Nd

Ja

b

c

a

ARRA

KSNNCBP

1

4tsThs

0d

Applied Soil Ecology 58 (2012) 66– 77

Contents lists available at SciVerse ScienceDirect

Applied Soil Ecology

journa l h o me page: www.elsev ier .com/ locate /apsoi l

ematodes as an indicator of plant–soil interactions associated withesertification

eremy R. Klassa,∗, Debra P.C. Petersb, Jacqueline M. Trojanc, Stephen H. Thomasc

New Mexico State University, Plant and Environmental Science, Las Cruces, NM 88003-8003, USAUSDA-ARS Jornada Experimental Range and Jornada Basin LTER, Las Cruces, NM 88003-8003, USANew Mexico State University, Entomology, Plant Pathology, and Weed Science, Las Cruces, NM 88003-8003, USA

r t i c l e i n f o

rticle history:eceived 5 October 2011eceived in revised form 14 February 2012ccepted 8 March 2012

eywords:emi-arid grasslandsematode communitiesematode diversityonnectivityouteloua eriopodarosopis glandulosa

a b s t r a c t

Conversion of perennial grasslands to shrublands is a desertification process that is important globally.Changes in aboveground ecosystem properties with this conversion have been well-documented, butlittle is known about how belowground communities are affected, yet these communities may be impor-tant drivers of desertification as well as constraints on the reversal of this state change. We examinednematode community structure and feeding as a proxy for soil biotic change across a desertificationgradient in southern NM, USA. We had two objectives: (1) to compare nematode trophic structure andspecies diversity within vegetation states representing different stages of desertification, and (2) to com-pare nematode community structure between bare and vegetated patches that may be connected via amatrix of endophytic fungi and soil biotic crusts. The gradient included a perennial grassland dominatedby Bouteloua eriopoda, the historic dominant in the Chihuahuan Desert, a duneland dominated by theshrub, Prosopis glandulosa, and the ecotone between them. We also sampled a relatively undisturbed,ungrazed B. eriopoda grassland at a nearby site to serve as an end member of our gradient. Nematodecommunities were sampled using soil cores to depth of 50 cm at each location in 2009 and 2010. Resultsshowed that grasslands and mesquite dunelands had different trophic groupings and herbivorous nema-tode communities with lower species diversity and evenness compared with the ecotone. Nematodetrophic structure and herbivore communities were significantly different in all vegetation states withthe highest observed diversity in the undisturbed, ungrazed B. eriopoda grassland in 2010. Vegetated andbare ground patches within the two grassland sites had similar herbivore communities, especially speciesfrom the family Tylenchinae. However, the mesquite duneland showed the lowest sampled diversity ofall sites, but had significantly larger nematode abundances in vegetated dunes than interdune areas that

are void of vegetation and soil biotic crusts where bacteriovores dominated. Decreased nematode trophicstructure and species diversity in the Jornada black grama grassland samples compared with the undis-turbed grassland illustrate the effect of desertification on the soil biotic community. Our results showthat nematodes can be used to identify changes in belowground community structure based on trophicinteractions. Large-scale disturbances like desertification can have consequences on the diversity andsoil biotic functioning at finer spatial scales.

. Introduction

Desertification of arid lands is a global problem, with nearly0% of the world’s land area either already converted or suscep-ible to a state change from productive grasslands to degradedhrublands or woodlands (Reynolds and Stafford Smith, 2002).

hese ‘state’ changes are difficult to restore because of the loss oferbaceous productivity, grass propagules, and modifications inoil surface properties that result in increases in runoff and erosion

∗ Corresponding author. Tel.: +1 575 571 6107; fax: +1 575 646 6041.E-mail address: jrklass@nmsu.edu (J.R. Klass).

929-1393/$ – see front matter © 2012 Elsevier B.V. All rights reserved.oi:10.1016/j.apsoil.2012.03.005

© 2012 Elsevier B.V. All rights reserved.

and losses in soil nutrients that favor woody plants (Schlesingeret al., 1990; Archer, 1994; Abrahams et al., 1995; Parsons et al.,1997; Whisenant, 1999; Wainwright et al., 2000; Peters, 2002a,b).Examining effects of drivers (e.g., multi-year drought, livestockovergrazing) of state changes on landscape-scale patterns oftenfocus on vegetation responses; much less is known about changesin belowground communities that can influence vegetationdynamics. However, shifts in belowground community structureassociated with desertification may provide a constraint on the

ability of grasses to recover if soil biota critical to grass success aremissing from the shrubland. Because arid ecosystems consist of amosaic of grass or shrub plants and bare soil interspaces (i.e., gaps),differences in soil biota between these two microhabitats may

Soil E

hdg

fnRtcPaOAib(acePbaoPaeashAliaeadtse

to(gledilmt1inepcl

towagtr

J.R. Klass et al. / Applied

ave important consequences for grass recovery. Our goal was toetermine how soil biota differs across a grassland-to-shrublandradient of increasing woody plant abundance.

The composition of soil organisms (e.g., nematodes, bacteria,ungi, etc.) determines the influence of the soil microbial commu-ity on plant assemblages, in general (Ogle and Reynolds, 2004;eynolds et al., 2004; Collins et al., 2008). Soil biota can affecthe establishment and abundance of individual plant species withonsequences for functional group and species diversity (van derutten et al., 1993; Molofsky, 1994; Bever et al., 1997; Millsnd Bever, 1998; Klironomos et al., 2000; Molofsky et al., 2001;’Hanlon-Manners and Kotanen, 2004; Reinhart et al., 2003, 2005).

disruption or disturbance to soil biotic consortia can play anntegral role in the dynamics of a system, and may be responsi-le for either promoting shrub invasion or limiting grass recoveryWardle et al., 2004). In some cases, interactions between soil biotand plants can generate feedback mechanisms among soil bioticrusts, fungi, and plants (Belnap, 1996; Collins et al., 2008; Greent al., 2008; Lucero et al., 2008; Porras-Alfaro et al., 2007, 2011;orra-Alfaro et al., 2008). Soil biotic crusts serve as an interfaceetween the soil and atmosphere capable of fixing N and C whileffecting the abundance, diversity, and successional maturity ofther soil biota (Evans and Belnap, 1999; Belnap, 2003; Garcia-ichel et al., 2003; Darby et al., 2006). Dark septate fungi (DSF)re the dominant colonizers of many native plants as well as B.riopoda in the arid southwest that have the ability to managend regulate carbon in plant communities, aid in nutrient acqui-ition and allocation, and provide increased drought tolerance toosts (Barrow, 1997; Barrow et al., 2004; Lucero et al., 2006; Porra-lfaro et al., 2008) The extension of fungal symbionts of B. eriopoda

eads us to expect that increased connectivity of vegetated patchesn grasslands harbor soil biotic communities linking vegetationnd soils biotic crusts that occupy bare ground patches. Shrubncroachment and increases in bare ground gaps that perpetu-te erosion/deposition processes causes vegetative production andecomposition of organic matter to be much more heterogeneoushat can have long term effects on nutrient availability and relegateoil biotic communities to isolated vegetated patches (Schlesingert al., 1990; Kéfi et al., 2007; Ravi et al., 2007; Scanlon et al., 2007).

The Chihuahuan Desert of North America provides an oppor-unity to evaluate these processes and feedbacks. The vegetationn sandy soils consists of perennial grasses, primarily black gramaBouteloua eriopoda) and the shrub, honey mesquite (Prosopislandulosa) that occur at variable proportions of cover across aandscape, yet total vegetation cover is similar. A key system prop-rty that differentiates grasslands from shrublands is the spatialistribution and density of plants and their associated bare soil

nterspaces (gaps) (Herrick et al., 2005; Okin et al., 2006). Theow diversity of soil organisms compared with mesic systems is

ost likely due to species operating close to their physiologicalolerance limits in the hot and dry abiotic environment (Whitford,996). Therefore, the limited functional redundancy of soil biota

n arid systems increases the susceptibility of soil biotic commu-ities to disturbances that can lead to profound impacts uponcosystem processes (Wall and Virginia, 1999). A study of plantarasitic nematodes revealed less herbivory within a duneland siteompared with grassland sites that was suggested to aid in estab-ishment of mesquite (Wall and Virginia, 1999).

Because of the ubiquity and abundance of nematodes and therophic specificity they exhibit, nematode community structureffers an efficient tool to assess soil biological function and foodeb quality (Bongers and Ferris, 1999). Specific nematode genera

re active within each heterotrophic food web level and can berouped by functional guilds whose members respond similarlyo food web enrichment and to environmental perturbation andecovery (Yeates et al., 1993; Bongers and Bongers, 1998; Bongers,

cology 58 (2012) 66– 77 67

1999; Ritz and Trudgill, 1999; Bongers and Ferris, 1999; Ferris et al.,2001; Neher, 2010).

This study relies on the quantification of nematodes with spe-cific, morphologically distinguishable feeding behaviors as a proxyfor characterizing changes in soil biota. We addressed the questionsto represent two spatial scales across a desertification gradient:(1) how do nematode communities, and presumably the soil bioticconsortia of vegetation states (black grama grasslands vs. mesquiteshrublands), change? and (2) how do nematode communities differbetween bare soil gaps and vegetated patches within each domi-nant vegetation type?

2. Methods

2.1. Study site

The study was conducted at the USDA-ARS Jornada Experimen-tal Range (JER) and Jornada Basin Long-Term Ecological ResearchSite in southern New Mexico, USA (32◦37′N, 106◦40′W, and 1260 ma.s.l.). The JER is a 100,000 ha research site located within the north-ern extent of the Chihuahuan Desert. Long-term (1915–2009) meanannual rainfall is 246 mm that is highly variable from year to yearwith more than half of the precipitation occurring between July andOctober. Mean monthly average temperatures peak in June (26 ◦C)and are lowest in January (4 ◦C). Upland grasslands on sandy soilswith a petrocalcic horizon at the JER are dominated by black grama(B. eriopoda) with mesa dropseed (Sporobolus flexuosus) and vari-ous three-awn species (Aristida spp.) as sub-dominate species. Overthe past century, large areas of the upland grasslands have con-verted to shrublands dominated by honey mesquite (P. glandulosa)(Fig. 1).

Sample locations were selected to represent both remnantblack grama patches and historic areas that were previously grass-lands and are currently dominated by mesquite. The black gramaand mesquite sites are the two end members of a desertificationgradient, with a mixture of grasses and shrubs in the ecotone(Bestelmeyer et al., 2006). The ecotone sampling area representsthe transition zone between the two end members as part ofthe desertification gradient. These transition zones, or ecotones,present optimal conditions because they are areas that have ahigh probability of response and are likely to experience thresh-old behavior, possess key processes that must be manipulated ormaintained in order to uphold current conditions or to reverse acurrent undesirable trend, and provide early warning indicatorsthat can be used to monitor and evaluate changes in critical pro-cesses (Peters et al., 2006). Because ecotones are defined by specificcharacteristics, they allow for the evaluation of the impact of deser-tification on a landscape scale interacting with fine-scale soil bioticcommunities specific to B. eriopoda and P. glandulosa.

Soils on the Jornada vary in texture from sandy loams to loamysands with an indurated calcium carbonate layer at 15 to >50 cmdepths (Hochstrasser et al., 2002). All geomorphic descriptions forthe Jornada Basin follow Monger et al. (2006). Sample locations alloccur on the same ancestral Rio Grande alluvium parent materialof the Camp Rice Formation, Fluvial Facies. These strongly devel-oped soils with thick petrocalcic horizons are components of theLa Mesa surface where wind erosion has been the main source ofquartzose sand accumulated on down-wind piedmont slopes andbedrock areas.

Nematode samples were also collected from Otero Mesa onJune 11, 2010 to be used as a complement to the desertifica-

tion gradient on the JER. Otero Mesa represents a reference siteof intact, relatively undisturbed B. eriopoda grasslands. This Chi-huahuan Desert site (32◦30′N, 105◦47′W) has the largest intact,public desert grassland in America (485,622 ha). Roughly half of

68 J.R. Klass et al. / Applied Soil Ecology 58 (2012) 66– 77

F na Ano ackag

OuPacwdlTo

2

s

ig. 1. Progression of shrub encroachment through time at the Jornada LTER site, Don the Jornada through time; courtesy of David Rachal. Created with GIS software p

tero Mesa is grassland, dominated by two species that occur onpland shallow soils, B. eriopoda and Bouteloua gracilis (blue grama).. glandulosa is virtually absent from the site and minimal wildnd domesticated large animal grazing occurs. Mean annual pre-ipitation is 255 mm/year with the majority occurring during thearm season (May–October). Soils of the sampled area are pre-ominantly silty, shallow calcareous soils derived from Paleozoic

imestone that also contains an indurated calcium carbonate layer.he mean annual precipitation is 255 mm/year with the majorityccurring during the warm season (May–October).

.2. Soil sampling and nematode extraction

Sampling at both sites occurred in two years during the dry sea-on (2009 June 11 2009; 2010 May 27) to ensure that soil biotic

a County, NM. Vegetation delimited is all grass species and all woody shrub speciese and datalayers obtained from Gibbens et al. (2005).

activity would be kept to a minimum. Each soil sample consistedof three cores randomly collected using a 7 cm diameter auger toa depth of approximately 50 cm, and composited in sterile plasticbags. A total soil volume of 1.1 l was collected from each samplinglocation within each vegetation site for nematode extraction, andto determine percentage soil moisture concentration at the timeof sampling. Soil moisture values are not reported because valuesdid not differ significantly among soil sampling sites or dates, andnematode numbers were not correlated to soil moisture, similarto previous studies (Freckman and Virginia, 1989). Grassland soilsamples were collected in three microsites: (1) beneath a black

grama patch that included the rooting zone; (2) from areas voidof aboveground vegetation that had visible soils crusts (identifiedas either of two cyannolichens, the rugose soil crust Peltula richard-sii or Collema sp.); and (3) from areas between vegetated patches

J.R. Klass et al. / Applied Soil Ecology 58 (2012) 66– 77 69

Table 1Soil sampling regime from both the Jornada and Otero Mesa sites indicating sample year, within site soil sample location, soil sampling scheme, and soil sampling strategy.Number in parentheses indicates the number of samples sampled in a giver year, numbers in sample location area breakdown of the total number of samples from 2009;2010 sampling years.

Site Jornada Otero Mesa

Sample year 2009 (33)/2010 (50) 2010 (15)Sample location Grassland (12; 15)/ecotone (9; 15)/duneland (12; 20) Undisturbed grassland (15)Within site sample scheme Bare ground/vegetated patch/crust/dune/interdune Bare ground/vegetated patch/crustSoil sampling strategies and analysis Nematode extraction and enumeration, species diversity

s, connNematode extraction and enumeration, species diversity

tpasmgpatgdfwpt

2

taasN(voOccisC(dwaTgtoamhat2

bwdait

and evenness, FB, HB, HT ratio analysianalysis.

hat had little visual evidence of soil biotic crusts. Ecotone soil sam-les were taken in areas where black grama and mesquite co-occurt varying amounts of vegetative cover. Soil samples in the ecotoneites were collected from three microsites: (1) directly beneath aesquite shrub canopy, (2) within a patch of co-occurring black

rama, and (3) from bare ground between a sampled black gramaatch and the nearest neighboring patch. Dune sampling mimicked

previously conducted pilot project that examined dune archi-ecture and identified which aspect of the dune would yield thereatest number of nematodes per sample. Based on the pilot-studyata, soil samples were taken from the top, the north and southaces of each dune as well as an interdune sample located on theindward, southern facing side of each dune sampled. All soil sam-les were placed in a cooler to regulate soil temperatures duringransport to the lab (Table 1).

.3. Nematode community analyses

Upon arrival in the laboratory, a portion of the soil sampleo be used for nematode extraction was rehydrated and storedt room temperature (23◦ C) for 48 h to hatch nematode eggsnd activate dormant, anhydrobiotic nematodes. The moistenedoil was then stored for 3–4 days at 15 ◦C prior to extraction.ematodes were extracted by elutriation-centrifugal flotation

Jenkins, 1964; Byrd et al., 1976) and quantified per 100-cm3

olume of soil. An aliquot of each extracted sample was placedn a chambered counting slide and examined on an invertedlympus IX51 compound microscope. Trophic groupings wereategorized as bacteriovores, omnivores, fungivore, herbivores,arnivores, and Tylenchinae. Herbivore species were categorizednto Tylenchinae (ectoparasite), Tylecnchorhynchus (ectopara-ite), Paratylenchus (ectoparasite), Helicotylenchus (ectoparasite),riconemella (ectoparasite), Trichodorus (ectoparasite), Hoplolaimusectoparasite), Pratylenchus (migratory endoparasite), and Meloido-era (sedentary endoparasite). In the 2010 samples, Tylenchinaeas grouped into two distinct groups based on morphological char-

cteristics and are indicated as Tylenchinae 1 and Tylenchinae 2.ylenchinae was categorized independently from other trophicroupings because of the niche breadth of this subfamily andhe multiple energy channels that constituent genera have beenbserved to feed upon. Debate has arisen in the scientific literatures to what trophic category is appropriate to use when categorizingembers of the subfamily Tylenchinae because different species

ave been observed to feed on algae, mosses, lichens (both fungalnd filamentous cyannobacterial components), and are often timeshe most abundant mycophagous species in the soil (Okada et al.,002; Okada and Kadota, 2003; Okada et al., 2005).

Herbivorous nematodes were further identified to genus foroth sampling years. A detrended correspondence analysis (DCA)as used to compare nematode counts from soils across the

esertification gradient established at the JER for both trophicnd herbivore groupings during both sampling years, a compos-te JER analysis, and in a combined analysis with samples fromhe Otero Mesa grassland soils. The DCA used downweighted rare

ectivity and evenness, FB, HB, HT ratio analysis, connectivityanalysis.

species as well as rescaled each significant axes’ with 26 segments(McCune and Mefford, 2006). To determine the significance of thenematode community loadings on individual axes derived fromeach DCA, a multiresponse permutation procedure was applied(MRPP) using Euclidean (Pythagorean) as the distance measure andc(I) = n(I)/sum(n(I)) as the weighting option (PC-ORD Version 5.32).The MRPP is a non-parametric procedure for testing the hypothesisthat no difference exists between two or more groups of entities,and provides a significance test of DCA scores for each axis. Theprocedure is therefore not held to the assumptions of multivari-ate normality and homogeneity of variance that seldom apply toecological data (Cai, 2006; McCune and Mefford, 2006). The MRPPanalyses also allow for pairwise comparisons that can be used tocompare nematode community groupings among the different veg-etation sites. DCA rescales and detrends the original axes, thereforethe traditional interpretation of eigenvalues is lost because thecorrespondence between the eigenvalue and the structure alongeach individual axis is lost. Therefore, an after-the-fact coefficientof determination between relative Euclidean distance in the unre-duced species space and Euclidiean distance in the ordination spacewas performed for each axis yielded by each individual DCA to eval-uate the effectiveness of each ordination in explaining variationwithin the data (McCune and Mefford, 2006).

The nematode fungivore/bacteriovore ratio (FB ratio) was usedto gain insight into nutrient cycling via detrital pathways and thesuccessional stage of the decomposer community. For the purposeof this analysis, we have categorized nematodes in the subfam-ily Tylenchinae as a fungivore based on previous studies that haveidentified the genera to feed upon the fungal energy pathway (e.g.,Okada et al., 2002; Okada and Kadota, 2003; Okada et al., 2005).The bacteriovore/bacteriovore + herbivore ratio (BH ratio) was cal-culated in order to gather information about the grazing nematodeand plant parasitic communities. Herbivorous nematode vs. thetotal nematode population (HT ratio) was also calculated in orderto assess the total proportion of the herbivore community amongthe three vegetation states that make up the desertification gra-dient. All three ratios were calculated for both 2009 and 2010 JERsamples. The calculated ratios were then subjected to univariate(ANOVA) analyses using LSD post hoc tests to look for differencesamong vegetation states. The FB ratio was log transformed to meetthe assumptions of ANOVA.

Shannon’s index of diversity (H) was used to assess the nema-tode diversity at each location across the established desertificationgradient and Otero Mesa soils for both trophic groups and herbi-vore genera. Index values were then averaged by vegetation state.Species evenness (E) was calculated using the same desertificationgradient on the JER and Otero Mesa soils in order to look at the equi-tability of nematode communities associated with each vegetationstate.

Multivariate (MANOVA) analyses using Wilks’ Lambda multi-

variate test statistic were conducted within sampled vegetationstates to compare bare areas with vegetated areas. For this analysis,both trophic groupings and herbivore genera from both samplingyears were grouped and analyzed separately Grassland samples

7 Soil E

cJvu

3

3

bc(tetrssvtTetgTvMnaAt(ctv(vlsvt(b

rwsi(tgnrntEEvpvpA2f

0 J.R. Klass et al. / Applied

ollected from Otero Mesa were included with the data from theER for this analysis. Post hoc tests were conducted to look at indi-idual species and trophic grouping differences within sample sitessing least significant differences (LSD).

. Results

.1. Desertification analysis

Each vegetation state along the desertification gradient har-ored distinct nematode communities (Fig. 2a and b). The MRPPonducted on the DCA trophic scores validated these differencesoverall: A = 0.237, p < 0.001; pairwise comparisons: dune vs. eco-one: A = 0.129, p = 0.004; dune vs. grassland: A = 0.279, p < 0.001;cotone vs. grassland: A = 0.108, p = 0.004). The effectiveness ofhe ordination decreased as each higher axis was added (Axis 1,2 = 0.230; Axis 2, r2 = −0.046; Axis 3, r2 = −0.015). Trophic diver-ity was found to be significantly different among vegetationtates (F = 5.289, p = 0.011, r2 = 0.261) and highest in the ecotoneegetation state (H = 1.106) followed by the mesquite vegeta-ion state (H = 0.772) and the black grama grassland (H = 0.680).rophic evenness was not found to be significant among veg-tation states (F = 1.703, p = 0.199, r2 = 0.102) but was showno be highest in the ecotone (E = 0.769) followed by the blackrama grassland (E = 0.644) and the mesquite duneland (E = 0.596).he herbivore analysis on the 2009 JER samples also showedery distinct groupings of species with vegetation (Fig. 2b). TheRPP confirmed that there were significant differences in the

ematode community associated with each vegetation state (over-ll: A = 0.430, p < 0.001; pairwise comparisons: dune vs. ecotone

= 0.174, p < 0.001; dune vs. grassland: A = 0.473, p < 0.001; eco-one vs. grassland: A = 0.430, p < 0.001). Axis 1 (r2 = 0.169) and Axis 2r2 = 0.063) cumulatively accounted for 23.1% of the variance asso-iated with the herbivore species data, where the addition of ahird axis (r2 = −0.014) provided no additional information. Herbi-ore diversity was significantly different among vegetation statesF = 7.836, p = 0.002, r2 = 0.359) with the highest in the ecotoneegetation state (H = 1.24) followed by the black grama grass-and (H = 0.639) and the mesquite duneland (H = 0.542). Herbivorepecies evenness was found to be significantly different amongegetation states (F = 4.235, p = 0.025, r2 = 0.232) and highest inhe ecotone (E = 0.850) and lowest in the black grama grasslandE = 0.558) with evenness in the mesquite duneland intermediateetween the other two vegetation states (E = 0.0603).

The desertification gradient sampled on the JER in 2010 showedesults similar to the 2009 sampling in that nematode communitiesere found to be significantly different among the three vegetation

tates except with the ecotone and mesquite duneland compar-son were not significantly different, as indicated by the MRPPoverall: A = 0.158, p < 0.001; pairwise comparisons: dune vs. eco-one: A = 0.030, p = 0.064; dune vs. grassland: A = 0.230, p < 0.001;rassland vs. ecotone: 0.088, p = 0.006). Trophic diversity was sig-ificantly different among vegetation states (F = 6.066, p = 0.005,

2 = 0.205) and highest in the ecotone (H = 0.952) as was even-ess (E = 0.740). B. eriopoda grassland trophic diversity was foundo be intermediate in both diversity and evenness (H = 0.742;

= 0.684) between the ecotone and mesquite duneland (H = 0.580; = 0.647) where there evenness was not significant among all threeegetation states (F = 0,228, p = 0.797, r2 = 0.010). Herbivore sam-ling also showed distinct nematode communities associated withegetation state (Fig. 4b; overall: A = 0.158, p < 0.001; pairwise com-

arisons: dune vs. ecotone A = 0.090, p = 0.001; dune vs. grassland:

= 0.214, p < 0.001; ecotone vs. grassland: A = 0.053, p = 0.012). The010 samples yielded two predominant genera within the sub-amily Tylenchinae that were classified separately. However, little

cology 58 (2012) 66– 77

information can be gathered on herbivore community assem-blages based on the amount variation explained by the first andadditional axes from the DCA analysis (Axis 1, r2 = 0.014; Axis2, r2 = −0.014; Axis 3, r2 = 0.000). High herbivore diversity wasobserved in both the ecotone (H = 0.972) and the black grama grass-land (H = 0.908) and substantially decreased within the mesquiteduneland sites (H = 0.219) and was significantly different amongvegetation states (F = 19.791, p < 0.001, r2 = 0.457). Herbivore even-ness was significantly different among vegetation states (F = 6.229,p = 0.004, r2 = 0.229) and highest in the ecotone (E = 0.845) followedby the black grama grasslands (E = 0.701) and substantially lowerin the dune vegetation states (E = 0.288).

When Otero Mesa black grama grassland samples were includedin the 2010 trophic and herbivore DCA, groupings aligned with theblack grama nematode communities taken from the JER withinordination space (Fig. 3a and b), however, differences emergewhen the MRPP data is considered. Trophic analysis shows theOtero Mesa (OM) nematode community structure to be signifi-cantly different when compared with the desertification gradientsampled on the JER, including the black grama grassland (overall:A = 0.232, p < 0.001; pairwise comparisons: dune vs. OM: A = 0.352,p < 0.001, grassland vs. OM: A = 0.035, p = 0.029, ecotone vs. OM:A = 0.178, p < 0.001). Except for the first axis, no cumulative varia-tion was explained by the successively higher axes (Axis 1, r2 = 0.12,Axis 2, r2 = −0.26, Axis 3, r2 = −0.24). Nematode trophic diversitywas significantly different among all vegetation states (F = 25.433,p < 0.001, r2 = 0.556) the highest recorded value among all samplelocations (H = 0.978) and species evenness was slightly lower thatthe ecotone vegetation state sampled on the Jornada (E = 0.714)but found not to be significant among sample locations (F = 0.194,p = 0.900, r2 = 0.010). Herbivore communities were found to be verysimilar among Otero Mesa and JER black grama grassland sam-ples (Fig. 3b) and the pairwise comparison of MRPP scores indicatethat these two communities are not significantly different (overall:A = 0.125, p = <0.001; pairwise comparisons: dune vs. OM: A = 0.179,p < 0.001; grassland vs. OM: A = 0.004, p = 0.301; ecotone vs. OM:A = 0.052, p = 0.017). Again, two distinct Tylenchinae genera wereobserved and enumerated separately. Variation increased with theinclusion of the Otero Mesa samples, where the first and higher axesaccounted for minimal variance (Axis 1, r2 = 0.07, Axis 2, r2 = 0.06,Axis 3, r2 = −0.04). The Otero Mesa soils had the highest measureof herbivore species diversity of all samples analyzed (H = 1.242)and species evenness was relatively high as well (E = 0.776). WhenOtero Mesa samples were compared with the Jornada herbivorespecies diversity and evenness analysis, all vegetation states werefound to significantly differ (H: F = 25433, p < 0.001, r2 = 0.556; E:F = 5.966, p = 0.001, r2 = 0.242).

When the two sampling years of the desertification gradienton the JER were combined to form a composite data set, the samegeneral patterns were observed as in the individual sampling yearanalyses—e.g., significant differences in both trophic and herbi-vore groupings among vegetation states (Fig. 4a and b; TrophicMRPP Analysis: overall: A = 0.190, p < 0.001; pairwise comparisons:dune vs. ecotone, A = 0.066, p = 0.001; dune vs. grassland, A = 0.254,p < 0.001; ecotone vs. grassland: A = 0.095, p < 0.001; HerbivoreMRPP Analysis: overall: A = 0.127, p < 0.001; pairwise comparisons:dune vs. ecotone: A = 0.084, p < 0.001, dune vs. grassland: A = 0.129,p < 0.001; ecotone vs. grassland: A = 0.075, p < 0.001). The variationexplained was only relegated to Axis 1 and Axis 2 (Axis 1, r2 = 0.06,Axis 2, r2 = 0.044, Axis 3, r2 = −0.03). There was some inter-annualvariability in nematode communities from one year to the next,with the 2009 sampling year yielding significantly more nematodes

in both trophic (Wilk’s Lambda, F = 3.937, Value = 0.514, p = 0.007)and herbivore (Wilk’s Lambda, F = 2.903, Value = 0.699, p = 0.040)analyses for the dune vegetation state as well as the herbivoreanalysis for the ecotone (Wilk’s Lambda, F = 2.714, Value = 0.409,

J.R. Klass et al. / Applied Soil Ecology 58 (2012) 66– 77 71

F cationb

pC(HnltphpM2

otomTswHug

the three vegetation states on the JER (Table 2). Post hoctests revealed that the FB ratio was significantly differentbetween the black grama grassland and the mesquite duneland(LSD, p = 0.006) with more fungivores present in the grassland

Table 2Univariate analysis of calculated nematode ratios from the Jornada for both 2009 and2010 sampling years combined as well as the Otero Mesa sampling from 2010: fungi-vore:bacteriovore ratio (FB), herbivore:bacteriovore ratio (HB), and herbivore:totalnematode population ratio (HT). All calculated ratios significantly differed among allvegetation states along the desertification gradient and the undisturbed grasslandsamples.

Ratio F p R2

ig. 2. DCA of nematode trophic (a) and herbivore sampling (b) of the JER desertifiottom panels for both trophic and herbivore groupings.

= 0.046) vegetation state where there were significantly morericonemoides spp. (Value = 7.508; p = 0.012) and Hoplolaimus spp.Value = 13.569; p = 0.001) were recovered in 2009 than 2010.owever, there were no significant differences observed in theematode communities associated with the black grama grass-

ands from one year to the next. The dune trophic analysis showedhat significantly greater numbers of bacteriovores (F = 15.451,

< 0.001, r2 = 0.340), tylenchus (F = 9.793, p = 0.004, r2 = 0.246), anderbivores (F = 18.040, p < 0.001, r2 = 0.376) in 2009 than 2010 sam-les and in the herbivore analysis there were significantly moreeloidodera spp. (F = 11.317, p < 0.001, r2 = 0.303) in 2009 than in

010.In the 2010 herbivore sampling, two distinctly different genera

f Tylenchinae were observed based on morphological characteris-ics, but have yet to be definitively classified. However, for purposesf this analysis these two genera were combined into the com-on subfamily Tylenchinae due to similarity in trophic behavior.

rophic and herbivore species diversity and evenness was also con-istent with results from the 2009 analyses in that the ecotone

as found to be the most diverse (Trophic, H = 1.010; Herbivore,

= 1.028) as well as having the most evenly distributed species val-es (Trophic, E = 0.751; Herbivore, E = 1.144) followed by the blackrama grassland (Trophic, H = 0.715, E = 0.749; Herbivore, H = 0.751,

gradient in 2009. Nematode species DCA scores are represented as vectors on the

E = 0.679) and lastly the mesquite duneland site (Trophic, H = 0.651,E = 0.628; Herbivore, H = 0.340, E = 0.406). Trophic diversity wasfound to be significantly different among vegetation states whenthe composite Jornada samples were compared (F = 9.236, p < 0.001,r2 = 0.190) but evenness was not (F = 1.238, p = 0.295, r2 = 0.030).

Overall, all three calculated ratios differed significantly among

FB 12.886 <0.001** .213HB 39.356 <0.001** .453HT 38.134 <0.001** .445

** Indicates significance at the = 0.001 level.

72 J.R. Klass et al. / Applied Soil Ecology 58 (2012) 66– 77

F cationa els fo

vTc(vrttgtaa

3

baTtsta

ig. 3. DCA of nematode trophic (a) and herbivore sampling (b) of the JER desertifis well. Nematode species DCA scores are represented as vectors on the bottom pan

egetation state (dune = 0.096; ecotone = 0.273; grassland = 0.373).he BH ratio differed among all vegetation states where signifi-antly more bacteriovores were found in the dune (0.225) > ecotone0.533) > grassland (0.714) (LSD, dune vs. grassland: p < 0.001, dunes. ecotone: p = 0.029, grassland vs. ecotone, p = 0.001). The HTatio also showed significant differences in herbivorous nema-odes across all vegetation states with the most herbivores inhe grassland (0.232) > ecotone (0.369) > dune (0.649) (LSD, dune vs.rassland: p < 0.001, dune vs. ecotone: p = 0.023, grassland vs. eco-one: p < 0.001). Because the data was log transformed to meet thessumptions of ANOVA, post hoc tests were not performed becauset least one group had fewer than two cases.

.2. Plant-interspace analysis

The MANOVA analysis was conducted on both trophic and her-ivore nematode groupings analyzed across microsite locationsmong the vegetation states along the desertification gradient.he grassland samples showed no significant differences in nema-

ode trophic and herbivore groupings among black grama patches,oil gaps with soil biotic crusts, and bare ground across loca-ions (Table 3). Multivariate tests showed no differences in trophicnd herbivore groupings among sampling locations within the

gradient in 2010 as well as grassland samples from Otero Mesa sampling in 2010r both trophic and herbivore groupings.

ecotone vegetation state (Table 3). The between subject effects ofthe herbivore analysis did reveal a differences in the distribution ofHoplolaimus species (F4,22 = 3.144, p = 0.040, r2 = 0.411) where posthoc tests showed significantly more species associated with sam-ples taken from soils with co-occurring mesquite and black grama(LSD, p = 0.048). No significant differences were detected in the eco-tone trophic analysis among sampling locations (Table 3). However,bacteriovores were found to be significantly more prevalent withinsoils associated with P. glandulosa than B. eriopoda or where thesespecies co-occurred (F4,23 = 5.904, p = 0.003, r2 = 0.550).

Only bacteriovore trophic groupings in dune vegetation weresignificantly different from interdune areas (F3,32 = 2.970, p = 0.049,r2 = 0.241), and no overall effect of sample location was observed(Table 3). However, when sampling location on the dune is notaccounted for and trophic groupings are compared between duneand interdune bare ground gap samples, numbers of bacteriovoresand Tylenchinae genera differed significantly with fewer num-bers in the interdune areas (Bacteriovores, F1,32 = 9.458, p = 0.004,r2 = 0.24; Tylenchinae, F1,32 = 7.081, p = 0.012, r2 = 0.191). No omni-

vores, fungivore, or carnivores were found in the interdune areasand herbivores were lower when compared with the mesquitedune nematode community. Herbivore communities were signif-icantly different among samples taken from the mesquite dune

J.R. Klass et al. / Applied Soil Ecology 58 (2012) 66– 77 73

F catioN oth tr

atcTtiba

TMLa

ig. 4. DCA of nematode trophic (a) and herbivore sampling (b) of the JER desertifiematode species DCA scores are represented as vectors on the bottom panels for b

nd the interdune areas as well (Table 3). Only four species ofhe nine genera of herbivores identified throughout the desertifi-ation gradient (Meloidodera, Tylenchorhynchus, Paratylenchus, andylenchinae) resided within the dune vegetation state. All nema-

ode herbivores associated with the dune vegetation state occurredn much lower numbers, if at all, within interdune areas. Num-ers of Tylenchorhynchus spp. were significantly lower in interdunereas (LSD, p = 0.034) as were Tylenchinae spp. (LSD, p = 0.020). No

able 3ANOVA analysis of the effects of sampling location for each of the three vegetation state

ambda. Numbers within parentheses in the dune vegetation are indicative of nematode cnd only relegated samples to either dune or interdune samples.

Vegetation Analysis F

GrasslandTrophic 0.810

Herbivore 0.720

EcotoneTrophic 1.348

Herbivore 1.435

DuneTrophic 1.162 (2.008)

Herbivore 2.123 (0.904)

* Indicates significance at the = 0.05 level.

n gradient in 2010 and grassland samples from the Otero Mesa sampling in 2010.ophic and herbivore groupings.

Paratylenchus spp. was found in the interdune areas, and low recov-ery of this genus within the dune itself failed to support significantdifferences among locations.

4. Discussion

The extensive biodiversity retained within soils is fundamen-tal to maintain plant productivity, yet soils as a living natural

s on nematode groupings along the desertification gradient on the JER using Wilk’sommunities where the multivariate analysis did not account for dune architecture

Value df p

0.716 15 0.6640.718 64 0.764

0.176 24 0.1840.066 32 0.135

0.457 (0.675) 18 (6) 0.318 (0.103)0.340 (0.717) 15 (10) 0.019* (0.537)

7 Soil E

rreBndbcIrwntta(

4

edpnfsLftf

rhrlthtaebt

vdgprwabbmi

abatpbt(if1

4 J.R. Klass et al. / Applied

esource are often ignored in issues of state changes where onlyudimentary knowledge of soil biotic diversity is known for a fewcosystem types (Brussaard et al., 1997; Wall and Virginia, 1999;ardgett et al., 2005; Sylvain and Wall, 2011). Our analysis ofematode community structure supports the hypotheses that theesertification of B. eriopoda grasslands through the encroachmenty mesquite shrubs has altered nematode communities, andaused them to become more heterogeneous and less diverse.ncreases in bare ground gaps that accompany desertification haveelegated the majority of nematodes to communities associatedith vegetated patches. Furthermore, the incidence of specificematode genera aligning with vegetation states along our deser-ification gradient support the “fungal loop” model proposed toranslocate C, N, P, and mineral nutrients among resources patchesnd sustain ecosystem productivity in similar grassland systemsCollins et al., 2008; Green et al., 2008).

.1. Desertification effect

The high nematode diversity observed in the ecotone can bexplained by ecological niche partitioning. Increases in resourcesue to multiple plant life forms (grasses and shrubs) may haverovided more microhabitats and energy channels for increasedematode diversity and subsequently, soil biota that has increased

unctional redundancy caused by decreases in competitive exclu-ion (Armstrong and McGehee, 1980; Freckman and Virginia, 1989).ess diverse, more specialized nematode communities divergerom ecotone community assemblages and align with the vegeta-ion specific to either B. eriopoda or P. glandulosa and occupy specificunctional roles resulting in increased competition.

Our results show that once P. glandulosa has encroached on, andeplaced areas previously dominated by grasses, the species guild oferbivorous nematodes is dominated by very specific species withelatively few remaining representatives of the B. eriopoda grass-and. Furthermore, trophic complexity becomes largely relegatedo first order grazers (bacteriovores) in the dunelands and becomesighly dependent on the spatial distribution of fragmented vege-ation patches. These shifts in dominance to a few genera signifylterations in plant–soil biotic associations within the differing veg-tation states. However, it is not clear how this change in the soiliotic community becomes initiated or what the direct effects tohe plant community.

Further support for nematode communities aligning withegetation distributions is apparent in the lack of significantifferences between herbivore communities from the JER blackrama sites and the Otero Mesa black grama grassland sam-les. High diversity observed within the Otero Mesa site may beelated to the lack of impact from long-term grazing combinedith a more stable geomorphic surface. Because these two sites

re geographically separated by approximately 160 km, yet har-or similar herbivore/plant-parasitic communities, illustrates theroad landscape-scale distribution patterns of nematode com-unities in the Southwest and are perhaps indicative of the

mportance of soil biofeedbacks in semi-arid grasslands.Native plants are more resistant and tolerant to nematode par-

sitism when compared to plants grown in agricultural settingsased on evolutionary histories and co-evolution, competition,rbuscular mycorrhizae and endophytic fungi that act as biocon-rol agents (Wallace, 1987; Neher, 2010). Furthermore, lateral rootroduction has been shown by plants responding to herbivoryy nematode populations along with increased root exudationhat can affect plant nutrient uptake as well as nutrient cycling

Freckman and Virginia, 1989; Moore et al., 2003). Also, wound-ng of roots by nematodes may provide increased infection sitesor symbiotic fungal infections to occur (Freckman and Virginia,989), thus perpetuating the connectivity among plant patches in

cology 58 (2012) 66– 77

black grama grasslands, aiding in drought tolerance, and enhancingnutrient uptake from shallow calcareous soils. Coupled with theseeffects, plant parasitic nematodes are involved in the transfer ofphotosynthetically fixed carbon and organic compounds into therhizosphere that affect nutrient dynamics and microbial biomassthat ultimately have the potential to have ecosystem level effects(Ingham and Coleman, 1983; Ingham et al., 1985; Bardgett et al.,1998; Yeates, 1999). The sheer number of plant parasitic nema-todes found to reside within the black grama grasslands sampledin this study does not appear to be severely impacting the produc-tion of these grasslands, especially those occurring on Otero Mesa.However, the coupled affect of large herbivorous nematode pop-ulations, drought, intermittent disturbances, and low fecundity ofblack grama (Peters, 2002b) could combine to have a detrimentaleffect on black grama and may be attributing to the desertificationof this arid grassland. Further information is needed to quantify thelarger impacts of plant parasitic nematodes on plant production inthis system and natural systems as a whole.

4.2. Plant-interspace effects

With the lowest observed diversity and reduced numbers of her-bivores genera, significant decline in bacteriovores, as well as theabsence of Tylenchinae, omnivores, fungivores, and carnivores withthe interdune areas compared with mesquite dunes themselvesillustrates the influence of vegetation on nematode communitystructure. The loss in connected vegetation and increases in inter-space gaps characteristic of the desertification of arid grasslandshas caused nematode abundance to become patchier. This frag-mentation has undoubtedly affected prey guilds as indicated bynematode community assemblages and their spatial distributionswithin the shrub dominated vegetation state (Fierer and Jackson,2006; Housman et al., 2007). Due to limitations in their coloniza-tion abilities, nematode community assemblages are indicativeof shifts in species dominance and because of their central rolein food-web dynamics, these changes in nematode communityassemblages serve as a corollary for changes in the soil biotic con-sortia once shrubs become dominant further altering ecosystemprocesses (Kardol et al., 2005).

The high proportion of bacterial feeding nematodes foundwithin the mesquite vegetation state has the potential to severelyalter the nutrient dynamics, especially nitrogen, both directlyand indirectly and may be accentuating the “islands of fertility”observed by Schlesinger et al. (1990). Studies have shown that dis-turbance has lead to shifts in the microbivorous nematodes frompredominantly fungivores to predominantly bacteriovores mirror-ing energy channel shifts (Hendrix et al., 1986; Moore and de Ruiter,1991; Darby et al., 2010). These shifts are important because nema-todes directly secrete ammonium as a byproduct of metabolismof prey guilds that possess lower C:N ratios than required bynematodes themselves (Neher, 2010). This excess nitrogen is read-ily available for plant uptake but also can stimulate increasedgrowth of microbial populations and lead to enhanced microbialturnover and dispersal where moderate grazing by nematodes caneven further stimulate microbial growth (Yeates, 1999; Neher,2010). Increased N mineralization rates in the bacterial decom-position channel compared to the fungal decomposition channellead to greater losses of nitrogen through volatilization. Therefore,systems with high bacterial populations may be prone to largenitrogen deficits when compared with other soil microbial systems(Peterjohn and Schlesinger, 1991; Darby et al., 2010; Neher, 2010).

The fungivore to bacteriovore ratio (FB ratio) has been widely

accepted to yield insight into the successional stage of the decom-poser community. However, the incidence of the translocation ofC from plant patches to adjacent crust patches via DSF and sapro-torphic fungi and with the biotic fixation, increased mineralization

Soil E

at(tbotClpbat

4

f2n2wutinrahcd

aTtblmtdeoshmvmbsa2coocch

aadihcpc

J.R. Klass et al. / Applied

nd mobilization of N by the cyannobacterial component of crusts,he FB ratio is indicative of other process outside of decompositionEvans and Ehleringer, 1993; Hawkes, 2003; Green et al., 2008). Ashe interactions of fungi and environmental conditions above andelow ground limit organic nutrient accumulation, the net effectsf a highly connected ecosystem exist in a tightly integrated sys-em of primary producers and heterotrophs (Collins et al., 2008;renshaw et al., 2008). While nutrient levels in black grama grass-

ands are much lower when compared with duneland vegetationatches, the increased connectivity among vegetation patches andiotic crusts in arid grasslands has the potential to retain nutrientsnd positively feedback amongst the biotic pathways classified byhe fungal loop model and distribution of soil biota.

.3. Interpretation of Tylenchinae

In this study, members of the subfamily Tylenchinae accountedor 16% and 17% of the soil nematode communities in 2009 and010, respectively. Other studies have shown that the Tylenchi-ae account for up to 30% of such communities (Okada and Kadota,003). Differing trophic classifications have important implicationshen summarizing the ecological roles of Tylenchinae or whensing indices based on arbitrarily assigned trophic classificationso assess soil quality, ecosystem function, nutrient cycling, and thempact of disturbance (Wright and Coleman, 2000). Tylenchinae areot generally considered to be significant root parasites and theirelative prominence in this study may provide more information asn indicator of specific ecosystem processes such as the fungal loopypothesis (Siddiqi, 2000; Collins et al., 2008). As nematodes play aentral role in soil food web dynamics, nematode assemblages areistributed in patches centered on available resources.

It appears that desertification may disrupt vital plant–soil inter-ctions that are characteristic of B. eriopoda grasslands whereylenchinae feed mainly upon fungi, algae, and lichens and wherehere are no differences in the distribution of this subfamilyetween vegetated patches and plant-interspaces in these grass-

ands (Siddiqi, 2000; Okada et al., 2005). However, in vegetatedesquite dunes Tylenchinae numbers were significantly lower

han in black grama grasslands and virtually absent from inter-unal areas in the mesquite dominated vegetation state that isssentially void of soil biotic crusts and provides further evidencef the disruption of vital soil biofeedbacks by desertification andhrub establishment. If Tylenchinae were feeding primarily onigher plants in these arid systems, we would expect the com-unities to be similar among vegetation states or at least among

egetated patches. Furthermore, mesquite dunes contain an assort-ent of associated plants such as saltbush (Atriplex canescens),

room snakeweed (Gutiereza sarothrae), dropseeds (Sporoboluspp.), threeawns (Aristida spp.), and a variety of forbs that serves food sources for herbivorous nematodes (Peters and Gibbens,006). The greater abundance of Tylenchinae where soil bioticrusts are present reinforces grouping Tylenchinae separately fromther conventional herbivore genera and supports their utilizationf photosynthetically active algal and cyannobacterial lichenousomponents as well as the fungal symbiont that comprise theserusts as substrate opposed to feeding largely as true parasites ofigher plants.

While some members of the Tylenchinae may occasionally feeds herbivores, as their low numbers within the mesquite dunes thatre void of soil biotic crusts may indicate, we believe that the abun-ance of the Tylenchinae community in sites lacking vascular plants

s indicative that they are typically not feeding as conventional

erbivores in this study. The consolidated matrices of soil bioticrusts characteristic of areas vegetated with black grama and thelant-interspaces between create a more homogenous, connectedover in the grasslands allowing for a more evenly distributed

cology 58 (2012) 66– 77 75

Tylenchinae community. Also, DSF endophyte extensions fromblack grama patches involved in the translocation of C and Nare linked to soil biotic crusts in plant-interspaces reinforcing theeven distribution of Tylenchinae among vegetated and bare-groundpatches (Green et al., 2008). We acknowledge that Tylenchinae arenot feeding on the traditional AMF and ectomycorrhizal energychannels as many indices suggest, but feed on the endophytic darkseptate fungi and soil biotic crust components of these semi aridgrasslands (Kohn and Stasovski, 1990; Barrow, 1997; Barrow et al.,2004; Porra-Alfaro et al., 2008; Herrera et al., 2010). These fac-tors led us to calculate the FB ratio in this study after classifyingmembers of the Tylenchinae as fungivores.

5. Summary and conclusion

While many unknowns exist in the evaluation of nematodepopulations within natural systems, previous research along withresults from this study provide information in the role of soil biotain desertification of arid grasslands. The shift in soil communitystructure and the subsequent change in soil biotic energy path-ways, associated with desertification can lead to ecosystem levelalterations of nutrient cycling and changes in aboveground pro-duction (Bardgett et al., 1998). As desertification increases, soilbiotic populations become less diverse and more heterogeneouslydistributed in vegetation patches that further alter nutrient dynam-ics on an ecosystem scale (Schlesinger et al., 1990). Furthermore,loss of the associated soil biotic consortia that is intimately tiedto ecosystem function of semi-arid grasslands may have real andlong-term consequences for the sustainability and recovery of B.eriopoda grasslands.

A more diverse trophic structure, as observed in the OteroMesa soils, may be indicative of stability in ecosystem interac-tions between roots and the fungal and bacterial energy channels(Moore et al., 2003). Dominance of any one energy channel mightbe indicative of dynamic instability as observed within the JERgrassland (fungal) and the mesquite duneland (bacterial), whereunimpeded vertical energy flow may provide increased nutrientsupplies, nutrient imbalance or losses, and a change in facilitativesoil biotic communities vital to plant production and resilience inarid systems (Moore et al., 2003; Darby et al., 2007, 2010). How-ever, the dominance of the fungal energy channel in black gramagrasslands on the Jornada may be indicative of a fully functioning,connected system of hyphal networks, plant patches and soil bioticcrusts.

Linking nematode populations and trophic structure to functionat individual plant scales is more complex than just enumera-tion and identification because of their central role in food websand connections to ecological processes in soil (Neher, 2010). Inorder for nematode community and trophic analyses to be moreuseful, a more complete understanding of feeding habits and tol-erance levels of each dominant nematode genus to abiotic stressesare needed before soil nematode community indices can be inter-preted and calibrated accurately as indicators of arid land soilcondition, function, and soil biotic interactions (Darby et al., 2007).This improved understanding will require improved species levelidentification using morphological and molecular criteria, wherelife history information relating to specific environments can beapplied to community level surveys.

Acknowledgements

We would like to thank the Nematology Lab at New MexicoState University for their assistance in data collection and sam-ple processing, Dr. Ernest Bernard, and Zachary Sylvain for criticalreview of our manuscript. Funding support was provided by the

7 Soil E

No0

R

A

A

AB

B

B

B

B

B

B

B

B

B

B

B

B

C

C

C

D

D

D

E

E

F

F

F

G

G

G

H

H

6 J.R. Klass et al. / Applied

ational Science Foundation to New Mexico State University as partf the Jornada Basin Long Term Ecological Research Program (DEB-618210) and the New Mexico Agricultural Experiment Station.

eferences

brahams, A.D., Parsons, A.J., Wainwright, J., 1995. Effects of vegetation changeon interill runoff and erosion, Walnut Gulch, southern AZ. Geomorphology 13,37–48.

rcher, S., 1994. Woody plant encroachment into southwestern grasslands andsavannas: rates, patterns, and proximate causes. In: Vacra, M., Laycock, W.,Pieper, R. (Eds.), Ecological Implications of Livestock Herbivory in the West.Society for Range Management, Denver, CO, pp. 13–68.

rmstrong, R.A., McGehee, R., 1980. Competitive exclusion. Am. Nat. 115, 151–170.ardgett, R.D., Wardle, D.A., Yeates, G.W., 1998. Linking above-ground and below-

ground interactions: how plant responses to foliar herbivory influence soilorganisms. Soil Biol. Biochem. 30, 1867–1878.

ardgett, R.D., Yeates, G.W., Anderson, J.M., 2005. Patterns and Determinants of SoilBiological Diversity. Cambridge University Press, Cambridge, UK.

arrow, J.R., 1997. Processing of mycrobacterial lipids and effects on host respon-siveness. Front. Biosci. 2, 387–400.

arrow, J.R., Osuna-Avila, P., Reyes-Vera, I., 2004. Fungal endophytes intrinsicallyassociated with micropropagated plants regenerated from native Bouteloua eri-opoda Torr. and Atriplex canescens (Pursh) Nutt. In Vitro Cell. Dev. Biol.-Plant 40,608–612.

estelmeyer, B.T., Ward, J.P., Havstad, K.M., 2006. Soil-geomorphic heterogeneitygoverns patchy vegetation dynamics at an arid ecotone. Ecology 87, 963–973.

elnap, J., 1996. Soil surface disturbances in cold deserts: effects on nitrogenaseactivity in cyanobacterial-lichen soil crusts. Biol. Fertil. Soils 23, 362–367.

elnap, J., 2003. The world at your feet: desert biological soil crusts. Front. Ecol.Environ. 1, 181–189.

ever, J.D., Westover, K.M., Antonovics, J., 1997. Incorporating the soil communityinto plant population dynamics: the utility of the feedback approach. J. Ecol. 85,561–573.

ongers, T., Ferris, H., 1999. Nematode community structure as a bioindicator inenvironmental monitoring. Trends Ecol. Evol. 14, 224–228.

ongers, T., 1999. The maturity index, the evolution of nematode life history traits,adaptive radiation and cp-scaling. Plant Soil 212, 13–22.

ongers, T., Bongers, M., 1998. Functional diversity of nematodes. Appl. Soil Ecol. 10,239–251.

russaard, L., Behan-Pelletier, V.M., Bignell, D.E., Brown, V.K., Didden, W., Folgarait,P., Fragoso, C., et al., 1997. Biodiversity and ecosystem functioning in soil. Ambio26, 563–570.

yrd, D.W., Barker, K.R., Ferris, H., Nusbaum, C.J., Griffin, W.E., Small, R.H., Stone,C.A., 1976. Two semi-automatic elutriators for extracting nematodes and certainfungi from soil. J. Nematol. 8, 206–212.

ai, L., 2006. Multi-response permutation procedure as an alternative to the analysisof variance: an SPSS implementation. Behav. Res. Methods 38, 51–59.

ollins, S.L., Sinsabaugh, R.L., Crenshaw, C., Green, L., Porras-Alfaro, A., Stursova, M.,Zeglin, L.H., 2008. Pulse dynamics and microbial processes in aridland ecosys-tems. J. Ecol. 96, 413–420.

renshaw, C., Lauber, C.L., Sinsabaugh, R.L., Stavely, L.K., 2008. Fungal control ofnitrous oxide production in semiarid grassland. Biogeochemistry 87, 17–27.

arby, B.J., Housman, D.C., Zaki, A.M., Shamount, Y., Adl, S.M., Belnap, J., Neher,D.A., 2006. Effects of altered temperature and precipitation on desert protozoaassociated with biological soil crusts. J. Eukaryot. Microbiol. 53, 507–514.

arby, B.J., Neher, D.A., Belnap, J., 2007. Soil nematode communities are ecologicallymore mature beneath late-than early-successional stage biological crusts. Appl.Soil Ecol. 35, 203–212.

arby, B.J., Neher, D.A., Belnap, J., 2010. Impact of biological soil crusts and desertplants on soil microfaunal community composition. Plant Soil 328, 421–431.

vans, R.D., Ehleringer, J.R., 1993. A break in the nitrogen cycle of aridlands: evidencefrom �15N of soils. Oecologia 94, 314–317.

vans, R.D., Belnap, J., 1999. Long-term consequences of disturbance on nitrogendynamics in an arid ecosystem. Ecology 8, 150–160.

erris, H., Bongers, T., De Goede, R.G.M., 2001. A framework for soil web diagnostics:extension of the nematode faunal analysis concept. Appl. Soil Ecol. 18, 13–29.

ierer, N., Jackson, R.B., 2006. The diversity of biogeogrpahy of soil bacterial com-munities. Proc. Natl. Acad. Sc. U.S.A. 103, 626–631.

reckman, D.W., Virginia, R.A., 1989. Plant-feeding nematodes in deep-rootingdesert ecosystems. Ecology 70, 1665–1678.

arcia-Pichel, F., Johnson, S.L., Youngkin, D., Belnap, J., 2003. Small-scale verticaldistribution of bacterial biomass and diversity in biological soil crusts from aridlands in the Colorado Plateau. Microbial Ecol. 46, 312–321.

ibbens, R.P., McNeely, R.P., Havstad, K.M., Beck, R.F., Nolen, B., 2005. Vegetationchange in the Jornada Basin 1858 to 1998. J. Arid Environ. 61, 651–668.

reen, L.E., Porras-Alfaro, A., Sinsabaugh, R.L., 2008. Translocation of nitrogen andcarbon integrates biotic crust and grass production in desert grasslands. J. Ecol.96, 1076–1085.

awkes, C.V., 2003. Nitrogen cycling mediated by biological soil crusts and arbus-cular mycorrhizal fungi. Ecology 84, 1553–1562.

endrix, P.F., Pamelee, R.W., Crossley Jr., D.A., Coleman, D.C., Odum, E.P., Groffman,P.M., 1986. Detritus food webs in conventional and no-tillage agroecosystems.Bioscience 36, 374–380.

cology 58 (2012) 66– 77

Herrera, J., Khidir, H.H., Eudy, D.M., Porras-Alfaro, A., Natvig, D.O., Sinsabaugh, R.L.,2010. Shifting fungal endophyte communities colonize Bouteloua gracilis: effectof host tissue and geographical distribution. Mycologia 102, 1012–1026.

Herrick, J.E., Bestelmeyer, B.T., Archer, S., Tugel, A.J., Brown, J.R., 2005. An inte-grated framework for science-based arid land management. J. Arid Environ. 65,319–335.

Hochstrasser, T.G., Kroel-Dulay, Peters, D.P.C., Gosz, J.R., 2002. Vegetation and cli-mate characteristics of arid and semiarid grasslands in North America and theirbiome transition zone. J. Arid Environ. 51, 55–78.

Housman, D.C., Yeager, C.M., Darby, B.J., Sanford, R.L., Kuske, C.R., Neher, D.A., Bel-nap, J., 2007. Heterogeneity of soil nutrients and subsurface biota in a drylandecosystem. Soil Biol. Biochem. 39, 2138–2149.

Ingham, R.E., Coleman, D.C., 1983. Effects of an ectoparasitic nematode on bacterialgrowth in gnotobiotic soil. Oikos 41, 227–232.

Ingham, R.E., Trofymow, J.A., Ingham, E.R., Coleman, D.C., 1985. Interactions of bac-teria, fungi, and their nematode grazers: effects on nutrient cycling and plantgrowth. Ecol. Monogr. 55, 119–140.

Jenkins, W.R., 1964. A rapid centrifugal flotation technique for separating nematodesfrom soil. Plant Dis. Rep. 48, 692.

Kardol, P., Bezemer, T.M., van der Wal, A., van der Putten, W.H., 2005. Successionaltrajectories of soil nematode and plant communities in a chronosequence ofex-arable lands. Biol. Conserv. 126, 317–327.

Kéfi, S., Rietkerk, M., Alados, C.L., Pueyo, Y., Papanastasis, V.P., Elaich, A., de Ruite,P.C., 2007. Spatial vegetation patterns and imminent desertification in Mediter-ranean arid ecosystems. Nature 44, 213–216.

Kohn, L., Stasovski, I., 1990. The mycorrhizal status of plants at Alexandra Fiord,Ellensmere Island, Canada, a high Arctic Site. Mycologia 82, 23–35.

Klironomos, J.N., McCune, J., Hart, M., Neville, J., 2000. The influence of arbuscularmycorrhizae on the relationship between plant diversity and productivity. Ecol.Lett. 3, 137–141.

Lucero, M.E., Barrow, J.R., Osuna, P., Reyes, I., 2006. Plant–fungal interactions in aridand semi-arid ecosystems: large-scale impacts from microscale processes. J. AridEnviron. 65, 276–284.

Lucero, M., Barrow, J., Osuna-Avila, P., Reyes-Vera, I., Duke, S., 2008. Enhancing nativegrass productivity by co-cultivating with endophyte-laden calli. Rangeland Ecol.Manage. 61, 124–130.

McCune, B., Mefford, M.J., 2006. PC-ORD. Multivariate Analysis of Ecological Data.Version 5.32. MjM Software, Gleneden Beach, OR, USA.

Mills, K.E., Bever, J.D., 1998. Maintenance of diversity within plant communities: soilpathogens as agents of negative feedback. Ecology 79, 1595–1601.

Molofsky, J., 1994. Population dynamics and pattern formation in theoretical popu-lations. Ecology 75, 30–39.

Molofsky, J., Bever, J.D., Antonovics, J., 2001. Coexistence under positive frequencydependence. R. Soc. 268, 272–277.

Monger, H.C., Mack, G.H., Nolen, B.A., Gile, L.H., 2006. Regional setting of the JornadaBasin. In: Havstad, K.M., Huenneke, L.F., Schlesinger, W.H. (Eds.), Structure andFunction of a Chihuahuan Desert Ecosystem. Oxford University Press, Inc., NewYork, pp. 15–43.

Moore, J.C., de Ruiter, P.C., 1991. Temporal and spatial heterogeneity of trophic inter-actions within below-ground food webs. Agric. Ecosyst. Manage. 34, 371–397.

Moore, J.C., McCann, K., Setala, H., de Rutter, P.C., 2003. Top-down is bottom-up:does predation in the rhizosphere regulate aboveground dynamics? Ecology 84,846–857.

Neher, D.A., 2010. Ecology of plant and free-living nematodes in natural and agri-cultural soil. Annu. Rev. Phytopathol. 48, 371–394.

O’Hanlon-Manners, D.L., Kotanen, P.M., 2004. Evidence that fungal pathogens inhibitrecruitment of shade-intolerant tree, white birch (Betula popyrifera) in under-story habitats. Oecologia 140, 650–653.

Ogle, K., Reynolds, J.F., 2004. Plant responses to precipitation in desert ecosys-tems: integrating functional types, pulses, thresholds, and delays. Oecologia 141,282–294.

Okada, H., Tsukiboshi, T., Kadota, I., 2002. Mycetophagy in Filenchus misellus(Andrássy, 1958) Lownsbery & Lownsberry, 1985 (Nematoda: Tylenchidae),with notes on its morphology. Nematology 4, 795–801.

Okada, H., Kadota, I., 2003. Host status of 10 fungal isolates for two nematode species,Filenchus misellus and Aphelenchus avenae. Soil Biol. Biochem. 35, 1601–1607.

Okada, H., Harada, H., Kadota, I., 2005. Fungal-feeding habits of six nematode isolatesin the genus Filenchus. Soil Biol. Biochem. 37, 1113–1120.

Okin, G.S., Gillette, D.A., Herrick, J.E., 2006. Multi-scale controls on and consequencesof Aeolian processes in landscape change in arid and semi-arid environments. J.Arid Environ. 65, 253–275.

Parsons, A.J., Wainwright, J., Abrahams, A.D., Simanton, J.R., 1997. Distributeddynamic modeling of interrill overland flow. Hydrol. Process. 11, 1833–1859.

Peters, D.P.C., Bestelmeyer, B.T., Herrick, J.E., Monger, H.C., Fredrickson, E., Havstad,K.M., 2006. Disentangling complex landscapes: new insights to forecasting aridand semiarid system dynamics. BioScience 56, 491–501.

Peters, D.P.C., Gibbens, R.P., 2006. Plant communities in the Jornada Basin: thedynamic landscape. In: Havstad, K.M., Huenneke, L.F., Schlesinger, W.H. (Eds.),Structure and Function of a Chihuahuan Desert Ecosystem. Oxford UniversityPress, Inc., New York, pp. 211–231.

Peters, D.P.C., 2002a. Plant species dominance at a grassland-shrubland ecotone: an

individual-based gap dynamics model of herbaceous and woody species. Ecol.Model. 152, 5–32.

Peters, D.P.C., 2002b. Recruitment potential of two perennial grasses with dif-ferent growth forms at a semiarid–arid transition zone. Am. J. Bot. 89,1616–1623.

Soil E

P

P

P

P

R

R

R

R

R

R

S

J.R. Klass et al. / Applied

eterjohn, W.T., Schlesinger, W.H., 1991. Factors controlling denitrification in a Chi-huahuan Desert ecosystem. Soil Sci. Soc. Am. J. 55, 1694–1701.

orra-Alfaro, A., Herrera, J., Sinsabaugh, R.L., Odenbach, K.J., Lowrey, T., Natvig, D.O.,2008. Novel root fungal consortium associated with a dominant desert grass.Appl. Environ. Microbiol. 74, 2805–2813.

orras-Alfaro, A., Herrera, J., Natvig, D.O., Sinsabaugh, R.L., 2007. Effect of long-termnitrogen fertilization on mycorrhizal fungi assocaited with a dominant grass ina semiarid grassland. Plant Soil 296, 65–75.

orras-Alfaro, A., Herrera, J., Natvig, D.O., Lipinski, K., Sinsabaugh, R.L., 2011.Diversity and distribution of soil fungal communities in a semiarid grassland.Mycologia 103, 10–21.

avi, S., D’Odorico, Okin, G.S., 2007. Hydrologic and aeolian controls onvegetation patterns in arid landscapes. Geophys. Res. Lett. 34, L24S23,doi:10.1029/2007GL031023.

einhart, K.O., Packer, A., Van der Putten, W.H., Clay, K., 2003. Plant–soil biota inter-actions and spatial distribution of black cherry in its native and invasive ranges.Ecol. Lett. 6, 1046–1050.

einhart, K.O., Royo, A., van Der Putten, W.H., Clay, K., 2005. Soil feedback andpathogen activity in Prunus serotina throughout its native range. J. Ecol. 93,890–898.

eynolds, J.F., Stafford Smith, M. (Eds.), 2002. Global Desertification: do humanscause deserts? Dahlem Workshop Report 88. Dahlem University Press, Dahlem,Germany.

eynolds, J.F., Kemp, P.R., Ogle, K., Fernández, R.J., 2004. Modifying the ‘pulse-reserve’ paradigm for deserts of North America: precipitation pulses, soil water,and plant responses. Oecologia 141, 194–210.

itz, K., Trudgill, 1999. Utility of nematode community analysis as an integrated

measure of the functional state of soils: perspectives and challenges. Plant Soil212, 1–11.

canlon, T.M., Caylor, K.K., Levin, S.A., Rodriguez-Iturbe, I., 2007. Positive feed-backs promote power-law clustering of Kalahari vegetation. Nature 449,209–213.

cology 58 (2012) 66– 77 77

Schlesinger, W.H., Reynolds, J.F., Cunningham, G.L., Huenneke, L., Jarrel, W.M., Vir-ginia, R.A., Whitford, W.G., 1990. Biological Feedbacks in global desertification.Science 247, 1043–1048.

Siddiqi, M.R., 2000. Tylenchida: Parasites of Plants and Insects, 2nd ed. CABI Pub-lishing, New York.

Sylvain, Z., Wall, D.H., 2011. Linking soil biodiversity and vegetation: implicationsfor a changing planet. Am. J. Bot. 98, 517–527.

van der Putten, W.H., van Dijk, C., Peters, B.A.M., 1993. Plant-specific soil-bornediseases contribute to succession in foredune vegetation. Nature 362, 52–56.

Wainwright, J., Parsons, A.J., Abrahams, A.D., 2000. Plot-scale studies of vegetation,overland flow, and erosion interactions: case studies from Arizona and NewMexico. Hydrol. Process. 14, 2921–2943.

Wall, D.H., Virginia, R.A., 1999. Controls on soil biodiversity: insights from extremeenvironments. Appl. Soil Ecol. 13, 137–150.

Wallace, H.R., 1987. Effects of nematode parasites on photosynthesis. In: Veech, J.A.,Dickson, D.W. (Eds.), Vistas on Nematology: a Commemoration of the Twenty-fifth Anniversary of the Society of Nematologists. Society of Nematologists,Hyattsville, MD, pp. 253–259.

Wardle, D.A., Bardgett, R.D., Klironomos, J.N., Setälä, van der Putten, W.H., Wall, D.H.,2004. Ecological linkages between aboveground and belowground biota. Science304, 1629–1633.

Wright, C.J., Coleman, D.C., 2000. Cross-site comparison of soil microbial biomass,soil nutrient status, and nematode trophic groups. Pedobiologia 44, 2–23.

Whisenant, S.G., 1999. Repairing Damaged Wildlands: a Process-oriented,Landscape-scale Approach. Cambridge University Press, Cambridge, U.K.

Whitford, W.G., 1996. The importance of the biodiversity of soil biota in arid ecosys-tems. Biodivers. Conserv. 5, 185–195.

Yeates, G.W., Bongers, T., de Goede, G.M., Freckman, D.W., Georgieva, S.S., 1993.Feeding habits in soil nematode families and genera-an outline for soil ecolo-gists. J. Nematol. 25, 315–331.

Yeates, G.W., 1999. Effects of plants on nematode community structure. Annu. Rev.Phytopathol. 37, 127–149.