Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Mycorrhizae and Succession in Plantings of Beachgrass in Sand Dunes Author(s): R. E. Koske and J. N. Gemma Source: American Journal of Botany, Vol. 84, No. 1 (Jan., 1997), pp. 118-130 Published by: Botanical Society of America Stable URL: http://www.jstor.org/stable/2445889 . Accessed: 19/07/2013 16:33 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. . Botanical Society of America is collaborating with JSTOR to digitize, preserve and extend access to American Journal of Botany. http://www.jstor.org This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions AmericanJournalof Botany 84(1): 118-130. 1997. MYCORRHIZAE AND SUCCESSION IN PLANTINGS OF BEACHGRASS IN SAND DUNES1 R. E. KOSKE2 AND J. N. GEMMA Departmentof Biological Sciences, Universityof Rhode Island, Kingston,Rhode Island 02881-0816 A survey of arbuscularmycorrhizalfungi (AMF), arbuscularmycorrhizae(AM), and hyphal networksof AMF was carriedout in sand dune sites of different successional stages in the Province Lands Area of Cape Cod National Seashore, Massachusetts.The studyfocused on large-scaleplantings(each of 12-20 ha) of Americanbeachgrass (Ammophilabreviligulata) aged 0-7 yr and fiveadjacent naturaldune areas. Sample sites rangedin vegetativecover frombarrento forested. Spores of 17 species of AMF were recoveredfromthe dunes. Over the successional sequence, therewere increases in the richnessand spore populationsof the AMF community,the extentof colonization of A. breviligulataroots, and the mycorrhizalinoculum potentialof the soil. Unvegetatedsites lacked propagules of AMF, but roots of planted culms of A. breviligulata(which carriedpropagulesof AMF) became mycorrhizalin <1 yr afterplanting.Spores were recoveredfrom previouslyAMF-free sites thathad been planted with beachgrass for47 wk, and five species of AMF sporulatedin sites <6 yr old. Significanthyphalnetworkswere not presentin any of the planted areas (<6 yr old at the time of sampling), but did occur in naturalareas. The rate of invasion of areas plantedto A. breviligulataby later successional plant species may in partdepend upon the establishmentof a vigorousnetworkof hyphaeof AMF in a site. Key words: Ammophilabreviligulata;arbuscularmycorrhizae;hyphalnetwork;plant succession; Poaceae; sand dunes. Since 1959, the possible importanceof arbuscularmy- inhabitantsof Cape Cod for more than threecenturies. Prior to the arrivalof European colonists in the 1600s, corrhizalfungi(AMF) in sand dune succession has been recognized,and numeroussurveyshave shown AMF to the area consisted of foresteddunes. Destructionof the forestand damage by grazinglivestockin the following be routineinhabitantsof temperateand tropicaldune sites centurypermittedthe formerlystabilized dunes to bein coastal and inland dunes throughoutthe world (e.g., Nicolson, 1960; Koske and Polson, 1984; Giovannetti, come mobile again and led to the currentproblemswith drifting sand (McCaffreyand Leatherman,1979). Human 1985; Puppi and Riess, 1987; Rose, 1988; Read, 1989; uses of the mobile dunes for recreationhave prevented Koske and Gemma, 1990, 1992, 1996; Sturmerand Bellei, 1994; Al-Agely and Reeves, 1995). In these primary re-establishmentof sufficientvegetation to slow the successional sites, AMF arriveearly,oftenwiththe first movementof the sand. Inundationof roads, buildings, and adjacent forestsand the fillingin of Provincetown plant colonizers (Nicolson, 1960; Sylvia and Will, 1988; Harbor and PilgrimLake by wind-blownsand are longGemma and Koske, 1989, 1992; Koske and Gemma, standingproblemsin the area. 1990). Vegetation is an effectivemeans of slowing sand The developmentof the belowgroundAMF commumovementin dunes (e.g., Jagschitzand Bell, 1966; Rannity is closely linked to aboveground vegetational changes duringcolonizationand succession. In a variety well, 1972; Woodhouse, 1982), and large plantingshave been made on Cape Cod for this purpose since at least of naturalsites,AMF have been foundto play a vitalrole 1825 (Thoreau, 1865; Strahler,1966). The most recent in plant succession, affectingthe success not only of obmajor plantingshave been made in the Province Lands ligatelymycotrophicspecies, but also of species thatare Dunes Area of Cape Cod National Seashore (CCNS). Defacultativelymycotrophicand even nonmycotrophic. 75 ha of American beachgrass (Ammopending on the level of disturbanceand existingplant Since 1985, phila breviligulataFern.) have been planted (in 1985, populationin a site,the abilityof plantspecies to become 1988, 1989, and 1990) in a programadministeredby the establishedafterdispersalto the site is determinedin part National Park Service (NPS). by thepopulationof AMF in the soil (Miller,1979, 1987; Careful record keeping of the historyof plantingsin Reeves et al., 1979; Janos,1980; Grimeet al., 1987; Read clearly identifiableareas by the NPS in The Province and Birch, 1988; Gemma and Koske, 1990, 1992; Koske Lands Dunes site of CCNS offereda unique opportunity and Gemma, 1990; Gange, Brown, and Sinclair, 1993; to studythe dynamicsof primarysuccession in the earFrancis and Read, 1994, 1995). Stabilizationof large,mobile dune fieldshas concerned liest seral stages (spanning 7 yr) in large plantingsof beachgrass.The purposeof the studiesreportedhere was 1 to documentthe species of AMF presentin thedunes and Manuscriptreceived 5 January 1996; revision accepted 22 July 1996. to identifychanges occurringin the AMF communityas The authorsthankMr. JimKillian and Dr. Charles Roman forlogiswell as theinoculumpotentialof the dune soil duringthe tical support,advice, and encouragementforthestudy;Dr. Eric Roberts, early stages of primarysuccession of the plant commuDr. David Berlinsky,Dr. Noel Jackson,Dr. Paul Godfrey,Dr. Alison nity. Roberts,ChristinLambiase, Rachel Maher,Ronnie LaBarbera, and Tom Bellohusen for assistance; and Dr. Keith Killingbeckfor assistance in the plants.This researchwas supportedby a grantfromthe identifying CooperativePark Studies Unit of the National Park Service. 2 Authorforcorrespondence. MATERIALS AND METHODS Study site-Studies were performedin an area measuring -4 X 1 km in the ProvinceLands Dunes regionof CCNS (70?8.5'W, 42?4.0'N) 118 This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions January 1997] KOSKE near Provincetown,MA. The vegetationand ecological processes occumng in theCape Cod dunes are similarto those in dunes along much* of the NorthAtlanticseaboard of the U.S. (Godfrey,1977; Godfrey, Leathermanand Zaremba, 1979; Woodhouse, 1982). The geology of the area is described by Strahler(1966). The Province Lands Dunes comprise a varietyof dune systemsin various stages of succession. Large mobile dunes up to 32 m high are interspersedamong smaller, more densely vegetatedareas (Strahler,1966). In the mobile dunes,the dominantplant species is A. breviligulata.Early invadersof A. breviligulata-dominatedsites are Polygonella articulata(L.) Meisn. and Solidago sempervirensL. Later successional sites are characterizedby the grass Deschampsia flexuosa (L.) Trin.,the shrubsMyricapensylvanica Loisel. and Prunus maritimaMarsh., the trees Pinus rigida Mill. and Quercus ilicifolia Wang., and otherspecies (Svenson and Pyle, 1979; Whatley,1988). A forestof mixed deciduous trees and shrubs is the most advanced successional stage in the area. Our studiesfocused on areas plantedin 1985, 1988, 1989, and 1990 (Figs. 1-4). Each plantingcovered 12-20 ha, and had been free of vegetationfor >5 yr priorto planting.Vegetationin the planted areas had a highlyregulardistribution.During NPS-supervised plantingon the dunes, threecleaned culms (tillerswith roots and dead leaves removed) of A. breviligulatacv. Cape were insertedin a hole 20 cm deep, and the soil was packed aroundthe culms. Such a groupof threeculms is termeda "hill." Hills in the dunes were planted46 cm apartin rows. The rows run in an east-westdirection(parallel to the shore) and are spaced 46 cm apart.Hills in adjacent rows were staggered.At thecom-* 36 kg/haof a pelletizedIO:10: IOV pletionof eachoftheNPS plantings, 119 IN SAND DUNES AND GEMMA-MYCORRHIZAE . -w. _ a __ t ii iL. j 1 - F ' 't _ "' 0iW+ f.4 ^ 5_aS .2 was add-Y fertilizer were applied by broadcasting.No additionalfertilizer pet of each of th N ed. The ProvinceLands plantingsare located 0.5-1.3 km fromthebeach, and separated fromit by a series of dune ridges. Movement of sand fromthe beaches inland to the plantingsites is uneven,and most sites appear to receive very little sand deposition (Godfrey,Fowler and Stack, 1991), resultingin a gradualdecline in the vigorof theplantings (Eldred and Maun, 1982; Disraeli, 1984; Godfrey,Fowler,and Stack, j'At J4 X *. . i v .. 2 1991). I In addition to the plantingsof 1985-1990, we also sampled some seralstages nearbynaturalareasof thedunes(Sites 1-5) in different forcomparison.These areas were naturallyvegetated,but had not been plantedby the NPS. Site 1 (Fig. 5) was a vigorous stand of A. breviligulata located between NPS plantingareas of 1988 and 1989. This area had been previously identifiedby Godfrey,Fowler, and Stack (1991) as a native stand of beachgrass.Sand accretionwas high at this site,organicmatterwas low, and invadingspecies were veryrare.Sites 2 and 3 (Fig. 6) were dominatedby decliningstandsof A. breviligulata and representedslightlylatersuccessional stages. In additionto A. brev- _ 3 iligulata,areas2 and 3 wereverysparselycolonizedbyAchilleamil- lefoliumL., Aster linariifoliusL., Deschampsia flexuosa, Myrica pensylvanica,Prunus maritima,Polygonella articulata,Quercus ilicifolia, Rosa rugosa Thunb., Solidago nemoralis Ait., S. odora Ait., and S. sempervirens.Site 4 (Fig. 7) had 98% cover by A. breviligulataand numerousvigorouslyinvadingspecies (e.g., Myricapensylvanica,Rosa rugosa,Artemisiaborealis Pall., Deschampsia flexuosa,Daucus carota L., Hieracium canadense Michx., and Kalmia angustifoliaL., Pinus rigida,and Q. ilicifolia).Site 5 was forestedwithAcer rubrum,Quercus spp., Sassafras albidum, and an understoryof shrubs (e.g., Clethera .4 r - s 4 .'_-', 4 S it . 3 ' '4w - . i . 4> A f . j4 r '. -- WK+f:# _ t = - _ Figs. 1-4. Plantingsof Ammophilabreviligulataon ProvinceLands Dunes. 1. 1990 planting(age 1 yr). Sparse area near centerwas planted withone culm per hill insteadof the usual threeper hill. 2. 1989 plant-'Jc ing (age 2 yr). This plantingsufferedextensivewinterkilla few months_i afterplanting.Note manydead individualsin some rows. 3. 1988 plant- ^ - : ing (age 3 yr). 4. 1985 planting(age 6.5 yr). Note low vigorand cover 4,. in thisdecliningstand. 3 - ;Lw trlt This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions 120 AMERICAN JOURNAL OF BOTANY [Vol. 84 Sampling-Sampling for AMF, arbuscularmycorrhizae(AM), vegwas carriedout in 1990 and 1991 using etation,and soil characteristics randomsamplingtechniques(Mueller-Domboisand Ellenberg, stratified 1974; Koske and Gemma, 1992). The most extensive sampling was carriedout in the planted areas thatwere of known age. Sampling of natural areas was less comprehensive,and those data were gathered mainlyforcomparisonto resultsfromthe plantedsites. To ensure thatany changes in vegetativecover, vigor,etc. were not the resultof pathogensor pests, we had Dr. Noel Jackson,a plant pathologistat the Universityof Rhode Island, examine A. breviligulata plantsfromthe NPS plantingsand fromsome naturalareas in November 1990, February 1991, March 1991, and August 1991. He looked specificallyforchinchbugs, rootcystnematodes,stripedsmut,and root aphids. No significantoccurrencesof any of these pests were found. ~~~~~~ . F I -'V 7~ Soil characteristics-Most soil samples were collected in October 1990 forphysicalanalysis. Samples at the 1990 plantingsite were collected a few days beforeplanting(October) and just afterplantingand fertilization(November). The upper 15 cm of sand at each sampling point was scraped away, and a sample (- 1 L) was collected by insertinga clean trowelto a depthof 15 cm and removingthe soil. Thus the sample includedonly soil thatoccurredin the zone between 15 and 30 cm below the originalsoil surface.Soil pH was measuredin a soil: waterslurry(1:2) witha pH meteraftera 20-minequilibration,organic at 400?C for6 h, and theother matterwas measuredby loss-on-ignition parameterswere measured by the Soil Testing Service, Universityof Rhode Island CooperativeExtensionService. AMF community-The AMF communitypresentin the soil at the time of sampling was assessed by directcounts of spores fromeach area in the dune system. Fifteen soil samples were collected on 12 October 1990 fromeach of the four plantingsites (1985-1990) and h'h fromarea 2. The collections fromthe 1990 plantingarea were taken beforethe plantingoccurred.In October 1991 additionalsamples were collected fromthefollowingsites: 1990 plantingarea (12 samples), site '1~~~~~~~~~~~~~~~~~~1 3 (eightsamples), site4 (13 samples), and site 5 (fivesamples). Samples were collected in the fall because spore populations in New England (Gemma and Koske, dunes typicallyreach theirmaximumin fall/winter 1988; Gemma, Koske and Carreiro,1989). Prior to collection of root zone samples, the upper 20-25 cm of :s~~~~~~~~~~~~~~~~~~~~~soil in which few roots occur were removed and discarded. A trowel was thenused to excavate soil (- 1 L) and roots fromthe bottomof the hole. Spores of AMF were recovered fromsoil samples by wet(Walker,Mize, and McNabb, 1982). All the sieving and centrifugation spores containedin an 85-mL soil sample were placed on a 5.5 cm filter paper -aftercentrifugationand then transferredto glass microscope slides. Spores were mountedin a polyvinylalcohol mountingsolution (Koske and Tessier, 1983) asndidentifiedat 400-1000X. Species idenspecimens.Vouchtitieswere confirmedby comparisonto authenticated ers have been preservedin the authors' collections. All spore counts were convertedto numberof spores per 100 mL soil. Figs. 5-7. Naturaldune areas in the ProvinceLands Dunes. 5. Site 1. A vigorous, floweringstand of Ammophilabreviligulatais shown. Lightband at top of dune indicatesthe numerousinflorescences.6. Site 2. Ammophilain decline, some invadingspecies present.7. Site 4. Note high cover.Ammophiladominates,but otherspecies are common. alnifoliaL., Gaylussacia baccata (Wang.) K. Koch). Soil organicmatter and diversitywere high. The naturalsites were ranked 1, 2, 3, 4, and 5 fromearliestto latest in a successional sequence on the basis of soil organic matter,incidence of invadingspecies, and vigor of A. breviligulata (Koske and Gemma, 1992). Mycorrfiizalstatus of plants-The extentof mycorrhizalcolonization in plantingsof A. breviligulataand in invadingspecies of plants was measured.Four fieldsurveyswere carriedout to assess the extent of AM in A. breviligulatain sites of different ages and to determine when plantingsformmycorrhizae.The firsttwo studies included only established plants, and the second two focused on seedlings. On 12 October 1990, fiveroot samples were collected fromeach of the 1985, 1988, and 1989 plantingareas and fromnativearea 2. At each sampling point,a root sample (50-80 cm long) was collected at a depthof 25 cm beneaththe surface.The second surveywas carriedout at the 1990 plantingsite on 18 October 1991 (47 wk afterplanting).Ten rootsamples were collected. In the seedling surveys,tenplants(3-5 cm high) were collectedfrom the 1988 area in October 1990, and ten fromthe 1990 plantingsite in This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions January 1997] KosKE AND GEMMA-MYCORRHIZAE IN SAND DUNES samples were collected June 1992. Because seedlings were infrequent, randomsampling haphazardlyfromthe areas and not by the stratified procedure. Root samples were cleared in hot (65?C) 2.5% KOH and stainedin an acidic glycerolsolutionof trypanblue (Koske and Gemma, 1989). Extentof arbuscularmycorrhizae(AM) developmentin the roots was quantifiedby calculating a MycorrhizalIndex (MI), in which colonization was assessed at 30-60X on a scale of 0-3, where 0 = no colonization,1 = up to 25% of the root lengthcontaininghyphae,arbuscules, vesicles, or coils of AMF, 2 = 25-50% colonized, and 3 = >50% of length colonized. Portions of each root system were observed at 400X. Roots were considered to have functionalvesicular-arbuscular (VA) mycorrhizaeonly if arbusculeswere observed (Gemma and Koske, 1990). To estimatemycorrhizaein invadingspecies, root samples were collected in April 1992 from13 species of plantsinvadingthe 1985, 1988, and 1990 plantingsor occurringin naturalareas 3 and 4. Care was takento ensurethatonly therootsof theinvadingspecies were assessed forcolonization.Invading species were rare in the youngestplantings, and the collectingreflectedthis.Roots were treatedas describedabove. Ectomycorrhizaewere ratedvisually only forpresenceor absence. Mycorrhizalinoculum potential of soils and assessment of the hyphal network-An assay of mycorrhizalinoculum potential (MIP) (Moorman and Reeves, 1979; Gemma and Koske, 1988) was employed to assess the relativepopulationof infectiveunits(i.e., spores,hyphae, pieces of mycorrhizalroots) in soils fromvarious sites. Assay of 1990 plantingarea prior to planting-A MIP assay of the soil fromthe 1990 plantingarea was performedto check forthe presence of viable propagulesof AME Fifteensoil samples were collected randomsamplingplan. At each on 12 October 1990 using a stratified collection point,the upper 15 cm of soil was removed and a 1-L soil sample (- 6 cm diameter)was collected. Soil samples were returned to thegreenhouseat theUniversityof Rhode Island, and 70 mL of each sample were mixed withan equal amountof Terra-Green?,a calcined clay (Oil Dri Corp., Chicago, IL 60611), and the mix was placed in a Cone-tainer?(a tapering,cylindricalplasticgrowthtube,20.5 X 4.0 cm of 150 mL capacity,Stuewe and Sons, Corvallis,OR 97331). Two seeds of Zea mays L. were sown in each of the 15 Cone-tainers,and the materialwas placed in a heated greenhouse(avg. 26?C). Sunlightwas Sodium vapor lamps giving supplementedfor 16 h/dwithhigh-pressure an intensityat theleaf surfaceof 350-1375 p1molphotons.m-2.s-2. After seed germination,Cone-tainerswere thinnedto containa single plant. Plants were wateredas required,but not fertilized,and harvestedafter 60 d. Roots were cleared, stained,and assessed forcolonizationby the method(Giovannettiand Mosse, 1980). line-intercept Assay of soils from vegetatedsites-Undisturbed soil samples (see below) were used for the MIP assays, and disturbedand undisturbed samples were used to estimatethe extentof the hyphalnetworkin the different sites. Soil samples were collected on 3 September1991 from plantingsof A. breviligulata(1985, 1988, 1989, and 1990 sites) and fromsites 1 and 4. Samples were collectedby drivinga PVC (polyvinyl chloride) tube (20 cm long X 5.1 cm inside,diameter;the top of the tube covered witha piece of fiberglasswindow screen glued to it) into the soil. The tube was thencarefullywithdrawn,withthe core of sand remaininginside. Before the tube was drivenintotheground,theupper 8 cm of soil had been removed.Aftertheirremoval fromthe soil, the tubeswere inverted(screennow at bottom)and used as potsforgrowing plantsin the greenhouse(see below). At each of the six collectingsites, fivehills (in the plantedareas) or locations(sites 1 and 4) were selected randomly,and two cores were taken.The distancebetween the center of the plant and the cores was 15-20 cm, so thatportionsof the root zone were included in each sample. The contentsof one memberof each pair of cores were emptied into a plastic bag. When five cores 121 froma site were in the bag, the bag was shaken vigorouslyfor 30 s, back intothefiveoriginalPVC tubes.These and thesoil was transferred shaken samples were the "disturbed" cores. The intact cores were termed"undisturbed."Thus, ten cores were obtainedfromeach of the six sites,and halfof thecores were undisturbedand halfwere disturbed. Disturbed and undisturbedcores were placed on the mist bench in the greenhouse,and one pregerminatedseed of Z. mays was added to each tube. After 1 wk, tubes were removed from the mist bed and transferred to the greenhousebenches and grown as described above. At 8.5 wk aftersowing,the entireroot systemwas removedfromeach tube,cleared,stained,and assessed formycorrhizaeas describedabove. In additionto the standardMIP assay whichmeasuresthe amountof rootcolonization,two otherparameterswere measured.When theplants were harvestedat 8.5 wk, the lengthof the longestleaf in each Conetainerwas recordedand shoot dry mass was determinedby dryingin a 60?C oven for48 h. Statistical methods-Data were analyzed using t tests,chi-squared tests,and one-wayANOVA. Means were separatedwithScheffe'stest. Spore count data and root colonizationdata typicallyare not normally distributed(Sylvia, 1986; St. Johnand Koske, 1988) and were transby adding 0.5 formedpriorto analysis. Spore countswere transformed to the abundance value and takingthe square root of the sum (Little and Hills, 1978), and colonizationvalues underwenta log (x+ 1) transformation(St. Johnand Hunt, 1983). A confidencelevel of P = 0.05 was used to determinestatisticalsignificance. RESULTS Physical characteristics of soils-Soils in all sites were nearlypure quartz sand. With few exceptions,all soils had verylow levels of organicmatter,nitrate,phosphorus,and potassium,and were acidic (Table 1). In the planted areas (of known age), organic matterand age of the stand were positively correlated(organic matter= 0.51 X age in yr + 0.123, r = 0.56; P = 0.02). When all sites were compared (planted and natural),organic matterwas higherin site 5 and nitratewas higherin site 4 thanelsewhere. AMF community-Spores of 17 species of AMF were isolated fromthe 98 samples collected fromroot zones (i.e., excludingthe 12 samples collected fromthe barren 1990 plantingarea beforethe plantinghad been made). Five of the species apparentlyare undescribedand are identifiedby numberand are informallydescribedbelow. Terminology follows that recommended by Walker (1983, 1986) and Morton(1986). No spores were present in the 1990 plantingarea priorto planting.Frequencyof occurrenceand spore abundance of the more important species are shown for each site in Fig. 8. In the notes thatfollow, referenceto the frequencyof occurrenceof particularspecies in all of the samples does not include the 12 samples collected fromthe 1990 plantingarea prior to planting. 1. Acaulospora lacunosa Morton Spores of A. lacunosa were recovered significantly in sites ' the 1985 plantingin age (P = more frequently 0.01) (Fig. 8). They occurredin 15% of all samples. This species is relativelycommonin dunes of theU.S. Atlantic coast fromMassachusettsto Virginia(R. E. Koske, personal observation).Until its descriptionin 1986, spores This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions 122 TABLE [Vol. 84 AMERICAN JOURNAL OF BOTANY 1. Some physicaland vegetationalcharacteristics of samplingsites. Site 1990c 1990c 1989 1988 1985 I 2 3 4 5 Age of planting(yr) 0 1 1 2 5 n.a.' n.a. n.a. n.a. n.a. Cover (%) OEd 33C 15D 34C 26C 59B 28C 25C 98A n.d. (N), pH 2 6 2 2 2 5 2 3 3 3 6.OA 5.4A 5.8A 6.1A 6.2A 5.8A 5.6A 5.8A 5.8A 5.2A NO, (mg/kg) l.5B 2.11B 1.4B 1.5B 1.6B 1.4B 1.4B 3.9B 9.1A 3.0B P (mg/kg) 14.5A 24.5A 13.9A 14.3A 15.1A 22.8A 13.2A 22.6A 30.6A 22.8A K (mg/kg) 5.OA 5.4A 5.OA 5.OA 5.OA 1O.5A 6.3A 13.3A 13.3A 25.8A Organic matter (%)b 0.12B (N = 14) n.d.c 0.12B (N 0.14B (N 0.15B (N 0.13B (N 0.13B (N 0.23B (N 0.38B (N 1.91A (N = = = = = = = = 15) 13) 13) 5) 13) 4) 12) 5) N = no. of samples analyzed forpH, nitrate,phosphorusand potassium. Mean percentageorganic matter.N = no. of samples analyzed fororganic matter. c The 1990 area was plantedand fertilizedin November 1990. It was sampled immediatelyafterand I yr later. d Mean of samples. Values in columns followed by the same uppercase letterdid not differsignificantly. Not determined. Age of the vegetationof the naturalareas was not determined. a b of this species would have been identifiedas A. scrobiculata Trappe. 2. Acaulospora mellea Spain & Schenck Spores of A. mellea were recoveredfrom 16% of all samples. Like A. lacunosa, it was absent fromthe 19881990 plantings,but occurredin all otherareas (Fig. 8). Maximumspore abundanceoccurredin the 1985 planting (avg. 4.9 spores/100mL). 3. Acaulospora scrobiculata Trappe A common species in dunes of the Atlanticcoast of the U.S. including other areas of Cape Cod, A. scrobiculata (e.g., Bergen and Koske, 1984; Koske, 1987; Gemma, Koske, and Carreiro,1989) was recoveredonly fromsite 2 in this study.It was presentin 20% of the samples fromthis site, averaging2.7 spores/100mL. 4. Acaulospora 3106A (Fig. 9) This species is similarto A. lacunosa, but spores possess a more finelydimpled laminatedwall. Spores were foundonly in threeof the samples fromsite C. Spores globose, 72-88 ,um diameter,yellow-brown. Spore wall structure consistingof six walls. Wall 1 membranous,wrinkling,hyaline,0.5 ,umthick,appressed to wall 2. Wall 1 was difficult to observe in mostspecimens, apparentlydisappearingduringexposureto the soil. Wall 2 laminated,yellow, 2.0-3.0 ,umthick,ornamentedwith hemisphericaldepressions2-3 ,umdiameterX 1-2 ,um deep, closely spaced, separatedfromeach otherby ridges - 2 ,umwide. Walls 3 and 4 paired, membranous,hyaline,each 0.5-0.6 ,umthick.Wall 5 membranous,beaded, 0.5-1.0 ,umthick.Wall 6 amorphous,hyaline,attached to wall 5, 1.5-2.0 ,umthick,up to 8 ,umthickin crushed spores, stainingdark red-brownin Melzer's reagent.Circatrix 12 ,umdiameter. Acaulospora 3106A is a a memberof the group of small-sporedAcaulospora species that include A. lacunosa and A. mellea, whichhave a commonwall structure of the laminatedwall (Morton,1994). The ornamentation is sufficientto distinguishA. 3106A fromother orna- quantitiesof mentedspecies in this group,but sufficient materialand stages of developmentwere not available to preparea formaldescription. 5. Acaulospora 7034 (Figs. 10, 11) Spores of this species were common in the midsuccessional sites in our study,and were especiallyabundant in the 1985 plantingand in site 2 (Fig. 8). Spores of A. 7034 were the most frequentlyrecoveredof any species, occurringin 31 % of all the rootzone samples. No spores were isolated fromthe 1-yr-old1990 plantingand sites 4 and 5 (i.e., the youngestand oldest sites). Spores globose to subglobose, 120-160 ,umdiameter, brownish-yellow.Spore wall structurecomplex, consisting of six walls. Walls 1 and 2 evanescent,hyaline,adherentto each otherand to wall 3. Wall 1 0.5-0.8 ,um thick;wall 2 1.5-2.0 ,umthick.Wall 3 laminated,colored, ornamentedby craniformfolds, 5.6-7.5 ,umthick.Wall 4 membranous,wrinkling,hyaline,1.0 ,umthick.Wall 5 coriaceous, hyaline, 1.0-2.2 ,umthick.Wall 6 appressed to wall 5, membranous,hyaline,0.5 ,umthick.Circatrix 12-14 pLmdiameter. Spores of A. 7034 are similar to those of A. rehmii Sieverding & Toro in their surface ornamentation,but differin wall structure. 6. Entrophosporainfrequens(Hall) Ames & Schneider Spores were foundin a single sample fromarea 4. This species has not previously been recovered from sand dunes of the Atlanticcoast of the U.S., although it is known fromdunes in the west coast (Koske and Halvorson, 1989). 7. Gigaspora gigantea (Nicol. & Gerd.) Gerd. & Trappe Spores of G. gigantea were recoveredfromthe 5-yrold site and naturalareas 2, 3, and 4 (Fig. 8). Spores were presentin 18% of all samples. The species is common in A. breviligulata-dominated dunes of the Atlantic coast of NorthAmerica fromQuebec to Virginia(Koske and Halvorson, 1981; Bergen and Koske, 1984; Koske, This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions AllSpeciesofAMF 1990 (O yr) Y 1990(I yr) 1989(19(I 1988(2 yr) 1986(6y198 site2 Zt B X _ ... site 4 site 5 I gm (O yr) _0 It990 (0 yr) y1 B y B Y B Y- B 1990(I yr) yr) B 1988(2 yr) 1985( yr) site2 site3 ------------- B XY XY X site 4 B AB _ 0 XY0 B site 5 |X 0 10 ., Y B Y B - 99(I 198(5(yr) Y B 1990(1 yr) 1989(1 r (2yYr)Y 1 2YrY-B18 g o ^ Y Y B 0., . 10 I1990(Oyr) 1990(1 yr) yr 2Y)Y-B1988 1985(5yr) fe2 i0sRe FrequencyAbundance AB AB margarita Gigaspora AB AB AB site A I 990(Oyr)| 1990(I yr) 1989(1 yr) 1988(1 yr) 1985(5yr) site2 site3 4 site5 Y-A XY XY _. X X X A A A X .. La Y Y B B -B1989 ---AB AB 4 3 ......................... sfte ~~AB I Scutellosporaerythropa calospora Scutellospora 2y- ~~~~~~~~site oRe L , Yt B Y B y B Y-- B Y y 1988(2 yr) 1985(5yr) site2 site3 site4 5 7243 Glomus _ C site 4 8?,?''? I 990(Oyr) 1990(I yr) ,. I1990 (Oyr) X B X B X B X A X AB AB X gigantea Gigaspora Y B 1990(Oyr) Y B 1990( yr) 1989(19 (I B Y; 1988(2 yr) 1985(5yr) ---- y s(te2 yE S B site3 Y B site4 B site YE 0b.. X--B 199 (I yr) 1989(1 yr) 1988(2 yr) 1985(5 yr) site2 site3 site 5 7034 Acaulospora ) meHea Acaulospora lacunosa Acaulospora B ? 'La 8 -~~~~~~~T . 123 KosKE AND GEMMA-MYCORRHIZAE IN SAND DUNES January 1997] 1990(Oyr) 1990(1 yr) (1 yr)----1985(5yr) site2 site45 sRO FrequencyAbundance X Y B - -YA --- FrequencyAbundance Fig. 8. Plots of frequencyof occurrenceand average spore abundance (spores/100mL soil) of the species of AMF presentin ?15% of all are not illustrated.Frequencyand abundanceof all 17 species are combined samples. Nine additionalspecies were recovered,but theirdistributions in the firstgraph. 1987; Gemma and Koske 1988; Gemma,Koske, and Carreiro, 1989; Dalpe, 1989) where it frequentlyis a codominantmemberof the AMF community. 8. Gigaspora margaritaBecker & Hall Spores of G. margaritawere recoveredfrom26% of all samples,includingthe 1-yr-old1989 plantingarea but were not significantlymore abundant at any particular site (Fig. 8). This was the second most frequently recovered species. 9. Glomus clarum Nicol. & Schenck This species was one of theearliestappearingpioneers, showing up in the 1-yr-oldplanting(1989), apparently originatingin theplantingstock(see below). Spores were presentin only 6% of the samples and were absentfrom the 1985 plantingas well as areas 3, 4, and 5. Glomus clarum also is knownfromsand dunes of easternCanada (Dalpe, 1989). 10. G. microaggregatumKoske, Gemma & Olexia Anothercommoninhabitantof coastal dunes of eastern NorthAmerica and theGreatLakes (Koske, Gemma,and Olexia, 1986; Koske, 1987; Koske and Tews, 1987; Dalpe, 1989), G. microaggregatumwas recoveredonly from the more advanced seral stages (sites 3, 4, and 5) in this study,and its abundance never exceeded 2 spores/100 mL. Spores of this species typicallyformwithindead spores of otherAMF, and we tended to findthemmore This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions 124 AMERICAN JOURNALOF BOTANY [Vol. 84 abundantlyin areas of the dunes where the A. breviligulata was in decline and organic matterwas higherthat near the dune crest(personal observation). 11. Glomus 7099 Spores of this undescribed species occurred in one sample fromsite 3 and one fromsite 4. Spores globose to subglobose,60-80 ,umdiameter,hyaline. Spore wall 2-3 pumthick,laminated. Spores surrounded by a hyaline peridium 10-30 ,um thick composed of loosely organized hyaline hyphae. Peridial hyphae 2-5 ,m broad,withwalls up to 1.0 ,umthick.Walls of thickesthyphaeroughened. 12. Glomus 7243 (Fig. 12) ~~~:. ~ ~ -, ^ This was anotherof the species that was limited to more advanced seral stage, and spores were recovered only fromsites 3, 4, and 5 (Fig. 8). In some samples, spores occurredin greatabundance(up to 600 spores/100 mL in a sample fromsite 4). Spores globose to subglobose, 80-90 (-115) ,umdiameter,red-brownto orange-brown.Spore wall relatively thick,8-12 ,umthick,laminated.Attachmenthyphayellow-brown,10-12 ,umwide, constrictedto 4-6 ,umat base of spore. The species is very similar to G. etunicatum Becker & Gerd. but differsin the smaller size of the spores and the more robustattachmenthypha.Spores of G. 7243 occurredonly in the soils withhigh organic matterand frequentlywere found inside dead spores of Gigaspora and Scutellospora spp. ^ 13. Scutellospora calospora (Nicol. & Gerd.) Walker& Sanders 12 4X! Fg. 91. Spors o nae AFfo roic ad Dunes 9. Surface of spore of Acaulospora 3 106A. Bar = 40 p.m. 10. Surface ofsoeo cuopr 04 oefn-txueoranainad Spores of S. calospora were presentonly in the older sites (i.e., the 1985 plantingand in all of the naturalareas) (Fig. 8). Highestabundance and frequencyoccurred in site 4. The species was recovered from 19% of all samples. Scutellospora calospora is a common memberof the dune communityof the Atlanticcoast of NorthAmerica and theGreatLakes (Koske and Halvorson,1981; Bergen and Koske, 1984; Koske, 1987; Gemma, Koske, and Carreiro,1989; Gemma and Koske 1988; Dalpe, 1989). 14. S. erythropa(Koske & Walker) Walker& Sanders Spores of S. erythropaoccurredin sites rangingfrom the 1-yr-oldplantingof 1990 to site 5 (Fig. 8). It seldom was presentin abundance,but was a consistentmember of most sites. It was found in 16% of all samples. This is anothercommon member of the sand dunes of the Atlanticcoast of the U.S. prominentcircatrix("C"). Bar = 60 ,um. 11. Spore of Acaulospora 7034. Walls 3-6 are indicated. Walls 1 and 2 are not visible in this photograph.Bar = 30 ,um. 12. Spores of Glomus 7243. Note thick laminatedwall of spore and stoutattachmenthypha.Bar = 50 ,um. This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions January 1997] KOSKE AND GEMMA-MYCORRHIZAE 1990 (Oyr) D ffi 1988(2yrs) A G 0 60 BC 1985 (5 yrs) BC site2 B site3 so 40t 0 INA site4 site5 100 80 CD CD 1989(1yr) 0L o ? CD 1990(1yr) 125 IN SAND DUNES I BC 16 14 12 10 8 6 4 2 0 Total Richness 1 2 E06 3 4 5 6 Average Richness Fig. 13. Species richnessin different successional sites. "Total richness" indicates the total numberof species recoveredfromeach site, and "average richness" is a measureof the average numberof species in each sample. The 1990 plantingwas sampled beforeplantingand 1 yr later. 15. S. pellucida (Nicol. & Schenck) Walker& Sanders Spores of S. pellucida occurredin the 1-yr-old1990 plantingand in sites 3 and 4. It was recoveredfrom6% of all samples and never in abundance. It is a common species in maritimedunes of the easternU. S. and Canada. 16. S. persica (Koske & Walker) Walker& Sanders This species was isolated only fromsites 3 and 4. It occurredin 7% of all samples. Scutellospora persica is anothercommonmemberof theAMF communityin sand dunes of the easterncoast of the U.S. 17. S. reticulata(Koske, Miller & Walker) Walker& Sanders Spores of this species were isolated only fromfourof the 13 soil samples collected at site 4. Its abundancewas low (avg. 0.6 spores/100mL at this site). Scutellospora reticulata is an infrequentlyrecovered species in U.S. dunes (Koske, Miller, and Walker,1983; Koske, 1987), and its spores seldom occur in abundance. Spores of S. reticulatatypicallyare foundin dunes of the NorthAtlantic coast of the U.S. where organic matteris higher thanin accretingareas (personal observation). Increases in the richness and complexityof the AM fungal communityaccompanied the development and succession in the plant community.Spores were absent priorto planting,but were recoveredfromareas thathad been vegetatedfor <1 yr (Fig. 8). The species richness (totalnumberof AMF species presentin a site) increased with increasingstabilizationof the dunes and then declined in theoak/mapleforest(Fig. 13). Average richness per sample at differentsites showed a similar pattern. Maximum richnessoccurredin site 4, where spores of 14 species were isolated. In some individual samples fromsite 4, spores of nine species of AMF were present. B 1 0 80-J 2 3 4 B 5 4 0 0 C') 06 0 1 2 3 4 5 Ageofplanting (yr) Fig. 14. Changes in the populations of AMF in association with plantingsof Ammophilabreviligulataover a 5-yr-period.The percentage of samples fromwhich spores were recovered(A) showed a significantlinear trendover the time period (r = 0.996, P = 0.003), and sporesabundance(B) showed a significant exponentialtrend(r = 0.964, P = 0.008). Average species richnessin site 4 was 5.3, a value significantlyhigherthan values fromothersites. Spores of only five species of AMF were isolated fromthe 1-yrold sites. In the planted areas (for which the age was known), the average species richness per sample was highlycorrelatedwith age of the planting(r = 0.699, P = 0.0001), as was the percentageof samples thatcontainedspores of AMF (r = 0.996, P = 0.003) (Fig. 14A). The abundanceof spores of AMF in the different sites (Fig. 8) generally mirroredthe species richness data. Youngest sites (1990, 1989, and 1988 plantings)had the fewestspores per 100 mL, and site 4 had themostspores (avg. 208.9 spores/100mL). In the planted areas, spore abundance was significantlycorrelated with age (r = 0.964, P = 0.008) (Fig. 14B). Based upon when their spores were firstrecovered fromthe sites,two groupsof AMF were identified.Early invaders(present1-2 yrafterplanting)includedA. 7034, GI. clarum,Gi. margarita,S. erythropa,and S. pellucida. Spores of the second group of species were isolated only fromlater successional sites (the 1985 plantingarea or older). This second groupincluded the majorityof AMF species, includingA. lacunosa, A. mellea, Gi. gigantea, GI. 7243, and S. calospora. Scutellospora erythropawas unique in sporulatingin all seral stages. This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions 126 TABLE 2. aged stands. Mycorrhizaein Ammophilarootsfromdifferent Mycorrhizal index Percentage of samples mycorrhizal Site and age of planting Plant type N 1990 (47 wk) 1990 (1.5 yr) Mature Seedling 9 10 I.OCa O.OD 78A OB 1989 (I yr) 1988 (2 yr) Mature Mature Seedling 5 5 10 1.2B 1.9A O.OD 80AB IOOA OB 1985 (5 yr) Naturalarea 2 (age unknown) Mature Mature 5 5 1.8A 2.OA IOOA IOOA a Values in columns followed by the same uppercase letterdid not differsignificantly. Mycorrhizalstatus of plantsA. breviligulata-Most plants became mycorrhizalin the 1st yr afterplanting,and all matureplantswere mycorrhizawithin2 yrafterplanting(Table 2). Greatestcolonizationoccurredin the 1985 and 1988 plantingsand in site 2 and least in the 1990 area (sampled 47 wk after planting). No mycorrhizaewere presentin any of the seedlings. Invadingspecies-The incidenceof mycorrhizaein invading species appeared to increase withsuccession,but numberof samples was recoveredto idenan insufficient tifytrendsin theplantedareas (Table 3). The majorityof species and specimensin naturalareas 3 and 4 were myTABLE [Vol. 84 AMERICAN JOURNAL OF BOTANY 3. corrhizal,and halfof thesix species fromthe 1985 planting were mycorrhizal.In the 1988 site,only one species, the nonmycorrhizalPolygonella articulata, was found. No invaderswere located in the 1989 area, and a single, nonmycorrhizalseedling of Prunus maritimawith red, puckeredleaves was presentin the 1.5-yr-old1990 planting. Mycorrhizal inoculum potential of soils-No AM were observed in any of the 15 roots systemsexamined from the MIP assays set up using soil fromthe 1990 plantingarea priorto planting,indicatingthatviable propagules of AMF were not presentin the samplingarea. In the planted sites the MIP of dune soils showed a significantlogarithmictrend(ln[x+ 1] = -0.308 + 0.526 x age in yr; r = 0.701, P = 0.0006), increasingwithage (Fig. 15A). Greatest root colonization occurred in the samples fromsite 4. Variationwas greaterforMIPs from soil collected fromyoungestplantedareas thanfromthe older, more established sites, indicatinga very uneven distributionof AMF propagules in the former.For example, rootcolonizationvalues in the fiveMIPs fromthe 1988 plantingsitewere 7, 52, 39, 4, and 5%, while values fromsite 4 were 33, 32, 30, 33, and 40%. The average MIP from the native areas (26.4%) was significantly higher(t test,P = 0.002) than the average MIP (8.6%) of the 1985-1990 plantingsites. Incidence of mycorrhizaein seedlingsof plant species invadingareas of the ProvinceLands dunes collected April 1992.' Naturalareas Plantingsof Ainit)ophzila Species 1990 (1.4 yr)b 1988 (3.4 yr) Asteraceae Achillea millefolium Artemisiaborealis Asterlinariifolius Solidago sempervirens Solidago odora Solidago nemoralis 1985 (7 yr) 0/2 0/2 2/2 1/3 Myricaceae Myrica pensylvanica Poaceae Deschampsia flexuosa 3/5 Area 3 2/4 1/5 2/2 0/1 0/1 2/3 5/8 3/3 Polygonaceae Polygonella articulata 0/4 Fagaceae Quercus ilicifolia Totals specimensc speciesc O/lAB 0/lA 0/4B 0/lA 6/15B 3/6A 3/3 0/5 0/1 0/ld All sites 0/3 0/2 2/4 1/7 4/4 1/3 0/3c Pinaceae Pinus rigida Rosaceae Prunus maritima Rosa rugosa Area 4 3/3 1/1 1/1 3/4 2/2 3/3 3/3 6/6 14/25AB 7/9A 7/7A 3/3A Mycorrhizae,if present,were of the arbusculartypeexcept in Quercus and Pinus, which had ectomycorrhizae. Age of plantingsite at time thatsamples were collected. C Firstvalue indicateshow many samples possessed mycorrhizae;second indicatesno. of samples examined at a site. dThis seedling had red, distortedleaves and appeared stunted. c Values in rows followedby the same uppercase letterdid not differsignificantly. This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions 27/52 9/13 January1997] KosKE AND GEMMA-MYCORRHIZAE A 1990C 1989C . ~root 1988 1985_ | site 1 0 10 ~~~ colonization 5; 2.5.$l$ X 2.|* 1990_ 30 40 _.._. ..........................._.. la length ......... . ...... 50 Fig. 0....... Efec 0 desrucio we ..... 00L a ' the 0 hypalnewok.n. 0ML oo. t0 .. clo Site ~~~~B 1989 _ _....._............_..... ........ shootmass *~0.5...... 20 ofroot Percentage length colonized C,) 3.5 2 site 4 127 IN SAND DUNES Fig. 16. Effectof destructionof the hyphalnetworkon root colonizationand growthof Zea mays plants grownin soils fromdifferent seral stages of the dunes. Y-axis values were calculated by dividingthe response of plants grownin undisturbedsoil cores by the response in disturbedcores. Values > 1.0 are indicativeof a hyphalnetworkpresent in undisturbedcores. Bars markedby a starindicatesignificantdifferences betweencolonizationor growthin disturbedand undisturbed samples. 198 site t site 4 0 0.2 0.4 0.6 0.8 1 1.2 Shoot mass(g) dry in disturbedsoil cores. Disturbance of cores from the plantedareas did notresultin significant reductionin root colonizationor growthof plants. DISCUSSION 1990 AMF are common membersof the biota of sand dune systemsthroughoutthe world where theyappear to play an important role in thedevelopmentof plantcommunity structure(e.g., Nicolson, 1960; Koske and Polson, 1984; 1988 Puppi and Riess, 1987; Dalpe, 1989; Koske and Gemma, 1985 1996). The simultaneousarrivalof AMF and plant colonists in primarysuccessional sites, representedby the site 1 planted areas in this study,allows the fungito interact with the developing plant communityfromthe earliest site 4 seral stage (Nicolson, 1960; Sylvia and Will, 1988; Gemma and Koske, 1989, 1992; Koske and Gemma, 1990). 0 5 10 15 20 25 30 35 40 In the Cape Cod dunes, most measures of mycorrhizal presence (e.g., spore abundance, species richness,root oflongest leaf(cm) Length colonization,and MIP) increased fromthe 1st yr in the plantedareas, and these increases generallycontinuedin Fig. 15. Measurementsof MIP fromundisturbedsoil cores from the later successional sites representedby naturallycolsites differing in successional stage. Cores were collectedin September 1991. Three parameterswere measured: root colonization,shoot dry onized areas 2-4. These increases in the early-and midmass, and lengthof the longest leaf. Narrow bars representstandard successional sites and theirsubsequentdecline in thefordeviation. Note great variationin the data fromthe youngestplanted ested site agree with previous studies of sand dunes in areas. Bars in 15A sharinga commonletterdid not differsignificantly. Scotland (Nicolson, 1960), the Pacific coast of the U.S. No significantdifferenceswere detectedin the data forFig. 15 B, C. (Rose, 1988), and Mexico (L. Corkidi and E. Rincon, Universityof Mexico, personalcommunication). Most of the species of AMF occurringin the Province Shoot growthand leaf lengthwere not correlatedwith the age of the site in the plantedsites,and no significant Lands have been previouslyreportedfromsand dunes of the mid-Atlanticcoast of the U.S. (e.g., Koske, 1987). differenceswere noted between growth made in soil Values for species richnessand spore abundance in the cores fromnaturalsites and plantedsites. (Fig. 15B, C). natural areas and the 1985 plantingsite are similar to Extent of the hyphal network-Soil disturbancerevalues reportedin earlier surveys of AMF in naturally duced the MIP of the soil cores fromsites 1 and 4, invegetateddunes on or adjacentto Cape Cod (Bergen and dicatingthatsignificanthyphalnetworkshad developed Koske, 1984; Gemma, Koske, and Carreiro,1989). only in these sites (Fig. 16). Similarreductionswere obThe identification of early-and later sporulatingspeserved in shoot mass and the lengthof the longest leaf cies of AMF in the dune sites suggestedan apparentsuc1989 .C This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions 128 AMERICAN JOURNAL OF BOTANY cession of AME While thismay be, the evidence is confoundedby the possibilitythat not all species of AMF presentin the roots had sporulatedat the time of collecof theoriginof thepioneerAME tion and by uncertainty Three of the fivespecies isolated fromtheyoungestsites (Gl. clarum, S. erythropa,and S. pellucida) apparently arrivedon theA. breviligulataplantingstock(Koske and Gemma, 1992), a common occurrencewithrhizomatous dune pioneers (Sylvia and Will, 1988; Gemma and Koske, 1989, 1992; Koske and Gemma, 1990). Because the plantingstock was raised in a nurseryin sandy agricultural soil (Church's Garden Center,Cape May, NJ), it cannotbe concluded thattheAMF species presenton the stock are necessarilypioneer species, althoughtheyare capable of sporulationin pioneer sites. Further,the primary vegetative succession on the plantedsites differsconsiderablyfromnaturalsuccession occurringin sand dunes that are not managed (Hewett, 1970; Ranwell, 1972). The very rapid and uniformcolonization of '12 ha of bare sand by a a single pioneer species thatresultsfroma plantingof beachgrasshas no parallel in nature.Thus, the rate of developmentand the complexityof the AMF communitydescribed in this studymay differfromthatoccurringin a naturalsuccession. The formationof the AMF communityis closely linked to changes in the aboveground vegetation.In a varietyof naturalsites,AMF have been foundto play a vital role in plant succession, affectingthe success not only of obligatelymycotrophicspecies, but also of those that are facultativelymycotrophicor nonmycotrophic. Essentially,obligatelymycotrophicplantspecies are prevented fromcolonizing a site if the fungiare absent or if theirabundance is below a criticallevel (Miller, 1979, 1987; Reeves et al., 1979; Janos, 1980; Grime et al., 1987; Read and Birch, 1988; Gemma and Koske, 1990, 1992; Koske and Gemma, 1990; Miller and Jastrow, 1992; Gange, Brown, and Sinclair, 1993). The stunted seedlingof Prunusmaritima(an obligatemycotroph[Koske and Gemma, 1992]) that was found in the 1990 CCNS plantingsite illustratesthis phenomenon.Some nonmycotrophicand facultativelymycotrophicspecies may be activelyexcluded when theinoculumpotentialis greaterthan a minimal level (Francis and Read, 1994, 1995). It is the hyphalnetworkof AMF thatfirstexertsthese beneficialor antagonisticeffectson the seedlings of potentialinvaders.Studies in pasturesoils and mine spoils have shown that the hyphal networkis responsiblefor the preponderanceof initialcontactswithroots of seedlings (Evans and Miller,1988; Fairchildand Miller,1988; Read and Birch,1988; Jasper,Abbottand Robson, 1989a, b; Gange, Brown, and Sinclair,1993; Francis and Read, 1994, 1995). A similarrole for the hyphal networkappears to exist in the succession on the dunes of CCNS. Of the two sites (1 and 4) in the CCNS studyarea that had significanthyphal networks,widespread invasion was occurringonly in the later succesional site (site 4). However,as the failureof plantsto invade site 1 shows, the formationof a hyphalnetworkis but one factorthat is necessaryduringsuccession to permitcolonizationby nonpioneer,obligatelymycotrophicplant species. Other site changes accompanyingsuccession in dunes include [Vol. 84 improved soil conditions (e.g., soil structure,organic and a largerpopulationof soil mimatter,soil nutrients, croorganisms) and an ameliorated microclimate(e.g., Oosting, 1954; Olson, 1958; Koske, Sutton,and Sheppard, 1975; Baldwin and Maun, 1983; Rose, 1988), and to the high these improvementscontributesignificantly plant diversityin site 4. The lack of invadingspecies in site 1 probablyresultsfromthe high rate of sand accretion in this area. Few plant species can tolerateburialby sand, and rapidly accretingdune sites (such as site 1) typicallyhave impoverishedsoils withvery low organic matterand are dominatedby a single species (Ranwell, 1972). The extentof of hyphalnetworksestimatedfromthe "undisturbed"cores in the presentstudymay have been underestimatedbecause of possible damage to the network when the cores were collected. However, the differencein root colonization between disturbedand undisturbedsamples fromsite4 were verysimilarto reports fromothersoils (e.g., Read and Birch 1987; Evans and Miller,1988), suggestingthatlittledisruptionoccurredin the undisturbedcores. The lack of AM in seedlingsof A. breviligulatacollectedfrom1- and 2-yr-oldplantingsites also indicates that no significantnetworkis presentin very young sites. All threeparametersused to assess the extent of the hyphal network (viz., root colonization, shootdrymass, and lengthof longestleaf) producedsimilar results.The lattertwo techniquespermita morerapid study. assessmentand warrantfurther In contrastto the slow rate of invasion by obligately mycotrophicspecies in highlydisturbedsecondary successional sites where the AMF populationhas been destroyed(Miller, 1979, 1987; Reeves et al., 1979; Allen, 1988; Miller and Jastrow,1992), dune sites are invaded relativelyquickly.The differenceis explained by thecodispersal of AMF with the plantingor colonizing stock (Gemma and Koske, 1989, 1992; Koske and Gemma, 1990) and the resultantrapid increase in populationsof spores and otherinocula of AME In the Cape Cod dunes, some of the earliestinvading species of stands of A. breviligulata are incapable of formingmycorrhizae(e.g., Polygonella articulata) or do not require mycorrhizaebut may benefitfromthe symbiosis if the fungiare present(e.g., Solidago sempervirens, Deschampsia flexuosa,Achillea millefolium)(Koske and Gemma, 1992). An importantfunctionsuch facultativelymycotrophicspecies may be in maintainingand increasingthe MIP in the dunes as the A. breviligulata stands lose vigor. These species can establish in areas where thereis a minimalhyphalnetwork(as seen in the older,planted areas). In time, theirroots can supporta populationof AMF, paving the way forlaterinvasion by plant species thatrequiremycorrhizae.The role of such mycorrhizal"facilitators"in secondarysuccessionalsites has been describedpreviously(e.g., Miller, 1979, 1987; Reeves et al., 1979; Janos,1980; Allen, 1988; Miller and Jastrow,1992). The great variationin the MIP data fromthe planted areas reflectedthe very uneven (aggregated)distribution of AMF propagules(e.g., Sylvia, 1986; Tews and Koske, 1986; St. Johnand Koske, 1988) in these young soils. limitthe numberof sites will further Such a distribution This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions January 1997] KOSKE AND GEMMA-MYCORRHIZAE thatare amenable to invasion by plantsthatrequireAM (Allen, 1991). Results fromthis studyagree withMiller's (1985) observationthatrestorationof the abovegroundplantcommunitymustincluderestoration of thebelowgroundcomponentsof the site. In primarysuccession in sand dunes AMF fungiare an importantcomponentof the complex belowgroundsystem. LITERATURE CITED A. K., AND F B. REEVES. 1995. Inland sand dune mycorrhizae: effectsof soil depth,moisture,and pH on colonizationof Oryzopsishymenoides.Mycologia 87: 54-60. ALLEN, M. F 1988. Re-establishment of VA mycorrhizasfollowing severe disturbance:comparativepatch dynamicsof a shrubdesert and a subalpine volcano. Proceedings of the Royal Societyof Edinburgh94B; 63-71. 1991. The ecology of mycorrhizae.Cambridge University Press, Cambridge. BALDWIN, K. A., AND M. A. MAUN. 1983. Microenvironment of Lake Huron sand dunes. Canadian Journalof Botany61: 241-255. BERGEN, M., AND R. E. KosKE. 1984. V-A mycorrhizal fungifromsand dunes of Cape Cod, Massachusetts.Transactionsof theBritishMycological Society 83: 157-158. DALPE, Y. 1989. Inventaireet repartition de la floreendomycorrhizienne de dunes et de rivages maritimesdu Quebec, du NouveauBrunswicket de La Nouvell&Ecosse. Le NaturalisteCanadien 16: 219-236. DISRAELI, D. J. 1984. The effectof sand deposits on the growthand morphologyof Ammophilabreviligulata.Journal of Ecology 72: 145-154. ELDRED, R. A., AND M. A. MAUN. 1982. A multivariateapproach to the problemof decline in vigourof Ammophila.Canadian Journal of Botany60: 1371-1380. EVANS, D. G., AND M. H. MILLER. 1988. Vesicular-arbuscular mycorrhizas and the soil-disturbance-induced reductionof nutrientabsorptionin maize. I. causal relations.New Phytologist110: 67-74. FAIRCHILD, G. L., AND M. H. MILLER. 1988. Vesicular-arbuscular mycorrhizasand the soil-disturbance-induced reductionof nutrient absorptionin maize. I. developmentof the effect.New Phytologist 110: 75-84. 1994. The contributionof mycorrhizal FRANCIS, R., AND D. J. READ. of plantcommunitystructure. fungito the determination Plant and Soil 159: 11-25. . 1995. Mutualismand antagonismin the mycor, AND rhizal symbioses,withspecial referenceto impactson plant communitystructure.Canadian Journalof Botany73(Suppl.): S1301S1 309. GANGE, A. C., V. K. BROWN, AND G. S. SINCLAIR. 1993. Vesiculararbuscularmycorrhizalfungi: a determinantof plant community structurein early succession. Functional Ecology 7: 616-622. GEMMA, J. N., AND R. E. KOSKE. 1988. Seasonal variationin spore abundance and dormancyof Gigaspora gigantea and in mycorrhizal inoculumpotentialof a dune soil. Mycologia 80: 211-216. . 1989. Field inoculationof Americanbeachgrass , AND (Ammophilabreviligulata)withVA mycorrhizalfungi.Journalof EnvironmentalManagement29: 173-182 . 1992. Are mycorrhizas presentin earlystages of , AND primarysuccession? In D. R. Read, D. H. Lewis, A. H. Fitter,and I. J. Alexander [eds.], Proceedingsof Third European Symposium on Mycorrhizas:Mycorrhizasin Ecosystems, 183-189. CAB International,Oxon. , AND M. M. CARREIRO. 1989. Seasonal dynamicsof ,1 fivespecies of VA fungiin a sand dune. Mycological Research 94: 27-32. GIOVANNErlI, M. 1985. Seasonal variationsof vesicular-arbuscular mycorrhizas and endogonaceous spores in a maritimesand dune. Transactionsof the BritishMycological Society 84: 679-684. , AND B. MOSSE. 1980. An evaluation of techniquesfor measuring vesicular arbuscular mycorrhizalinfectionin roots. New Phytologist84: 489-500. AL-AGELY, IN SAND DUNES 129 GODFREY, P. J. 1977. Climate,plantresponseand developmentof dunes on the barrierbeaches of the U.S. east coast. InternationalJournal of Biometeorology21: 203-215. , A. M. FOWLER, AND L. STACK. 1991. Dune revegetationmonitoring/CapeCod National Seashore. Universityof Massachusetts/ NPS CSU. Amherst,MA. , S. P. LEATHERMAN,AND R. ZAREMBA. 1979. A geobotanical approach to classificationof barrierbeach systems.In S. P Leatherman [ed.], Barrierislands fromthe Gulf of St. Lawrence to the Gulf of Mexico, 99-126. Academic Press, New York, NY. GRIME, J. P., J. M. L. MACKEY, S. H. HILLIER, AND D. J. READ. 1987. Floristic diversityin a model systemusing experimentalmicrocosms. Nature 328: 420-422. HEWEFr,D. G. 1970. The colonizationof sand dunes afterstabilization withmarramgrass (Ammophilaarenaria). Journalof Ecology 58: 653-668. JAGSCHITZ,J. A., AND R. S. BELL. 1966. Americanbeachgrass(establishment-fertilization-seeding). Bulletin 383. AgriculturalExperimentStation,Universityof Rhode Island, Kingston,RI. JANOS,D. P 1980. Mycorrhizaeinfluencetropicalsuccession. Biotropica 12: 56-64. JASPER,D. A., L. K. ABBOTrr,AND A. D. ROBSON. 1989a. Soil disturbance reduces the infectivityof externalhyphae of vesicular-arbuscularmycorrhizalfungi.New Phytologist112: 93-99. . 1989b. Hyphae of a vesicular-arbuscular , AND 11 in dry soil, except when mycorrhizalfungusmaintaininfectivity the soil is disturbed.New Phytologist112: 101-107. JEHNE,W., AND C. H. THOMPSON. 1981. Endomycorrhizaein plantcolonizationon coastal sand dunes at Cooloola, Queensland. Australian Journalof Ecology 6: 221-230. KOSKE,R. E. 1987. Distributionof VA mycorrhizalfungialong a latitudinaltemperature gradient.Mycologia 79: 55-68. , AND J. N. GEMMA. 1989. A modified procedure for staining roots to detectV-A mycorrhizas.Mycological Research 92: 486488. . 1990. VA mycorrhizaein vegetationof the Ha, AND waiian coastal strand:evidence forcodispersalof fungiand plants. AmericanJournalof Botany77: 466-474. . 1992. Restorationof early and late successional , AND dune communitiesat ProvinceLands, Cape Cod NationalSeashore. Tech. Rep. NPS/NARURI/NRTR-92/03.Coop. NPS Studies Unit, Universityof Rhode Island, NarragansettBay Campus. Narragansett,RI. . 1996. Arbuscular-mycorrhizal , AND fungiin Hawaiian sand dunes: Island of Kauai. Pacific Science 50: 36-45. , AND P D. OLEXIA. 1986. Glomus microaggregatum, 11 a new species in the Endogonaceae. Mycotaxon26: 125-132. , AND W. L. HALVORSON. 1981. Ecological studiesof vesiculararbuscularmycorrhizaein a barriersand dune. Canadian Journal of Botany59: 1413-1422. . 1989. Mycorrhizalassociationsof selected plant , AND species fromSan Miguel Island, Channel Islands National Park, California.Pacific Science 43: 32-40. , D. D. MILLER, AND C. WALKER. 1983. Gigaspora reticulata: a newlydescribedendomycorrhizal fungusfromNew England.Mycotaxon 16: 429-435. , AND W. R. POLSON. 1984. Are VA mycorrhizae required for sand dune stabilization?BioScience 34: 420-424. , J. C. SurroN, AND B. R. SHEPPARD. 1975. Ecology of Endogone in Lake Huron sand dunes. Canadian Journalof Botany 53: 87-93. , AND B. TESSIER. 1983. A convenient, permanent slide mounting medium.Mycological SocietyofAmericaNewsletter34(2): 59. , AND L. L. TEWS. 1987. Vesicular-arbuscular mycorrhizalfungi of Wisconsinsandy soils. Mycologia 79: 901-905. John LITTLE,T. M., AND F. J.HILLS. 1978. Agricultural experimentation. Wiley & Sons, New York, NY. MCCAFFREY, C., AND S. P LEATHERMAN. 1979. Historical land use practices and dune instabilityin the Province Lands. In S. P Leatherman [ed.], Environmentalgeologic guide to Cape Cod National Seashore, 207-219. Universityof Massachusetts. National Park Service CooperativeResearch Unit,Amherst,MA. MILLER,R. M. 1979. Some occurrencesof vesicular-arbuscular my- This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions 130 AMERICAN JOURNAL OF BOTANY corrhiza in naturaland disturbedecosystemsof the Red Desert. Canadian Journalof Botany 57: 619-623. 1985. Mycorrhizae.RestorationManagementNotes 3: 14-20. . 1987. Mycorrhizaeand succession. In W. R. Jordan,M. E. Gilpin,and J. D. Aber [eds.], Restorationecology,205-219. CambridgeUniversityPress, Cambridge. , AND J. D. JASTROW. 1992. The applicationof VA mycorrhizae to ecosystemrestorationand reclamation.In M. F Allen [ed.], Myprocess,438-467. corrhizalfunctioning: an integrative plant-fungal Chapman and Hall, New York, NY. 1979. The role of endomycorrhizae MOORMAN, T., AND F B. REEVES. in revegetationpractices in the semi-aridwest. II. a bioassay to popdeterminethe effectof land disturbanceon endomycorrhizal ulation.AmericanJournalof Botany 66: 14-18. MORTON, J. B. 1986. Three new species of Acaulospora (Endogonalow-pH soils in West Virginia.Mycolceae) fromhigh-aluminum, ogia 78: 641-648. . 1994. A problemgroup of five species in Acaulospora. INVAM Newsletter4(2): 8. MUELLER-DoMBoIs, D., AND H. ELLENBERG. 1974. Aims and methods of vegetationecology. JohnWiley & Sons, New York,NY. NIcOLSON, T H. 1960. Mycorrhizain the Gramineae.II. development in different habitats,particularlysand dunes. Transactionsof the BritishMycological Society43: 132-145. OLSON, J. S. 1958. Rates of succession and soil changes on southern Lake Michigan sand dunes. Botanical Gazette(Chicago) 119: 125170. OOSTING, H. J. 1954. Ecological processes and vegetationof the maritime strandin the southeasternUnited States. Botanical Reviews 20: 226-262. PuppI,G., AND S. RIESS. 1987. Role and ecology of VA mycorrhizae in sand dunes. AngewandteBotanik 61: 115-126. RANWELL, D. S. 1972. The ecology of salt marshes and sand dunes. Chapman and Hall, London. cyclingin sand dune ecoREAD, D. J. 1989. Mycorrhizasand nutrient systems.Proceeding of the Royal Society of Edinburgh96B: 89110. , AND C. P. D. BIRCH. 1988. The effectsand implicationsof disturbanceof mycorrhizalmycelial systems.Proceeding of the Royal Societyof Edinburgh94B: 13-24. REEVES, F B., D. WAGNER, T. MOORMAN, AND J. KIEL. 1979. The role in revegetationpracticesin the semi-aridwest. of endomycorrhizae [Vol. 84 I. A comparisonof incidenceof mycorrhizaein severelydisturbed AmericanJournalof Botany 66: 6-13. vs. naturalenvironments. ROSE, S. L. 1988. Above and belowgroundcommunitydevelopment in a marinesand dune ecosystem.Plant and Soil 109: 215-226. ST. JOHN, T. V., AND H. W. HUNT. 1983. Frequency distributionsof mycorrhizalinfectiondata and theirstatisticalanalysis. Plant and Soil 73: 307-313. of endogonaceous , AND R. E. KosKE. 1988. Statisticaltreatment spore counts. Transactionsof the BritishMycological Society 91: 117-121. STRAHLER,A. N. 1966. A geologist's view of Cape Cod. NaturalHistoryPress, Garden City,NY. STURMER, S. L., AND M. M. BELLEI. 1994. Compositionand seasonal variationof spore populationsof arbuscularmycorrhizalfungiin dune soils on the island of Santa Catarina,Brazil. Canadian Journal of Botany72: 359-363. SVENSON, H. K., AND R. W. PYLE. 1979. The floraof Cape Cod. Cape Cod Museum of NaturalHistory,West Brewster,MA. SYLVIA, D. M. 1986. Spatial and temporaldistributionof vesiculararbuscularmycorrhizalfungiassociated with Uniola paniculata in Florida foredunes.Mycologia 78: 728-733. , AND M. E. WILL. 1988. Establishmentof vesicular-arbuscular mycorrhizalfungiand othermicroorganismson a beach replenishmentsite in Florida. Applied and EnvironmentalMicrobiology54: 348-352. TEWS, L. L., AND R. E. KosKE. 1986. Toward a samplingstrategyfor vesicular-arbuscular mycorrhizas.Transactionsof the BritishMycological Society87: 353-358. THOREAU,H. D. 1865. Cape Cod. Ticknorand Fields, Boston, MA. WALKER,C. 1983. Taxonomic conceptsin the Endogonaceae. I. Spore wall characteristicsin species descriptions.Mycotaxon 18: 443455. . 1986. Taxonomic concepts in the Endogonaceae. II. A fifth morphologicalwall typein endogonaceous spores. Mycotaxon25: 95-99. , C. W. MIZE, AND H. S. McNABB. 1982. Populationsof endogonaceous fungiat two locationsin centralIowa. Canadian Journal of Botany 60: 2518-2529. WHATLEY, M. E. 1988. Common trailsideplantsof Cape Cod National Seashore. EasternNational Park and MonumentAssociation,Eastham,MA. WOODHOUSE, W. W. 1982. Coastal sand dunes of the U.S. In R. R. Lewis [ed.], Creationand restorationof coastal plantcommunities, 1-44. CRC Press, Boca Raton, FL. This content downloaded from 128.193.8.24 on Fri, 19 Jul 2013 16:33:39 PM All use subject to JSTOR Terms and Conditions