Arbuscular mycorrhizae promote establishmentof prairie species in a tallgrass prairierestoration

M.R. Smith, I. Charvat, and R.L. Jacobson

Abstract: The effect that arbuscular mycorrhizal (AM) inoculum has on the development of an early successionaltallgrass prairie restoration was investigated in field plots of a recently disturbed area in Minnesota, U.S.A.Mycorrhizal inoculum reproduced from a native prairie was placed below a mix of prairie seed. Two sets of controlplots were established, those with seed only and those with seed and a sterilized soil. By the end of 15 months, plantsin the inoculated plots had a significantly greater percentage of roots colonized by AM fungi. While inoculation had noeffect on total percent cover of plants, percent cover of native planted grasses was significantly greater in theinoculated plots than in the two sets of controls. The increase in percent cover of native grasses may increase the rateof succession by allowing these grasses to outcompete the ruderal species also present at the site. Our findings suggestthat inoculation with arbuscular mycorrhizae promotes the development of early successional tallgrass prairiecommunities.

Key words: mycorrhizae, prairie, reclamation, plant community, inoculation, restoration.

Résumé: Les auteurs ont examiné les effets exercés par l’inoculation arbusculo-mycorhizienne (AM) sur ledéveloppement du début de la succession d’une prairie à herbes hautes en restauration, à partir de parcelles auxchamps, sur une surface récemment perturbée au Minnesota, aux États-Unis. L’inoculum mycorhizienne, reproduit àpartir d’une prairie indigène, a été placé sous un mélange de graines provenant d’une prairie. Les auteurs ont établisdeux groupes de parcelles témoins : avec graines seulement; avec graines et sol stérilisé. Après 15 mois, les plantesdes parcelles inoculées montraient un pourcentage significativement plus grand de racines mycorhizées par leschampignons AM. Alors que l’inoculation est demeurée sans effet sur le pourcentage total de couverture par lesplantes, le pourcentage de couverture des herbes indigènes implantées était significativement plus grand dans lesparcelles inoculées que dans les deux types de parcelles témoins. L’augmentation du pourcentage de couverture par lesherbes indigènes pourrait augmenter le taux de succession en permettant à ces herbes d’éliminer par compétition lesespèces rudérales aussi présentes sur le site. Ces constatations suggèrent que l’inoculation avec des champignonsarbusculo-mycorhiziens stimule le développement des communautés d’herbes hautes de prairie en début de succession.

Mots clés: mycorhizes, prairie, revégétalisation, communautés végétales, inoculation, restauration.

[Traduit par la Rédaction] Smith et al. 1954

The mutualistic relationship between arbuscular mycor-rhizal (AM) fungi and host plant has traditionally been stud-ied in terms of benefits to individual plants and fungi (Safir1987; Fitter 1991). AM fungi are known to benefit plants byincreasing the uptake of nutrients, especially phosphorus(Marschner and Dell 1994), increasing drought tolerance(Sylvia et al. 1993; Ruiz-Lozano and Azcon 1995), and po-tentially protecting roots from plant pathogens (Perrin 1990;

Newsham et al. 1995). More recently, some attention hasbeen paid to the role that AM fungi may have in shaping thestructure of plant communities (Francis and Read 1994).This role may be most pronounced in restorations of highlydisturbed areas where secondary succession is taking place.

The restoration of prairie communities in North Americahas received considerable attention in recent years (Ander-son and Roberts 1993; Wilson and Gerry 1995). These resto-ration efforts often involve converting old field to prairie(Cook et al. 1988) or planting prairie species in highly dis-turbed habitats (Noyd et al. 1995). Highly disturbed habitats,however, usually have altered soil communities. For exam-ple, Harris et al. (1993) found that total microbial biomassand number of fungal propagules decreased significantly asa result of topsoil storage. The amount of viable AMpropagules is also affected by size and severity of distur-bance (Miller 1979). Most work that has been conducted onmycorrhizae in disturbed habitats has taken place in extremeenvironments such as mine tailings (Reeves et al. 1979; Al-len and Allen 1980; Lambert and Cole 1980; Kiernan et al.

Can. J. Bot.76: 1947–1954 (1998) © 1998 NRC Canada

1947

Received December 18, 1997.

M.R. Smith and I. Charvat. 1 Department of Plant Biology,University of Minnesota, 1445 Gortner Avenue, St. Paul,MN 55108, U.S.A.R.L. Jacobson.Minnesota Department of Transportation,Office of Environmental Services, 3485 Hadley Avenue N.,Oakdale, MN 55128, U.S.A.

1Author to whom all correspondence should be addressed.

1983; Noyd et al. 1995) or sites of volcanic eruption (Allen1987).

Some research suggests that non-mycorrhizal or faculta-tive arbuscular mycorrhizal plants are effective colonizers ofdisturbed habitats and often dominate early successional en-vironments (Reeves et al. 1979; Miller 1987; Allen and Al-len 1990). Late successional environments, on the otherhand, are often dominated by obligate or facultative mycor-rhizal species (Allen and Allen 1990). The tallgrass prairieregion in North America has been shown to be dominated bystrongly obligate C4 species likeAndropogon gerardiiVit-man. but also contain C3 facultative species likeElymuscanadensisL. (Hartnett et al. 1993; Hetrick and Wilson1994; Anderson et al. 1994). In cases where plants of differ-ing mycorrhizal status compete with each other, especiallynon-mycorrhizal plants competing with obligate mycorrhizalplants, the presence of mycorrhizae can have a profound ef-fect on the outcome of the competition and therefore theplant community (Reeves et al. 1979; Allen and Allen 1990;Hartnett et al. 1993; Wilson and Hartnett 1997).

The presence and infectivity of mycorrhizal fungi in thesoil can be greatly reduced by soil disturbance (Moormanand Reeves 1979; Jasper et al. 1989). Thus, when consider-ing the restoration of prairie communities in disturbed habi-tats where succession in the early seral stages is dependenton mycorrhizae, it may be necessary to increase the amountof AM fungi in the soil. This may allow the mycorrhizal-dependent species to outcompete the non-mycorrhizal spe-cies thereby speeding up the rate of succession (Reeves et al.1979; Allen and Allen 1988). In an early successional plantcommunity, Gange et al. (1990) found that eliminating AMfungi using fungicide decreased total cover of the vegeta-tion. They also found fewer plant species in the fungicide-treated plots compared with nontreated plots and attributedthese differences to the positive effects that AM may haveon seedling establishment and (or) competition.

While many studies have monitored the effects ofmycorrhizae in greenhouse experiments, few have attemptedto determine their significance under natural conditions. Ofthese, most of the studies using mycorrhizal inoculation ofdisturbed sites have occurred in the semiarid west (Reeves etal. 1979; Allen and Allen 1980; Williams and Allen 1984)and involve the use of one or two plant species and a limitednumber of fungal species (Aldon 1975; Lindsey et al. 1977;Stahl et al. 1988). There have been very few studies thathave inoculated soils with mycorrhizae in the tallgrass prai-rie region (Noyd et al. 1995) and none that have monitoredthe effects of inoculation on a diverse plant community. Theobjectives of this research were to determine if mycorrhizalinoculation of a disturbed soil seeded to prairie (i) increasesroot colonization by AM fungi, (ii ) increases the percentcover of native prairie species, (iii ) increases the number ofprairie grasses that reach reproductive maturity, and (iv) af-fects the percent cover of ruderal species.

The study site was located in Isanti County, Minnesota, near thecity of Cambridge (hereafter referred to as the Cambridge studysite), 56 km north of St. Paul, Minn., on the Anoka Sand Plain(45°60′N, 93°21′W). The soils are mapped as Anoka fine loamy

sands: deep glacial outwash sands that are somewhat excessivelydrained. The average yearly precipitation and temperature are71 cm and 18.5°C, respectively. The site was part of a road con-struction project that was completed in September 1993. Topsoilthat was stockpiled was spread over the site after construction. Acover crop consisting ofLolium perennevar. aristatum Willd.,Avena sativaL., and Re-green (sterile hybrid ofTriticum aesti-vumL., andElymus trachycaulus(Link.) Gould.) (HybriTech Inc.,Wichita, Kan.) was planted to stabilize the soils in spring of 1994.All nomenclature follows Gleason and Cronquist (1991). The ex-perimental plots were established in mid-June 1995 over a totalarea of 364 m2. All plots received approximately 2.84 g of prairieseed (27 lb/acre; 1 lb/acre = 1.121 kg/ha) in a 25:1 grass:forb mix.(Table 1). All seed was obtained from Prairie Restorations Inc.,Princeton, Minn., approximately 29 km from the study site.

Three treatments were applied to twenty-four 13 2 m plots:plots that received mycorrhizal inoculum, plots that received asterile soil (control), and plots that received seed only (control)(hereafter referred to as inoculated, sterile soil control, and control,respectively). The mycorrhizal inoculum consisted of soil contain-ing mycorrhizal spores and colonized root segments. A control ofsteam-sterilized soil (see Methods below) was used to account forany effect that adding soil to the site might have. A buffer zone of2 m was placed between each plot. Twelve of the plots were usedto measure belowground parameters, and 12 were used to measureaboveground parameters resulting in four replicate plots for eachtreatment. The plots were randomly placed in the study area.Within each plot, five furrows were made at 20 cm spacing. Fur-rows were made using a belt seeder (custom made by M. Moore,Department of Plant Pathology, University of Minnesota, St. Paul).To get even distribution in the furrows, inoculum, sterile soil, andseed were spread evenly in a 2 mlong, 4 in. (1 in. = 2.54 cm) di-ameter PVC pipe, which was cut in half lengthwise. Inoculum,sterile soil, and seed were then dumped into the furrows. Threehundred fifty grams of inoculum or sterile soil (approximately875 g/m2) were placed in the furrows, seed was placed on top andthe furrow was then covered with soil from the site. In all cases,furrows were dug so that final seed depth was 0.625 cm. Inoculum,sterile soil, and seed were weighed prior to placing in the field todeliver consistent amounts among furrows and plots.

Arbuscular mycorrhizal inoculumTo produce mycorrhizal inoculum, field soil was obtained from

a native, undisturbed prairie in the Twin Cities metropolitan areaand used in trap culture withSchizachyrium scoparium(Michx.)Nash. At least 20 species of mycorrhizal fungi have been identifiedfrom this prairie (H. Agwa and I. Charvat, in preparation). The trapculture for production of inoculum consisted of steam-sterilizedgreenhouse soil (steamed twice for 12 h with a 24-h interval be-tween steamings) with a 1-cm3 layer of prairie soil placed betweenlayers of sterilized greenhouse soil. The resulting inoculum wastherefore primarily composed of greenhouse soil. ThreeSchiza-chyrium scopariumseedlings were planted in each 5.5-in. pot. Af-ter transplanting, pots were maintained in a growth chamber with a12 h light : 12 h dark photoperiod and thermoperiod (25:20°C) andwatered with 50 mL 10% Hoagland’s solution once a week andwith tap water when needed. After 1 month, pots were transferredto the greenhouse and watered once a week with 100 mL 10%Hoagland’s solution and with tap water daily. Plants were grownfor a total of 19 weeks, at which point watering was stopped.Aboveground plant mass was clipped, and the dried soil and rootmass was refrigerated for 1 month before use. Inoculum wassieved through 1-cm hardware cloth, roots were cut to 1 cm pieces,and roots from all of the pots were homogenized. To determinespore number present in the inoculum, spores were isolated follow-ing methods in Daniels and Skipper (1982). Along with root andhyphal fragments, the inoculum contained 38 AM spores per gram

© 1998 NRC Canada

1948 Can. J. Bot. Vol. 76, 1998

of dry soil. Sterilized soil used as a control consisted of the samegreenhouse soil that was used in the trap cultures (inoculum) andwas steam sterilized using the same method. This sterile soil con-trol was treated with a microbial rinse obtained from inoculum soiland rinsed over a 38-µm sieve.

Mycorrhizal–soil parametersAt the time of planting, soil cores were taken randomly through-

out the study site and used for measuring background percent colo-nization levels and spore numbers at the site. In September 1996,soil cores were taken directly adjacent to the furrows to a depth of5 cm using a 2 cmdiameter soil probe. By taking soil cores di-rectly adjacent to the furrows, we assumed that the majority of theroots obtained from these cores originated from the planted prairiespecies. Five cores were taken at random locations along each fur-row resulting in 25 cores per plot. These 25 cores were thoroughlyhomogenized to obtain a composite sample for each plot and usedfor nutrient analyses and the determination of root colonization.

Roots were isolated over a 250-µm sieve, washed free of debrisand preserved in 50% ethanol. Clearing and staining proceduresfollowed methods modified from Phillips and Hayman (1970),Koske and Gemma (1989), and Kormanik et al. (1980). All rootsobtained from the site were cleared overnight in 10% KOH, acidi-fied for 1 h in 1%HCl, stained overnight with 0.05% trypan blue,and destained with acidic glycerol. A randomly selected subsampleof the stained roots from each plot were mounted on microscopeslides. Percent AM colonization was determined using the magni-fied intersection method (McGonigle et al. 1990). Roots were ex-amined at 160 and 4003. Four replicates of 100 intersections eachwere counted for each plot. Total mycorrhizal colonization wascalculated by adding vesicular, arbuscular, and AM hyphal coloni-zation. AM hyphae were distinguished from hyphae of unknownorigin by the presence of mycorrhizal structures attached to them.If the eyepiece crosshair intersected a hypha that was attached to avesicle or arbuscule within the field of view, this hypha wascounted as an AM hypha.

Nutrient analysis was performed on soil from the study site (col-lected from the top 12 cm at the time of planting), inoculum soil,and sterile soil. Analysis of soil NO3-N and Bray-P was conductedat the University of Minnesota Research and Analytical Labora-tory. Soil pH was determined using 0.01 M CaCl2 with asoil:CaCl2 ratio of 1:2 (Hendershot et al. 1993).

Percent coverPercent cover was measured during the second week of Septem-

ber in 1995 and 1996 using the point frame method (Goodall1957). The frame was constructed to fit over the 1 m wide plots.Ten flags were evenly spaced along this 1-m distance on the frame.The entire frame was placed 10 times in each plot at 20-cm inter-

vals, and averages for each plot were taken from these 10 place-ments. This insured that percent cover estimates were taken overthe entire plot. Each time an individual leaf, flower, or stemtouched the flag, its presence was recorded, even if it was from thesame individual. If the plant touching the flag was dead, this wasrecorded next to the species code. If the plant could not be identi-fied, it was either collected from outside the plot for future identi-fication or recorded as an unknown. The senesced basal rosette ofleaves often present at the base of warm-season prairie grasseswere recorded but were not recorded as dead because they are partof a living plant. This method of measuring percent cover often re-sults in measurements exceeding 100% because the overlappinglayers of vegetation are taken into account.

Many of the warm-season native grasses were too immature toidentify to species and were placed in their own general category.This category consisted ofAndropogon gerardii, Panicum virga-tum L., and Sorghastrum nutansL. Nash. If one of these plantscould be identified to species (i.e., if it were found flowering), itwas recorded in its own category. For statistical analysis, however,all native grasses were grouped into a single category. The numberof individual flowering grasses and forbs of each species were re-corded from all plots. A plant was also recorded as flowering if ithad previously flowered during the current season. Many of theforbs, for example, had already set seed or had withered flowers bySeptember 1996 and were recorded as flowering.

Statistical analysisDifferences in mycorrhizal colonization were analyzed using

ANOVA at α = 0.05 followed by Fisher’s protected least signifi-cant difference test (Statview; Abacus Concepts Inc., 1992) to sep-arate means. All populations of percent colonization were normal.Variances were tested for equality using Bartlett’sF test for homo-geneity of variances atα = 0.01. An arcsine transformation is com-mon for percent cover data but was not used here because thepopulations were normal, the variances were equal using the Bart-lett’s F test, and there were data points exceeding 1.0. Comparisonof percent cover and root colonization from the first and secondyears was performed using pairedt test atα = 0.05. Comparison ofplant life-history groups in first and second years was done usingthe Wilcoxon signed rank test because populations were not nor-mally distributed.

Mycorrhizal–soil parametersSoil was collected from the site, from the inoculum and

from the sterile soil for analysis at the time of planting inJune 1995. The results from these tests are presented in

© 1998 NRC Canada

Smith et al. 1949

Plant species Common name Percentage in seed mix*

Andropogon gerardiiVitman Big bluestem 35Schizachyrium scoparium(Michx.) Nash. Little bluestem 24Sorghastrum nutansL. Nash. Indian grass 24Bouteloua curtipendula(Michx.) Torr. Side-oats gramma 8.6Elymus canadensisL. Canada wild rye 2.4Panicum virgatumL. Switch grass 2.4Rudbeckia hirtaL. Black-eyed susan 0.5Dalea purpureaVent. Purple prairie clover 0.5Dalea candidaMichx. White prairie clover 0.3Heliopsis helianthoidesL. Common ox-eye 0.2

*Percentages for grasses are percent pure live seed. Percentages for forbs are by bulk weight. Species with less than 0.2% are not listed.

Table 1. List of major species and percentages in seed mix used at the Cambridge, Minn. prairie restoration site, June 1995.

Table 2. This soil from the site was also used to determinetotal background AM root colonization levels and sporenumbers. Mycorrhizal fungi colonized 0.4 ± 0.2% (mean ±SE) of roots at that time and the soil contained 1.47 ± 0.12spores per gram of dry soil.

In September of 1996, total root colonization was signifi-cantly greater than in 1995 in all treatments (p = 0.018). To-tal mycorrhizal colonization in 1996 was 44.2 ± 1.9% in theinoculated treatments, 35.3 ± 1.7% in the sterile soil controltreatments and 37.7 ± 1.4% in the control treatments. Totalcolonization was comprised of arbuscular, vesicular, andAM hyphal colonization. Of these, arbuscular and vesicularcolonization were significantly greater in the inoculatedtreatments than either control treatment (Fig. 1). There wasno difference in colonization of AM hyphae among treat-ments.

Percent coverThe total percent cover of plants measured in September

1995 was 177 ± 8.8% for the inoculated treatments, 190 ±16.2% for the sterile soil control treatments, and 189 ±20.4% for the control treatments. No significant differenceamong treatments was observed. Total cover was then segre-gated into three groups: planted prairie species, unplantedweedy species, and cover crop. In 1995 all of the treatmentswere dominated by unplanted weedy species and cover crop

species; the planted prairie species had not yet germinated.The percent cover of the unplanted weedy species were asfollows: 116 ± 3.4% in the inoculated treatment, 120 ±11.1% in the sterile soil control treatment, and 125 ± 16.4%in the control treatment. The percent cover for the covercrop species were as follows: 55.3 ± 4.7% in inoculatedplots, 61 ± 10.4% in sterile soil control plots, and 56 ±18.2% in control plots. There was no significant differencein cover of these groups between treatments. The plant com-munity was further described in terms of plant life history(Table 3) and was characteristic of recently disturbed sites:the majority of the cover was comprised of annual and bien-nial forbs. Conyza canadensis(L.) Cronq. comprised ap-proximately 50% of the total cover. Perennial grasses alsomade up a significant portion of the cover and consisted al-most exclusively ofPoa pratensisL. Annual and biennialgrasses and perennial forbs were present in smaller amountsand consisted ofDigitaria ischaemum(Schreber) Muhl.,Setariasp., Plantago majorL., Silene latifoliaPoiret., andAmbrosia artemisiifoliaL. There were no significant differ-ences in percent cover between the plots for any of thesegroups in the first year (Table 3).

In 1996, total percent cover of plants was 164 ± 19.2% inthe inoculated treatment, 140 ± 8.1% in the sterile soil con-trol treatment, and 133 ± 7.3% in the control treatment. Nostatistical difference in total percent cover of plants was ob-served among treatments. Figure 2 illustrates total cover seg-regated into three groups: planted prairie species, unplantedweedy species, and cover crop species. The percent cover ofunplanted species and of cover crop species was not statisti-cally different among treatments. The percent cover ofplanted prairie species, however, was significantly greater inthe inoculated treatment than in the two control treatments.

This increase in percent cover of native species in the in-oculated treatments can be attributed mainly to the nativeplanted grasses (an average of 77% per plot forAndropogongerardii, Panicum virgatum, and Sorghastrum nutansand

© 1998 NRC Canada

1950 Can. J. Bot. Vol. 76, 1998

Soil origin Soil type Bray-P (ppm) NO3-N (ppm) pH

Study site Fine loamysand

69 1.3 6.4

Inoculum Loam 101 0.08 6.9Sterile soil Loam 128 198 na*

*na, not available.

Table 2. Soil characteristics for the Cambridge, Minn. study site;the inoculum; and the sterile soil.

Fig. 1. Mean percentage colonization of roots (± SE), September1996 at the Cambridge, Minn. restoration site. Bars withdifferent letters are significantly different between the samemycorrhizal structures atα = 0.05. Arbuscular colonization,p = 0.038; vesicular colonization,p = 0.017.

Fig. 2. Mean percent cover (± SE) of plants by type at theCambridge, Minn. restoration site in September, 1996. Barswith different letters are significantly different atα = 0.05.

21% for Schizachyrium scopariumand B. curtipendula)(Table 3). The planted forbs (mainlyRudbeckia hirtaL. andHeliopsis helianthoidesL.) also contributed a small amountof cover (3%). The cover of unplanted weedy perennialforbs was significantly greater in the inoculated treatmentsthan in the two controls. This consisted ofPotentillaargenteaL., Trifolium hybridumL., andMedicago sativaL.(Table 3). The percent cover of plants in the remaining life-history groups did not differ significantly among treatmentsin the second year.

There was no statistical difference in the number of nativeprairie grasses or forbs that were flowering in mid-September in any of the treatments in the second year(Table 4).

The total percent cover of plants did not differ signifi-cantly between any of the treatments between the first andsecond years. The makeup of the plant community, however,changed dramatically (Table 3). The cover of perennialweedy grasses decreased significantly in the second year inthe inoculated and sterile soil control treatments. All treat-ments had a significantly greater amount of planted grassesin the second year as compared with the first year. All treat-ments also had significantly fewer annual and biennial forbsin the second year. This was mainly due to the loss ofC. canadensisin the second year.

Many factors have resulted in a loss of natural communi-ties across North America. The restoration of native plantcommunities has been important in attempting to counteractthat loss. Our results show that mycorrhizae may play an im-portant role in the restoration of tallgrass prairie plant com-munities. We found that inoculating a disturbed site withmycorrhizae while seeding to prairie increased mycorrhizalcolonization of native prairie grasses. Furthermore, inocu-lated plots had significant increases in the percent cover ofnative grasses.

InoculationThe very low root colonization levels and spore numbers

found at the site in the spring of 1995 indicate that the dis-turbance greatly affected the mycorrhizal populations, atleast temporarily. Colonization found in the control plots in1996 indicates that either the low levels of mycorrhizae ini-

tially present at the site grew in the presence of host species(native prairie grasses), and (or) the rate of influx ofmycorrhizal propagules into the site was high. Working ondisturbed mine sites in western Kentucky, Gould et al.(1996) found that mycorrhizae were present in the first yearat very low levels and naturally recolonized the area overthe following 2.5 years, after which their numbers stabilized.The same pattern of colonization may be taking place in thepresent study.

Inoculation with mycorrhizae resulted in a significant in-crease in total, vesicular, and arbuscular root colonization.These significant differences, however, were not as great asexpected and may not fully explain the differences observedin percent cover of native plants. In our study, it is unclearwhether it was solely the amount of mycorrhizae availableto the plants that increased percent cover or whether thisacted in concert with the AM species composition of theinoculum. AM species vary in their efficiency as mutualists,and some species may even act as parasites (Johnson et al.1992; Newsham et al. 1995; Francis and Read 1995). It istherefore possible that the inoculum provided a mycorrhizalcommunity substantially different than that initially presentin the study site, shifting community composition towardsmore beneficial species. Comparisons of the mycorrhizalcommunities in the inoculum versus the restored site arecurrently underway.

FloweringThe presence of mycorrhizae has been shown to have

variable effects on the incidence of flowering warm-seasonprairie grasses. Wilson and Hartnett (1997) found that thepresence of mycorrhizal fungi decreased the ratio of repro-ductive to vegetative biomass. Hartnett et al. (1994), work-ing with the fungicide Benomyl to reduce mycorrhizae in

© 1998 NRC Canada

Smith et al. 1951

Plantedgrasses

Plantedforbs Total

Inoculated 7.63±1.71 3.38±0.8 11.0±1.69Sterile soil control 4.00±1.52 3.13±1.64 7.13±2.88Control 4.63±1.69 5.75±2.56 10.4±2.92

Note: No significant differences among treatments were found.

Table 4. Mean (± SE) number of flowering planted grasses andforbs at the Cambridge, Minn. restoration site, September 1996.

Planted native species Unplanted weedy species

TreatmentPerennialgrasses

Mixedforbs

Annual orbiennial grasses

Annual orbiennial forbs

Perennialgrasses

Perennialforbs Other

Covercrop

1995 Inoculated 0.5±0.03* 0.08±0.05 6.8±3.7 51±4.6* 52±3.0* 6.3±1.8 1.3±0.6 55±4.7*Sterile soil control 0* 0.3±0.3* 8.8±6.5 43±3.2* 50±3.0* 16±7.5* 2.8±0.8 61±11Control 0* 0 24±9.2 40±5.3* 40±14 11±6.4 6.5±3.3 56±18

1996 Inoculated 98±8.7a 5.5±2.4 12±4.6 14±6.6 12±10 5.8±2.0a 3.8±1.0 17±6.6Sterile soil control 68±10.0ab 2.5±0.9 7.5±2.9 0.8±0.5 28±11 0.5±0.3b 2.8±1.0 31±11Control 55±10.4b 0.8±0.5 26±11 11±10 17±8.3 2.0±1.2ab 7.5±6.9 16±3.8

Note: Values are mean percent cover values ± SE. Values in 1996 with different letters are significantly different atα = 0.05.*Significant difference between years atα = 0.06 (α = 0.06 is the greatest level of significance possible using the Wilcoxon signed rank test with this

sample size). Between-year comparisons are made within treatment and plant life history groups only.

Table 3. Plant community composition among treatments at Cambridge, Minn.

the field, found that mycorrhizae had no effect on the flow-ering of warm-season grasses in an unburned prairie. Similarto Hartnett et al. (1994), we found that increased levels ofmycorrhizae had no significant effect on flowering of thewarm-season prairie grasses. However, most of the informa-tion concerning flowering of these species and mycorrhizaecomes from studies where mycorrhizae is suppressed withfungicide to levels of colonization less than 10% (Hartnettet al. 1994; Wilson and Hartnett 1997). Background coloni-zation levels in the control plots in our study were 37% andmay explain the lack of possible response to inoculation.

Plant community compositionWhile the total percent cover of plants from the first year

to the second did not change significantly in any of theplots, the makeup of the plant community shifted consider-ably. Overall, inoculation had no effect on the percent coverof unplanted weedy species probably because of the fact thatthe inoculum was placed in furrows and not broadcast overthe plots. A significant decrease in percent cover of weedyannual or biennial forbs was observed from 1995 to 1996 inall plots. The majority of the cover of this group in the firstyear consisted ofC. canadensis. Keever (1950) found thatC. canadensisoften dominates old fields the first year aftercultivation but loses dominance by the second year. Thisloss of dominance occurs because the growth of the second-year seedlings is inhibited by the decaying root materialfrom the first-year plants.

The percent cover of the warm-season prairie grasses in-creased significantly from the first year to the second. Thisincrease consisted mainly ofAndropogon gerardii, Panicumvirgatum, andSorghastrum nutans. The impact that mycor-rhizae may have on the survival and competition of theseprairie grasses has been given considerable attention (Het-rick et al. 1988; Hetrick et al. 1990; Wilson and Hartnett1997). The presence or absence of mycorrhizae can alter thecompetitive outcome between plants of different mycorrhi-zal dependence. The presence of AM allows the obligatemycorrhizal Andropogon gerardiiand Sorghastrum nutansto outcompete facultative or non-mycotrophic cool-seasongrasses (Hetrick et al. 1989; Hetrick et al. 1994; Wilson andHartnett 1997). The dominance thatAndropogon gerardiiand Sorghastrum nutanshold in the tallgrass prairie, there-fore, is highly dependent on their mycorrhizal status (Hart-nett et al. 1993; Hetrick et al. 1994). Our data suggest thatthe presence of mycorrhizae is essential for the establish-ment of these species and therefore the tallgrass prairie com-munity.

It is likely that the increased cover of the planted prairiegrasses in the inoculated plots is a result of the competitiveadvantage that AM fungi bestow on these plants. Grimeet al. (1987), working with a microcosm dominated by cool-season (i.e., facultative and non-mycorrhizal) turf grasses,showed that the presence of mycorrhizae resulted in an in-crease of plant species diversity. This occurred presumablybecause the subordinate, obligate mycorrhizal species expe-rienced competitive release. In ecosystems where obligatemycorrhizal species dominate, however, the opposite may betrue (Bergelson and Crawley 1988; Allen, 1991). Workingwith experimental microcosms containing native prairie

grasses and forbs, Wilson and Hartnett (1997) showed thatwhen mycorrhizal activity was reduced, the warm-seasongrasses became relatively less abundant and species diver-sity increased. The dominant unplanted species at our sitebefore planting prairie species were the facultative mycor-rhizal speciesC. canadensis(Harley and Harley 1987),Lo-lium perenne(Lambert and Cole 1980), andPoa pratensis(Wilson and Hartnett 1987). The restoration of prairie com-munities, therefore, more resembles the Grime et al. (1987)model in that the dominant plants are facultative and non-mycorrhizal weedy species. The competitive advantagebrought about by inoculation may speed up the process ofsuccession by allowing the warm season prairie grasses tooutcompete the weedy species also present at the site.

At our disturbed study site, the mycorrhizal population(measured as percent colonization) increased over a 15-month period regardless of treatment. In spite of this, wefound that placing mycorrhizal inoculum directly below na-tive prairie seed significantly increased percent mycorrhizalcolonization. We also found greater cover of native prairiespecies in the inoculated plots, suggesting that mycorrhizalinoculation promotes establishment of prairie species. Thus,our results support the Grime et al. (1987) model whereinthe addition of mycorrhizae can lead to a more rapid rate ofsuccession.

The authors thank Jenny White for her helpful editing,Elizabeth Gould and others for assistance with field work,and Dwayne Stenlund for technical and field assistance. Thiswork represents a portion of a thesis submitted to the Gradu-ate School of the University of Minnesota in partial fulfill-ment of the requirements for the degree Master of Science.Funding was provided by the Carolyn Crosby Fellowship,Minnesota Department of Transportation, and the MinnesotaLegislature (ML 95, Chp. 220, Sec. 19, Subd. 13(C)) as rec-ommended by the Legislative Commission on Minnesota re-sources from the Minnesota Environment and NaturalResources Trust Fund.

Aldon, E.F. 1975. Endomycorrhizae enhance survival and growthof fourwing saltbrush on coal mine spoils. USDA For. Serv.Rocky Mount. For. Range Exp. Stn. Res. Note No. RM-294.

Allen, E.B., and Allen, M.F. 1980. Natural re-establishment ofvesicular–arbuscular mycorrhizae following stripmine reclama-tion in Wyoming. J. Appl. Ecol.17: 139–147.

Allen, E.B., and Allen, M.F. 1988. Facilitation of succession by thenon-mycotrophic colonizerSalsola kali(Chenopodiaceae) on aharsh site: effects of mycorrhizal fungi. Am. J. Bot.75: 257–266.

Allen, E.B., and Allen, M.F. 1990. The mediation of competitionby mycorrhizae in successional and patchy environments.InPerspectives on plant competition.Edited byD. Tilman and J.P.Grime. Academic Press, San Diego, Calif. pp. 367–389.

Allen, M.F. 1987. Re-establishment of mycorrhizas on Mount St.Helens: migration vectors. Trans. Br. Mycol. Soc.88: 413–417.

© 1998 NRC Canada

1952 Can. J. Bot. Vol. 76, 1998

Allen, M.F. 1991. The ecology of mycorrhizae. Cambridge Univer-sity Press, Cambridge, U.K.

Anderson, R.C., and Roberts, K.J. 1993. Mycorrhizae in prairierestoration: response of three little bluestem (Schizachyriumscoparium) populations to mycorrhizal inoculum from a singlesource. Restor. Ecol.1: 83–87.

Anderson, R.C., Hetrick, B.A.D., and Wilson, G.W.T. 1994. Mycor-rhizal dependence ofAndropogon gerardiiand Schizachyriumscopariumin two prairie soils. Am. Midl. Nat.132: 366–376.

Bergelson, J.M., and Crawley, J.M. 1988. Mycorrhizal infectionand plant species diversity. Nature (London),334: 202.

Cook, B.D., Jastrow, J.D., and Miller, R.M. 1988. Root andmycorrhizal endophyte development in a chronosequence of re-stored tallgrass prairie. New Phytol.110: 355–362.

Daniels, B.A., and Skipper, H.D. 1982. Methods for the recoveryand quantitative estimation of propagules from soil.In Methodsand principles of mycorrhizal research.Edited byN.C. Schenck.American Phytopathological Society, St. Paul, Minn.

Fitter, A.H. 1991. Costs and benefits of mycorrhizae: implicationsfor functioning under natural conditions. Experientia,47: 350–362.

Francis, R., and Read, D.J. 1994. The contributions of mycorrhizalfungi to the determination of plant community structure. PlantSoil, 159: 11–25.

Francis, R., and Read, D.J. 1995. Mutualism and antagonism in themycorrhizal symbiosis, with special reference to impacts onplant community structure. Can. J. Bot.73(Suppl 1): S1301–S1309.

Gange, A.C., Brown, V.K., and Farmer, L.M. 1990. A test ofmycorrhizal benefit in an early successional plant community.New Phytol.115: 85–91.

Gleason, H.A., and Cronquist, A.C. 1991. Manual of vascularplants of northeastern United States and adjacent Canada. 2nded. New York Botanical Garden, New York.

Goodall, D.W. 1957. Some considerations in the use of point quad-rats for the analysis of vegetation. Aust. J. Biol. Sci.5: 1–41.

Gould, A.B., Hendrix, J.W., and Ferriss, R.S. 1996. Relationship ofmycorrhizal activity to time following reclamation of surfacemine land in western Kentucky. I. Propagule and spore popula-tion densities. Can. J. Bot.74: 247–261.

Grime, J.P., Mackey, J.M.L., Hillier, S.H., and Read, D.J. 1987.Floristic diversity in a model system using experimental micro-cosms. Nature (London),328: 420–422.

Harley, J.L., and Harley, E.L. 1987. A checklist of mycorrhiza inthe British flora. New Phytol.105(Suppl): 1–102.

Harris, J.A., Birch, P., and Short, K.C. 1993. The impact of storageof soils during opencast mining on the microbial community: astrategist theory interpretation. Restor. Ecol.1: 88–100.

Hartnett, D.C., Hetrick, B.A.D., Wilson, G.W.T., and Gibson, D.J.1993. Mycorrhizal influence on intra- and interspecific neigh-bour interactions among co-occurring prairie grasses. J. Ecol.81: 787–795.

Hartnett, D.C., Samenus, R.J., Fischer, L.E., and Hetrick, B.A.D.1994. Plant demographic responses to mycorrhizal symbiosis intallgrass prairie. Oecologia,99: 21–26.

Hendershot, W.H., Lalande, H., and Duquette, M. 1993. Soil reac-tion and exchangeable acidity.In Soil sampling and methods ofanalysis.Edited byM.R. Carter. Canadian Society of Soil Sci-ence, Lewis Publishers, Boca Raton, Fla. pp. 141–145.

Hetrick, B.A.D., and Wilson, G.W.T. 1994. Mycorrhizal activity inwarm and cool season grasses: variation in nutrient-uptake strat-egies. Can. J. Bot.72: 1002–1008.

Hetrick, B.A.D., Kitt, D.G., and Wilson, G.W.T. 1988. Mycorrhizaldependence and growth habit of warm-season and cool-seasontallgrass prairie plants. Can. J. Bot.66: 1376–1380.

Hetrick, B.A.D., Wilson, G.W.T., and Hartnett, D.C. 1989. Rela-tionship between mycorrhizal dependence and competitive abil-ity of two tallgrass prairie grasses. Can. J. Bot.67: 2608–2615.

Hetrick, B.A.D., Wilson, G.W.T., and Todd, T.C. 1990. Differentialresponses of C3 and C4 grasses to mycorrhizal symbiosis, phos-phorus fertilization and soil microorganisms. Can. J. Bot.68:461–467.

Hetrick, B.A.D., Hartnett, D.C., Wilson, G.W.T., and Gibson, D.J.1994. Effects of mycorrhizae, phosphorus availability, and plantdensity on yield relationships among competing tallgrass prairiegrasses. Can. J. Bot.72: 168–176.

Jasper, D.A., Abbott, L.K., and Robson, A.D. 1989. Soil distur-bance reduces the infectivity of external hyphae of vesicular–arbuscular mycorrhizal fungi. New Phytol.112: 93–99.

Johnson, N.C., Copeland, P.J., Crookston, R.K., and Pfleger, F.L.1992. Mycorrhizae: possible explanation for yield decline withcontinuous corn and soybeans. Agron. J.84: 387–390.

Keever, C. 1950. Causes of succession on old fields of thePiedmont, N.C. Ecol. Monogr.20: 229–250.

Kiernan, J.M., Hendrix, J.W., and Maronek, D.M. 1983. Endomy-corrhizal fungi occurring on orphan strip mines in Kentucky.Can. J. Bot.61: 1798–1803.

Kormanik, P.P., Bryan, W.C., and Shultz, R.C. 1980. Proceduresand equipment for staining large numbers of plant roots forendomycorrhizal assay. Can. J. Microbiol.26: 535–538.

Koske, R.E., and Gemma, J.N. 1989. A modified procedure forstaining roots to detect VA mycorrhizas. Mycol. Res.94(4):486–488.

Lambert, D.H., and Cole, H., Jr. 1980. Effects of mycorrhizae inestablishment and performance of forage species in mine spoil.Agron. J.72: 257–260.

Lindsey, D.L., Cress, W.A., and Aldon, E.F. 1977. The effects ofendomycorrhizae on growth of rabbitbrush, fourwing saltbushand corn in coal mine spoil material. USDA For. Serv. RockyMount. For. Range Exp. Stn. Res. Note No. RM-343.

Marschner, H., and Dell, B. 1994. Nutrient uptake in mycorrhizalsymbiosis. Plant Soil,159: 89–102.

McGonigle, T.P., Miller, M.H., Evans, D.G., Fairchild, G.L., andSwan, J.A. 1990. A new method which gives an objective mea-sure of colonization of roots by vesicular–arbuscular mycor-rhizal fungi. New Phytol.155: 495–501.

Miller, M.R. 1979. Some occurrence of vesicular–arbuscular my-corrhiza in natural and disturbed ecosystems of the Red Desert.Can. J. Bot.57: 619–623.

Miller, M.R. 1987. Mycorrhizae and succession.In Restorationecology: a synthetic approach to ecological research.Edited byW.R. Jordon III, M.E. Gilpin, and J.D. Aber. Cambridge Univer-sity Press, Cambridge, U.K. pp. 205–219.

Moorman, T., and Reeves, F.B. 1979. The role of endomycorrhizaein revegetation practices in the semi-arid west. II. A bioassay todetermine the effect of land disturbance on endomycorrhizalpopulations. Am. J. Bot.66: 14–18.

Newsham, K.K., Fitter, A.H., and Watkinson, A.R. 1995. Multi-functionality and biodiversity in arbuscular mycorrhizas. TrendsEcol. Evol. 10: 407–411.

Noyd, R.K., Pfleger, F.L., and Russelle, M.P. 1995. Interactions be-tween native prairie grasses and indigenous arbuscular mycor-rhizal fungi: implications for reclamation of taconite iron oretailing. New Phytol.129: 651–660.

© 1998 NRC Canada

Smith et al. 1953

© 1998 NRC Canada

1954 Can. J. Bot. Vol. 76, 1998

Perrin, R. 1990. Interactions between mycorrhiza and diseasescaused by soil-borne fungi. Soil Use Manage.6: 189–195.

Phillips, J.M., and Hayman, D.S. 1970. Improved procedures forclearing roots and staining parasitic and vesicular–arbuscularmycorrhizal fungi for rapid assessment of infection. Trans. Br.Mycol. Soc.55: 158–161.

Reeves, F.B., Wagner, D., Moorman, T., and Kiel, J. 1979. The roleof endomycorrhizae in revegetation practices of the semi-aridwest. I. A comparison of incidence of mycorrhizae in severelydisturbed vs. natural environments. Am. J. Bot.66: 6–13.

Ruiz-Lozano, J.M., and Azcon, R. 1995. Hyphal contribution towater uptake in mycorrhizal plants as affected by the fungal spe-cies and water status. Physiol. Plant.95: 472–478.

Safir, G.R. (Editor). 1987. Ecophysiology of VA mycorrhizalplants, CRC Press, Boca Raton, Fla.

Stahl, P.D., Williams, S.E., and M. Christensin. 1988. Efficacy ofnative vesicular–arbuscular mycorrhizal fungi after severe soildisturbance. New Phytol.110: 347–354.

Sylvia, D.M., Hammond, L.C., Bennett, J.M., Haas, J.H., andLinda, S.B. 1993. Field response of maize to a VAM fungus andwater management. Agron. J.85: 193–198.

Williams, S.E., and Allen, M.F. 1984. VA mycorrhizae and recla-mation of disturbed arid and semiarid lands. Sci. Rep. No. 61.Wyoming Agricultural Experiment Station, Laramie, Wyo.

Wilson, S.D., and Gerry, A.K. 1995. Strategies for mixed-grassprairie restoration: herbicide, tilling, and nitrogen manipulation.Restor. Ecol.3: 290–298.

Wilson, S.D., and Hartnett, D.C. 1997. Effects of mycorrhizae onplant growth and dynamics in experimental tallgrass prairie mi-crocosms. Am. J. Bot.84: 478–482.