Effects of mycorrhizae, phosphorus availability, and

Effects of mycorrhizae, phosphorus availability, and plant density on yield relationships among competing tallgrass prairie grasses1 B.A.D. HE TRICK^ Department of Plant Pathology, Kansas State University, Manhattan, KS 66506, U.S.A.

D.C. HARTNETT Division of Biology, Kansas State University, Manhattan, KS 66506, U.S.A.

G.W.T. WILSON Department of Plant Pathology, Kansas State University, Manhattan, KS 66506, U.S.A.

Can. J. Bot. Downloaded from www.nrcresearchpress.com by 74.121.191.133 on 06/06/13 For personal use only.

AND

D.J. GIBSON Department of Plant Biology, Southern Illirlois University, Carbondale, IL 62901, U.S.A. Received April 16, 1993 HETRICK,B.A.D., HARTNETT, D.C., WILSON,G.W.T., and GrssoN, D.J. 1994. Effects of mycorrhizae, phosphorus availability, and plant density on yield relationships among competing tallgrass prairie grasses. Can. J. Bot. 72: 168- 176. A replacement series experiment was used to investigate the effects of mycorrhizae, phosphorus availability, and plant density on competitive relationships between three tallgrass prairie species of varying mycorrhizal dependencies. Under mycorrhizal conditions, the obligately mycorrhizal dependent warm-season grass Andropogon gerardii (big bluestem) was a better competitor in mixture with the nonmycorrhiza-dependent cool-season grass Koeleria pyrarnidata (Junegrass). In the absence of mycorrhizae, however, competitive effects of big bluestem were greatly reduced and Junegrass experienced competitive release. Relative yield totals increased when mycorrhizae were suppressed, suggesting greater intensity of interspecific competition in the presence of mycorrhizae. Thus, the competitive dominance of big bluestem in tallgrass prairie is strongly related to its mycorrhizal status. Elymus canadensis (Canada wild rye) outcompeted big bluestem both with and without mycorrhizae. Relative yield totals of this species mixture were also lower under mycorrhizal conditions, indicating that mycorrhizae increase the intensity of interspecific competition between them. Relative yields of wild rye competing with big bluestem increased in the absence of mycorrhizae, suggesting that it also experiences competitive release when big bluestem are not mycorrhizal. The outcomes of competition were generally similar among the three total plant density treatments and between P-fertilized and nonfertilized treatments. However, interactions between mycorrhizal effects and plant density confirm that outcomes of interspecific competitive interactions may be density dependent in some cases. Key words: arbuscular mycorrhizae, de Wit replacement series, Andropogon gerardii, Elytnus canadensis, Koeleria pyrarnidata. HETRICK, B.A.D., HARTNETT, D.C., WILSON,G.W.T., et GIBSON,D.J. 1994. Effects of mycorrhizae, phosphorus availability, and plant density on yield relationships among competing tallgrass prairie grasses. Can. J. Bot. 72 : 168- 176. Les auteurs ont Ctudit les effets des mycorhizes, de la disponibilitk du phosphore et de la densitt des plantes sur les relations de compttitivitC entre trois espkces d'herbes hautes possedant divers degrks de dkpendance mycorhizienne. A cette fin, ils ont conduit une experience de remplacements series. Sous des conditions mycorhiziennes, l'herbe de saisons chaudes et obligatoirement dependante des mycorhizes, llAndropogon gerardii (barbon de GeFard), s'avkre meilleur competiteur en melange avec l'herbe de saisons froides non dependante des mycorhizes, le Koeleria pyrarnidata (koelerie ?i cretes). En absence de mycorhizes, cependant, les effets competitifs du barbon de GCrard sont grandement reduits et la koelkrie ?i cr&tes tchappe ?i la competition. Les rendements totaux relatifs augmentent lorsqu'on supprime les mycorhizes, ce qui suggkre l'existence d'une plus grande intensitt de la competition interspecifique en presence des mycorhizes. Ainsi la dominance competitive du barbon de Gtrard dans les prairies d'herbe haute est fortement relike ?i son statut mycorhizien. L'Elymus canadensis (agropyre du Canada) dtplace le barbon de GCrard en absence aussi bien qu'en presence de mycorhizes. Les rendements totaux relatifs de ce melange d'espkces sont tgalement plus faibles sous des conditions mycorhiziennes, ce qui indique que les mycorhizes augmentent l'intensite de la competition interspecifique entre elles. Les rendements relatifs de l'agropyre du Canada en compttition avec le barbon de GCrard augmentent en absence de mycorhizes, ce qui suggkre qu'il serait Cgalement libere de la competition lorsque le barbon de GCrard n'est pas mycorhize. De facon gtntrale, les issues de la competition sont semblables entre les trois traitements de densitt totale, et entre les traitements de fertilisation et de nonfertilisation au P. Cependant, les interactions entre les effets mycorhiziens et la densite des plantes confirment que les issues aux interactions de compttition intersptcifique pourraient dependre de la densite dans certains cas. Mots elks : mycorhizes arbusculaires, remplacements strits de de Wit, Andropogon gerardii, Elyrnus canadensis, Koeleria pyrarnidata. [Traduit par la redaction]

'Contribution No. 93-405-J from the Kansas Agricultural Experiment Station, Kansas State University, 2Author to whom all correspondence should be addressed. Printed in Canada i Imprim6 nu Canada

anh hat tan,^^ 66506-5502, U.S.A.

169

HETRICK ET AL.


Introduction Mycorrhizal fungi are known to facilitate plant nutrient acquisition from soil and to form hyphal networks among plants through which nutrients can flow (e.g., Chiariello et al. 1982; Whittingham and Read 1982; Miller 1988; Koide 1991). The symbiosis can also significantly affect the structure and dynamics of natural plant communities (e.g., St. John and Coleman 1983; Connell and Lowman 1989; Allen 1991; Brundrett 1991). While these effects are expressed in a wide variety of habitats (Fitter 1977; Hall 1978; Buwalda 1980; Allen and Allen 1984; Hetrick et al. 1989), the specific mechanisms and processes by which mycorrhizae influence plant community composition and species diversity remain poorly understood and controversial (Grime et al. 1987; Bergelson and Crawley 1988). Plant species vary along a continuum in their dependency on mycorrhizal fungi, from obligate through facultative mycotrophs to nonmycotrophs (Janos 1980; Hetrick et al. 1988a, 1990, 1992). Because of these differences, mycorrhizal fungi may mediate plant competition (Allen and Allen 1990), and changes in rnycorrhizal activity may alter plant competitive hierarchies. It seems likely that mycorrhizal fungi alter the ability of plant species to compete for nutrients by altering the capacity of the root system for P acquisition and uptake (Fitter 1987; Hetrick et al. 19886, 199 1; Brundrett 1991). Interplant nutrient transfer via mycorrhizae hyphal links between the root systems of neighboring plants may be another mechanism by which mycorrhizal associations influence plant competition. Competition is an important process influencing plant growth dynamics, demography, and community structure in grasslands (Parrish and Bazzaz 1979; Berendse 1981; Fowler 1986; Tilman 1986; Turkington and Mehrhoff 1990). In a previous greenhouse experiment, Hetrick et al. (1989) showed that the degree of mycorrhizal dependence of host plants affected interspecific competitive relationships of Andropogon gerardii and Koeleria pyramidata, two tallgrass prairie grasses. This indicated a potentially important role for mycorrhizae in shaping the structure of tallgrass prairie plant communities. In a previous target plant - neighbor competition study, Hartnett et al. (1993) showed that mycorrhizae influenced individual plant competitive effects and responses (growth and tillering) of Andropogon gerardii and Elymus canadensis. The study reported here extends this earlier work by examining competitive relationships of three species that vary along the continuum of mycorrhizal dependency. In addition, the replacement series experimental design used in this study allowed us to assess the effects of mycorrhizae on the intensities and relative importance of intra- versus inter-specific competition in mixed-composition stands of grasses gnd to determine whether outcomes of competition among these species were density dependent. In addition, analyses of relative yields of species in a replacement series can provide useful predictions of changes in relative abundances of competing species in the field and patterns of coexistence. The objective of this study was to assess the influence of mycorrhiial fungi on interspecific competition in three tallgrass prairie species that vary in mycorrhizal dependency. Using a de Wit replacement series (de Wit 1960; de Witt and Van Der Bergh 1965) we examined competitive interactions of species pairs with and without mycorrhizae and with and without additional soil phosphorus. To determine whether outcomes

of interspecific competition were density dependent, we examined pairwise species competition over a range of total plant densities. We hypothesized that the effect of mycorrhizae on interspecific competition would be related to the mycorrhizal dependency of the competing species, such that the competitive dominance of mycorrhiza-dependent warm-season grasses would be diminished when mycorrhizal activity was reduced. We also predicted that if enhanced soil P acquisition was the primary benefit of mycorrhizal associations in tallgrass prairie, the addition of P to the soil should strongly diminish the benefits of mycorrhizae in obligate and facultative mycotrophs, thus further altering outcomes of competition.

Materials and methods Seeds of big bluestem ( A . gerurdii) were obtained from the Soil conservation Service Plant Materials Center, Manhattan, Kans. Seeds of Junegrass (K. pyramidata) were provided by the Upper Colorado Environmental Plant Center, Meeker, Colo., and Canada wild rye (E. canadensis) seeds were supplied by CRM Ecosystems, Inc., Prairie Ridge Nursery, Mt. Horeb, Wis. Following germination in vermiculite, 2-week-old seedlings of each species were transplanted into 2 1.5 x 21.5 cm pots containing 5.22 kg (dry weight) of nonsterile prairie soil freshly collected from the Konza Prairie Research Natural Area in northeast Kansas (Chase silty clay loam: fine, montmorillonitic, mesic Aquic Arguidoll) containing 6.0 pglg plant available phosphorus (Bray test 1). Based on taxonomic criteria of Schenck and Perez (1 988), this soil contained approximately 136, 109, 30, 26, 12, 6, and 3 spores of Glonzus etunicaturn Becker & Gerd., Glomus aggregatum Schenck & Smith, Glomus constrictum Trappe, Sclerocystis coremioides Berk. & Broome, Enrropospora irzfrequens (Hall) Ames & Schneider, Glomus mosseae (Nicol. & Gerd.) Gerdemann & Trappe, and Gigaspora gigantm (Nicol. & Gerd.) Gerdemann & Trappe, respectively, in 100 g (dry weight) soil. Any hyphal networks existing in the soil were probably disturbed during soil collection. A replacement series design (de Wit 1960) was used for the competition experiment in which big bluestem was grown in mixture with either Junegrass or wild rye in big bluestem to competitor planting ratios of 1.0, 0.66:0.33, 0.33:0.66, and 0: 1. These different species ratios were each established at three different total plant densities (3, 6, and 12 plants per pot). In addition, nonmycorrhizal and amended phosphorus treatments were established for each species pair, ratio, and density combination. The nonmycorrhizal treatment was established by applying 100 mL of a 50 pg . g-' (active ingredient) Benomyl solution (Benlate SODF; E. I. duPont de Nemours & Co., Wilmington, Del.) to the soil. While Benomyl is not specific to mycorrhizal fungi and will inhibit growth of some other soil fungi, it provides a more natural environment for comparison of mycorrhizal responses than soil from which all microbes have been removed by steaming. Although Benomyl application can result in release from pathogens or other mutualists, the former is distinguishable from a mycorrhizal response because benomyl-treated plants would be larger. In contrast, Benomyl inhibition of mycorrhizal symbiosis usually results in reduced 'plant vigor or biomass production. Seedlings grown in nonsterile prairie soil not amended with Benomyl solution became mycorrhizal due to the native mycorrhizal spores in nonsterile prairie soil. The amended phosphorus treatment was obtained by applying 90 pglg P to the soil surface in 100-mL aliquots of a KH2P0, solution. Thus, the overall experiment involved two mycorrhizal treatments (nonmycorrhizal and mycorrhizal) x two P treatments (amended and nonamended) x three total plant densities (3, 6, and 12 plants per pot) x four ratios of each of the two species pairs, with four replicate potsper treatment. The P level used in this experiment is higher than in other studies (e.g., Hetrick et al. 19886, 1990, 1992, etc.) because the pot size was


170

CAN. J . BOT. VOL. 72, 1994

larger in the present study and we wished to ensure that P would reduce mycorrhizal colonization and would not be limiting at any time during the experiment. Several recent studies critically evaluated de Wit type replacement series designs for competition experiments (e.g., Jolliffe et al. 1984; Connolly 1986; Snaydon 1991; Grace et al. 1992). A replacement series design was used in this study because it offers the opportunity to provide a direct measure of yield responses of competing species in mixture under different conditions (Firbank and Watkinson 1990). One of the most persistent criticisms of this method, that results may vary depending upon the total plant density used (Firbank and Watkinson 1985), was circumvented in this study by using three different total pot densities that encompassed field densities of the component species. Pots containing the seedlings were arranged in a randomized complete block design in a greenhouse maintained at 20-25°C. This temperature was selected because it allows growth of both warm- and cool-season grasses (Hetrick et al. 1 9 8 8 ~ ) .Plants were harvested after 15 weeks, and their roots, shoots, and seed heads were ovendried and weighed. The dried roots were subsampled, stained in trypan blue (Phillips and Hayman 1970), and examined microscopically to assess percent root area colonized using a Petri plate scored in 1-mm squares (Daniels et al. 1981). Relative yields (RY) and relative yield totals (RYT) were calculated for the pairs of species in each mixture. RYT was calculated as

[I] RYT = (Y'J/Yii) + (YJi/Y*) where Yii and Yjj are the yields (biomass) per unit area of species i and j in monoculture and and Yji are the yields of species i in mixture with j and vice versa. YGIYii and YjilP are the relative yields of species i and j , respectively (Snaydon 1991; de Wit and Van Der Bergh 1965). Of the many indices suggested for the analysis of replacement series, RYT is the most stable and independent of the replacement line (Connolly 1986). The RY curves (relative yield of a species as a function of its proportional initial contribution to the mixture relative to monoculture) and RYT provide an indicator of the competitive relationships between species. RY curves can be compared with the linear null model of no significant competitive interactions in which the relative yield of each species is directly proportional to its original relative contribution to the sown mixture. RYT values > 1.0 indicate divergence in resource use between species and weaker interspecific than intraspecific effects, whereas RYT < 1.0 indicates relatively stronger interspecific interference (Harper 1977). Yield totals were initially analysed using ANOVA to assess the significance of total plant density, soil mycorrhizae, and soil phosphorus as treatments, followed by Fisher's least significant difference test on main effect means. Total plant density, rather than the number of either one of the two competing species, was included as a treatment in the ANOVA to reflect the intensity of competition because it is a better representation of stand yield relationships and because there were too few degrees of freedom in the analysis to include the numbers of the two competing species as additional treatments. In this and all other analyses, data from plants that had been assigned to the nonmycorrhizal treatment but had < 4 % root area colonized were discarded from the analysis. However, no treatment contained fewer than three replications. ANOVA of total, shoot, root, and seed head dry weights, and root to shoot ratio gave the same results with respect to the direction and magnitude of differences among treatments (Table 1). Thus, only the results of the further analyses of total dry weights are reported here. A jackknifing procedure (Sokal and Rohlf 1981) was used to reduce the bias in the estimates of RY and RYT. Because RY and RYT have a lower bound of zero and an unknown upper bound, the distribution of these statistics is unknown. Previous approaches to assessing the statistical significance of RY and RYT have either not considered these distributional problems (e.g., Trenbath 1975) or have used a less powerful nonparametric approach (Adee et al. 1990). In con-

TABLE1. Results of ANOVA on the yield components (g dry wt.) of prairie grasses in pairwise competition Treatment"

Total mass

Shoot mass

Root mass

Reproductive mass

Rootto shoot ratio

Big bluestem with Junegrass D M P DxM DxP M XP DxMxP Big bluestem with wild rye D M P DxM DxP MXP DXMXP D M P DxM DXP MXP DxMXP

Junegrass with big bluestem ** ** ns ns ns ns ns ns ns ns ns

**

Wild rye with big bluestem D M P DxM DxP MXP DxMxP NOTE: Dry weight data and root lo shoot ratios were log and arcsine transformed, respectively. *, statistically significant at the p 5 0.05 level; **, statistically different at the p 5 0.0 1 level. "D, total plant density; _M, mycorrhizae; P, phosphorus.

trast, for various applications in ecological studies, the jackknife procedure provides a normally distributed estimate of the mean and standard error of observed variates that are themselves not normally distributed (e.g., Meyer et al. 1986; Briggs and Knapp 1991). Jackknife estimates of the means for these data were calculated by sampling all combinations of three of the four replications per experimental treatment. Two standard errors around the jackknifeestimated means for RYT provided 95% confidence limits that were then used to assess whether a given RYT was significantly different from 1.0; the null value gave no significant interspecific interaction (or equal intensities of inter- and intra-specific effects). Similarly, the null RY for each species corresponds to its original proportional contribution to the seedling mixture established in the pot. ANOVA on jackknifed RYTs of big bluestem - Junegrass and big bluestem wild rye mixtures investigated the effects of total plant density, mycorrhizal status, and additional soil P treatments.

Results and discussion Root colonization responses Benomyl-treated plants of all species were less than 5 % colonized at harvest and in no case differed from zero (P =

HETRICK ET AL. -M.P A

D

.M+P +M.P +M+P

CD

D

CD

A 0 A-C CD A A.C A-C

D

D

BC A

AB B

A0 0.D

BC C D

AB

B

0.D

CD

C

C

BC AB A BC A

BC

.M-P .M+P A +M.P 0 +M+P

C

0

A0 BC

0

BC

EF D.F F D-F

C.E 8.E 0.E DE 8.E A E C.E BC E DE B

1

2

3

B A B A0

C C C C

BC A BC A C A A 0 AB

NUMBER O F WILD RYE

NUMBER OF BIG BLUESTEM 0

C.E BC D.E D.F

0

2

4

6

0 4 8 1 2

6

4

2

0

1 2 8 4

0

6

4

2

0

12 8


loo

0.1

3

2

1

0

0

NUMBER OF BIG BLUESTEM

NUMBER OF JUNEGRASS

o -MP 0 -M*P A + M P 0 +M+P

CD B.D A B

CD B - D C D BC 0-D D 0-D 0-D

B C A-C A BC A-C BC A.C B C

A-C

A-D A 0

CD

C BC BC

0 - D A-C D A D D AB CD 0 - D

+

-M-P -M*P ++MP +M+P

D D CD

D D D

D B A

C C BC

C C BC

A0

C

CD 0 - D B

BC

BC

BC

AB

BC

A

CD D D CD CD

D A B

FIG. 1. The influence of plant density (total number of plants per pot), mycorrhizal symbiosis ( + M or -M), and P-fertilization treatments ( + P or -P) on individual plant biomass of big bluestem when grown alone or in combination with Junegrass. To simplify the figures, significant differences between treatments are indicated for big bluestem above the figure and for Junegrass below the figure. For each plant species, within each overall density (3 plants, 6 plants, or 12 plants per pot), treatments with the same letters are not significantly different at the p 5 0.05 level. Note that individual plant biomass is shown on a logarithmic scale.

FIG.2. The influence of plant density per pot, mycorrhizal symbiosis ( + M or -M), and P-fertilization treatments (+P or -P) on individual plant biomass of wild rye when grown alone or in combination with big bluestem. To simplify the figures, significant differences between treatments are indicated for wild rye above the figure and for big bluestem below the figure. For each plant species, within each overall density (3 plants, 6 plants, or 12 plants per pot), treatments with the same letters are not significantly different at the p r 0.05 level. Note that individual plant biomass is shown on a logarithmic scale.

0.05). There were no effects of plant density or mixture on root colonization by mycorrhizal fungi. Percent mycorrhizal root colonization of big bluestem grown with Junegrass ranged from 15 to 22 % in nonfertilized pots but declined to 6 - 12 % in P-fertilized pots. Junegrass without fertilizer ranged from 8 to 15% colonization and was reduced in the presence of P to 2-7%. Colonization of big bluestem paired with wild rye ranged from 10 to 20% in nonfertilized pots but declined to 5 - 12 % in fertilized pots. Wild rye colonization ranged from 10 to 17% in nonfertilized and 4-7% in fertilized pots.

increasing total density and for big bluestem, with increasing number of interspecific competitors (Figs. 1 and 2). In the big bluestem - wild rye mixture, individual plant biomass of wild rye exceeded that of big bluestem, and wild rye biomass generally increased as the number of conspecific competitors decreased and the number of big bluestem increased (Fig. 2). These responses will not be discussed in further detail, however, because the focm of this study and the following discussion is the assessment of mycorrhizal effects on competitive relationships through interpretation of relative yields.

Dry weight yield responses

Relative yields-big bluestem with Junegrass RY for each species in the big bluestem - Junegrass mixture and RYT differed in most cases from null values expected if no significant competition occurred (Fig. 3). Without P fertilization at all densities, the RY of Junegrass exceeded the expected values when mycorrhizae were absent, and the presence of the symbiosis reduced its RY. Conversely, the RY of big bluestem exceeded the null values if mycorrhizae were present and exceeded the RY obtained under nonmycorrhizal conditions (Fig. 3). These data indicate that under mycorrhizal conditions, the clear winner was big bluestem, the species with greater mycorrhizal dependence. Since Junegrass was generally not competitively suppressed by increasing numbers of big bluestem except when plants were mycorrhizal, big bluestem's competitive dominance in tallgrass prairie appears to be strongly related to its mycorrhizal status, and the cool-season nonmycotrophic species experiences competitive release under nonmycorrhizal conditions.

Total plant density, mycorrhizal status, P amendment, and the interactions of these factors significantly influenced the dry weight yields in the big bluestem - Junegrass mixture (Table I). In the big bluestem - wild rye mixture, these factors also significantly affected yields, but there were no significant factor interactions (Table 1). Both mycorrhizal status and P amendment had a significant effect on the yield of big bluestem and wild rye, regardless of their competitor species (Table l), but they had no significant effect on yields of Junegrass that is very weakly to nonmycotrophic (Hetrick et al. 1988a, 1990). Total plant density in the mixture strongly influenced yields of Junegrass and wild rye but not big bluestem (Table 1). Only big bluestem competing with Junegrass showed a significant interaction between total plant density and response to mycorrhizae or between P amendment treatment and response to mycorrhizae. In general, individual plant biomass of both species in the big bluestem - Junegrass mixture decreased with

172

CAN. J. BOT. VOL. 72, 1994

Low density / Added phosphorus


Low density / No phosphorus

3:O

2:l 1:2 Big bluestern : Junegrass

0:3

12:O

4:2

2:4

2: 1 1:2 Big bluestem : Junegrass

0:6

Big bluestern : Junegrass

Big bluestem : Junegrass

High density / No phosphorus

High density / Added phosphorus

8:4

4:8

0:3

Medium density / Added phosphorus

Medium density / N o phosphorus

0 6:O

3:O

0:12

Big bluestern : Junegrass

12:O

8:4 4:8 Big blueslern : Wild Rye

0:12

FIG.3. Relative yield totals ( A ) and relative yields for big bluestem (0) grown in various proportions with Junegrass (0) at low, medium, or high overall plant densities (a) without added phosphorus or (b) with additional phosphorus. Relative yields of species and relative yield totals of mycorrhizal treatments are denoted by solid lines, while relative yields and relative yield totals of nonmycorrhizal treatments are denoted by broken lines. The light dotted lines denote the null model expectations (predicted relationships if components of the mixture showed no competitive interactions). All values are significantly different from the null model at the p I0.05 level unless indicated by an asterisk.

The RYT values varied depending on the particular ratio of the two species and showed few consistent trends. Without mycorrhizae, when RYT values differed from null predictions, they were consistently greater than 1.0, indicating reduced interspecific interference in the absence of mycorrhizal fungi (Fig. 3). This further indicates that removal of mycorrhizae removes the strong interspecific competitive effects of big bluestem. When mycorrhizal and fertilized with P, the RY of Junegrass was not as strongly reduced as it had been without added

P at low or medium plant density, but this beneficial effect of P was not evident a t the highest total plant density (Fig. 3). While the competitive success of Junegrass was improved in the absence of mycorrhizae and P fertilizer at all densities, when fertilized with P the RY of Junegrass were less consistent and were increased in the absence of mycorrhizae only at high total plant densities (Fig. 3). Thus, P fertilization appears to mitigate the strong effect of mycorrhizae on competitive superiority of big bluestem over Junegrass and to reduce the competitive advantage that Junegrass had under nonmycoru


HETRICK ET AL.

rhizal conditions. The latter is probably an indirect result of improved growth and RY of big bluestem following P fertilization. Mycorrhizae had weaker and less consistent effects on RY of fertilized big bluestem. Addition of P did not substantially alter the patterns observed for RYT of mycorrhizal plants (Fig. 3). In general, in both nonfertilized- and P-fertilized treatments, RYT values for the mixture were greater when mycorrhizal fungi were suppressed, indicating that suppression of mycorrhizae significantly reduced the intensity of interspecific competition between these two grasses and reduced the strong asymmetry of competition (dominance by big bluestem). This strong competitive suppression of Junegrass by big bluestem conferred by mycorrhizal symbiosis was evident at all densities assessed in the present studies. This leads to the general hypothesis that coexistence of highly mycorrhizadependent and nondependent species with similar phenologies would only occur in the field if environmental or edaphic conditions limited mycorrhizal symbiosis. Given that mycorrhizal fungi are abundant in and indigenous to tallgrass prairie (Hetrick and Bloom 1983), the phenologic differences between these species may, in fact, be a critical prerequisite for their continued coexistence. Inhibition of mycorrhizal activity during the cool season may limit the competitive advantage of big bluestem such that Junegrass would grow unsuppressed. Alternatively, under cooler conditions the mycorrhizal fungi may still be active, and the competitive advantage may then be conferred upon the cool-season plant species. This latter hypothesis is supported by Bentivenga and Hetrick (1992) and B.A.D. Hetrick and G.W.T. Wilson (unpublished data) who demonstrated greater mycorrhizal activity and more 32P uptake in cool-season grasses than in warm-season grasses at low temperature. Relative yields-big bluestem with wild rye In all cases, the RY of wild rye exceeded null model values and exceeded big bluestem RY values whether or not mycorrhizae were present, indicating its general competitive advantage over big bluestem (Fig. 4). At medium and high plant densities, RY of nonmycorrhizal wild rye plants significantly exceeded that of mycorrhizal plants. Apparently, at higher total densities, mycorrhizae reduced the relative competitive success of wild rye plants against big bluestem. The RY of big bluestem when mycorrhizal was lower than null model expectations at all densities. When nonmycorrhizal, the RY curves of big bluestem were generally above the null model expectations (Fig. 4). Apparently, while mycorrhizae did not substantially improve competitiveness of wild rye, mycorrhizae generally reduced the competitiveness of big bluestem grown with wild rye. At the highest density, the presence of mycorrhizae reduced RY of both species. The general RYT and RY patterns indicate that the presence or absence of mycorrhizae altered the relative intensity of intraand inter-specific competitive effects but did not appreciably change the predicted outcome of competition (winner versus loser) in the big bluestem - wild rye mixture. The RYT of the nonmycorrhizal mixture was consistently much higher than 1.0 and higher than the mycorrhizal mixture at all plant densities, indicating that without mycorrhizae these two species experience significantly reduced interspecific competitive effects. In the presence of P fertilizer, as in the nonfertilized treatments, the RY of wild rye was consistently well above the null

173

model values whether the plants were mycorrhizal or not. P amendment resulted in even greater wild rye RY than had been observed without it. For big bluestem, fertilized plants had RY below the null model expectations at all densities whether or not mycorrhizae were present. At medium and high densities the relative yields of mycorrhizal big bluestem plants exceeded nonmycorrhizal. The addition of P appears to heighten competitive asymmetry between wild rye and big bluestem such that wild rye is an even stronger winner. Although RYT values were significantly higher for nonmycorrhizal as opposed to mycorrhizal plants in the nonfertilized treatment, in the Ertilized treatment RYT values were much closer to the null value of 1.0 and at medium and high densities RYTs were generally greater in the mycorrhizal treatments (Fig. 4), suggesting that P fertilization increases the relative intensity of interspecific competition. The opposite was evident for mycorrhizal plants. RYT values of nonfertilized plants were consistently close to the null value of 1.O, while fertilized plants displayed significantly higher RYT values than expected at the two higher densities. Apparently for mycorrhizal plants, P fertilization increases the intensity of interspecific competition between big bluestem and wild rye and the relative advantage of wild rye. Since mycorrhizal symbiosis confers such a strong competitive advantage on big bluestem when it is paired with Junegrass, it was surprising that the opposite relationship was observed when big bluestem was paired with wild rye. The competitive dominance of wild rye over big bluestem is obvious when either individual ~ l a n biomass t of RY are examined. There was no significant effect of mycorrhizae on biomass of wild rye when this species was grown intraspecifically or when paired with big bluestem. The RY of both big bluestem and wild rye and the RYT were considerably higher when plants were mycorrhizal, suggesting that the absence of the mycorrhizal symbiosis reduces the competitive intensity between these two species and reduces the interspecific interference relatively more than intraspecific effects. Removing mycorrhizae increased the competitive advantage of wild rye against big bluestem because although wild rye is a facultative mycotroph, eliminating mycorrhizae disproportionately reduced the competitive effect of-the obligate mycotroph big bluestem. Apparently, like Junegrass, wild rye experiences competitive release in the absence of mycorrhizae. Reso~rrcecompetition and coexistence It could be hypothesized that a mycorrhiza-dependent plant growing with a nondependent one coexist because they do not compete for the same pool of phosphorus. The mycorrhizal plant may have access to a P supply that is unavailable to the nonmycorrhizal species resulting in belowground niche partitioning, much like that resulting in stable grass-legume mixtures due to N-fixing symbiosis (Harper 1977). Jayachandran et al. (1992) showed that mycorrhizal plants are able to mineralize organic P sources that were not available to nonmycorrhizal plants. Since organic P compounds are the principal source of P in the native prairie soil studied here, the importance of this mechanism cannot be overlooked. However, nutrient resource partitioning among mycorrhizal and nonmycorrhizal species would be predicted to result in RYT values significantly greater than 1.0 in replacement series mixtures, and RYT values greater than 1.0 occurred only under nonmycorrhizal and nonfertilized conditions in our experiment. Berendse (1981) similarly observed RYT values greater

CAN. 1. BOT. VOL. 72, 1994

174

Low density / Added phosphorus


Low density / No phosphorus

3:O

2:l 1 :2 Big bluestern : Wild rye

0: 3

3:O

2:l

Medium density

Medium density / N o phosphorus

0 6:O

12:O

1 :2

/ Added phosphorus

Big bluestem : Wild rye

4:2 2:4 Big bluestern : Wild rye

High density / No phosphorus

High density I Added phosphorus

8:4 4:8 Big bluestem : Wild rye

0:3

Big blueslem : Wild rye

0:12

0 12:O

8:4 4:s Big bluestern : Wild Rye

0:6

0:12

FIG. 4. Relative yield totals (A) and relative yields for wild rye (0) grown in various proportions with big bluestem (0) at low, medium, or high overall plant densities (a)without added phosphorus or (b) with additional phosphorus. Relative yields and relative yield totals of mycorrhizal treatments are denoted by solid lines, while relative yields and relative yield totals of nonmycorrhizal treatments are denoted by broken lines. The light dotted lines denote the null model expectations for relative yields and relative yield totals (predicted relationships if components of the mixture showed no competitive interactions). All values are significantly different from the null model at the p 5 0.05 level unless indicated by an asterisk.

than 1.0 for species mixtures under nutrient-poor conditions. These results are in contrast with those of Turkington and Klein (1991) who found no trends in RYTs in competing species with and without soil microorganisms (including mycorrhizae). Trenbath (1975) attributed RYT values greater than 1.0 to niche partitioning through different rooting depths of competing species. The clear competitive superiority of wild rye over big bluestem observed in these experiments, in contrast with the typi-

cal dominance of big bluestem in the field, may be explained by the life-history stages of the competing plants. In the present experiment, plants were transplanted into pots soon after germination. The wild rye seedlings that did not require the mycorrhizal symbiosis grew rapidly after transplanting, while the big bluestem seedlings did not grow appreciably in the first several weeks until the mycorrhizal symbiosis was established. Thus, the wild rye had at least a short-term competitive advantage in its lack of requirement for mycorrhizae


HETRlCK ET AL.

and may have been able to preempt a large enough quantity of soil resources to win over big bluestem. An experiment of longer duration involving established plants might have revealed a long-term advantage to the mycorrhiza-dependent big bluestem. The latter would explain the dominance of big bluestem in the field. In addition, under changing temperature regimes in the field, phenological differences in the growth habits of the two plant species may limit the advantage of the rapidly growing cool-season grass and allow big bluestem to increasingly dominate as the season progresses and temperatures increase, although plants with divergent but partially overlapping growth phenology may still exert strong competitive effects (e.g., earlier growing species can preempt resources o r continue to reduce light availability, resulting in reduced growth of later species). Also, under field conditions, the rapid attachment i f mycorrhiza-dependent plants to the existing network of mycorrhizal hyphae would allow rapid nutrient acquisition and establishment of the newly growing plant. Further research will b e necessary to determine whether competitive dominance shifts when big bluestem is already mycorrhizal when forced to compete with wild rye, o r whether planting into a soil with previously established mycorrhizal networks also alters the dominance of these two species. Effects of density In general, the outcomes of competition demonstrated here were similar among the three density levels and were generally similar between the P-fertilized and nonfertilized treatments. These results are consistent with recent work of Cousens and O'Neill (1993) who showed that altering total plant densities in replacement series experiments may change the magnitude of competitive effects but not the qualitative outcomes (e.g., identity of winners and losers). ~ d w e v e r ,in our experiment total density of the mixture did affect some of the results of competition. For example, there was a significant interaction between mycorrhizal effects and total plant density on yields of big bluestem competing with Junegrass. Changes in response at different densities are also suggested in the big bluestem wild rye mixture in which mycorrhizal effects on the relative yield patterns of wild rye changed with increasing density. These mycorrhizae x density interactions support earlier workers' assertions (Firbank and Watkinson 1985, 1990; Connolly 1986) that replacement series experiments need to incorporate multiple total plant densities to draw general conclusions about competitive relationships. It is noteworthy that while mycorrhizae and P had by far the strongest effects on relative yields of the generally mycorrhiza-dependent big bluestem and wild rye and no effect on the weakly mycorrhizal Junegrass, plant density had the greatest effect on the competitively weaker Junegrass and no effect on the mycorrhizal dependent species (Table 1). These effects together argue strongly for inclusion of density comparisons when the replacement series approach is used to assess competitive interactions between plants (Silvertown 1987; Taylor and Aarssen 1989).

Conclusions Mycorrhizal symbiosis and plant density exerted strong influence on the competitive interactions among these three prairie grasses. The competitive dominance of big bluestem in tallgrass prairie is strongly related to its mycorrhizal status, and less mycorrhiza-dependent subdominant grasses experience competitive release in the absence of the symbiosis. Furthermore, differences between competitive relationships

175

among established seedlings in this experiment and patterns of relative abundance of adult plants in the field strongly suggest that effects of mycorrhizal symbiosis on plant competition may vary considerably among the different life-history stages of prairie grasses.

Acknowledgments W e thank Laura Fischer and A.P. Schwab for assistance in preparation of this manuscript. Geoff Henebry provided advice on the jackknifing analysis and Les Firbank provided advice on the experimental design. This research was supported by National Science Foundation fNSF) grant BSR-9020426 and NSF grant 901 1662 for Long-Term Ecological Research at the Konza Prairie Research Natural Area. Adee, S.R., Pfender, W.F., and Hartnett, D.C. 1990. Competition between Pyrenophora tritici-repentis and Septoria tzodorutn in the wheat leaf as measured with de Wit replacement series. Phytopathology, 80: 1177-1182. Allen, M.F. 1991. The ecology of mycorrhizae. Cambridge University Press, Cambridge. Allen, E.B., and Allen, M.F. 1984. Competition between plants of different successional stages: mycorrhizae as regulators. Can. J. Bot. 62: 2625-2629. Allen, E.B., and Allen, M.F. 1990. The mediation of competition by mycorrhizae in successional and patchy environments. In Perspectives on plant competition. Edited by J.B. Grace and D. Tilman. Academic Press, San Diego. pp. 376-389. Bentivenga, S.P., and Hetrick, B.A.D. 1992. Seasonal and temperature effects on mycorrhizal activity and dependence of cool- and warm-season tallgrass prairie grasses. Can. J. Bot. 70: 15961602. Berendse, F. 1981. Competition between plant populations with different rooting depths. 11. Pot experiments. Oecologia, 48: 334-341. Bergelson, J.M., and Crawley, J.M. 1988. Mycorrhizal infection and plant species diversity. Nature (London), 334: 202. Briggs, J.M., and Knapp, A.K. 1991. Estimating aboveground biomass in tallgrass prairie with the harvest method: determining proper sample size using jackknifing and Monte Carlo simulations. Southwest. Nat. 36: 1-6. Brundrett, M. 1991. Mycorrhizas in natural ecosystems. Adv. Ecol. Res. 21: 171-311. Buwalda, J.G. 1980. Orowth of a clover-ryegrass association with vesicular arbuscular mycorrhizas. N.Z. J. Agric. Res. 23: 379-383. Chiariello, N., Hickman, J.C., and Mooney, H.A. 1982. Endomycorrhizal role for interspecific transfer of phosphorus in a community of annual plants. Science (Washington, D.C.), 217: 941 -943. Connell, J.H., and Lowman, M.D. 1989. Low-diversity tropical rain forests: some possible mechanisms for their existence. Am. Nat. 134: 88- 119. Connolly, J. 1986. On difficulties with replacement-series methodology in mixture experiments. J. App. Ecol. 23: 125- 137. Cousens, R., and O'Neill, M.O. 1993. Density dependence of replacement series experiments. Oikos, 66: 347-352. Daniels, B.A., McCool, P.M., and Menge, J.A. 1981. Comparative inoculum potential of spores of six vesicular-arbuscular mycorrhizal fungi. New Phytol. 89: 385-391. de Wit, C.T. 1960. On competition. Versl. Landbouwkd. Onderz. 66: 1-82. de Wit, C.T., and Van Der Bergh, J.P. 1965. Competition between herbage plants. Neth. J. Agric. Sci. 13: 212-221. Firbank, L.G., and Watkinson, A.R. 1985.,.Onthe analysis of competition within two-species mixtures of plants. J. App. Ecol. 22: 503-517.


176

CAN. J. BOT.

Firbank, L.G., and Watkinson, A.R. 1990. On the effects of competition; from monocultures to mixtures. In Perspectives on plant competition. Edited by J.B. Grace and D. Tilman. Academic Press, San Diego. pp. 165 - 192. Fitter, A.H. 1977. Influence of mycorrhizal infection on competition for phosphorus and potassium by two grasses. New Phytol. 79: 119- 125. Fitter, A.H. 1987. An architectural approach to the comparative ecology of plant root systems. New Phytol. lO6(Suppl.): 61 -77. Fowler, N.L. 1986. Density-dependent population regulation in a Texas grassland. Ecology, 67: 545-554. Grace, J.B., Keough, J., and Guntenspergen, G.R. 1992. Size bias in traditional analyses of substitutive competition experiments. Oecologia, 90: 429-434. Grime, J.P., Mackey, J.M.L., Hillier, S.H., and Read, D.J. 1987. Floristic diversity in a model system using cxperimental microcosms. Nature (London), 328: 420 -422. Hall, I.R. 1978. Effects of endomycorrhizas on the compctitive ability of white clover. N.Z. J. Agric. Res. 21: 505-519. Harper, J.L. 1977. Population biology of plants. Academic Press, San Diego. Hartnett, D.C., Hetrick, B.A.D., Wilson, G. W.T., and Gibson, D.J. 1993. Mycorrhizal influence on intra- and inter-specific neighbour interactions among co-occurring prairie grasses. J. Ecol. 81: 787-795. Hetrick, B.A.D., and Bloom, J. 1983. Vesicular-arbuscular mycorrhizal fungi associatcd with native tallgrass prairie and cultivated winter wheat. Can. J. Bot. 61: 2140-2146. Hetrick, B.A.D., Kitt, D.G., and Wilson, G.T. 1988a. Mycorrhizal dependence and growth habit of warm-scason and cool-season tallgrass prairie species. Can. J. Bot. 66: 1376- 1380. Hetrick, B.A.D., Leslie, J.F., Wilson, G.T., and Kitt, D.G. 19886. Physical and topological assessment of effects of a vesiculararbuscular mycorrhizal fungus on root architecture of big bluestem. New Phytol. 110: 85-96. Hetrick, B. A.D., Wilson, G. W.T., and Hartnett, D.C. 1989. Relationship between mycorrhizal dependence and competitive ability of two tallgrass prairie grasses. Can. J. Bot. 67: 2608-2615. Hetrick, B.A.D., Wilson, G.W.T., and Todd, T.C. 1990. Differential responscs of C, and C, grasses to mycorrhizal symbiosis, phosphorus fertilization, and soil microorganisms. Can. J. Bot. 68: 461 -467. Hetrick, B.A.D., Wilson, G.W.T., and Todd, T.C. 1992. Relationships of mycorrhizal symbiosis, rooting strategy, and phenology among tallgrass prairie forbs. Can. J. Bot. 70: 1521 - 1528. Janos, D .P. 1980. Mycorrhizae influence tropical succession. Biotropica, 12: 56-64. Jayachandran, K., Schwab, A.P., and Hetrick, B.A.D. 1992. Mineralization of organic phosphorus by vesicular-arbuscular mycorrhizal fungi. Soil Biol. Biochem. 24: 897 -903.

Jolliffe, P.A., Minjas, A.N., and Ruenckles, V.C. 1984. A reinterpretation of yield relationships in replacement series experiments. J . Appl. Ecol. 21: 227-243. Koide, R.T. 1991. Density-dependent response to mycorrhizal infection in Abutilon theophrasti Medic. Oecologia, 85: 389-395. Meyer, J.S., Igersoll, C.G., McDonald, L.L., and Boyce, M.S. 1986. Estimating uncertainty in population growth rates: jackknife vs. bootstrap techniques. Ecology, 67: 1156- 1166. Miller, R.M. 1988. The ecology of vesicular-arbuscular mycorrhizae in grass and shrublands. In Ecophysiology of VA mycorrhizal plants. Edited by G.R. Safir. CRC Prcss, Inc., Boca Raton, Fla. pp. 136- 170. Parrish, J.A.D., and B a z z z , F.A. 1979. Differenccs in pollination niche relationships in early and late successional plant communities. Ecology, 60: 587-610. Phillips, J . M . , and Hayman, D.S. 1970. Improved procedures for clearing roots and staining parasitic and vesicular-arbuscular mycorrhizal fungi for rapid assessment of infection. Trans. Br. Mycol. Soc. 55: 158-160. Schenck, N.C., and Perez, Y. 1988. Manual for the identification of VA mycorrhizal fungi. 2nd ed. International Culture Collection of Arbuscular and Vesicular-arbuscular Mycorrhizal Fungi: (INVAM), Univcrsity of Florida, Gainesville, Fla. p. 241. Silvertown, J. 1987. Introduction to plant population ecology. 2nd ed. Longman Scientific & Tcchnical, Harlow, U.K. Snaydon, R.W. 1991. Replacement or additive designs for competition studies? J. Appl. Ecol. 28: 930-946. Sokal, R.R., and Rohlf, F.J. 1981. Biometry. 2nd cd. W.H. Freeman and Co., New York. St. John, T.V., and Coleman, D.C. 1983. The role of mycorrhizae in plant ecology. Can. J. Bot. 61: 1005-1014. Taylor, D.R., and Aarssen, L.W. 1989. On the density dependence of replacement-series competition experiments. J. Ecol. 77: 975-988. Tilman, D. 1986. Nitrogen-limited growth in plants from different succcssional stages. Ecology, 67: 555 - 563. Trenbath, B.R. 1975. Neighbour effects in the genus Avena 111. A diallel approach. J. Appl. Ecol. 12: 189-200. Turkington, R., and Klein, E. 1991. Competitive outcome among four pasture species in sterilized and unsterilized soils. Soil Biol. Biochem. 23: 837 - 843. Turkington, R., and Mehrhoff, L.A. 1990. Thc role of competition in structuring pasture communities. In Perspectives on plant competition. Edited by J.B. Grace atid D . Tilman. Academic Press, San Diego. pp. 308_-340. Whittingham, J., and Read, D.J. 1982. Vesicular-arbuscular mycorrhiza in natural vegetation systems. 111. Nutrient transfer between plants with mycorrhizal interconnections. New Phytol. 90: 277-284.