lable at ScienceDirect

Journal of Arid Environments 111 (2014) 91e97

Contents lists avai

Journal of Arid Environments

journal homepage wwwelsevier comlocate jar idenv

Soil sterilization alters interactions between the native grass Boutelouagracilis and invasive Bromus tectorum

Taraneh M Emam a Erin K Espeland b Matthew J Rinella c

a Department of Plant Sciences University of California Davis Mail Stop 1 One Shields Ave Davis CA 95616 USAb Pest Management Research Unit US Department of Agriculture Agricultural Research Service 1500 N Central Avenue Sidney MT 59270 USAc US Department of Agriculture Agricultural Research Service 243 Fort Keogh Road Miles City MT 59301 USA

a r t i c l e i n f o

Article historyReceived 24 February 2014Received in revised form7 August 2014Accepted 25 August 2014Available online

KeywordsCompetitionGreat PlainsPlantesoil feedbackSoil microbes

Corresponding author Tel thorn1 530 752 1701 faxE-mail addresses tmemamucdavisedu (TM Em

gov (EK Espeland) mattrinellaarsusdagov (MJ Ri

httpdxdoiorg101016jjaridenv2014080060140-1963copy 2014 Published by Elsevier Ltd

a b s t r a c t

The invasive grass Bromus tectorum negatively impacts grass and shrublands throughout the westernUS particularly in arid and semiarid regions We asked whether soil microbes associated with a nativegrass (Bouteloua gracilis) affect growth of Bromus and competition between Bromus and Bouteloua Wealso examined whether plant responses varied between soils from 15 sites in the Northern Great PlainsBromus and Bouteloua were grown in media with sterilized or unsterilized soil alone and together Soilsterilization reduced biomass of Bouteloua and Bromus grown alone by an estimated 50 and 48respectively Additionally results provided evidence that sterilization increased the effect of competitionon Bromus and may have reduced the effect of competition on Bouteloua Bouteloua likely had a strongernegative effect on Bromus in sterilized soils because sterilization reduced Bromus biomass by a greaterabsolute amount Response to sterilization varied appreciably by site for Bromus but not Bouteloua Ourresults support the hypothesis that invasive species such as Bromus often have positive responses to soilbiota in the invaded range Soil microbes are one factor that may be important in determining dynamicsof plant invasions and plant responses to new sites and competition with natives

copy 2014 Published by Elsevier Ltd

1 Introduction

The non-native grass Bromus tectorum L (hereafter referred to asBromus) currently occupies over 22 million hectares in the westernUnited States (Duncan et al 2009) and during the first decades ofinvasion from 1889 to 1929 it spread at one of the fastest docu-mented rates of invasion for plants (Pysek and Hulme 2005) Bro-mus has greatly altered ecosystem processes in the arid westernUS by increasing fire frequency (Balch et al 2013) and out-competing native plants (Mack 1981) Much research effort hasbeen expended to determine the traits of Bromus that lead to itsdominance in arid and semiarid lands Explanations include prolificseed production a phenology which enables early access to springmoisture and being adapted to large grazers (Harris 1967 Hulbert1955 Mack 1989) Relationships with soil biota may be anotherdeterminant of Bromus invasion dynamics Past research hasdemonstrated effects of Bromus invasion on resident soil biota andnutrient cycling (eg Hawkes et al 2006 Schaeffer et al 2012)However the effects of resident soil biota on Bromus remain

thorn1 530 752 4361am) erinespelandarsusdanella)

unclear as past studies have shown positive negative or nodetected effects of soil biota on Bromus growth (Al-Qarawi 2002Rowe et al 2009 Wilson and Hartnett 1998) Factors such asvariation in abiotic conditions across sites may influence the neteffect of soil biota on Bromus and interspecific competition mayalter responses to soil biota (eg Callaway et al 2004) Examiningthe role of soil biota in Bromus invasion may assist in identifyingnewmethods for preventing or controlling Bromus (eg Meyer andNelson 2006 Rowe et al 2009) In addition interactions betweenplants and soil biota influence multiple processes including suc-cession plant community diversity and productivity (Bever et al2010 Inderjit and van der Putten 2010) but further study isneeded to understand how planteplant interactions andgeographic variation shape these interactions

11 Plantesoil feedbacks influence invasion

Plantesoil feedbacks e plant effects on soil biotic andor abioticfactors which affect subsequent plant growth e can influence in-vasion processes (Inderjit and van der Putten 2010 Klironomos2002) Most native species exhibit a negative conspecific feed-back a species fitness is often lower in soil previously occupied byconspecifics than in sterilized soil or soil previously occupied by

Table 1Characteristics of the 15 sites used in this study Sampling area names are followedby total annual precipitation (Precip) in mm andmean annual temperature (MAT) inC averaged from 1983 to 2012 Aspect of ldquoUndrdquo indicates undulating terrain whereaspect could not be determined Soil texture C frac14 clay F frac14 fine L frac14 loam S frac14 sandSites with no mining history have a reclamation year of ldquoNArdquo Missing data areindicated by ldquoerdquo Climate data were obtained from PRISM Climate Group (availableonline at wwwprismclimateorg)

Sampling area Site Slope()

Aspect Soiltexture

Soil pH Reclamationyear

Caballo MinePrecip 384MAT 74

1 1 Und SCL 74 19902 1 Und SCL 74 19903 2 S C 76 19984 2 S SCL 74 1998

Eagle Butte MinePrecip 378MAT 75

1 3 SE FSCL 72 19922 2 NW FSCL 80 19993 2 N SCL 74 19994 3 NW SL 76 1992

Fort KeoghPrecip 348MAT 77

1 4 NNW SCL 74 NA2 7 SSE SL 73 NA

Spring Creek MinePrecip 378MAT 70

1 1 Und L 78 2001

Thunder Basin NorthPrecip 384MAT 74

1 0 Neutral C 61 NA2 3 NNE SL 60 NA

Thunder Basin SouthPrecip 325MAT 81

1 e e e e NA2 e e e e NA

TM Emam et al Journal of Arid Environments 111 (2014) 91e9792

heterospecifics (Brinkman et al 2010 Kulmatiski et al 2008)These negative plantesoil feedbacks are thought to be due to theaccumulation of specialist pathogens parasites and herbivores(Reynolds et al 2003) Negative feedbacks are particularly preva-lent among grasses Kulmatiski et al (2008) suggest that charac-teristics of grasses adapted to competing for water in semiaridlands may lead to greater exposure to soil enemies

The prevalence of negative conspecific feedbacks may promotecoexistence of competing plant species and thereby contribute tomaintenance of biodiversity (Bever 2003) However among inva-sive plant species many studies have found positive (or failed todetect negative) conspecific feedbacks in the invaded range(Inderjit and van der Putten 2010) Positive feedbacks with soilbiota occur when the benefits of soil mutualists outweigh thenegative effects of natural enemies which are thought to lead toincreased interspecific competition and competitive exclusion ofnative plant species by invasive species (Bever 2003) Severalstudies have found that while native plant species tend to generatefeedbacks that benefit community diversity and coexistence ofmultiple species (ie negative conspecific effects but positive ef-fects on others) invasive species tend to generate feedbacks thatpromote their own growth this phenomenon has been docu-mented in arid lands as well as through meta-analysis of manysystems (Kulmatiski et al 2008 Perkins and Nowak 2013) Usingthe context of feedbacks between plants and soil biota weattempted to examine how interactions between an invasive spe-cies (Bromus) and a native species (Bouteloua gracilis) responded tothe soil community across a range of sites

12 Study aims

We conducted a greenhouse study comparing the responses ofBromus and a native grass (Bouteloua gracilis HBK Lag ex Steudhereafter referred to as Bouteloua) to growthmedia inoculated withsoil gathered beneath Bouteloua in the field We chose Boutelouagracilis as a study species because it is a key late-seral species ofGreat Plains mixed and shortgrass prairie and was historically thedominant species in terms of frequency and biomass (Costello1944) Moreover Bouteloua is common throughout western NorthAmerica and occurs throughout much of the invasive range ofBromus Bromus interacting with Bouteloua-associated soil mi-crobes therefore represents a realistic invasion scenario in theGreat Plains We expected Bouteloua and Bromus to differ appre-ciably in their response to soil microbes Bromus is a non-nativecool-season annual grass while Bouteloua is a native warm-season perennial C4 grass C4 grasses are known to respond morepositively to mycorrhizal fungi (Hoeksema et al 2010) and dif-ferences in photosynthetic pathways between these two speciesmay affect species compositions of associated soil bacteria(Porazinska and Bardgett 2003)

We collected soil samples from beneath patches of Bouteloua at15 sites in the semiarid Northern Great Plains to serve as a source ofmicrobial inoculum Bromus and Bouteloua were grown alone andtogether with soil sterilized by autoclaving and unsterilized in afactorial experiment to examine whether soil sterilization affectedplanteplant interactions and whether there was substantial vari-ation in the responses of Bromus and Bouteloua among sites of soilcollection

We tested the following hypotheses

(1) Unsterilized soil from beneath Bouteloua plants will have anegative effect on Bouteloua biomass and a positive effect onBromus biomass because native species tend to generatenegative conspecific feedbacks and positive heterospecificfeedbacks (Perkins and Nowak 2013)

(2) If Bouteloua biomass is reduced in the unsterilized soil Bro-mus presence will have a greater negative effect on Boutelouawhen the two species are grown together in unsterilized soilas opposed to sterilized soil

Testing these two hypotheses provides information regardinghow Bromus might invade established communities of the GreatPlains or those undergoing restoration if Bromus has an advantagein unsterilized soils this would suggest that soil microbes maycontribute to competitive exclusion of Bouteloua by BromusBecause we studied multiple sites we were also able to examinevariation in response to soil sterilization

2 Materials and methods

21 Collection of soil samples and seeds

In summer 2012 soil samples were collected from 15 sites ineastern Montana or northeastern Wyoming US (AppendixFig A1) Sites were located within 6 sampling areas 1) Caballo2) Eagle Butte and 3) Spring Creek mines 4) North and 5) Southregions of the Thunder Basin National Grassland and 6) Fort KeoghLivestock and Range Research Laboratory Climate landscape soiland site history are described in Table 1 Landscape and soil data formined sites were collected by the US Department of Agriculture aspart of a study of these sites At Thunder Basin and Fort Keogh soilpH and texture were analyzed by AampL Western Laboratories Inc(Modesto CA) and site slope and aspect were assessed inNovember 2013 Sites with a mining history had been strip-minedfor coal Reclamation activities at mines included soil replacementmulching and fertilizing at some sites and seeding with Boutelouaand a variety of other species Vegetative cover at sites was pre-dominantly made up of native perennial grasses Agropyron cris-tatum (L) Gaertn and annual Bromus species Total cover wasrelatively low across sites averaging between approximately 5and 30 Though cover was low and soil was sampled from directlybeneath Bouteloua it is possible that any neighboring plant speciespresent may have influenced the soil community as well

TM Emam et al Journal of Arid Environments 111 (2014) 91e97 93

Soil and Bouteloua root fragment samples were collected frombeneath 4 to 11 Bouteloua patches per site using an 11-cm tallconical bulb planter (base and top diameter 6 and 75 cm respec-tively) to collect one soil core per Bouteloua patch Samples werethen pooled by site yielding between approximately 15 and 4 L persite (the number of cores taken and consequently soil volume waslimited by the number of Bouteloua patches available at each site)The pooled samples were oven-dried for 4 days at 60 C Ovendrying is likely to have excluded soil organisms aside from mi-crobes (eg nematodes arthropods etc) and may have altered themicrobial community to favor heat-tolerant species However thisis the optimal method for preserving mycorrhizal propagules(Habte and Byappanhalli 1998) allowing soils to retain moistureduring storage would have also resulted in an altered communitydue to the creation of anaerobic conditions and activity by de-composers (Harris et al 1993) Soils were coarsely ground to~4 mm to homogenize them and allow mixing with artificialgrowth media and stored at ambient temperature in a laboratoryUtilizing field-collected soils as opposed to greenhouse-conditioned soils (ie growing a certain plant species to cultureits associated soil microbes in the greenhouse prior to the start of afeedback experiment) enabled us to examine among-site variationSeeds from individual Bromus plants were collected from pop-ulations in eastern Montana Seeds of Bouteloua Bad River varietywere purchased from Granite Seed Company (Lehi UT)

22 Experimental design

To create a ldquosterilized soilrdquo treatment half of the soil collectedfrom each sitewas sterilized by autoclaving 3 times at 121 C for 1 hwith 36e48 h between each autoclaving in order to remove soilbiota (modification of Meiman et al 2006) Autoclaving can affect anumber of soil properties and soil characteristics are affected by allcurrently available sterilization methods (Perkins et al 2013) Asdescribed below we diluted the soils with artificial media andapplied fertilizer in an attempt to diminish the effects of auto-claving that were unrelated to soil biota The effect of autoclavingon soil nutrient availability is typically much smaller than effect ofapplying fertilizer (eg He and Cui (2009) detected a change of5 mgkg NH4 with autoclaving while we applied a fertilizer con-taining 200 mgL N in this study) Sterilized and unsterilized soilwas mixed with PROFILE Field and Fairwayreg Natural a ceramic par-ticle soil replacement (PROFILEreg Products LLC Buffalo Grove IL) at arate of 12 soil to 88 PROFILE

reg (vv) and this growth media wasadded to pots Although using artificial media led to less realisticgrowing conditions mixtures comprising between 1 and 15 fieldsoil are widely used for studying effects of soil microbes on plantswhile reducing the effects of abiotic soil characteristics (Brinkmanet al 2010) PROFILEreg was used to make up the remainder of thegrowth media because it has higher water and nutrient retentionthan sand (McCoy and Stehouwer 1999) The growth media wasplaced in 64-cm wide by 25-cm deep conical Deepotstrade (Stueweand Sons Corvallis OR) Approximately 1 cm of PROFILEreg was addedon top of the media to reduce transmission of soil microbes acrosspots We collected one sample of growth media (PROFILEreg and soilmixture) from sterilized and unsterilized treatments for 8 of thesites to determine whether soil sterilization had significant effectson soil chemical properties after dilution with PROFILEreg Sampleswere analyzed for plant-available P K Mg Ca Na S pH and cationexchange capacity (CEC) at AampL Western Laboratories Inc (Mod-esto CA) Growth media samples submitted for analysis did notinclude the fertilizer that was later added

Pots were arranged within 5 blocks in the greenhouse (ie 5replicates at start of experiment however this changed due tomortality e see Table A1) Unsterilized and sterilized treatments

were separated by approximately 15 cm to reduce cross-contamination during watering Bromus individuals which tendto have a very high germination rate were planted as seedsBecause Bouteloua germination can be low seeds were pre-germinated in the lab on wet paper towels in ambient naturaland artificial light and then the 1e2 day old germinants placed inthe pots Bouteloua germinants that died were replaced by newgerminants for up to a week following the initial planting Withinsoil sterilization treatments in each block we applied two differentcompetition treatments to each species for each of the 15 sites 1pot was planted with a single Bromus individual 1 pot with a singleBouteloua individual and 2 pots containing an individual of eachspecies Pots containing one live individual of each species after 3weeks of growth were defined as ldquowith competitionrdquo or growntogether pots with only one species present at this point weredefined as ldquowithout competitionrdquo or grown aloneWe use the termsldquocompetitionrdquo and ldquocompetitive effectrdquo throughout this texthowever we did not implement a replacement series to determinethe effects of inter-versus intra-specific competition and intendonly to describe the effects of Bouteloua and Bromus grown togetherversus alone

Plants were grown in a greenhouse from November 8th 2012 toJanuary 14th 2013 (approximately 9 weeks) at 32 C (day) and 16 C(night) with 13 h of supplemental daytime lighting and daily wa-tering Pots were fertilized with a P-free nutrient solution prior toplanting in order to prevent excessive nutrient deficiency The so-lution contained 02 gL N 004 gL K 002 gL S and 0002 gL FeOn Dec 6th 2012 a more dilute solution containing 008 gL N001 gL K 0007 gL S and 00007 gL Fe was added Any otherspecies that germinated from the existing seed bank were removedupon detection Aboveground biomass was harvested from pots byblock from Jan 7th 2013 to Jan 14th 2013 Biomass was oven-driedto constant weight at 60 C then weighed

23 Statistical analysis

Because some plants failed to establish some combinations ofsite soil sterilization and competition treatments did not occur inall blocks When Bromus and Bouteloua were planted together andone of the two plants failed to establish within three weeks of theexperiment start date the surviving plant served as a subsample ofthe corresponding ldquogrown alonerdquo treatment in the analysis Thistype of establishment failure led to 0e3 subsamples per treatmentcombination within blocks Subsamples were averaged within thefive blocks Replication of treatment combinations is described inthe Appendix Table A1

According to initial analyses the effects of soil sterilizationsite soil sterilization competition treatment soil sterilizationand site competition treatment soil sterilization onmortality ofBromus and Bouteloua were not substantial therefore we do notpresent mortality results here Our two response variables werenatural-log transformed Bromus and Bouteloua abovegroundbiomass per plant We fit the same linear model to both responsevariables Initial models indicated no substantial difference be-tween plant response to soils from non-mined versus mined sitesand this termwas not included in subsequent models Additionallyincluding a term for sampling area did not substantially improvethe model when site was also included therefore sampling areawas excluded from the final model

The final model had terms indicating block site competitiontreatment (plants grown alone or with an individual of the otherspecies) and whether or not soil was sterilized Site competitiontreatment site soil sterilization competition treatment soilsterilization and site competition treatment soil sterilizationinteractions were also included in the model Site and all

TM Emam et al Journal of Arid Environments 111 (2014) 91e9794

interactions involving site were modeled as random effects due tothe high number of levels for this factor (Gelman and Hill 2007)Block was also modeled as random and all other parameters weremodeled as fixed We wrote a FORTRAN program to fit the models(Intel Corporation 2013) Inferences were based on back-transformed parameter point estimates and 95 confidence in-tervals and P values calculated from uncertainty estimates on theregression parameters using methods outlined by Gelman and Hill(2007)

Effects of site and sterilization treatment on properties of thegrowth media were analyzed using the GLM procedure in SAS 9(SAS Institute Inc) using untransformed response variables for soilP K Mg Ca Na S pH and CEC

3 Results

When grown alone soil sterilization decreased abovegroundbiomass of Bromus by an estimated 48 (P frac14 001 confidence in-terval 30e61 Fig 1a) Similarly Bouteloua grown alone had 50less biomass when soil was sterilized (P frac14 001 confidence interval10e71 Fig 1b) In unsterilized soil Bromus biomass was notsubstantially different when grown with or without a Boutelouaindividual (Pfrac14 022) However soil sterilization increased the effectof competition on Bromus (P frac14 004 Fig 1a) Conversely in un-sterilized soils competition reduced Bouteloua biomass by 61(P frac14 001 confidence interval 25e79) but there was some evi-dence that sterilization lessened the effect of competition on Bou-teloua (P frac14 013 Fig 1b) Bouteloua biomass decreased less inresponse to competition when soil was sterilized than when soilwas unsterilized

Fig 1 Aboveground biomass of Bromus (a) and Bouteloua (b) in unsterilized andsterilized soils when grown alone or with a competitor of the other species (barsshown with standard errors)

The effect of soil sterilization on Bromus biomass varied appre-ciably by site (Fig 2a) Bromus had less biomass in response tosterilization of soil from most but not all sites (Fig 2b) Comparedto the mean across sites sterilization reduced Bromus biomassmore drastically at one Eagle Butte mine site and one Fort Keoghsite and less drastically at one Eagle Butte mine site and twoThunder Basin National Grassland sites (Fig 2a) Interestingly site-to-site variation in Bromus response to sterilization appeared to besimilar in magnitude within and among sampling areas The site ofsoil collection appeared to affect Bromus biomass both when soilwas sterilized and unsterilized (Fig 2b) Bouteloua responses to soilsterilization did not differ substantially by site and the effect ofcompetition in sterilized or unsterilized soils did not vary sub-stantially by site for either species See Appendix Table A2 fordetailed results

Some properties of the growth media (soil mixed with PROFILEreg)

were altered by soil sterilization (Table 2) Sterilization increasedCEC by 11 (P frac14 003 confidence interval 2e21) while steriliza-tion decreased pH by 5 (P frac14 0004 confidence interval 2e7) Siteof soil origin affected CEC S Ca Mg and pH (all confidence in-tervals not overlapping zero and P lt 005 results not shown) Pointestimates of site effects tended to be higher than sterilization ef-fects for example point estimates of the effect of sitewere 46 and23 on CEC and pH in sites with the largest effects compared toestimates of 11 and 5 for sterilization

4 Discussion

41 Soil sterilization decreased Bromus tectorum and Boutelouagracilis biomass when grown alone

We found that overall soil sterilization decreased abovegroundbiomass of Bromus grown alone This result is consistent with ourhypothesis that Bromus would benefit from unsterilized soil due tosoil microbes associated with a native plant similar to invasiveplants in other studies (Inderjit and van der Putten 2010) Howeverour results show that Bouteloua biomass also decreased when soilwas sterilized which runs counter to our hypothesis that Boutelouawould have a positive response to soil sterilization This could bebecause late-seral or climax species such as Boutelouamay promotetheir own dominance through positive feedback mechanisms(Connell and Slatyer 1977) despite the accumulation of naturalenemies Meta-analysis has shown that overall native species tendto have negative plantesoil feedbacks but later successional spe-cies such as Boutelouamay have less negative plantesoil feedbacks(Kulmatiski et al 2008)

Relationships between plants and soil microbes may explain ourresults One key way that soil microbes affect plants is by influ-encing quantities and availabilities of soil nutrients For examplesoil bacteria perform the processes of N mineralization andimmobilization which heavily affect soil N availability (Myrold andBottomley 2008) Past research has found that Bromus increasesrates of microbial N immobilization and N cycling relative to nativegrassland plants (Schaeffer et al 2012) AM fungi can also increaseplant biomass by providing nutrients particularly P (Smith andRead 2008) Biomass of Bromus has been shown to increase (Al-Qarawi 2002) or remain similar (Wilson and Hartnett 1998) inresponse to AM fungi C4 grasses such as Bouteloua tend to havemore positive responses to AM fungi than C3 grasses (Hoeksemaet al 2010) and Bouteloua has been shown to have high levels ofAM colonization and substantial increases in carbon fixation rateswith AM symbiosis (Allen et al 1984) In addition soil microbes canhave positive non-nutritive effects on plants For example AM fungihave been shown to reduce effects of soil pathogens (Wehner et al2010) and plant growth promoting rhizobacteria can increase plant

Fig 2 a) Point estimates (dots) and 95 confidence intervals (lines) estimating site-to-site differences in the effects of soil sterilization on Bromus grown alone Confidence intervalsthat do not overlap zero (dashed line) indicate sites that differed from the mean of all sites in terms of the effect of soil sterilization on Bromus biomass b) Aboveground biomass(solid bars shown with standard errors) of Bromus grown alone with unsterilized or sterilized soil from each site Asterisks indicate biomass was substantially different betweenpaired bars ie 95 confidence intervals of the relative effect of sterilization (not shown) did not overlap zero

TM Emam et al Journal of Arid Environments 111 (2014) 91e97 95

biomass by producing hormones such as auxin which stimulatesroot growth (Lugtenberg and Kamilova 2009) Another importantconsideration is that plant responses to soil biota are not uniformover the life of plants Past research has shown that feedback re-sponses of native grasses tend to become more negative over timeas deleterious soil microbes accumulate while this may not be truefor non-native congeners (Hawkes et al 2012) While our resultssuggest positive effects of soil biota on both a native and a non-native grass our experiment was conducted over a 2-monthperiod and the response of Bouteloua to soil microbes may havebecome negative over a longer time period In addition our soilpreparation methods may have favored heat-tolerant microbesparticularly fungi and we did not attempt to identify which taxa orfunctional groups of soil biota may have affected Bromus andBouteloua or the mechanism(s) underlying the positive responsesto unsterilized soils

An alternative explanation is that plants may have responded tosoil changes imposed by autoclaving (Perkins et al 2013) despitesteps taken to minimize these effects Soil sterilization had effectson soil CEC and pH Increased CEC enables the soil to retain more

Table 2Growth medium properties in sterilized and unsterilized treatments CEC frac14 Cation exch

Treatment pH P ppm K ppm M

Unsterilized 63 plusmn 02 25 plusmn 2 446 plusmn 20 3Sterilized 60 plusmn 02 30 plusmn 2 391 plusmn 10 3

nutrients but may have somewhat reduced plant-available nutri-ents However levels of the nutrients we assessed did not changesubstantially with sterilization Decreased pH may have affectednutrient availability as well however the effect of sterilization onpH was relatively small particularly in comparison to the level ofvariation across sites

42 Soil sterilization altered competition between Boutelouagracilis and Bromus tectorum

Both species had greater biomass in unsterilized soil but inabsolute terms Bromus size increased to a greater degree thanBouteloua This likely conferred greater competitive ability to Bro-mus in unsterilized soil lessening the effect of competition onBromus and increasing the effect of competition on Boutelouawhensoil was not sterilized The influence of soil sterilization oncompetition may reflect one or more of several possible mecha-nisms One possibility is that Bromus size increased to a greaterdegree than Bouteloua in response to soil microbes conferringgreater competitive ability to Bromus in unsterilized soil If soil

ange capacity averages are given plusmnstandard error

g ppm Ca ppm Na ppm CEC meq100 g

05 plusmn 21 1847 plusmn 216 22 plusmn 1 145 plusmn 0904 plusmn 28 2000 plusmn 267 205 plusmn 08 16 plusmn 1

TM Emam et al Journal of Arid Environments 111 (2014) 91e9796

microbes increased nutrient availability then these results mayreflect the demonstrated ability of Bromus to better exploit soilnutrients compared to native grasses such as Bouteloua (Lowe et al2003 Vasquez et al 2008) Past research has shown that non-native species are often able to rapidly use added or fluctuatingresources while native species are better able to compete withnon-native species in nutrient-poor soils (Davis et al 2000) Inaddition changes in soil pH and CEC imposed by autoclaving mayhave also affected nutrient dynamics Competition between Bromusand Bouteloua could have shifted in response to these abiotic effectsas well as in response to microbes We noted that Bouteloua had alower survival rate than Bromus (62 compared to 93 across alltreatments) had Bouteloua survival been higher the effect ofcompetition on Bromus may have been stronger overall HoweverBouteloua mortality was not affected by soil sterilization or in-teractions between soil sterilization and other treatments (resultsnot shown) so this does not explain differences between compe-tition in sterilized and unsterilized soils

If other non-nutritive mechanisms of increasing biomass dis-cussed in the previous section (eg growth stimulation) had astronger positive effect on Bromus than Bouteloua this could alsoexplain how microbes increased Bromus size and thus competitiveability In addition because Bromus can alter microbial communitycomposition as a dynamic process during growth (Schaeffer et al2012) Bromus may have changed microbial communities in un-sterilized soils in our experiment thereby reducing Boutelouagrowth indirectly

43 Response of Bromus tectorum to soil sterilization variesappreciably by site

Bromusbiomass variedby site of soil origin as did the response ofBromus to soil sterilization (Fig 2) Site-to-site variation in the effectof soil sterilization on Bromus (ie the difference in biomass be-tween sterilized and unsterilized soils) may reflect differences intaxa (or abundances) of certain soilmutualists orpathogens that hadstronger effects on Bromus than Bouteloua Abiotic conditions suchas climate and soil pH and biotic factors such as plant taxa areknown to shape soil microbial communities (Fierer and Jackson2006 Kivlin et al 2011) It is uncertain which environmental fac-tors were most important in determining Bromus response to soilsterilization across sites in our study However regardless of thesource of the variability our results indicate that Bromus is moresensitive to it than Bouteloua in terms of aboveground biomassresponse Variation in Bromus biomass among siteswhen soils weresterilized likely reflects differences in soil properties that were notfully compensated for by diluting the soil with PROFILEreg and applyingfertilizer Site of soil origin affected chemical properties of thegrowth media such as CEC S Ca Mg and pH The effects of site onsoil properties tended to be larger than the effect of soil sterilization

5 Conclusions

Our results support prior work showing that invasive speciessuch as Bromus often have a positive (or non-negative) response tosoil biota in their invasive range (Inderjit and van der Putten 2010Perkins and Nowak 2013) Conversely our hypothesis that thenative grass Boutelouawould respond negatively to conspecific soilmicrobes was not supported e this species also responded posi-tively to unsterilized soil The increase in biomass of both species inunsterilized soils could be due to microbial-mediated increases innutrient availability non-nutritive effects of soil microbes such asprotection from pathogens or stimulation of plant growth effects ofautoclaving on some abiotic soil properties or a combination offactors

Soil sterilization affected competition between the invasivegrass Bromus tectorum and the native grass Bouteloua gracilis Whensoil was sterilized Bromus biomass was reduced by competitionfrom Bouteloua but in unsterilized soil Bromus was not substan-tially affected by Bouteloua This implies that soil microbes mayfacilitate Bromus growth and competition in some instances andmay contribute in part to the invasiveness of this species In addi-tion our findings show that Bromus responded differently to ster-ilization of soil from different sites while there was no appreciableeffect of site on the response of Bouteloua The microbial commu-nity composition may have varied across these sites in a way thatstrongly affected Bromus but not Bouteloua possibly due to a longerhistory of Bouteloua presence at these sites Bromus growth may befacilitated by the microbial community at some sites and Boutelouamay not be able to compete as effectively with Bromus at these sitesas a result However because we did not examine the compositionof the soil microbial community we cannot say for certain thatmicrobes alone were driving this effect Bromus biomass differed bysite of soil origin in sterilized soils as well likely due to differencesamong sites in soil properties such as pH CEC Ca Mg and S

Plant biomass responses to soil biota are likely to be influencedby many factors such as environmental conditions and abiotic soilcharacteristics which shape themicrobial community Responses tosoil sterilization in one context (eg using soil from one site or onegreenhouse conditioning environment) may not accurately repre-sent responses under other conditions and this should be takeninto account when studying relationships between plants and soilbiota In addition soil sterilization and site of soil origin may affectabiotic soil properties even after diluting soil in other growth me-dia in our study variation in soil properties was greater across sitesthan between sterilized and unsterilized soil treatments While soilmicrobial communities are clearly important for plant growth andcompetitive outcomes other soil characteristics may play roles aswell Determining the importance of microbial communities oninvasion dynamics requires an approach that makes distinctionsbetween the effect of experimental methods and the effect of soilbiota themselves Our data suggest soil microbes play a role in in-teractions between Bromus and Bouteloua however the soil mi-crobial community is just one of many factors that can influenceplant growth and invasiveness

Acknowledgments

This material is based upon work supported by the National Sci-enceFoundationGraduateResearchFellowshipProgramunderGrantNo 1148897 We thank Valerie Eviner Lora Perkins Kevin Rice KateScow and four anonymous reviewers for comments on previousversions of this work and Lian Rother JamesMizoguchi Kao SaeleeMaureen OMara Darcy Hammond Ming-Yu Stephens AnnalisaBryant Bruce Moffat Bob Haynes and Brian Kozar for researchassistance Mention of trade names or commercial products in thispublication is solely for thepurpose of providing specific informationand does not imply recommendation or endorsement by the USDepartment of Agriculture or the National Science Foundation

Appendix A Supplementary data

Supplementary data related to this article can be found at httpdxdoiorg101016jjaridenv201408006

References

Allen M Allen E Stahl P 1984 Differential niche response of Bouteloua gracilisand Pascopyrum smithii to VA mycorrhizae Bull Torrey Bot Club 111 361e365httpdxdoiorg1023072995917

TM Emam et al Journal of Arid Environments 111 (2014) 91e9796

microbes increased nutrient availability then these results mayreflect the demonstrated ability of Bromus to better exploit soilnutrients compared to native grasses such as Bouteloua (Lowe et al2003 Vasquez et al 2008) Past research has shown that non-native species are often able to rapidly use added or fluctuatingresources while native species are better able to compete withnon-native species in nutrient-poor soils (Davis et al 2000) Inaddition changes in soil pH and CEC imposed by autoclaving mayhave also affected nutrient dynamics Competition between Bromusand Bouteloua could have shifted in response to these abiotic effectsas well as in response to microbes We noted that Bouteloua had alower survival rate than Bromus (62 compared to 93 across alltreatments) had Bouteloua survival been higher the effect ofcompetition on Bromus may have been stronger overall HoweverBouteloua mortality was not affected by soil sterilization or in-teractions between soil sterilization and other treatments (resultsnot shown) so this does not explain differences between compe-tition in sterilized and unsterilized soils

If other non-nutritive mechanisms of increasing biomass dis-cussed in the previous section (eg growth stimulation) had astronger positive effect on Bromus than Bouteloua this could alsoexplain how microbes increased Bromus size and thus competitiveability In addition because Bromus can alter microbial communitycomposition as a dynamic process during growth (Schaeffer et al2012) Bromus may have changed microbial communities in un-sterilized soils in our experiment thereby reducing Boutelouagrowth indirectly

43 Response of Bromus tectorum to soil sterilization variesappreciably by site

Bromusbiomass variedby site of soil origin as did the response ofBromus to soil sterilization (Fig 2) Site-to-site variation in the effectof soil sterilization on Bromus (ie the difference in biomass be-tween sterilized and unsterilized soils) may reflect differences intaxa (or abundances) of certain soilmutualists orpathogens that hadstronger effects on Bromus than Bouteloua Abiotic conditions suchas climate and soil pH and biotic factors such as plant taxa areknown to shape soil microbial communities (Fierer and Jackson2006 Kivlin et al 2011) It is uncertain which environmental fac-tors were most important in determining Bromus response to soilsterilization across sites in our study However regardless of thesource of the variability our results indicate that Bromus is moresensitive to it than Bouteloua in terms of aboveground biomassresponse Variation in Bromus biomass among siteswhen soils weresterilized likely reflects differences in soil properties that were notfully compensated for by diluting the soil with PROFILEreg and applyingfertilizer Site of soil origin affected chemical properties of thegrowth media such as CEC S Ca Mg and pH The effects of site onsoil properties tended to be larger than the effect of soil sterilization

5 Conclusions

Our results support prior work showing that invasive speciessuch as Bromus often have a positive (or non-negative) response tosoil biota in their invasive range (Inderjit and van der Putten 2010Perkins and Nowak 2013) Conversely our hypothesis that thenative grass Boutelouawould respond negatively to conspecific soilmicrobes was not supported e this species also responded posi-tively to unsterilized soil The increase in biomass of both species inunsterilized soils could be due to microbial-mediated increases innutrient availability non-nutritive effects of soil microbes such asprotection from pathogens or stimulation of plant growth effects ofautoclaving on some abiotic soil properties or a combination offactors

Soil sterilization affected competition between the invasivegrass Bromus tectorum and the native grass Bouteloua gracilis Whensoil was sterilized Bromus biomass was reduced by competitionfrom Bouteloua but in unsterilized soil Bromus was not substan-tially affected by Bouteloua This implies that soil microbes mayfacilitate Bromus growth and competition in some instances andmay contribute in part to the invasiveness of this species In addi-tion our findings show that Bromus responded differently to ster-ilization of soil from different sites while there was no appreciableeffect of site on the response of Bouteloua The microbial commu-nity composition may have varied across these sites in a way thatstrongly affected Bromus but not Bouteloua possibly due to a longerhistory of Bouteloua presence at these sites Bromus growth may befacilitated by the microbial community at some sites and Boutelouamay not be able to compete as effectively with Bromus at these sitesas a result However because we did not examine the compositionof the soil microbial community we cannot say for certain thatmicrobes alone were driving this effect Bromus biomass differed bysite of soil origin in sterilized soils as well likely due to differencesamong sites in soil properties such as pH CEC Ca Mg and S

Plant biomass responses to soil biota are likely to be influencedby many factors such as environmental conditions and abiotic soilcharacteristics which shape themicrobial community Responses tosoil sterilization in one context (eg using soil from one site or onegreenhouse conditioning environment) may not accurately repre-sent responses under other conditions and this should be takeninto account when studying relationships between plants and soilbiota In addition soil sterilization and site of soil origin may affectabiotic soil properties even after diluting soil in other growth me-dia in our study variation in soil properties was greater across sitesthan between sterilized and unsterilized soil treatments While soilmicrobial communities are clearly important for plant growth andcompetitive outcomes other soil characteristics may play roles aswell Determining the importance of microbial communities oninvasion dynamics requires an approach that makes distinctionsbetween the effect of experimental methods and the effect of soilbiota themselves Our data suggest soil microbes play a role in in-teractions between Bromus and Bouteloua however the soil mi-crobial community is just one of many factors that can influenceplant growth and invasiveness

Acknowledgments

This material is based upon work supported by the National Sci-enceFoundationGraduateResearchFellowshipProgramunderGrantNo 1148897 We thank Valerie Eviner Lora Perkins Kevin Rice KateScow and four anonymous reviewers for comments on previousversions of this work and Lian Rother JamesMizoguchi Kao SaeleeMaureen OMara Darcy Hammond Ming-Yu Stephens AnnalisaBryant Bruce Moffat Bob Haynes and Brian Kozar for researchassistance Mention of trade names or commercial products in thispublication is solely for thepurpose of providing specific informationand does not imply recommendation or endorsement by the USDepartment of Agriculture or the National Science Foundation

Appendix A Supplementary data

Supplementary data related to this article can be found at httpdxdoiorg101016jjaridenv201408006

References

Allen M Allen E Stahl P 1984 Differential niche response of Bouteloua gracilisand Pascopyrum smithii to VA mycorrhizae Bull Torrey Bot Club 111 361e365httpdxdoiorg1023072995917

TM Emam et al Journal of Arid Environments 111 (2014) 91e97 97

Al-Qarawi AA 2002 Relationship Among Nitrogen Availability Vesicular-Arbsucular Mycorrhizae and Bromus tectorum in Disturbed Rangeland Sites inColorado (Dissertation) Colorado State University pp 1e111

Balch JK Bradley BA DAntonio CM Gomez-Dans J 2013 Introduced annualgrass increases regional fire activity across the arid western USA (1980e2009)Glob Change Biol 19 173e183 httpdxdoiorg101111gcb12046

Bever JD 2003 Soil community feedback and the coexistence of competitorsconceptual frameworks and empirical tests New Phytol 157 465e473 httpdxdoiorg101046j1469-8137200300714x

Bever JD Dickie IA Facelli E Facelli JM Klironomos J Moora M Rillig MCStock WD Tibbett M Zobel M 2010 Rooting theories of plant communityecology in microbial interactions Trends Ecol Evol 25 468e478 httpdxdoiorg101016jtree201005004x

Brinkman EP Van der Putten WH Bakker E-J Verhoeven KJF 2010 Plant-soilfeedback experimental approaches statistical analyses and ecological in-terpretations J Ecol 98 1063e1073 httpdxdoiorg101111j1365-2745201001695x

Callaway R Thelen G Barth S 2004 Soil fungi alter interactions between theinvader Centaurea maculosa and North American natives Ecology 851062e1071 httpdxdoiorg10189002-0775

Connell JH Slatyer RO 1977 Mechanisms of succession in natural communitiesand their role in community stability and organization Am Nat 111 1119e1144httpdxdoiorg101086283241

Costello DF 1944 Important species of the major forage types in Colorado andWyoming Ecol Monogr 14 107e134 httpdxdoiorg1023071961633

Davis MA Grime JP Thompson K Davis A Philip J 2000 Fluctuating re-sources in plant communities a general theory of invasibility J Ecol 88528e534 httpdxdoiorg101046j1365-2745200000473x

Duncan C Jachetta J Brown M 2009 Assessing the economic environmentaland societal losses from invasive plants on rangeland and wildlands 1 WeedTechnol 18 1411e1416

Fierer N Jackson RB 2006 The diversity and biogeography of soil bacterialcommunities Proc Natl Acad Sci U S A 103 626e631 httpdxdoiorg101073pnas0507535103

Gelman A Hill J 2007 Data Analysis Using Regression and MultilevelHierarchicalModels Cambridge University Press New York NY

Habte M Byappanhalli B 1998 Influence of pre-storage drying conditionsand duration of storage on the effectiveness of root inoculum of Glomusaggregation J Plant Nutr 21 37e41 httpdxdoiorg10108001904169809365490

Harris G 1967 Some competitive relationships between Agropyron spicatum andBromus tectorum Ecol Monogr 37 89e111 httpdxdoiorg1023072937337

Harris JA Birch P Short KC 1993 The impact of storage of soils during opencastmining on the microbial community a strategist theory interpretation RestorEcol 1 88e100

Hawkes CV Belnap J DAntonio C Firestone MK 2006 Arbuscular mycorrhizalassemblages in native plant roots change in the presence of invasive exoticgrasses Plant Soil 281 369e380 httpdxdoiorg101007s11104-005-4826-3

Hawkes CV Kivlin SN Du J Eviner VT 2012 The temporal development andadditivity of plant-soil feedback in perennial grasses Plant Soil 369 141e150httpdxdoiorg101007s11104-012-1557-0

He W-M Cui Q-G 2009 Manipulation of soil biota in ecological research WebEcol 9 68e71 httpdxdoiorg105194we-9-68-2009

Hoeksema JD Chaudhary VB Gehring CA Johnson NC Karst J Koide RTPringle A Zabinski C Bever JD Moore JC Wilson GWT Klironomos JNUmbanhowar J 2010 A meta-analysis of context-dependency in plantresponse to inoculation with mycorrhizal fungi Ecol Lett 13 394e407 httpdxdoiorg101111j1461-0248200901430x

Hulbert L 1955 Ecological studies of Bromus tectorum and other annual brome-grasses Ecol Monogr 25 181e213 httpdxdoiorg1023071943550

Intel Corporation 2013 Intel Visual Fortran Compiler Professional Edition 140Inderjit van der Putten WH 2010 Impacts of soil microbial communities on exotic

plant invasions Trends Ecol Evol 25 512e519 httpdxdoiorg101016jtree201006006

Kivlin SN Hawkes CV Treseder KK 2011 Global diversity and distribution ofarbuscular mycorrhizal fungi Soil Biol Biochem 43 2294e2303 httpdxdoiorg101016jsoilbio201107012

Klironomos JN 2002 Feedback with soil biota contributes to plant rarity andinvasiveness in communities Nature 417 67e70 httpdxdoiorg101038417067a

Kulmatiski A Beard KH Stevens JR Cobbold SM 2008 Plant-soil feedbacks ameta-analytical review Ecol Lett 11 980e992 httpdxdoiorg101111j1461-0248200801209x

Lowe P Lauenroth W Burke I 2003 Effects of nitrogen availability on compe-tition between Bromus tectorum and Bouteloua gracilis Plant Ecol 176 247e254httpdxdoiorg101023A1023934515420

Lugtenberg B Kamilova F 2009 Plant-growth-promoting rhizobacteria Annu RevMicrobiol 63 541e556 httpdxdoiorg101146annurevmicro62081307162918

Mack R 1981 Invasion of Bromus tectorum L into Western North America anecological chronicle Agro-Ecosystems 7 145e165 httpdxdoiorg1010160304-3746(81)90027-5

Mack R 1989 Temperate grasslands vulnerable to plant invasions characteristicsand consequences In Drake JA (Ed) Biological Invasions a Global Perspec-tive John Wiley amp Sons Ltd pp 155e179

McCoy EL Stehouwer RC 1999 Water and nutrient retention properties ofinternally porous inorganic amendments in high sand content root zonesJ Turfgrass Manage 2 49e69 httpdxdoiorg101300J099v02n04_05

Meiman PJ Redente EF Paschke MW 2006 The role of the native soil com-munity in the invasion ecology of spotted (Centaurea maculosa auct non Lam)and diffuse (Centaurea diffusa Lam) knapweed Appl Soil Ecol 32 77e88httpdxdoiorg101016japsoil200502016

Meyer S Nelson D 2006 Cheatgrass (Bromus tectorum) biocontrol using indige-nous fungal pathogens In Kitchen SG Pendleton RL Moncaco TAVernon J (Eds) Proceedings Shrublands Under Fire Disturbance and Recov-ery in a Changing World pp 61e67

Myrold D Bottomley P 2008 Nitrogen mineralization and immobilization InSchepers JS Raun WB Follett RF et al (Eds) Nitrogen in AgriculturalSystems Agronomy Monograph 49 American Society of Agronomy MadisonWI pp 157e172

Perkins LB Blank RR Ferguson SD Johnson DW Lindemann WC Rau BM2013 Quick start guide to soil methods for ecologists Perspect Plant Ecol EvolSyst 15 237e244 httpdxdoiorg101016jppees201305004

Perkins LB Nowak RS 2013 Native and non-native grasses generate commontypes of plant-soil feedbacks by altering soil nutrients and microbial commu-nities Oikos 122 199e208 httpdxdoiorg101111j1600-0706201220592x

Porazinska D Bardgett R 2003 Relationships at the aboveground-belowgroundinterface plants soil biota and soil processes Ecol Monogr 73 377e395httpdxdoiorg1018900012-9615(2003)073[0377RATAIP]20CO2

Pysek P Hulme PE 2005 Spatio-temporal dynamics of plant invasions linkingpattern to process Ecoscience 12 302e315 httpdxdoiorg102980i1195-6860-12-3-3021

Reynolds HL Packer A Bever JD Clay K 2003 Grassroots ecologyplantemicrobeesoil interactions as drivers of plant community structure anddynamics Ecology 84 2281e2291 httpdxdoiorg10189002-0298

Rowe HI Brown CS Paschke MW 2009 The influence of soil inoculum andnitrogen availability on restoration of high-elevation steppe communitiesinvaded by Bromus tectorum Restor Ecol 17 686e694 httpdxdoiorg101111j1526-100X200800385x

Schaeffer SM Ziegler SE Belnap J Evans RD 2012 Effects of Bromus tectoruminvasion on microbial carbon and nitrogen cycling in two adjacent undisturbedarid grassland communities Biogeochemistry 111 427e441 httpdxdoiorg101007s10533-011-9668-x

Smith S Read DJ 2008 Mycorrhizal Symbiosis third ed Academic Press NewYork NY

Vasquez E Sheley R Svejcar T 2008 Nitrogen enhances the competitive abilityof cheatgrass (Bromus tectorum) relative to native grasses Invasive Plant SciManage 1 287e295 httpdxdoiorg101614IPSM-08-0621

Wehner J Antunes PM Powell JR Mazukatow J Rillig MC 2010 Plantpathogen protection by arbuscular mycorrhizas a role for fungal diversityPedobiologia 53 197e201 httpdxdoiorg101016jpedobi200910002

Wilson GW Hartnett DC 1998 Interspecific variation in plant responses tomycorrhizal colonization in tallgrass prairie Am J Bot 85 1732e1738 httpdxdoiorg1023072446507