Edge effects and intraguild predation in native and introduced centipedes: evidence from the field and from laboratory microcosms by Cari-Ann Hickerson | Papers by Cari-Ann

Oecologia (2005) 146: 110–119 DOI 10.1007/s00442-005-0197-y C O N S E R V A T I O N EC O L O G Y Cari-Ann M. Hickerson Æ Carl D. Anthony B. Michael Walton Edge effects and intraguild predation in native and introduced centipedes: evidence from the field and from laboratory microcosms Received: 25 October 2004 / Accepted: 24 June 2005 / Published online: 23 July 2005 Ó Springer-Verlag 2005 Abstract Human alteration of habitat has increased the proportion of forest edge in areas of previously continuous forest. This edge habitat facilitates invasion of exotic species into remaining fragments. The ability of native species to resist invasion varies and may depend on intrinsic variables such as dispersal and reproductive rates as well as external factors such as rate of habitat change and the density of populations of introduced species in edge habitat. We examined the distributional and competitive relationships of two members of the class Chilopoda, Scolopocryptops sexspinosus, a centipede native to the eastern US, and Lithobius forficatus, an exotic centipede introduced from Europe. We found that L. forficatus was most abundant in edge habitat and S. sexspinosus was most abundant in the interior habitat at our field sites. Although L. forficatus was present in habitat interiors at 11 of 12 sites, there was no correlation between fragment size and numbers of L. forficatus in interior habitat. The native centipede was rarely found occupying fragment edges. We used laboratory microcosms to examine potential competitive interactions and to indirectly assess prey preferences of the two species. In microcosms both species consumed similar prey, but the native centipede, S. sexspinosus, acted as an intraguild predator on the introduced centipede. Native centipedes were competitively superior in both intraspecific and interspecific pairings. Our results suggest that intraguild predation may aid native centipedes in resisting invasion of introduced centipedes from edge habitat. Keywords Chilopoda Æ Competition Æ Invasive Æ Lithobius Æ Scolopocrytops Introduction Two major threats to global biodiversity are humaninduced habitat alteration and the introduction of nonnative species (Vitousek 1990). The most obvious negative consequences of habitat alteration occur when certain habitats required of species are destroyed and those species are then lost from a community. Biodiversity also suffers when non-native species expand their ranges and populations at the expense of native species (Gurnell et al. 2004; Rooney et al. 2004) resulting in homogenization or the replacement of native species by invasive exotics (McKinney and Lockwood 1999; Olden et al. 2004). Habitat alteration and the introduction of nonnative species have the potential to act synergistically because habitat alteration often favors the growth and spread of non-native species. As habitat is cleared for agriculture and urban development, edge habitat is created which can support the establishment and spread of introduced species (Laurance and Yensen 1991; Hobbs and Huenneke 1992; Didham et al. 1996; Suarez et al. 1998; Hill and Curran 2001). Once introduced species become established, they have the potential to interact with native species in a number of different ways. They may compete directly with native species for contested resources. For example, Argentine ants (Linepithema humile) have competitively displaced their native counterparts in riparian areas of northern California (Holway 1999). The introduced common house gecko, Hemidactylus frenatus, is displacing native gekkonids on many tropical islands through a combination of exploitation and interference competition (Case et al. 1994; Petren and Case 1998). Intraguild predation, the killing and eating of prey species by a predator that can also utilize the resources of those prey (Polis et al. 1989), is also likely to play a role in some of these interactions. In laboratory trials, Communicated by Scott Robinson C.-A. M. Hickerson (&) Æ B. M. Walton Department of Biological, Geological, and Environmental Sciences, Cleveland State University, Cleveland, OH 44115, USA E-mail: chickerson2@msn.com Tel.: +1-216-6872407 C. D. Anthony Department of Biology, John Carroll University, University Heights, OH 44118, USA 111 adults of H. frenatus preyed upon juveniles of a native competitor (Bolger and Case 1992) and the well-documented displacement of the green anole (Anolis carolinensis) by the introduced brown anole (A. sagrei) in the southeastern United States may be mediated by predation of juvenile green anoles by adult brown anoles (Gerber and Echternacht 2000). Recent studies have focused on the role that invasive species play in forest floor ecosystems. For example, introduced earthworms can influence colonization rates by mycorrhizal fungi (Lawrence et al. 2003) and are associated with the decline of a rare fern (Gundale 2002). Introduced detritivores such as isopods (Kalisz and Powell 2004) and millipedes (Griffin and Bull 1995) have successfully invaded forest ecosystems, but the long-term effects of these invasions remain unclear. The introduction of predators, such as the terrestrial flatworm Bipalium adventitium, also illustrates how exotic species may alter food webs. Bipalium is a voracious earthworm predator that is apparently distasteful to and is avoided by vertebrate predators (Ducey et al. 1999). It has the potential to exert top-down forces on detrital food webs by reducing numbers of earthworms, which could affect the rate of soil formation and leaf litter decomposition. Introduced predators, especially those that occupy the upper levels of trophic pyramids, have the greatest potential to alter food web structure by outcompeting native predators (Vitousek 1990). Centipedes are top predators in terrestrial detrital food webs. They have large biomasses (Lewis 1981), are generalist predators, and are potentially important in determining community composition of the detrital macrofauna and mesofauna. Several species of centipedes are of European origin (Williams and Hefner 1928; Shelley 2002) and one introduced species, Lithobius forficatus, is often associated with areas of human impact (Lee 1980) and is found under debris in edge habitats (Auerbach 1951). This species prefers more xeric habitats than the other members of the genus Lithobius (Auerbach 1951; Roberts 1956) and evidence suggests that it has invaded edge habitats from urban areas (Frund et al. 1997). Little published information is available regarding the geographic spread or stability of introduced populations of L. forficatus, but we suspect that the introduction of L. forficatus is not a recent one. McNeill (1888) reports the species in the midwestern United States (Indiana) by 1886. By 1925, the species was widespread in the eastern half of the United States, excluding the southeastern states, and had reached the Rocky Mountains (Colorado, Utah, Idaho) (Chamberlin 1925). In a three-part study, we examined the interactions between L. forficatus and a native species of centipede, Scolopocryptops sexspinosus, whose geographic ranges overlap extensively in the eastern United States. Field data were collected on distribution and abundance of these two species in forest edge and forest interior habitats. We hypothesized that the introduced centipede, L. forficatus, would be more abundant at the forest edge than in the interior of the fragments because colonization by exotic species is positively correlated with human disturbance. In a microcosm experiment, we used changes in centipede mass over a 9-week period to assess the competitive ability of each species. We hypothesized that differences in competitive ability would be reflected in differential mass gain. Finally, we assessed the effects that each species had on diversity and abundance of leaf litter invertebrates in microcosms. Materials and methods Field distribution—edge habitat use We collected field data 19–29 May 2003 from 12 forested sites in northeast Ohio (Table 1). Both species of Table 1 Numbers of centipedes and means per 50 cover objects sampled in edge and interior habitats of the 12 forest fragments Fragment size (ha) Lithobius Interior 23 24 36 45 49 59 63 65 97 152 174 326 Total Mean (SE) a Scolopocryptops Edge 0 9 2 2 2 14 4 6 7 5 3 10 64 5.33 1.18 Interior 3 1 3 0 0 1 4 3 2 2 2 2 23 1.92 0.35 Edge 2 0 1 1 0 0 0 0 0 0 0 0 4 0.33 0.19 Distance (km) to urban areaa 1 8 2 1 3 2 2 0 3 6 1 4 33 2.75 0.66 4.2 1.4 4.5 21.6 1.7 9.2 20.8 8.2 0.8 4.6 7.7 7.0 Distance from site center to nearest urban area is indicated Urban areas were defined as ‘‘built-up areas’’ in USGS 1:100,000 scale maps 112 centipede are active at the surface during the spring months (Auerbach 1951). The forested sites were mixed deciduous forest dominated by red (Acer rubrum) and sugar maple (A. saccharum), American beech (Fagus grandifolia), and red oak (Quercus rubra). Forest fragments were not younger than 50 years of age and were characterized by deeply shaded forest interiors that lacked subcanopies. We chose forest fragments that varied in area (23–326 ha) to test the hypothesis that increasing fragment size would correlate with decreased invasion of the introduced centipede into forest interiors. We expected complete invasion by L. forficatus in our smaller fragments (Frund et al. 1997) but we posited that larger fragments would present more of a challenge to invasion. The largest fragments available for sampling in the study area were 174 and 326 ha (Table 1). Because our fragments varied in their distances to urban areas and in time since invasion, persistence of invasion by Lithobius may have varied as well. We hypothesized that abundance of introduced centipedes would be highest in fragments that were closest to urban areas and that abundance of native centipedes would be lower in these fragments. Forest fragment boundaries were delineated on topographic maps using Terrain Navigator 5.0 and the geographic center of each fragment was identified and located using a GPS unit. We visited the sites in random order and overturned 100 cover objects (rocks, logs, and bark >25 cm2) at each site, 50 along the fragment edge, and 50 in the fragment interior. In an attempt to minimize bias caused by differences in cover object preferences between the two species, we excluded sites that had a large disparity in the numbers and kinds of cover objects between edge and interior habitats and we made a careful effort to sample a wide variety of cover object types within each habitat. We defined edge habitat as a 10-m deep strip at the forest edge and interior habitat as the area surrounding the fragment center in a 15-m radius. Approximately equal sized areas of each habitat type were searched at each locality. A total of 1,200 cover objects were sampled. We recorded numbers of L. forficatus and S. sexspinosus found beneath each cover object for both habitat types. Species were identified according to Shelley (2002) for Scolopocryptops and Williams and Hefner (1928) for Lithobius. We excluded small specimens of Lithobius from analysis because of the difficulty in keying these individuals to species. Specimens were collected, preserved and brought back to the lab for identification before being placed in the invertebrate collection at the Cleveland Museum of Natural History. Comparisons of the abundance of S. sexspinosus in edge habitat versus interior were made using two-tailed, paired t-test (a=0.05). Comparisons of the abundance of L. forficatus in edge habitat versus interior were made using a two-tailed, Wilcoxon signed-ranks test because the data failed to meet the assumptions of parametric statistics (a=0.05). We used regression analysis (Zar 1999) to examine the relationship between numbers of centipedes and distance to urban areas and between fragment size and centipede abundance in forest interiors. The relationship between size of fragment and distance to urban area was explored with a Pearson’s correlation. Statistical analyses were done using SPSS for Windows, Version 11.5. Microcosms—change in centipede mass and effects on invertebrate communities Adult centipedes (L. forficatus and S. sexspinosus,N=36 of each species) and leaf litter used in the microcosm experiment were collected on 29 September and 1 October 2003 from a forested site (Lake County) in northeastern Ohio (41°35¢56¢¢ N; 81°21¢22¢¢ W) and brought to the laboratory. We mixed leaf litter thoroughly by shaking and by transferring portions among two large garbage bags. Leaf litter was then used to create 46 microcosms (Ziploc boxes, 20-cm long · 12-cm wide · 6cm deep) containing 110 g wet mass of leaf litter. The leaf litter in microcosms approximated litter depth in the field. After every fourth microcosm was created, a litter sample was set aside to assess initial (pre-experiment) litter invertebrate abundance and diversity (N=12). We measured and weighed centipedes prior to placing them in microcosms (mean mass and length of Lithobius=0.100 g, SE=0.006 and 23.41 mm, SE=0.478, mean mass and length of Scolopocryptops=0.228 g, SE=0.007 and 40.77 mm, SE=0.568). Centipedes were randomly assigned to one of the three experimental treatment microcosms. Size asymmetries were reduced by randomly pairing centipedes within treatments to lessen the potential competitive advantage of larger size. Centipedes within a treatment were separated into three size-classes and then randomly paired with one another within those size classes. This approximated the natural size differences between species while minimizing differences within species. Treatment 1 consisted of boxes with two individuals of Lithobius (intraspecific pairs, L/ L, N=12; mean mass and length differences=0.015 g (SE=0.004) and 0.38 mm (SE=0.053)). Treatment 2 consisted of boxes with one Scolopocryptops and one Lithobius (interspecific pairs, S/L, N=12; mean mass and length differences=0.127 g (SE=0.007) and 17.22 mm (SE=0.310)). Treatment 3 consisted of boxes with two individuals of Scolopocryptops (intraspecific pairs, S/S, N=12; mean mass and length differences=0.030 g (SE=0.009) and 0.79 mm (SE=0.363)). The fourth treatment was a control to assess arthropod abundance and diversity in the absence of centipede predators (post-experiment, N=12). Microcosms were maintained at 10±1°C on a 12-h light/12-h dark photoperiod. Both species normally gain mass when held individually under these conditions and Summer and Uetz (1979) report intermediate to high levels of surface activity at soil temperatures of 10°C. Contents of all microcosms were emptied weekly into a large stainless steel tray and hand sifted until the two 113 experimental centipedes were located. Controls were treated similar to the experimental treatments. Mass was recorded for each centipede in experimental treatments and all contents (litter, inverts, etc.) were placed back into the microcosms. Each microcosm was dampened with spring water (approximately 1.5 ml) before being placed back in the environmental chambers. Microcosms were moved each week among rack positions within the chamber to reduce positional effects. The experiment ran 9 weeks (1 October–28 November). Focal centipedes were removed from all microcosms and invertebrates were extracted from the microcosm litter at the end of the 9-week experiment to examine the effect of the two centipede species on the invertebrate community and to indirectly assess any differences in the diets of L. forficatus and S. sexspinosus. Invertebrates were separated from leaf litter by Berlese extraction into 70% ethanol. Berlese funnels were run for 4 days and invertebrates were separated from ethanol by pouring flask contents over paper coffee filters. Invertebrates were counted and identified to order. Invertebrates were not extracted from litter in treatment 2 (interspecific pairs) because a high occurrence of predation on L. forficatus reduced the sample size. We used changes in centipede body mass to determine if one species was a better competitor in microcosms than the other. We reasoned that in microcosms where no additional prey were added, individuals would be more likely to lose mass when compared with competitive dominants. To assess whether a species was a stronger intraspecific or interspecific competitor in laboratory microcosms, we compared the frequency of individuals that lost weight in intraspecific pairings to the frequency that lost weight in interspecific treatments. Here, we predicted that competitive dominants would be less likely to lose mass in interspecific trials. To assess the intensity of competition within intraspecific trials, we recorded instances where centipedes gained mass at the expense of their box-mate and compared these frequencies across intraspecific treatments. We predicted that in competitively dominant species intense competition would prevent both individuals from gaining mass. We used two-tailed chi-square tests for these comparisons. We analyzed the effect of intraguild predation on mass gain using a one-tailed, Mann–Whitney U-test where we compared the change in mass of Scolopocryptops that ate Lithobius (N=7) to a random sample of non-predatory Scolopocryptops (N=7) in the weeks that a Lithobius was consumed. To directly assess the effect of intraguild predation on mass gain, we compared mass of individual Scolopocryptops before and after eating box-mates (Lithobius) in interspecific treatments using a one-tailed, Wilcoxon signed-ranks test. We used onetailed tests for these analyses because we predicted that intraguild predators would gain rather than lose weight. Pre-experimental (N=12) versus post-experimental controls (N=9; three replicates were lost due to experimenter error) were compared to quantify changes in the invertebrate community over the 9-week period in the absence of adult centipedes. We compared total numbers of each taxon using two-tailed Mann–Whitney Utests. A three-way comparison of the abundance of common taxa was conducted between post-experimental control, and the two intraspecific treatments (L/L and S/ S) using a Kruskal–Wallis test. Taxa selected for analysis were those that occurred frequently enough in samples that there were relatively few zero cases. No Berlese funnel extractions were conducted on the interspecific treatment (S/L) due to intraguild predation and a resulting small sample size. All statistical analyses were conducted with SPSS for Windows, Version 11.5 with alpha set at 0.05. Results Field distribution—edge habitat use We observed 27 Scolopocryptops and 97 Lithobius under the 1,200 cover objects sampled (Table 1). Mean numbers of L. forficatus differed significantly in interior compared to edge habitats (Z=2.10; P=0.036; twotailed test). We found a mean of 5.33 (SE=1.18) L. forficatus in edge habitat and only 2.75 (SE=0.66) in fragment interiors. S. sexspinosus was found in fragment interiors significantly more often than in edge habitat (T=4.18; P=0.002; two-tailed test). A mean of 1.92 (SE=0.35) S. sexspinosus were collected from forest interiors and only 0.33 (SE=0.19) from fragment edges. There was no relationship between fragment size and distance to nearest urban area (r=0.149; P=0.644) indicating that smaller fragments were not necessarily closer to urban areas. We found no relationship between fragment size and numbers of either species in fragment interiors (Fig. 1; L. forficatus, R2 =0.024, P=0.633; S. Fig. 1 Numbers of centipedes observed in forest fragments of varying size. Centipedes (Lithobius—exotic, open circle and Scolopocryptops—native, filled circle) were observed in forest interiors during constrained searches. No significant effect of forest fragment size on abundance of either species was detected. The 12 forested sites were visited 19–29 May 2003 114 sexspinosus,R2 =0.0037, P=0.85) and no effect of site distance to urban area on the abundance of either species in the interior habitats was detected (Lithobius in edge, R2 =0.015, P=0.70; Scolopocryptops in edge, R2 =0.006, P=0.80). Microcosms—change in centipede mass and effects on invertebrate communities Over the course of the experiment, the mean mass of S. sexspinosus remained higher in the interspecific treatment compared to the intraspecific treatment (Fig. 2a). For Lithobius, the reverse was true (Fig. 2b). By week 8, 27% of surviving Scolopocryptops lost mass when paired with Lithobius (Fig. 3a) and 52% lost mass when paired with conspecifics (Fig. 3b). However, treatment had no statistically significant effect on mass lost by Scolopocryptops (v2=1.30; d.f.=1; P=0.3). In contrast, significantly more Lithobius lost mass when paired with Scolopocryptops than when paired with conspecifics (v2=6.67; d.f.=1; P<0.025). By week 8, 67% of surviving Lithobius lost mass when paired with Scolopocryptops (Fig. 3c) but only 14% lost mass when paired with conspecifics (Fig. 3d). In intraspecific pairings, we recorded instances where paired centipedes either gained mass, lost mass, or one gained and one lost mass. In six of nine surviving pairs of Lithobius, both centipedes gained mass or showed no change and in only three cases did a Lithobius gain mass while its box-mate lost mass (v2=6.0; d.f.=2; P=0.05). However, in 8 of 11 pairs of Scolopocryptops, one centipede gained mass while its box-mate lost mass and in only one instance did both centipedes gain mass (v2=6.26; d.f.=2; P=0.04). Mortality rates differed by treatment (v2=22.89; P<0.001; week 5); the highest rates were observed in interspecific pairings. In interspecific trials, 7 of 12 Lithobius were killed and consumed by Scolopocryptops; a single Scolopocryptops died but was not consumed by the Lithobius with which it was paired. Individuals of Scolopocryptops were significantly heavier (mean increase of 0.017 g or 6.8%) the week following consumption of Lithobius (N=7; paired Wilcoxon signedranks test; P=0.018) and Scolopocryptops that ate Lithobius (N=7) tended to gain more mass than those that did not (N=5; Mann–Whitney U-test; P=0.073). We assumed that mortality was due to predation because no deaths occurred for 10 weeks post-experiment when centipedes were held separately. Eighteen common taxa were identified in the four treatments and a total of 10,148 invertebrates were counted (Table 2). Regardless of treatment, microcosms were dominated numerically by mites, Collembola, and Coleoptera (Table 2). We compared numbers of invertebrates in leaf litter at the beginning of the experiment (pre-control) to the numbers at the end of the experiment (post-control) in the absence of adult centipede predators. There were no significant increases in any prey taxa, suggesting that emergence and/or reproduction of prey was not occurring in microcosms. At the end of the 9-week experiment there were fewer mites (Mann–Whitney U-test, P=0.011) and fewer larval beetles (P=0.042) in post-control microcosms compared to pre-controls (Table 2). Remaining comparisons were made between the post-control and centipede treatments. We found significantly more small centipedes and spiders in controls than in either experimental treatment (Fig. 4a, b). Similar numbers of pseudoscorpions were found in control and Lithobius treatments, but there were significantly fewer pseudoscorpions in Scolopocryptops treatments compared to control and Lithobius treatments (Fig. 4c). Discussion Land use by humans in recent decades has resulted in an increase in forest fragmentation. One result of fragmentation is an increase in edge habitat that often influences the distribution and abundance of species (Didham et al. 1996). Rapid anthropogenic habitat alteration has the potential to cause a reversal in the competitive advantage that previously well-adapted native species have over non-native species, and in extreme cases can drive native species locally extinct (Petren and Case 1998; Byers 2002). We used two common species of Fig. 2 Mass of a Scolopocryptops and b Lithobius centipedes in intraspecific (squares) and interspecific (circles) pairings (mean ± SE). Centipedes were paired in microcosms containing leaf litter. The centipedes were weighed weekly 115 Fig. 3 Equal probability plots illustrating mass gain and loss in the four microcosm treatments. Points above the line indicate a mass gain by the end of the 9-week experimental period. a S/L treatment where the majority of Scolopocryptops (S) gained mass when paired with Lithobius (L). b S/S treatment where approximately equal numbers of Scolopocryptops (S) lost and gained mass when paired with conspecifics. c L/S treatment where the majority of Lithobius (L) lost mass when paired with Scolopocryptops (S). d L/L treatment where the majority of Lithobius (L) gained mass when paired with conspecifics forest-dwelling centipedes, one native and one introduced by humans, to examine the effects of forest edge on species interactions. At our field sites, Lithobius was more abundant in edge habitat than interior habitat and was 16 times more abundant in edge habitat than was S. sexspinosus. Even within forest interiors, Lithobius outnumbered Scolopocryptops, but to a much lesser degree. No effects of forest fragment size or distance of fragments to urban areas were detected, suggesting that since its introduction in the 1800s (or earlier), Lithobius has spread throughout the region. Had we sampled fragments smaller than 20 ha we may have detected an increase in abundance of Lithobius, but we have seldom found Scolopocryptops in fragments this small. Our findings are consistent with Auerbach (1951) who noted that L. forficatus is associated with areas of human disturbance, and with the findings of Summers and Uetz (1979) who found L. forficatus to be the most abundant in clear-cut habitats. Similarly, in a survey of North American woodland centipedes by Lee (1980), L. forficatus was only found along the edge habitat created by road cuts. Likewise, in German forests, the abundance of L. forficatus decreased with increased distance to forest edges (Frund et al. 1997). In contrast to the distribution of L. forficatus, we found the native centipede, S. sexspinosus, to be most abundant in forest interiors and rarely occupying microhabitats in forest edge. Most workers have described S. sexspinosus as a deciduous forest or moist pine forest species (Lee 1980; Shelley 2002) but we know of no other studies that have established its lack of abundance in edge habitat. There are several possible explanations for the observed distributions of L. forficatus and S. sexspinosus at our field sites. Non-random spatial patterns can result from abiotic factors and differences in microhabitat use (Auerbach 1951; Lee 1980; Blackburn et al. 2002), competition for contested resources (Hairston 1980; Blackburn et al. 2002; Hickerson et al. 2004), predator avoidance (Murray et al. 2004), and intraguild predation (Suutari et al. 2004). 116 Table 2 Mean (SE) numbers of invertebrates in control and experimental treatments Control Taxa Nematoda Gastropoda Oligochaeta Isopods Symphyla Diplopoda Chilopoda Pseudoscorp Acari Aranae Collembola Hemiptera Thysanoptera Pscoptera Coleoptera (adult) Coleoptera (larva) Hymenoptera Diptera (adult) Diptera (larva) Lepidoptera (larva) Unidentified larvae Pre 0.42 (0.19) 0.25 (0.18) 1.00 (0.56) 2.67 (0.79) 0 (0) 0.58 (0.37) 0.75 (0.35) 4.25 (0.71) 142.67 (18.99) 2.25 (0.57) 13.75 (3.58) 0.58 (0.23) 0.67 (0.19) 0.58 (0.23) 6.25 (1.19) 3.67 (0.72) 1.67 (0.54) 0.42 (0.19) 1.67 (0.53) 1.42 (0.38) 0.67 (0.14) Post 0.33 (0.24) 0.22 (0.15) 0.44 (0.34) 3.67 (0.96) 0.11 (0.11) 1.00 (0.37) 1.11 (0.35) 3.11 (0.86) 66.67 (19.19) 1.67 (0.50) 7.56 (1.56) 0.67 (0.24) 1.11 (0.59) 0 (0) 5.11 (0.89) 1.89 (0.92) 1.00 (0.33) 0.22 (0.15) 1.33 (0.69) 0.44 (0.24) 1.67 (0.78) Predator Lithobius 0.14 (0.14) 0 (0) 0 (0) 1.14 (0.56) 0.14 (0.14) 0.43 (0.30) 0.14 (0.14) 3.43 (0.81) 108.14 (20.39) 0.43 (0.30) 6.71 (1.61) 0.57 (0.30) 0.43 (0.20) 0 (0) 5.86 (1.16) 4.43 (1.69) 1.71 (0.47) 0 (0) 2.29 (0.68) 0.14 (0.14) 1.14 (0.26) Scolopocryptops 0 (0) 0.27 (0.14) 0 (0) 2.09 (0.68) 0.27 (0.14) 0.90 (0.28) 0.18 (0.12) 0.91 (0.34) 109.55 (21.23) 0.45 (0.21) 11.09 (3.99) 0.36 (0.15) 0.09 (0.09) 0 (0) 5.00 (1.11) 6.36 (2.26) 0.91 (0.46) 0.36 (0.15) 2.64 (1.29) 0.27 (0.14) 0.73 (0.36) The outcomes of competitive interactions can be strongly affected by the physical attributes of the environments in which they occur (Park 1954; Connell 1961). Because edge habitats are often characterized by drier, warmer conditions (Murcia 1995), they may provide advantages to species adapted to such conditions by tipping the competitive balance in favor of such species. For example, Holway et al. (2002) illustrated how physical conditions of habitat fragments interacted with competitive interactions between introduced and native ant species. They found that the introduced Argentine ant (L. humile) had limited success in invading hot, xeric habitat fragments, but successfully invaded warm, mesic ones. We did not directly assess the role of the abiotic environment on competitive interactions between Scolopocryptops and Lithobius, but we recognize that these factors have the potential to reverse competitive outcomes and subsequently affect species distributions. Forest edge may provide a preferred habitat and/or a competitive advantage to introduced Lithobius. A number of authors have characterized L. forficatus as desiccation-resistant relative to other members of the genus (Auerbach 1951; Roberts 1956). However, owing to its larger surface to volume ratio, adults of S. sexspinosus are more resistant to desiccation than are adults of L. forficatus in laboratory experiments (Auerbach 1951). Thus it is unclear which species would benefit most from the physical conditions present in edge habitat and this question awaits further study. Our use of microcosms allowed us to indirectly assess use of shared resources and the potential role of intraguild predation in interactions of these two species. In our microcosms, 7 of 12 L. forficatus were preyed on by S. sexspinosus in interspecific pairings. Thus we view S. sexspinosus as an intraguild predator of L. forficatus. Predation on guild members is a potentially costly behavior because of the risk of physical injury resulting from capturing and subduing prey that are similar in size and fighting ability. An additional cost may be incurred by transfer of pathogens to intraguild predators that prey on phylogenetically similar guild members (Pfennig 2000). These fitness costs are similar to those incurred by cannibalistic species (Dawkins 1976; Polis 1981; Elgar and Crespi 1992; Pfennig et al. 1998). One way these costs can be offset is through the immediate nutritional benefit of consumption (Polis et al. 1989). Although we interpret the observed increase in mass of S. sexspinosus that preyed on L. forficatus as an example of a direct benefit of intraguild predation, it is not clear if the behavior is adaptive in this context. L. forficatus is a relatively new introduction to North America so there may be no reason to expect populations of Scolopocryptops to respond adaptively. Alternatively, native species of Lithobius have probably coexisted with Scolopocryptops over long periods. If Scolopocryptops responds similarly to L. forficatus as it does to native Lithobius then the observed behavior may be an example of an exaptation (sensu Gould and Vrba 1982). In a recent study, Snyder et al. (2004) examined interactions between native and introduced species of ladybird beetles (Coccinelidae) and found that in laboratory microcosms, intraguild predation played a significant role in competitive outcomes. In contrast with our results, native ladybird beetles were at a significant disadvantage, as both larvae and adults, in interactions with introduced species. The reported declines in native ladybird abundance (Day et al. 1994; Brown and Miller 1998) may be mediated by intraguild predation in these 117 Fig. 4 Numbers of invertebrates in the three treatments (Postcontrol, Scolopocryptops, Lithobius) at the end of the 9-week microcosm experiment (mean ± SE). Post-control microcosms were maintained without adult centipedes for the duration of the experiment. a Number of centipedes (of juvenile size or smaller) in each treatment. b Number of spiders in each treatment. c Number of pseudoscorpions in each treatment. Lowercase letters above bars indicate statistical differences at P=0.05 species (Obrycki et al. 1998). Not all introduced species cause declines in their native counterparts. For example, Bolger et al. (2000) found that native spiders and carabid beetles increased in abundance with increasing fragment age, despite increases in exotic species in these habitat fragments. Our results were similar in that native centipedes did not appear to decline with decreasing fragment size. Instead, we found that Scolopocryptops coexists with Lithobius in forest interiors. Size asymmetries between introduced and native guild members are likely to affect competitive outcomes. For example, introduced ladybirds are two to three times heavier than their native counterparts (Obrycki et al. 1998), and in our study, larger native centipedes appear to be at a competitive advantage. Holt and Polis (1997) suggest that intraguild predators are more likely to coexist if the intraguild prey (L. forficatus in our study) is a superior exploitative competitor for shared prey resources. Thus in our system we might expect coexistence only if S. sexspinosus (the intraguild predator) is an inferior exploitative competitor. Otherwise theory would predict the eventual competitive exclusion of L. forficatus through intraguild predation and competition. We have only indirect evidence for shared prey from our microcosm study. Centipedes, regardless of species, consumed smaller predators present in their microcosms as indicated by a significant reduction in spiders and small centipedes. Additionally, no statistical differences in species composition of prey remaining in microcosms were detected, with the exception of a reduction in the number of pseudoscorpions in the Scolopocryptops treatment. Thus centipedes in our microcosms appear to be entering into competition for limited resources. However, in contrast with the predictions of Holt and Polis (1997), interactions that occurred within microcosms suggest that S. sexspinosus is a stronger competitor (both intraspecifically and interspecifically) than is L. forficatus. In our study, L. forficatus lost mass when paired with S. sexspinosus and more S.sexspinosus lost mass when paired with conspecifics than when paired with L. forficatus. In forest fragments, Lithobius is not confined to edge habitat, though we did find significantly fewer Lithobius in fragment interiors. Thus, Lithobius apparently can coexist with the competitively dominant Scolopocryptops, despite serving as intraguild prey for this species. Temperate forest floor centipedes have historically been thought of as predators on other small soil-dwelling invertebrates and insects, but Lewis (1965) found that plant material in the form of dead leaf fragments, fungal hyphae, rootlets and spores made up approximately 50% of the diet of L. forficatus during the winter and spring months. The remaining gut contents consisted of common arthropods such as aphids, Collembola, and mites. We hypothesize that access to resources at lower trophic levels by Lithobius increases the likelihood of coexistence between it and its intraguild predator through reduction in competition for detrital mesofauna. In edge habitat, Lithobius was 16 times more abundant than was Scolopocryptops, yet Scolopocryptops preys on Lithobius and appears to be competitively dominant in laboratory microcosms. Edge habitat can promote the proliferation of shade intolerant plant species, alter microclimate, light regimes, moisture levels and facilitate exotic species invasions (Laurance and Yensen 1991; Hill and Curran 2001). Additionally, exotic species can substantially change aspects of their environment (e.g. soil development, nutrient cycling, hydrology, and primary or secondary productivity) and may intensify the effects of other invaders causing ecosystem level changes (Vitousek and Walker 1989; Vitousek 1990). Thus, edge habitat can change so drastically as to become inhospitable to native species, while becoming ideal habitat for introduced species. Indeed, we found populations of Lithobius in this altered habitat to be two to three times as dense as Scolopocyptops populations in forest interiors. Janzen (1983) described how influx of weedy species from fragment edges and the surrounding habitat could result in local extinction of competitively dominant species in fragment interiors. 118 Although we found no evidence for such an effect in our study, the failure to find a significant relationship between fragment size and abundance for introduced centipedes in interior habitat suggests that Lithobius can disperse from edge habitat where it is abundant. We hypothesize that intraguild predation by Scolopocryptops serves to maintain its competitive footing in forest interiors despite continual influx of Lithobius from the surrounding edge habitat. Acknowledgements Our manuscript benefited from the comments of two anonymous reviewers. The Ohio Conservation Alliance provided funding for this study. We thank J. Keiper and the Cleveland Museum of Natural History for additional support. CDA was supported by a JCU summer faculty research grant. All experiments complied with current laws of the country in which they were performed. References Auerbach SI (1951) The centipedes of the Chicago area with special reference to their ecology. Ecol Mono 21:97–124 Blackburn J, Farrow M, Arthur W (2002) Factors influencing the distribution, abundance and diversity of geophilomorph and lithobiomorph centipedes. J Zool 256:221–232 DOI:10.1017/ S0952836902000262 Bolger DT, Case TJ (1992) Intraspecific and interspecific interference behavior among sexual and asexual geckos. Anim Behav 44:21–30 Bolger DT, Suarez AV, Crooks KR, Morrison SA, Case TJ (2000) Arthropods in urban habitat fragments in southern California: area, age, and edge effects. Ecol Appl 10:1230–1248 Brown MW, Miller SS (1998) Coccinellidae (Coleoptera) in apple orchards of eastern West Virginia and the impact of invasion by Harmonia axyridis. Entomol News 109:136–142 Byers JE (2002) Impact of non-indigenous species on natives enhanced by anthropogenic alteration selection regimes. Oikos 97:449–458 DOI: 10.1034/j.1600-0706.2002.970316.x Case TJ, Bolger DT, Petren K (1994) Invasions and competitive displacement among house geckos in the tropical pacific. Ecology 75:464–477 Chamberlin RV (1925) The genera Lithobius, Neolithobius, Gonibius, and Zinapolys in America north of Mexico. Bull Mus Comp Zool Har 57:441–504 Connell JH (1961) The influence of interspecific competition and other factors on the distribution of the barnacle Chthamalus stellatus. Ecology 42:710–723 Dawkins R (1976) The selfish gene. Oxford University Press, Oxford Day WH, Procrym DR, Ellis DR, Chianese RJ (1994) The known distribution of the predator Propylea quatuordecimpunctata (Coleptera: Coccinellidae) in the United States, and thoughts on the origin of this species and five other exotic lady beetles in eastern North America. Entomol News 105:244–256 Didham RK, Ghazoul J, Stork NE, Davis AJ (1996) Insects in fragmented forests: a functional approach. Trends Ecol Evol 11:255–290 DOI: 0169-5347(96)20047-3 Ducey PK, Messere M, Lapoint K, Noce S (1999) Lumbricid prey and potential herpetofaunal predators of the invading terrestrial flatworm Bipalium adventitium (Turbellaria: Tricladida: Terricola). Am Midl Nat 141:305–314 Elgar MA, Crespi BJ (1992) Ecology and evolution of cannibalism. In: Elgar MA, Crespi BJ (eds) Cannibalism, ecology and evolution in diverse Taxa. Oxford University Press, New York, pp 1–12 Frund HC, Balkenhol B, Ruszkowski B (1997) Chilopoda in forest habitat-islands in north-west Westphalia, Germany. Entomol Scand 51:107–114 Gamboa GJ, Greig EI, Thom MC (2002) The comparative biology of two sympatric paper wasps, the native Polistes fuscatus and the invasive Polistes dominulus (Hymenoptera, Vespidae). Insect Sociaux 49:45–49 DOI: 0020-1812/02/010045-05 Gamboa GJ, Noble MA, Thom MC, Togal JL, Srinivasan R, Murphy BD (2004) The comparative biology of two sympatric paper wasps in Michigan, the native Polistes fuscatus and the invasive Polistes dominulus (Hymenoptera, Vespidae). Insect Sociaux 51:153–157 DOI: 10.1007/s00040-003-0721-1 Gerber GP, Echternacht AC (2000) Evidence for asymmetrical intraguild predation between native and introduced Anolis lizards. Oecologia 124:599–607 Gould SJ, Vrba ES (1982) Exaptation—a missing term in the science of form. Paleobiology 8:4–15 Griffin TT, Bull CM (1995) Interactions between introduced and native millipede species in South Australia. Aust J Zool 43:129– 140 Gundale MJ (2002) Influence of exotic earthworms on the soil organic horizon and the rare fern Botrychium mormo. Conserv Biol 16:1555–1561 Gunn A, Cherrett JM (1993) The exploitation of food resources by soil meso-macro invertebrates. Pedobiologia 37:303–320 Gurnell J, Wauters LA, Lurz PWW, Tosi G (2004) Alien species and interspecific competition: effects of introduced eastern grey squirrels on red squirrel population dynamics. J Anim Ecol 73:26–35 DOI:10.1111/j.1365-2656.2004.00791.x Hairston NG (1980) Evolution under interspecific competition: field experiments on terrestrial salamanders. Evolution 34:409– 420 Hickerson CM, Anthony CD, Wicknick JA (2004) Behavioral interactions between salamanders and centipedes: competition in divergent taxa. Behav Ecol 15:679–686 DOI:10.1093/beheco/ arh064 Hill JL, Curran PJ (2001) Species composition in fragmented forests: conservation implications of changing forest. Appl Geo 21:157–174 DOI: S0143-6228(01)00002-9 Hobbs RJ, Huenneke LF (1992) Disturbance, diversity, and invasion: implications for conservation. Conserv Biol 6:324– 337 Holt RD, Polis GA (1997) A theoretical framework for intraguild predation. Am Nat 149:745–764 DOI:0003-0147/97/4904-0007 Holway DA (1999) Competitive mechanisms underlying the displacement of native ants by the invasive Argentine ant. Ecology 80:238–251 Holway DA, Suarez AV, Case TJ (2002) Role of abiotic factors in governing susceptibility to invasion: a test with Argentine ants. Ecology 83:610–619 Janzen DH (1983) No park is an island: increase in interference from outside as park size decreases. Oikos 41:402–410 Kalisz PJ, Powell JE (2004) Exotic isopods (Crustacea: Isopoda) in disturbed and undisturbed forest soils on the Cumberland Plateau of Kentucky, USA. Nat Areas J 24:54–56 Laurance WF, Yensen E (1991) Predicting the impacts of edge effects in fragmented habitats. Biol Conserv 55:77–92 DOI:0006-3207/90 Lawrence B, Fisk MC, Fahey TJ, Suarez ER (2003) Influence of nonnative earthworms on mycorrhizal colonization of sugar maple (Acer saccharum). New Phytol 157:145–153 Lee RE Jr (1980) Summer microhabitat distribution of some centipedes in a deciduous and coniferous community of central Ohio (Chilopoda). Entomol News 91:1–6 Lewis JGE (1965) The food and reproductive cycles of the centipedes Lithobius variegates and L. forficatus in a Yorkshire woodland. Proc Zool Soc Lond 144:269–283 Lewis JGE (1981) The biology of centipedes. Cambridge University Press, London McKinney ML, Lockwood JL (1999) Biotic homogenization: a few winners replacing many losers in the next mass extinction. Trends Ecol Evol 14:450–453 DOI:S0169-5347(99)01679-1 McNeill J (1888) A list, with brief descriptions of all species, including one new to science, of myriapoda of Franklin, Co, Ind. Bull Brookville Soc Nat Hist 3:1–20 119 Murcia C (1995) Edge effects in fragmented forests: implications for conservation. Trends Ecol Evol 10:58–62 DOI:S01695347(00)88977-6 Murray DL, Roth JD, Wirsing AJ (2004) Predation risk avoidance by terrestrial amphibians: the role of prey experience and vulnerability to native and exotic predators. Ethology 110:635–647 DOI:10.1111/j.1439-0310.2004.01004.x Obrycki JJ, Giles KL, Ormord AM (1998) Interactions between an introduced and indigenous coccinellid species at different prey densities. Oecologia 117:279–285 Olden JD, Poff NL, Douglas MR, Douglas ME, Fausch KD (2004) Ecological and evolutionary consequences of biotic homogenization. Trends Ecol Evol 19:18–24 DOI:10.1016/ j.tree.2003.09.010 S0169-5347(03)00288-X Park T (1954) Experimental studies of interspecific competition II: temperature, humidity, and competition in two species of Tribolium. Physiol Zool 27:177–238 Petren K, Case TJ (1998) Habitat structure determines competition intensity and invasion success in gecko lizards. Proc Natl Acad Sci USA 95:11739–11744 Pfennig DW (2000) Effect of predator-prey phylogenetic similarity on the fitness consequences of predation: a trade-off between nutrition and disease?. Am Nat 155:335–345 DOI:0003-0147/ 2000/15503-0004 Pfennig DW, Ho SG, Hoffman EA (1998) Pathogen transmission as a selective force against cannibalism. Anim Behav 55:1255– 1261 Pimm SL, Lawton JH (1978) On feeding on more than one trophic level. Nature 275:542–544 Polis GA (1981) The evolution and dynamics of intraspecific predation. Ann Rev Ecol Syst 12:225–251 Polis GA, Myers CA, Holt RD (1989) The ecology and evolution of intraguild predation: potential competitors that eat each other. Ann Rev Ecol Syst 20:297–330 Roberts RB (1956) An ecological study of the arthropods of a mixed beech-oak woodland, with particular reference to Lithobiidae. PhD dissertation, University of Southampton, Highfield Rooney TP, Wiegmann SM, Rogers DA, Waller DM (2004) Biotic impoverishment and homogenization in unfragmented forest understory communities. Conserv Biol 18:787–798 DOI:10.1111/j.1523-1739.2004.00515.x Shelley RM (2002) A synopsis of the North American centipedes of the order Scolopendromorpha (Chilopoda). Virginia Museum of Natural History, memoir no.5, Martinsville, VA, USA Snyder WE, Clevenger GM, Eigenbrode SD (2004) Intraguild predation and successful invasion by introduced ladybird beetles. Oecologia 140:559–565 DOI:10.1007/s00442-004-1612-5 Suarez AV, Bolger DT, Case TJ (1998) Effects of fragmentation and invasion on native ant communities in coastal southern California. Ecology 79:2041–2056 Summers G, Uetz GW (1979) Microhabitats of woodland centipedes in a streamside forest. Am Midl Nat 102:346–352 Suutari E, Rantala MJ, Salmela J, Suhonen J (2004) Intraguild predation and interference competition on the endangered dragonfly Aeshna viridis. Oecologia 140:135–139 DOI:10.1007/ s00442-004-1559-6 Vitousek PM (1990) Biological invasions and ecosystem processes: towards an integration of population biology and ecosystem studies. Oikos 57:7–13 Vitousek PM, Walker LR (1989) Biological invasion by Myrica faya in Hawaii: plant demography, nitrogen fixation, and ecosystem effects. Ecol Mono 59:247–265 Williams SR, Hefner RA (1928) The millipedes and centipedes of Ohio. Ohio Biol Sur Bull 33:140–143 Zar JH (1999) Biostatistical analysis, 4th edn. Prentice-Hall, Upper Saddle River, NJ