Above- and belowground insect herbivores differentially affect soil nematode communities in species-rich plant communities

Above- and belowground insect herbivores differentially affect soil nematode communities in species-rich plant communities
of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Above- and belowground insect herbivores differentially affectsoil nematode communities in species-rich plant communities Gerlinde B. De Deyn, Jasper van Ruijven, Ciska E. Raaijmakers, Peter C. de Ruiter andWim H. van der Putten G. B. De Deyn (, C. E. Raaijmakers and W. H. van der Putten, Dept of Multitrophic Interactions, Centre  for Terrestrial Ecology, Netherlands Inst. of Ecology (NIOO-KNAW), P.O. Box 40, NL-6666 ZG, Heteren, the Netherlands. Present address for GBDD: Dept of Biological Sciences, Inst. of Environmental and Natural Sciences, Lancaster Univ., Lancaster, UK,LA1 4YQ. Present address forWHP: Laboratory of Nematology, Wageningen Univ. and Research Centre, PO Box 8123, NL-6700 ES Wageningen, the Netherlands.     J. van Ruijven, Nature Conservation and Plant Ecology Group, Wageningen Univ., the Netherlands.     P. C. de Ruiter, Soil Science Center, Wageningen UR, P.O. Box 47, NL-6700 AA Wageningen, the Netherlands. Interactions between above- and belowground invertebrate herbivores alter plant diversity, however, little isknown on how these effects may influence higher trophic level organisms belowground. Here we explorewhether above- and belowground invertebrate herbivores which alter plant community diversity and biomass, inturn affect soil nematode communities. We test the hypotheses that insect herbivores 1) alter soil nematodediversity, 2) stimulate bacterial-feeding and 3) reduce plant-feeding nematode abundances. In a full factorialoutdoor mesocosm experiment we introduced grasshoppers (aboveground herbivores), wireworms (belowgroundherbivores) and a diverse soil nematode community to species-rich model plant communities. After two years,insect herbivore effects on nematode diversity and on abundance of herbivorous, bacterivorous, fungivorous andomni-carnivorous nematodes were evaluated in relation to plant community composition. Wireworms did not affect nematode diversity despite enhanced plant diversity, while grasshoppers, whichdid not affect plant diversity, reduced nematode diversity. Although grasshoppers and wireworms causedcontrasting shifts in plant species dominance, they did not affect abundances of decomposer nematodes atany trophic level. Primary consumer nematodes were, however, strongly promoted by wireworms, whilecommunity root biomass was not altered by the insect herbivores. Overall, interaction effects of wireworms andgrasshoppers on the soil nematodes were not observed, and we found no support for bottom-up control of thenematodes. However, our results show that above- and belowground insect herbivores may facilitate root-feeding rather than decomposer nematodes and that this facilitation appears to be driven by shifts in plantspecies composition. Moreover, the addition of nematodes strongly suppressed shoot biomass of several forbspecies and reduced grasshopper abundance. Thus, our results suggest that nematode feedback effects on plantcommunity composition, due to plant and herbivore parasitism, may strongly depend on the presence of insectherbivores.  Above- and belowground herbivores are key determi-nants of plant community productivity and diversity and may thus be crucial for the functioning of many terrestrial ecosystems (Hooper et al. 2000, 2005, Wardle 2002). While impacts of vertebrate herbivoresare well documented, effects of invertebrate herbivoreson plant community composition and particularly onhigher trophic levels remain understudied (Wardleet al. 2004, Hooper et al. 2005). In temperate grass-lands not grazed by vertebrate herbivores, shoot- androot-feeding insects are major drivers of ecosystemprocesses (Masters 2004). Shoot feeders may stimulateroot exudation (Paterson et al. 2005) and the abun-dance and activity of soil microorganisms (Guitian andBardgett 2000), especially soil bacteria and bacterialfeeders (Ingham and Detling 1984, Seastedt et al. 1988,Bardgett et al. 1997, Mikola et al. 2001). Also rootfeeders can affect soil resource supply to secondary  Oikos 116: 923    930, 2007 doi: 10.1111/j.2007.0030-1299.15761.x,Copyright # Oikos 2007, ISSN 0030-1299Subject Editor: Heikki Seta ¨la ¨, Accepted 16 February 2007 923  consumers in soil by stimulating microbial biomass(Denton et al. 1999), through increased root exudation(Bardgett et al. 1999a, 1999b), plant productivity (Bardgett et al. 1999b, Scha¨dler et al. 2004) and plantdiversity (De Deyn et al. 2003, 2004a). In contrast tosecondary consumers, primary consumers (i.e. rootherbivores) are supposed to be suppressed by shoot-feeding invertebrate herbivores (Masters et al. 1993)due to aboveground control of belowground resourceavailability.Shifts in the soil food web may in turn feed back tothe plant community through altered plant nutrientavailability (bottom-up) (Patra et al. 2005) and patho-gen and parasite pressure (top-down) (Wardle et al.2005). In order to fully understand and predict theeffects of above- and belowground herbivores on plantcommunity dynamics and ecosystem properties, theresponses of the soil biota and their effect on ecosystemprocesses need to be studied (Bardgett and Wardle2003, De Deyn et al. 2004b). To date, most experi-ments aiming at elucidating the effects of shoot androot herbivorous insects on grassland ecosystems usebiocides to exclude either group of invertebrates. Sincebiocides may cause hidden treatments by affecting decomposer invertebrates and plant nutrient availability (Scha ¨dler et al. 2004, Siemann et al. 2004) alternativeexperimental approaches are needed (Rogers andSiemann 2004). Here we explore effects of above- andbelowground insect herbivores on trophic groups of soilnematodes in an addition experiment. Grassland com-munities were subjected to consistently replicatedexposure to grasshoppers and wireworms.In a previous study we observed that adding above-and belowground invertebrate herbivores resulted innon-additive shifts in plant community compositionand diversity. Grasshoppers reduced plant productivity aboveground without altering plant diversity, whilewireworms stimulated plant diversity without affecting plant productivity. However, simultaneous exposure tograsshoppers and wireworms led to a reduction of plantdiversity (van Ruijven et al. 2005). Here, we investigatehow these non-additive effects on the plant community (van Ruijven et al. 2005) may impact on various trophicpositions the soil food web by studying the soilnematode community composition. We hypothesizedthat in our model system the abundance of nematodesin the soil food web is bottom-up controlled (Yeates1979, Seastedt et al. 1988, Mikola and Seta ¨la ¨ 1998,Mikola 1999, Wardle et al. 1999). To test thishypothesis we analyzed the taxonomical compositionand abundance of bacterivorous, fungivorous, herbivor-ous, and omni-carnivorous nematodes, as well as plantcommunity root biomass. We hypothesized that 1)altered plant species dominance by insect herbivores(van Ruijven et al. 2005) causes qualitative shifts in thenematode communities, 2) herbivory stimulates bacter-ial-feeding nematode abundance and 3) shoot-feeding grasshoppers suppress root-feeding nematodes. Material and methods Experimental design  We established 48 outdoor mesocosms of 1 m 2 , withplant communities of 16 plant species. Each mesocosmwas isolated from the surrounding field soil by woodenframes of 40 cm deep that went down to the mineralsoil layer. Mesocosms were filled with a soil mixture of 3:1 white sand: black soil srcinating from an old field,resulting in soil with 1.05% organic matter, pH CaCl 2 of 7.6 and mineral-N and phosphorus concentration of 0.6 and 0.2 mg kg   1 . No insect larvae were observedduring the soil processing. In order to reduce a potentialnematode legacy effect of the old field soil, the soil inthe mesocosms was treated with a non-persistentnematicide (Nemacur, Phenamiphos) before planting of the seedlings in September, which was seven monthsbefore the start of our nematode addition treatment.Each mesocosm was planted with 144 seedlings in a grid of 12  12. The seedlings were germinated andpre-grown in a greenhouse for three to four weeks. After planting, the experimental treatments wereestablished in a full factorial design. There were six randomized blocks and each block contained onereplicate of eight treatments: aboveground herbivores(grasshoppers:  Chorthippus parallelus  ; hereafter referredto as A), belowground herbivores (wireworms: Elater-idae,  Agriotes lineatus   larvae; hereafter referred to asB) and an assemblage of soil nematodes from a plantspecies rich grassland (including bacterivores, fungi-vores, herbivores, omnivores and carnivores; hereafterreferred to as N), their combinations (AB, NA, NB,NAB), and mesocosms without any introduction(None).The plant communities were mixtures of four grassspecies ( Holcus lanatus  ,  Festuca rubra  ,  Anthoxanthum odoratum   and  Agrostis capillaris  ) and 12 forb species( Plantago lanceolata  ,  Leucanthemum vulgare  ,  Rumex acetosa  ,  Centaurea jacea  ,  Achillea ptarmica  ,  Prunella vulgaris  ,  Hypochaeris radicata  ,  Cerastium fontanum  , Galium mollugo  ,  Leontodon autumnalis  ,  Ranunculus acris   and  Senecio jacobea  ). Every mesocosm wasenclosed aboveground by a 1 m 3 gauze cage (meshsize 300  m m), in order to keep the grasshoppers in andother insects out.The introduced insect herbivores and the nematodeswere collected from natural grassland communitiescomposed of the same plant species as used in theexperiment. The grasshoppers and wireworms wereintroduced in spring, seven months after planting the seedlings. Grasshoppers were collected with sweep924  nets, sexed and five males and five females were addedper mesocosm. Wireworms were introduced into 10holes (1 per hole) evenly distributed over the meso-cosm. Nematodes were extracted from soil by decanta-tion and from roots in funnels that were placed in a mist chamber. The collected nematodes were inoculatedas a mixture, so that the different feeding types wereintroduced in their natural proportions. The nematodecommunity was dominated by plant feeders (Table 1).Nematodes were inoculated as 10 ml of inoculum ineach of 25 inoculation points per mesocosm. Due toplanting error three mesocosms had to be excludedfrom the analyses: two N replicates and one NABreplicate. Measurements  After 11 (year 1) and 23 months (year 2), plant shootsfrom the inner zone (60  60 cm) of the mesocosmswere clipped at 2.5 cm above soil surface, sorted intospecies, dried and weighed. In both years immediately after shoot clipping, soil samples were collected fornematode community and soil nutrient analysis. Sam-pling holes were filled with invertebrate-free dry sand. All nematode analyses and data presented are based onsamples collected after 23 months, in analogy to theanalysis of the plant community data by van Ruijvenet al. (2005). Twelve soil cores of 1 cm diameter and 20cm deep were collected per mesocosm according to a stratified random pattern. Samples were pooled afterseparation in to 0    10 cm and 10    20 cm depthfractions. Nematodes were extracted from 100 cm 3 of soil by Oostenbrink elutriators, and from the rootspresent in that soil volume using a mist chamber(Oostenbrink 1960). Nematodes from the roots andfrom the soil of each individual mesocosm were pooledbefore counting and averaged over both depth fractionsbefore analysis because after data exploration depthproved to have no effect, so that further analysis wereperformed on total average numbers present in 100 cm 3 of soil plus roots in the top 20 cm of soil. The presenceof entomopathogenic nematodes was assessed by ex-posing meal-worms ( Tenebrio molitor   larvae) as hostbait to soil samples of the mesocosms, as described by Solomon et al. (1999).Nematodes were counted and identified to family/genus level and sorted into feeding groups according toBongers (1988) and Yeates et al. (1993). Biomass of extracted soil and roots were determined in order toexpress nematode densities per soil and root mass. Rootdry mass was obtained by drying extracted roots for48 h at 70 8 C. After 23 months we also recorded thetotal number of grasshoppers in each mesocosm. Calculations and data analyses Nematode abundance and taxonomic diversity  To verify the establishment of nematodes from theinoculum we analyzed nematode abundances in meso-cosms without (  N) and in those with only nematodesinoculated (  N) by univariate general linear models(GLM), with nematode inoculation as a fixed andblock as a random factor. Subsequent analyses wereperformed on the data from the mesocosms withnematodes, including nematodes only (N), nematodesand grasshoppers (NA), nematodes and wireworms(NB), nematodes and wireworms and grasshoppers(NAB). Effects of grasshoppers, wireworms and their Table 1. Nematode abundance and taxa/functional group composition in the nematode inoculum (mean total 9 1 SE per mesocosmor per m 2 ).Taxon TaxonBacterial-feeding Rhabditida e   237 9 90 plant-feeding  Meloidogyne   3838 9 285 Acrobeles   17 9 3  Heterodera  67 9 13Other Cephalobidae 1625 9 172 Pratylenchidae 328 9 38 Plectus/Anaplectus   41 9 14 Paratylenchidae 419 9 90 Wilsonema  37 9 8  Helicotylenchus   1361 9 132 Prismatolaimus   18 9 5  Rotylenchus   8 9 5 Rhabdolaimus   10 9 10 Dolichodoridae 188 9 28Bunonematidae 4 9 2  Criconema  46 9 8Bastianiidae 3.5 9 2.6 Trichodoridae 32 9 9Alaimidae 3.8 9 1.3 Longidoridae 5 9 3Teratocephalidae 1.7 9 1.7  Macroposthonia  6 9 1Diplogasteridae 1.3 9 0.9  Ecphyadophora  19 9 19Monhysteridae 0.4 9 0.4  Tylolaimophorus   5 9 2Total 1998 9 178 Total 6317 9 490Fungal-feeding Aphelenchidae 288 9 177 plant Tylenchidae 514 9 76Aphelenchoididae 37 9 34 associated Ditylenchus   0.4 9 0.4 omnivores Other  Dorylaimida  98 9 17Diphtherophoridae 1 9 1 carnivores Mononchidae 18 9 7Total 325 9 61 Total 9315 9 724 925  interaction on root biomass, and on nematode diversity and nematode abundances (total and per feeding-group) were tested by GLM with grasshoppers, wire-worms and their interaction as a fixed and block as a random factor. Post-hoc tests on treatment effects wereperformed using modified Tukey HSD test for unequalnumbers of replicates. Prior to analysis, in order to meetthe assumption of normality and homoscedasticity, data were log  10  (1  x) (plant-feeding nematode abundance)or square root (nematode taxa) transformed. Treatmenteffects on number of nematodes per taxon were testedby Kruskall     Wallis and Mann     Whitney U non-para-metric tests. Nematode diversity was measured as thenumber of taxa and as Shannon     Wiener diversity index H ? a (p i ) (log  2  p i ) where p i  is the proportionalrelative abundance of individuals belonging to the ithtaxon (Magurran 1988). Grasshopper abundance  Effects of wireworms, nematodes and their interactionon grasshopper abundance were analyzed using GLMwith wireworms, nematodes and their interaction asfixed and block as random factors. Normality of thedata was verified via Shapiro     Wilk’s test and equality of variances via Levene’s test. Tukey’s HSD multiplecomparisons of means test, for equal or unequal groupsizes, was used to identify the significant differencesbetween individual treatment groups. Nematode establishment  Nematode establishment of the introduced commu-nities as compared to re-colonization from surrounding sources was investigated first. Two years after nematodeinoculation nematode communities had more taxa (8.6 9 0.6 vs 12.9 9 0.9, F 1,3  8.5, p  0.05) andwere more abundant (F 1,3  25.8, p B 0.05, Table 2)with inoculation compared to nematode communitiessrcinating from external sources only. Plant-feeding nematodes were most abundant and both plant- andbacterial-feeding nematode abundances differed be-tween the treatments (Table 2). Results Grasshopper and wireworm effects on rootbiomass and on nematodes Grasshoppers or wireworms did not affect root biomass(F 1,12  0.22 and F 1,12  1.25, p  0.05), which rangedfrom 144 9 20 to 204 9 37 (mg dry weight per 100 cm 3 of top 20 cm soil) in the treatments with grasshoppers,wireworms and nematodes and with grasshoppers andnematodes, respectively. After two years, grasshopperstended to have suppressed the number of nematodetaxa from 12.9 9 0.6 taxa without to 10.6 9 0.8 taxa with grasshoppers (F 1,12  4.43, p  0.06). Shannon     Wiener diversity of the nematode communities was notsignificantly affected by the insect herbivores, althoughthere was a trend for communities with only nematodesto show a higher diversity (1.6 9 0.1) than communitieswith also above- and belowground invertebrate herbi-vores (1.2 9 0.2). After two years, grasshoppers did not affect totalnematode abundance of the introduced nematodecommunities (F 1,12  0.49, p  0.05), while wirewormsdid stimulate total nematode abundance (F 1,12  10.70,p B 0.01; Fig. 1). Especially the numbers of plant-feeding nematodes was increased with wireworms(F 1,12  12.11, p B 0.01; Fig. 1), mainly due to thestimulation of the dominant plant-feeding nematodespecies  M. hapla   (Table 3). The effects of wirewormswere not affected by the presence or absence of grasshoppers. Numbers of bacterial-feeding, fungal-feeding and omni-carnivorous nematodes, which repre-sent the soil decomposer subsystem, were not signifi-cantly affected by grasshoppers, wireworms or theircombination. The ratio of bacterial-feeding to bacterial-and fungal-feeding nematodes was not affected by grasshoppers or wireworms. The fraction indicated Table 2. Nematodes per feeding group after 23 months in plantcommunities without (  N) and with (  N) nematode inoculum(mean over 0    10 and 10    20 cm 9 1 SE per 100 g soil, n  6 for  N and n  4 for 9 N). Significant treatment differences areindicated by *p B 0.05; ns indicates not significant at p B 0.05with HSD test for unequal n.  N   NPlant-feeders 102 9 56 567 9 82*Bacterivores 41 9 8 64 9 6*Plant-associated 13 9 5 16 9 4 nsFungivores 5 9 2 19 9 8 nsOmni-carnivores 104 9 13 149 9 20 nsTotal 269 9 51 819 9 102* 020040060080010001200140016001800 N NA NB NAB    N  e  m  a   t  o   d  e  a   b  u  n   d  a  n  c  e   (  p  e  r   1   0   0  g  s  o   i   l   ) TotalPlant-feeding axyabyxbxyab Fig. 1. Total and plant-feeding nematode abundances innematode inoculated mesocosms with nematodes only (N),nematodes and grasshoppers (NA), nematodes and wireworms(NB) and nematodes, grasshoppers and wireworms (NAB)after 23 months. Bars represent means 9 1 SE and differentletters within total and plant-feeding nematodes denotesignificant differences at p B 0.05. 926  bacterial feeder dominance, which varied between 81 9 6% and 73 9 8% with treatment. Nematode effects on grasshoppers In year two, numbers of grasshoppers were lower with(9 9 3 individuals m  2 ) than without nematodes added(25 9 3 individuals m  2 ) (F 1,13  15.23 p B 0.01).These results suggested increased mortality of grass-hopper offspring perhaps due presence of entomo-pathogenic nematodes, although these had not beenobserved in the soil and root extracts. Therefore, weperformed a standard assay with larvae of   Tenebrio molitor  , as host bait, and exposed the larvae to soilsamples from nematode-inoculated mesocosms. In thisassay, we indeed recovered entomopathogenic nema-todes. Discussion In the outdoor mesocosms interactive effects of grass-hoppers and wireworms did not influence nematodediversity, as was shown for the aboveground compart-ment of the plant community (van Ruijven et al. 2005).Grasshoppers reduced the number of nematode taxa,whereas wireworms did not, despite their strong effecton qualitative changes in the vegetation aboveground.Therefore our results do not support the proposedbottom-up control of soil nematodes (Yeates 1979,Seastedt et al. 1988, Mikola and Seta ¨la ¨ 1998, Mikola 1999, Wardle et al. 1999). Moreover, insect herbivoresdid not affect abundances of bacterial-, fungal-feeding or omni-carnivorous nematodes. This indicates no top-down effect to become bottom-up effect of insectherbivores on nematodes in the decomposer soil foodweb, as proposed for plant growth stimulation by soilpredators (Moore et al. 2003).Root-feeding nematodes responded to the insectherbivores, albeit different from what we had expected. Wireworms stimulated root-feeding nematode abun-dance while grasshoppers did not affect them. Accord-ing to the aboveground    belowground herbivoreinteraction model of Masters et al. (1993) we expectedthat shoot-feeding grasshoppers would suppress root-feeding nematodes. The discrepancy between the resultsand the model may be explained by the fact that themodel was developed for above- and belowgroundherbivores on a shared host plant. In our experiment,however, we used multi-species plant communities,adding to the level of complexity and aiming for a closerrepresentation of reality. In these mixed plant commu-nities the grasshoppers only fed on grasses, whereas thedominant plant-feeding nematode species  M. hapla  selectively feeds on a wide range of forb species(Goodey et al. 1965, Karssen 1999), so that theinteraction between both above-and belowground her-bivore groups through a shared host was limited.Interestingly, wireworms had a selective enhancing effect on the root-feeding nematodes in our plantspecies-rich communities. Specific enhancement of root-feeding nematodes by shoot vertebrate herbivoreswas shown by Ingham and Detling (1984), however, toour knowledge, our study is the first to report a facilitative interaction between belowground herbivor-ous insects and nematodes. Above ground, facilitationamong herbivores is a common phenomenon knownfor vertebrate grazers. This type of facilitation has beenattributed to complementarity of grazing height andforage preferences (Huisman and Olff 1998, Arsenaultand Owen-Smith 2002). Facilitation among foliage-feeding insects involves enhanced availability of specificplant phenological stages (Damman 1989), whilefacilitation between vertebrates and invertebrates may be caused by altered plant chemical composition(Martinsen et al. 1998). Below ground, herbivorefacilitation may be explained by enhanced root quality  Table 3. Plant-feeding nematode taxa abundance after 23 months in nematode treated mesocosms (N) with grasshoppers (NA),wireworms (NB) or both (NAB) (mean 9 1 SE per 100 g soil average over 0    10 and 10    20 cm). Different letters denote significanttreatment differences at p B 0.05, ns indicates no significant treatment effects at p B 0.05 using non-parametric tests.Taxon N (n  4) NA (n  6) NB (n  6) NAB (n  5) Meloidogyne hapla  478 9 78 ab 416 9 65 b 952 9 241 a 858 9 234 abParatylenchidae 22 9 9 ab 10 9 6 b 26 9 6 a 24 9 12 abTrichodoridae 17 9 4 ns 19 9 6 ns 20 9 2 ns 23 9 13 ns Hemicycliophora  19 9 3 ns 19 9 7 ns 44 9 17 ns 21 9 10 ns Heterodera  10 9 7 ns 10 9 2 ns 5 9 2 ns 7 9 2 nsDolichodoridae 8 9 5 ns 7 9 2 ns 20 9 5 ns 13 9 8 ns Rotylenchus   4 9 2 ns 3 9 2 ns 6 9 4 ns 3 9 2 nsPratylenchidae 3 9 2 ab 0 b 4 9 2 a 2 9 1 ab Criconema  3 9 3 ns 0 ns 1 9 1 ns 0 ns Helicotylenchus   1 9 1 ns 0 ns 0 ns 2 9 1 ns Macroposthonia  2 9 2 ns 0 ns 0 ns 0 nsTotal 567 9 83 b 485 9 72 b 1080 9 225 a 954 9 227 ab 927
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks