uninfected (U) and infected (I) Datz

tlzonia spicnta plants under high and low fertility and moisture

treatments in the greenhouse (N = 26-28 plants per treatment).

Moisture    Infcction    High        Low
High    U    79        48
I    56        46
Low    u    59        52
I    59        39
ences in the greenhouse and common garden to the distribution of fungal infection in natural field populations.

Studies in Indiana and North Carolina have found that in- fected D. spicnta plants consistently grew better than unin- fected plants in both field plantings and in the greenhouse (Clay, 1984; Kelley and Clay, 1987; Leuchtmann and Clay, 1988). In contrast, we found that although infected plants were common in areas of the field with high ammonia supply, areas with high soil moisture and low ammonia supply had very few infected vlants. Across all three field sites. occurrence of

A. hypoxylon was unrelated to D. spicnta density, suggesting that the distribution of infection in these sites cannot be ex- plained by increased potential for infectious spread in fertile areas as a result of higher plant density. In two of three field populations, areas where D. spicata was infected by A. hypoxylon had lower soil moisture and higher ammonia than ar- eas with only uninfected plants. This result suggests that in these Michigan sites, soil resource heterogeneity may influence the distribution of A. hjpoxylon infected D. spicata plants. It also implies that the impact of A. hypoxylon on D. spicata depends on environmental conditions at the microsite where the vlant is located.

One possible explanation for differences between our results and those in Indiana and North Carolina is that all three field sites we surveved have low-fertilitv soils and the substrate we used to grow plants in the greenhouse (silica sand) was ex- tremely nutrient poor compared to the substrates used in other experiments (standard greenhouse soil; K. Clay, personal com- munication). It also suggests that the consistent performance advantage of infected over uninfected plants observed in pre- vious studies (Clay, 1984; Kelley and Clay, 1987; Leuchtinann and Clay, 1988) may, in part, be due to the higher fertility of the inedium used in their greenhouse studies.

garden experiment. Although infected reproductive stalks and vegetative parts both had higher nitrogen concentration than 3icorresponding parts of uninfected plants, the nitrogen differ-


Vegetative Reproductive

Plartt material sampled

Fig. 4. Plant biomass production and leaf nitrogen concentration in veg- etative and reproductive tissue for Danthonin spicatn plants uninfected (open bars) and infected (black bars) by Atk~~isoi~elln

11jpo.xjloti grown in a common garden. Tissue sampled from infected plants included fungal hyphae on veg- etative surfaces and the fungal sclerotiurn on reproductive tissue. All values are means 5 1 SE; N = 3 genotypes for infected plants and uninfected repro- ductive tissue; N = 10 genotypes for uninfected biomass and vegetative tissue nitrogen measurement.

I11 our three field populations, D. rpicata occurred primarily in areas of the field where soil moisture generally was 5-12% and was rarely found in sites where soil molsture was >20%. Moister areas of these fields were dominated by Polytricurn sp., Rubu~sp., and other grasses, ecpecially Pocr compressa and P. pmtcnsi~ (McCormick, unpublished data), which may hahe excluded D. spicnta (Kelley and Clay, 1987). Soil mois- ture in our low-frequency watering treatments was maintained at -6%, while in the high frequency watering treatments soil moisture was -25%. Consequently, responses of D. spicntn to our low-moisture greenhouse treatments are most relevant to distribution of infected and uninfected plants in our three field populations. In the low-moisture, low-fertility treatment we found that infected plants performed less well than unin- fected plants, but in the low-moisture, high-fertility treatment infected and uninfected plants performed equally well.

A possible reason for the poor performance of infected plants in the low-fertility, low-moisture treatment ic suggested by the differences in tissue nitxogen concentration we mea- sured in infected and uninfected plants grown in the cornnlon ence for reproductive stalks was substantially greater than for vegetative tissue. Infected reproductive stalks included fungal sclerotia, so they had a substantially higher proportion of fun- gal tissue relative to plant tissue than vegetative parts did. Hence, the higher tissue nitrogen in infected plants was most likely a result of high nitrogen concentrations in fungal tissue, rather than increased nitrogen in plant tissues. Higher nitrogen concentration in fungal biomass may put a high nitrogen acquisition demand on infected plants. An increased nitrogen denland by A. 1zj~po.xylorz may have caused infected plants to grow less well than uninfected plants in the low-fertility, low- moisture treatment in our greenhouse experiment and could limit infected plants to field locations with higher relative am- monia supply. However, other studies have shown that D. spicntn is a poor competitor for light with other grass species (Kelley and Clay, 1987). suggesting that con~petition with plant species that are better light co~npetitors might exclude

D. .spicat~~

from more fertile areas of a field. Thus, light com- petition together with the high nitrogen demand of infection may limit infected D. spicntn to high nitrogen, low-moisture patches in the field.

Infected D. spicutct might also be especially sensitive to ni- trogen availability if infected individuals are less infected by mycorrhizal fungi than uninfected plants, as has been shown for endophyte-infected fescue by Chu-chou et al. (1992). Dnn- thonia spicatn are considered to be obligately lnycorrhizal (Darbyshire and Cayouette, 1989) and decreased access to my- corrhizal nitrogen and other nutrients could also limit D. spicata infected by A. hypo.xyLon to higher fertility locations than occupied by uninfected plants. It is also possible that differ- ences in root structure could allow infected plants greater ac- cess to soil nitrogen, as has been shown for some endophyte- infected grasses (e.g., Richardson et al., 1999). However. this effect has not been shown for epiphyte-infected plants and preliminary studies indicate that neither lnycorrhizal coloni- zation nor root structure differs substantially between infected and uninfected D. spicata plants (McCormick. unpublished data).

Infected D. spicnfn plants are common and often dominate low-moisture, high-ammonia areas of fields in southern Mich- igan. However, in the greenhouse, infected plants did not have a clear performance advantage under these, or any, conditions. A high incidence of infection in some areas of the field, de- spite 110 indication of advantage in the greenhouse or common garden, suggests that other factors may be influencing the field distribution of A. 1~yl~o~x)~lon-infected

D. spicntn in these sites. Competition (intra- and interspecific) and herbivory were not included in our greenhouse study, but could influence the dis- tribution of infected plants in the field. Kelley and Clay (1987) found that infected D. spicnfn were better interspecific com- petitors than uninfected D. spicata. Infected plants in our com- mon garden study did not differ from uninfected plants in total biomass, but they did have greater vegetative biomass. Weed- ing and spacing of plants in the common garden precluded lnost competition, but greater allocation to vegetative components (albeit at the expense of seed production) could in- crease the competitive ability of infected plants in field envi- ronments where interaction with other plants is common. Fur-

the^; such increased allocation to vegetati~~e components may


explain the increased competitive ability found by Kelley and taceae) on the reproducti\re system and demography of the grass Danthonia spicara. A1ei.v Phytologist 98: 165-175.

Clay (1987).

Several studies have shown that infected plants are less sus-


1990a. Fungal endophytes of grasses. Annual Review of Ecology

arld Systenzaric.~ 21: 275-297.

ceptible to herbivory (e.g., Bacon et al., 1986; Read and

1990b. The impact of parasitic and mutualistic fungi on competitive

Camp, 1986; Clay, Marks, and Cheplick, 1993). We saw little

interactions among plants. Iiz J. B. Grace and D. Tilman [eds.], Perspec-

evidence of herbivory in our common garden study even though herbivores (e.g., rabbits, deer. mice) were common in this field and were not excluded from the plots. Increased com- petitive ability or herbivory resistance could convey a perfor- mance advantage to infected plants in the field. Thus, biotic components of the field environment may convert the D. spicata-A. hjpo.xjllon relationship from parasitic or mutualistic (as in our greenhouse study) to beneficial in some environ- ments.

If a symbiotic relationship can change from parasitic to ben- eficial as a function of environmental conditions, as seems to be the case for the D. sl~icntn-A. Izjp~rxylonsymbiosis and has been shown for many mycorrhizal associations (Johnson, Gra- ham, and Smith. 1997). then environmental conditions can in- fluence the distribution of infected and uninfected plants in a population. Depending on the range of conditions present in a site and the extent of spatial structuring, infected and unin- fected plants could be either interspersed (as in Population R) or distributed in relatively discrete patches (as in Populations L, and L,). If infected plants have a higher nitrogen demand than uninfected plants, as our observation of tissue nitrogen concentration suggests, then an increased demand for nitrogen to support fungal tissue may cause the exclusion of infected plants from dry, low-fertility field locations.


BACON, C. W. 1993. Abiotic stress tolerances (moisture, nutrients) and pho-

tosynthesis in endophyte-infected tall fescue. Agriculture, Ecosystenzs

and Environment 44: 123-141. ------, P. C. LYONS, J. K. PORTER,AND J. D. ROBBINS. 1986. Ergot toxicity

from endophyte-infected grasses: a review. Agrononzy Joririzul 78: 106-


BELESKY,D. P.: 0. J. DEVISE, J. E. PALLAS, JR., AND W. C. STRINGER. 1987. Photosynthetic activity of tall fescue as influenced by a fungal endophyte. Photosynthetica 21: 82-87.

BRADSHAW, 1959. Population differentiation in Agrostis renuis Sibth.

A. D.

11. The incidence and significance of infection by Epichloe t)plzina. New Phytologist 58: 310-3 15. CARROLL,G. C. 1988. Fungal endophytes in stems and leaves: from latent pathogen to mutualistic symbiont. Ecology 69: 2-9.

CHEPLICK,G. P. 1997. Effects of endophytic fungi on the phenotypic plas- ticity of Loliunz perenne (Poaceae). Ainerican Jo~irnal of Botany 84: 34-

40. -, K. CLAY,AND S. MARKS. 1989. Interactions between infection by

endophytic fungi and nutrient limitation in the grasses Lolizim pei-enne

and Festirca arunclirzacea. .Me>t, Pliytologist 11 1: 89-97.

CHU-CEIOU,M., B. GUO, Z.-Q. AN, J. \V. HENDRIX,R. S. FEKRISS, M. R. SIECTEL, AND P. B. BURRUS. 1992. Suppression of

C. T. DOUGHERTY, mycorrhizal fungi in fescue by the Acrer~~oniumcoenophiczlurn endophyte. Soil Biology and Riocl~enzistiy 24: 633-637.

CLAY,K. 1982. Environmental and genetic determinants of cleistogamy in a natural population of the grass Danthor7ia spicata. Evolufioiz 36: 734-

741. . 1984. The effect of the fungus Atkin,sonella hy,r~oxylor~ (Clavicipi

tives on plant competition, 391-412. Academic Press, San Diego, Cali- fornia. USA. . 1994. Hereditary symbiosis In the grass genus Danthoizia. A'ew Phy- tologist 126: 223-23 1. -----, AND J. HOLAH. 1999. Fungal endophyte symbiosis and plant diver- sity in successional fields. Science 285: 1742-1744.

-, S. MARKS, AND G. P. CHEPLICK.1993. Effects Of insect herbivory and fungal endophyte infection on competitive interactions among grass- es. Ecology 74: 1767-1777.

DARBYSHIRE, J. CAYOUETIE. 1989. The biology of Canadian

S. J., ASD weeds. 92. Danthonia spicata (L) Beauv. in Roem. and Schult. Canaclian Joz~rizal of Plarlt Science 69: 12 17-1233.

DIEHL,W. hr.1950. Ualansia and the Balansiae in America. U.S. Department of Agriculture, Agricrilt~rml Moiujgraphs 4: 1-82. D~BSON,

A,, AND M. CRAWLEY. 1997. Pathogens and the structure of plant communities. Treizds in Ecology and Evolirtion 9: 393-397. GROSS,K. L., K. S. PREGITZER.

AND A. J. BURTON. 1995. Spatial variation in nitrogen availability in three successional plant communities. Jozlrnal of Ecologv 83: 357-367.

JACKSON,R. B., AND h/I. M. CALDWELL. 1993. The scale of nutrient hetero- geneity around individual plants and its quantification with geostatistics. Ecology 74: 61 2-6 14.

JOHNSON,N. C., J. H. GRAHAM,AND E A. SMITH. 1997. Functioning of mycorrhizal associations along the mutualism-parasitism continuum. h'etv Phytologist 135: 575-585.

KELLEY. S. E., AND K. CLAY. 1987. Interspecific competitive interactions and the maintenance of genotypic variation within the populations of two perennial grasses. Evol~ition41: 92- 103.

KOVER, P. X.,D. E. Trro~As,AND K. CLAY. 1997. Potential versus actual contribution of vertical transmission to pathogen fitness. Proceedings of tlze Royal Society uf Loiidoiz 264: 903-909.

LATCH, G. C. M. 1993. Physiological interactions of endophytic fungi and their hosts: biotic stress tolerance impasted to grasses by endophytes. Agriculrz~re. Ecosysrems and Enliroiiment 44: 143-156.

LEUCHTMANN,A,, AND K. CLAY. 1988. Experimental infection of host grass- es and sedges with Atkinsonella hypoxylon and Balansitr cyperi (Balansiae, Clavicipitaceae). Mycologia 80: 291-297. , AND -. 1989. Isozyme variation in the fungus Atkiizsonella i~ypoxylonwithin and among populations of its host grasses. Carladian Journal of Botany 67: 2600-2607.

MARKS, S., AND K. CLAY. 1990. Effects of CO, enrichment, nutrient addi- tion, and fungal endophyte-infection on the growth of two grasses. Oecologia 84: 207-214.

MCCORMICK, M. K. 1999. Spatial environmental variation: within and among population effects in Darzthoizia .spicara. Ph.D. dissertation, Mich- igan State University, East Lansing, Michigan, USA.

READ. J. C., AND B. J. CAMP. 1986. The effect of fungal endophyte Acremonium coenophinlriin in tall fescue on animal performance, toxicity, and stand maintenance. Agronorny Journal 78: 848-850.

RICIIARDSON,M. D., R. I. CABRERA,J. A. MURPHY, AND D. E. ZAUROV. 1999. Nitrogen-form and endophyte-infection effects on growth, nitro- gen uptake, and alkaloid content of Chewings fescue turfgrass. Jo~irizal of Plant Nutrition 22: 67-79.

ROBERTSON,G. P., AND K. L. GROSS. 1994. Assessing the heterogeneity of belowground resources: quantifying pattern and scale. In M. M. Caldwell and R. W. Pearcy [eds.], Exploitation of environmental heterogeneity by plants, 237-253. Academic Press, San Diego, California, USA. SCIIE~INER,

S. M. 1989. Variable selection along a successional gradient. Evolution 43: 548-562.

THRALL.F! H., AND J. J. BG'RDON. 1997. Host-pathogen dynamics in a nie- tapopulation context: the ecological and evolutionary consequences of being spatial. .lo~rrt~alof Ecology 85: 743-753.

  • Recommend Us