Title: Facilitation and predation structure a grassland detrital food web: the responses of soil nematodes to isopod processing of litter
Abstract: Journal of Animal EcologyVolume 80, Issue 5 p. 947-957 Free Access Facilitation and predation structure a grassland detrital food web: the responses of soil nematodes to isopod processing of litter Justin L. Bastow, Corresponding Author Justin L. Bastow Present address: Biology Department, Eastern Washington University, 258 SCI, Cheney, WA 99004, USA. Correspondence author. E-mail: [email protected] for more papers by this author Justin L. Bastow, Corresponding Author Justin L. Bastow Present address: Biology Department, Eastern Washington University, 258 SCI, Cheney, WA 99004, USA. Correspondence author. E-mail: [email protected] for more papers by this author First published: 11 May 2011 https://doi.org/10.1111/j.1365-2656.2011.01853.xCitations: 15 AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Summary 1. Detritus can support successive consumers, whose interactions may be structured by changes in the condition of their shared resource. One model of such species interactions is a processing chain, in which consumers feeding on the resource in a less processed state change the resource condition for subsequent consumers. 2. In a series of experiments, the hypothesis was tested that a common detritivore, the terrestrial isopod Porcellio scaber, affects soil nematodes through the processing of plant litter. Different detrital resources were added to soil from a California coastal prairie in order to simulate litter processing by the detritivore. Treatments that included only whole grass litter corresponded to detrital food webs lacking detritivores, while treatments that included mixtures of P. scaber faeces and grass litter corresponded to different densities or feeding rates of P. scaber. 3. Simulated litter processing by P. scaber increased the abundance of bacterivorous nematodes by between 32% and 202% after 24–44 days in laboratory experiments, but had no effect on fungivorous or predaceous nematodes. 4. In a subsequent field experiment, however, fungivorous nematodes were suppressed by isopod litter processing while bacterivores showed no response. Instead, P. scaber processing of litter increased the abundance of predaceous nematodes in the field experiment by 176%. 5. When simulated litter processing of litter was crossed in laboratory experiments with predaceous nematode addition (comparable to the response of predators in the field experiment), the abundance of bacterivores was increased by isopod processing of litter (by an average of 122%), but suppressed by elevated densities of predaceous nematodes (by an average of 41%). 6. This suggests that litter processing by P. scaber facilitates the bacterial channel of the soil food web, but that predaceous nematodes suppress the response of bacterivores in the field. Processing chain interactions may, therefore, be important in understanding the relative importance of bacterial and fungal channels in the soil food web, while top-down effects of predators determine the resulting changes in population abundance and biomass. Introduction Decomposition is one of the fundamental ecosystem processes and is the process through which nutrients in dead biomass are recycled. Dead biomass, or detritus, is broken down by detritivores and decomposers, which feed on the detritus, assimilate a fraction of what they ingest, and modify the quality and condition of the unassimilated fraction (Hunter et al. 2003). From the point of view of consumers, one of the distinctive characteristics of detritus is that its resource quality and condition are continually changing as a result of other consumers in the community. It is therefore not surprising that the assemblage of consumers feeding on detritus undergoes succession as the basal resource decomposes. Carrion feeding insects (Payne 1965), wood-degrading fungi (Renvall 1995), dung beetles (Gittings & Giller 1998) and consumers of whale carcasses (Smith & Baco 2003) are just a few examples of the species assemblages for which such succession on a resource is well documented. Such systems, in which a resource is shared by successive consumers that feed on the resource in different conditions, have been termed 'processing chains' (Heard 1994a, 1995). Consumers in a processing chain may positively or negatively affect later successional consumers by altering the rate or efficiency with which the resource is processed. Thus, processing chain interactions occur when early successional consumers of a resource alter the rate with which the resource becomes available to subsequent consumers (Heard 1995) or the total quantity of resource which becomes available to them (Heard 1994a). Resource processing is, therefore, one mechanism through which species in a community may have positive interactions with other species, although facilitation is by no means the inevitable outcome of a processing chain interaction. Heard's processing chain model has primarily been applied to aquatic container communities, including pitcher plants (Heard 1994b; Hoekman, Winston & Mitchell 2009), bromeliad tanks (Starzomski, Suen & Srivastava 2010) and tree holes (Paradise 1999; Daugherty & Juliano 2003). These studies have shown that litter processing by aquatic detritivores can have facilitative effects on particle-feeding insects (Heard 1994b; Daugherty & Juliano 2003; Starzomski, Suen & Srivastava 2010) and bacteria (Hoekman, Winston & Mitchell 2009), but that resource availability (Paradise 1999) and predation (Hoekman, Winston & Mitchell 2009; Starzomski, Suen & Srivastava 2010) are important in understanding how processing chain interactions are manifest in real food webs. Although resource processing is believed to be a general feature of detrital food webs (Moore et al. 2004), processing chains have been difficult to demonstrate in more open aquatic systems (Heard & Buchanan 2004) and have rarely been invoked to explain species interactions in terrestrial or marine systems (but see Tiunov & Scheu 2005 for an exception). There are, however, a number of apparently facilitative interactions among terrestrial detritivores and decomposers that appear to be structured by the processing of detritus, including the facilitation of soil microbes by invertebrate detritivores (Hättenschwiler & Bretscher 2001; Zimmer, Kautz & Topp 2005), sugar fungi by lignin-degrading fungi (Osono & Takeda 2001), and nitrifying bacteria by ammonifying bacteria (Torres, Abril & Bucher 2005). I tested the hypothesis that an abundant detritivore in California coastal prairie, the isopod Porcellio scaber, positively affects soil-dwelling nematodes through its processing of grass litter. Terrestrial isopods often increase litter decomposition rates (Kautz & Topp 2000; Hättenschwiler & Bretscher 2001; Zimmer, Kautz & Topp 2005; Bastow, Preisser & Strong 2008) and soil microbial abundance or activity (Hanlon & Anderson 1980; Hassall, Turner & Rands 1987; Kayang, Sharma & Mishra 1996; Kautz & Topp 2000; Hättenschwiler & Bretscher 2001; Zimmer, Kautz & Topp 2005), but their interactions with microbivorous soil fauna have not previously been studied. Nematodes were chosen as focal soil fauna because of their importance in nutrient cycling and ubiquity in soils (Bongers & Bongers 1998). Because the nematode assemblage includes bacterivores, fungivores, herbivores and predators of other nematodes, its structure provides information about the importance of detrital and rhizal energy sources, bacterial and fungal channels, and higher trophic levels in the soil food web (Neher 2001). As a detritivore feeding at the soil surface, isopods likely act as early successional consumers (upstream consumers, sensu Heard 1994a) in a litter-processing chain. By increasing decomposition rates, isopod processing of litter may increase the rate at which organic matter is incorporated into the soil, thus having short-term facilitative effects on soil microbes (i.e. bacteria and fungi) as well as microbivorous soil fauna (downstream consumers). Fungi are generally better able to utilize litter on the soil surface than bacteria, because of their hyphal growth form and ability to translocate water and nutrients (Beare et al. 1992). Bacteria and bacterivores are therefore likely to be more strongly facilitated by isopod processing of litter than fungi and fungivores, because in addition to making organic matter available to fungi in the soil, isopods likely compete with fungi for surface litter. Comparing the responses of bacteria and fungi to isopod processing of litter may provide insights into the ecosystem consequences of detritivores; the bacterial channel of the soil food web is generally associated with rapid decomposition and nutrient turnover and the fungal channel with slower decomposition and greater nutrient retention (Moore & Hunt 1988; Wardle 2002). Responses of nematode populations to simulated litter processing by isopods were measured in laboratory and field microcosm experiments. In these experiments, I added grass litter and P. scaber faeces (grass litter that was consumed but not assimilated) to soil microcosms to simulate different activity levels (i.e. different densities or feeding rates) of isopods. These experiments isolated the effects of litter processing by P. scaber from other effects the isopod may have on soil fauna (e.g. through soil turbation) in order to clarify the importance of this particular mechanism through which detritivores may affect soil food webs. In two of the experiments, isopod litter processing was crossed with different densities of predaceous nematodes, which are known to exert strong top-down control of microbivorous nematodes (Allen-Morley & Coleman 1989; Mikola & Setälä 1998). The following hypotheses were tested: (i) litter processing by P. scaber increases the abundance of microbivorous nematodes, (ii) because fungi are better able to utilize litter on the soil surface than bacteria, bacterivorous nematodes are more strongly facilitated by P. scaber than fungivorous nematodes, and (iii) predaceous nematodes suppress the positive response of microbivores to isopod processing of litter. Materials and methods Study Site All of the soil, litter and animals used in these experiments came from the University of California Bodega Marine Reserve which was also the site of the field experiment. The Bodega Marine Reserve is a 146-ha reserve in Sonoma Co., CA, USA (38° 19′ N, 123° 4′ W). The reserve has a coastal Mediterranean climate, with a cool, rainy winter (November–March, average precipitation of 71 cm per season, Bodega Ocean Observing Node 2008) and a dry, but foggy, summer (average precipitation 14 cm per season). The coastal prairie of the Bodega Marine Reserve is an annual-dominated grassland with loamy sand soil. The vegetation comprises both native Californian and introduced European grasses and forbs. Porcellio scaber (Latreille), a terrestrial isopod native to Europe (Harding & Sutton 1985), is the most abundant macrodetritivore at the Bodega Marine Reserve. Pitfall trap data suggest that P. scaber is relatively inactive during the winter at the Bodega Marine Reserve, but that their numbers and activity increase steadily through the spring and summer (J.L. Bastow unpublished data). Fall densities average 350 P. scaber per m2 (based on trapping isopods out of enclosed 0·25 m2 plots, N = 8). Similarly, P. scaber has little effect on litter decomposition in the winter, but accounts for most of the litter lost from litterbags during the summer (Bastow, Preisser & Strong 2008). Overall, P. scaber increases litter mass loss at the Bodega Marine Reserve by c. 29% (Bastow, Preisser & Strong 2008), consuming between 90 and 126 g m−2 of grass litter and producing between 45 and 63 g m−2 of faeces (based on laboratory measurements of consumption and assimilation rates, Bastow 2007 and J.L. Bastow unpublished data). Laboratory Microcosm Experiments The responses of microbivorous nematodes (i.e. bacterivores and fungivores) to detrital resources were measured in four laboratory microcosm experiments. In the first two of these experiments (Processing Experiments 1 and 2), the response of nematodes to four different levels of litter processing by isopods was measured at multiple points in time (10–59 days). In the next two experiments (Predation Experiments 1 and 2), the response of nematodes to two different levels of litter processing at two different densities of predaceous nematodes was measured. In the two processing experiments, microcosms received one of five resource treatments, although only four of the treatments were included in each of the two experiments. Both experiments included a treatment in which the soil received no added resource ('soil control') and a treatment that received 0·18 g of grass litter (dry mass, 'grass addition'). Each experiment also included two treatments of simulated isopod processing of litter. These treatments simulated the conversion of 25%, 50% or 100% of the grass litter to isopod faeces with a 0·50 assimilation efficiency, assuming that uneaten litter is unaltered by isopods. These treatments correspond to different densities or feeding rates of isopods. The 0·50 assimilation efficiency was determined gravimetrically in a preliminary experiment (Bastow 2007) and is within the range of values reported for terrestrial isopods (Zimmer 2002). These treatments included the following: 0·135-g grass litter and 0·0225-g isopod faeces ('25% processing'); 0·09-g grass litter and 0·045-g faeces ('50% processing'); and 0·09-g faeces ('100% processing') (all masses expressed in dry mass). Processing Experiment 1 included the 25% and 50% processing treatments, while Processing Experiment 2 included the 50% and 100% processing treatments. The grass litter used in each experiment is listed in Table 1. Isopod faeces were produced in the laboratory by feeding P. scaber on 1·5-mm mesh screen and collecting the faeces that fell through the screen. The species of grass litter used differed between experiments because of the limited availability of grass litters of particular species in certain seasons, but all species were of similar resource quality (i.e. carbon : nitrogen ratio) and all were common species at the Bodega Marine Reserve. Within each experiment, all grass litter and isopod faeces were derived from the same batch of grass litter, collected from the field on the same date. Table 1. Litter resources and sampling schedule of microcosm experiments Experiment Grass litter C : N grass litter C : N isopod faeces Soil collected Start date Duration (days) Processing 1 Bromus diandrus 47·2 22·0 27/9/2005 19/10/2005 58 Processing 2 Mixed annuals 47·4 24·0 20/4/2005 29/4/2005 59 Field Mixed annuals 56·3 25·6 6/2/2006 13/2/2006 111 Predation 1 Calamagrostis nutkaensis 52·6 27·4 8/6/2008 13/6/2008 24 Predation 2 Bromus diandrus 45·5 25·6 22/6/2008 2/7/2008 23 C : N, carbon-to-nitrogen ratio by mass. The two predation experiments included grass addition and 100% processing treatments but, due to limited growth chamber capacity and processing time constraints, omitted the intermediate levels of isopod processing. These two resource treatments were crossed in the predation experiments with two different levels of predaceous nematodes: ambient (no predators added) and elevated (10 Mononchida predators added). The density of additional predators (1 per g dry soil) was based on the response of predaceous nematodes in the field experiment (see below). Predaceous nematodes were extracted from raw soil using Baermann funnels (Coleman et al. 1999), removed from samples using a pipette and stored individually at 8 °C in vials until experimental set-up (less than a week). Soil for all laboratory experiments was collected to a depth of 20 cm and stored at room temperature until microcosm construction (Table 1). Soil was passed through a sieve (1·6 mm) immediately prior to microcosm construction and wetted to 0·19 gravimetric water content (g water per g dry weight soil, SD 0·017). Each microcosm consisted of a polystyrene sample vial (79 mm height × 27 mm diameter) to which was added 12·00 g of soil (wet weight, ±0·20 g). Microcosms were randomly arranged in a growth chamber, which cycled between 11 and 15 °C on a 24-h cycle (12 h light, 12 h dark), approximating field conditions at the Bodega Marine Reserve in the spring (Bastow 2007). Microcosms were watered every 5–10 days in an effort to maintain constant soil moisture. Soil moisture nonetheless declined over the course of the experiments, to a mean gravimetric water content of 0·14 (SD 0·064). Microcosm Sampling and Processing There were six replicate microcosms of each treatment at each sampling point within each experiment. Processing Experiments 1 and 2 were destructively sampled four times, first at 10 or 12 days, then at 26 or 24 days, at 44 or 39 days, and finally at 58 or 59 days. Predation Experiments 1 and 2 were sampled once, at 24 and 23 days, respectively. Sampling dates differed slightly between experiments because of access to the growth chambers. Nematodes were extracted from six soil samples for day 0 data in each experiment. In addition to microcosms sampled for nematodes, four microcosms were used at each sampling time to measure gravimetric water content of the soil. Microcosms were destructively sampled for nematodes by placing all soil and litter into a Baermann funnel. Nematodes were extracted for 2 days at 22 °C. The total numbers of nematodes were then counted, and a subsample of nematodes (10% of the total) was identified to functional group on the basis of stylet and oesophagus features (Freckman & Baldwin 1990; Yeates et al. 1993). The functional groups identified were bacterivore, fungivore, predator (i.e. consuming nematodes, enchytraeids, protists, and rotifers), omnivore, plant parasite, and 'tylenchus type' (i.e. Tylenchida of ambiguous feeding habit, most likely feeding on roots and fungal hyphae). Plant parasitic, tylenchus type, and omnivorous nematodes did not respond to treatments in any experiment and were generally at low abundance in laboratory microcosms; only the data on microbivorous and predaceous nematodes are presented here. Field Microcosm Experiment The field microcosm experiment was similar in design to the processing experiments, but measured the response of nematodes to simulated processing of litter in the field rather than the laboratory. Each microcosm consisted of a cylinder of 0·5-mm mesh nylon screen (12 cm height × 4 cm diameter), rolled up at the bottom and stapled shut, to which 55 g (wet weight, +5 g) of soil was added. The soil had an initial gravimetric water content of 0·20 (SD 0·005), and no additional water was added before microcosm construction. Resource treatments were the same as in the processing experiments, scaled to the larger microcosm size: 0·5 g grass litter ('grass addition'); 0·375 g grass litter and 0·0675 g faeces ('25% processing'); 0·25 g grass litter and 0·125 g faeces ('50% processing'); 0·25 g faeces ('100% processing') (all masses expressed in dry mass). Resources were placed on the soil surfaces. The microcosms were placed in the field and the experiment began in February 2006. There were six replicate microcosms for each resource treatment at each of five sampling times (16, 31, 62, 87 and 111 days). The microcosms were arranged in a fully randomized 10 × 12 grid in the field with 0·5 m of undisturbed soil between adjacent rows and columns. The microcosms were placed in holes, so that the soil surface within the sleeve was flush with the surrounding soil, and the entire experiment was watered immediately after set up so that water films in the soil cores would be reconnected to those in the surrounding soil. The microcosms were not watered again for the remainder of the study. Six microcosms were collected for initial (day 0) data. Beginning on the second sampling time (day 31), six soil samples were collected from the prairie surrounding the experimental grid at the same time that microcosms were collected ('ambient' samples, from within 1 m of the edge of the grid). Microcosms and ambient soil samples were processed within twelve hours of collection. Two subsamples were removed from each of the microcosms and ambient soil samples. One subsample was used to measure nematode abundances, as in the laboratory experiments, and the other was used to measure gravimetric moisture content. Although it would have been possible for P. scaber to climb into microcosms, none were found during microcosm processing. Soil temperature was measured from day 31 to 62 and from day 68 to 111 of the experiment using a HOBO temperature data logger (Onset Computer Corporation, Bourne, MA, USA) buried at a depth of 10 cm. Daily minimum and maximum soil temperatures between day 31 and 62 were 10·5 ± 1·3° and 15·7 ± 2·5 °C (mean ± SD) and rose to 14·4 ± 1·4° and 22·1 ± 2·3 °C between day 68 and 111. Precipitation was measured by the Bodega Ocean Observing Node (2008). Soil moisture increased from 0·20 ± 0·002 to 0·43 ± 0·013 g water per g dry soil during the first half of the experiment due to frequent rains (Appendix S1, Supporting information), but subsequently declined to less than 0·10 g water per g dry soil. Data Analysis The responses of nematode abundance (per g dry soil) to treatments and sampling time were analysed in factorial anovas. The initial nematode abundances (day 0) were not included in the analyses, because the initial sampling was not crossed with treatment. In the processing experiments, treatment and sampling time were included as fixed factors, along with their interaction. Time was regarded as a fully crossed factor because experimental units were destructively sampled. In the predation experiments, resource treatment and predator treatment were included as fixed factors, along with their interaction. The abundances of nematodes in the field experiment were analysed using ANCOVAs with treatment and time as fixed factors and soil moisture content as a covariate. Because ambient soil for nematode extraction was not collected at the first sampling time (day 16), treatment and time were not fully crossed, and the analysis of the full data set could not test for an interaction between the two factors. An initial analysis was performed excluding the ambient soil treatment to see whether the interaction between time and treatment was significant for the other treatments. The interaction term was not significant for any functional group of nematodes. The results of analysis of the full data set, without the interaction term in the model, are reported here. Separate anovas were used for each functional group of nematodes (bacterivores, fungivores and predators) in each experiment and Bonferroni corrected for performing three anovas (i.e. α = 0·016). Tukey tests were used to separate means when factors were significant in the anova and had more than two levels. Nematode abundances were log transformed to meet anova assumptions (normally distributed residuals and homogeneity of variances). In the case of predaceous nematodes, which were absent from some replicates, a constant was added to all data points prior to transformation. All means and standard errors presented in text and on figures are of raw data. Analyses were performed in JMP IN 4.0.3. Results Processing Experiments Bacterivorous nematodes were the most abundant functional group of nematodes in all experiments and accounted for two-thirds of all nematodes recovered from laboratory microcosm experiments. In Processing Experiment 1, bacterivore abundances increased in the grass addition treatment while remaining low in the soil control (Fig. 1a). Bacterivore abundances increased more rapidly, however, in the two isopod processing treatments than they had in the grass addition treatment (Fig. 1a, time × treatment interaction, Table 2). Although overall nematode abundances were lower in Processing Experiment 2, the response of bacterivores to treatments was similar; bacterivore abundances in the 50% and 100% processing treatments increased to between 110 and 260 nematodes per g soil at the second and third sampling times, while increasing more slowly in the grass addition and remaining unchanged in the soil control (Fig. 1b, Table 2). Figure 1Open in figure viewerPowerPoint The responses of bacterivorous (a, b) and fungivorous (c, d) nematode density (per g dry soil, mean ± SE) to simulated isopod processing of litter in Processing Experiments 1 (a, c) and 2 (b, d). The grass addition treatment simulates the absence of terrestrial isopods, while the 25%, 50% and 100% processing treatments simulate different densities or feeding rates of isopods. Only four of the five treatments were included in each of the experiments. Different letters indicate treatments that are significantly different according to Tukey post-hoc test. Bacterivores increased in density in response to simulated isopod processing of litter. Table 2. Results of anovas on nematode abundance in the five experiments. Separate anovas were performed on each of the three functional groups of nematodes (bacterivores, fungivores and predators) and an α of 0·016 was used to determine statistical significance (i.e. α = 0·05 Bonferroni corrected for performing three anovas on each experiment) Experiment Source d.f. Bacterivorous nematodes Fungivorous nematodes Predaceous nematodes MS F P MS F P MS F P Processing Experiment 1 Treatment 3 21·21 123·91 <0·0001* 14·75 68·75 <0·0001* 0·017 1·46 0·23 Time 3 26·35 153·97 <0·0001* 37·56 175·09 <0·0001* 0·0046 0·39 0·76 Treatment × Time 9 2·26 13·21 <0·0001* 3·37 15·71 <0·0001* 0·0096 0·82 0·60 Error 79 0·17 0·21 0·012 Processing Experiment 2 Treatment 3 25·31 65·84 <0·0001* 6·51 15·58 <0·0001* 0·14 1·74 0·17 Time 3 9·79 25·47 <0·0001* 20·80 49·78 <0·0001* 0·074 0·92 0·44 Treatment × Time 9 1·39 3·62 0·0008* 1·98 4·73 0·0008* 0·075 0·93 0·51 Error 80 0·38 0·42 0·080 Field experiment Treatment 4 0·23 0·46 0·77 2·16 5·01 0·0009* 1·07 5·44 0·0004* Time 4 2·12 4·26 0·0028* 7·21 16·71 <0·0001* 1·35 6·86 <0·0001* Soil moisture 1 2·13 4·29 0·04 2·89 6·71 0·011* 0·79 4·02 0·047 Error 133 0·50 0·43 0·20 Predation Experiment 1 Resource 1 14·68 22·55 <0·0001* 0·68 1·47 0·24 1·05 7·37 0·013* Predator 1 5·17 7·95 0·011* 3·08 6·63 0·018 6·18 43·51 <0·0001* Resource × Predator 1 2·30 3·52 0·075 2·71 5·83 0·025 0·030 0·21 0·65 Error 20 0·65 0·46 0·14 Predation Experiment 2 Resource 1 2·92 13·78 0·0014* 0·42 2·15 0·16 0·12 0·49 0·49 Predator 1 1·72 8·11 0·0099* 0·37 1·91 0·18 3·16 13·37 0·0016* Resource × Predator 1 0·12 0·54 0·47 0·34 1·71 0·21 0·0034 0·015 0·91 Error 20 0·21 0·20 0·24 *Statistical significance after Bonferroni correction. d.f., degrees of freedom; MS, mean square. Fungivorous nematodes increased in abundance in the grass addition and 25%, 50% and 100% processing treatments, but there were no differences in fungivore abundance between these treatments in either experiment (Fig. 1c,d; Table 2). Fungivore abundances were considerably higher in all treatments in Processing Experiment 1 than in Processing Experiment 2. Predaceous nematodes were scarce in both processing experiments (0·026 ± 0·016 and 0·21 ± 0·043 per g soil in processing experiments 1 and 2, respectively) and did not respond to treatments (Table 2). Field Experiment Although, still the most abundant functional group of nematodes, bacterivorous nematodes were much less abundant in the field experiment than in the Processing Experiments. Bacterivore abundances declined during the first 16 days of the experiment and then fluctuated between 6 and 14 per g soil (Fig. 2a, Table 2). There was no effect of treatment on bacterivore abundance (Fig. 3a, Table 2). Fungivorous nematodes generally increased in abundance throughout the field experiment (Fig. 2b, Table 2). The highest level of isopod processing (100% processing treatment) suppressed the abundance of fungivores relative to the 25% and 50% processing and grass addition treatments (Fig. 3b, Table 2). Figure 2Open in figure viewerPowerPoint The responses of bacterivorous (a), fungivorous (b), and predaceous (c) nematode density (per g dry soil, mean ± SE) to simulated isopod processing of litter in the field experiment over 111 days. Treatments are the same as in Fig. 1, except that ambient soil was used instead of a soil control. Figure 3Open in figure viewerPowerPoint The mean density (per g dry soil, ±SE) of bacterivorous (a), fungivorous (b), and predaceous (c) nematodes in the field experiment (averaging over all sampling dates). Different letters indicate treatments that are significantly different according to Tukey post-hoc test. Treatments are the same as in Fig. 2. High levels of simulated isopod processing of litter increased the densities of predaceous nematodes and suppressed fungivorous nematodes. Bacterivore densities did not differ among treatments. In contrast to bacterivores and fungivores, predaceous nematodes were more