Trophic dynamics in a simple experimental ecosystem: Interactions among centipedes, Collembola and introduced earthworms
Introduction
Biological invasions induced by human activities have forced ecologists to quantify the consequences of these invasions for ecosystem structure, ecological function, and ecosystem services. Less attention has been paid to the underlying processes that determine how invasive species interact and co-exist in the food web of a given ecosystem. Biotic interactions have long been considered one potential mechanism to explain species co-existence at a local scale (Wardle, 2006), but the contribution of an invading species to the preexisting assemblage of biotic interactions is not always clear, especially in belowground ecosystems.
Regulation of prey populations by predators is a well-known phenomenon in soil food webs (de Ruiter et al., 1995, Wardle, 2002), and many studies have found that populations of soil animals are controlled by their predators, emphasizing the role of top-down control (Wardle, 2002). Populations of soil animals occupying lower trophic levels (e.g., detritivores and fungivores such as Collembola) can be directly controlled through predator-prey interactions, i.e. density constrained by their predators, such as spiders (Lawrence and Wise, 2000) and centipedes (Poser, 1988). On the other hand, bottom-up regulation of trophic dynamics has also been demonstrated, either through direct availability of food resources (e.g., Scheu and Schaefer, 1998, Chen and Wise, 1999) or through non-trophic interactions, that alter the strength of top-down control through predation. For example, in a mesocosm experiment, the predator-prey interaction effect was altered by manipulating habitat complexity, with decreased habitat complexity resulting in greater control of collembolan populations by their centipede predators (Kalinkat et al., 2013). These results suggested that top-down control of predators on microbi-detritivore prey is less effective in litter-rich (i.e., complex habitat) ecosystems (Kalinkat et al., 2013). These examples reveal the importance of top-down vs. bottom-up controls in soil food-webs and the complexity of these interactions, supporting the overall view endorsed by Hunter and Price (1992) that both forces can act individually or simultaneously within ecosystems depending on conditions. Indeed, interpretation of results from experiments designed to examine the relative importance of top-down vs. bottom-up forces can change depending upon the duration of the experiments themselves (Lawrence and Wise, 2017). In light of the apparent context dependency of predator-prey outcomes in soil ecosystems, we wished to examine the influence of adding an “engineering” variable, such as the behavior of a novel invasive earthworm, and evaluate the strength of top-down control for a predator-prey food chain.
Non-native earthworms promote leaf litter decomposition in forest ecosystems across North America (Liu and Zou, 2002, Hale et al., 2008, Fahey et al., 2013) and therefore likely possess the ability to alter environmental complexity for other soil and litter dwelling organisms. Among these introduced species, Amynthas agrestis is a widespread invasive species that lives in the litter layer of forested ecosystems (Ishizuka, 2001) and is known to substantially promote leaf litter decomposition and decrease litter horizon depth (Qiu and Turner, 2016). Amynthas agrestis are native to Japan, but have expanded into forests in 17 states in the eastern United States since being introduced (Callaham et al., 2003, Reynolds and Wetzel, 2008), and it has been suggested that A. agrestis may have potential to further expand their distribution in the USA (Ikeda et al., 2015).
If invasive worms like A. agrestis consume substantial amounts of leaf litter in the forest, habitat structure for native microarthropods, like Collembola, should be reduced and simplified. We therefore expect that a “concentration effect” (as in Kalinkat et al., 2013) may be observed if Collembola are forced to live in such habitats. Identifying the influence of perturbations (such as invasive earthworms) on food web dynamics can help understand the mechanisms of species co-existence and inform management of biotic resources in forest ecosystems.
In this study, we sought to examine the effects of an invasive earthworm (A. agrestis) on the interaction between a generalist predator (centipede) and its putative prey (Collembola: Sinella curviseta). Invasive earthworm activity (specifically earthworms in the genus Amynthas) has been shown to decrease the complexity of the leaf litter habitat in mesocosms (Greiner et al., 2012, Snyder et al., 2013), as well as in field settings (Snyder et al., 2011). Thus, we hypothesized that invasive earthworm activity would strengthen the top-down control of centipedes on collembolan populations due to the assumed decrease in habitat complexity, resulting in lower collembolan populations in treatments with both the earthworm and centipede predator present. A corollary hypothesis was that if the collembolan abundance were adequate to meet the needs of centipede consumption, then centipede biomass should increase over the course of the experiment when incubated with Collembola alone, and that this increase in centipede biomass would be more pronounced in treatments with earthworms.
Section snippets
Collection and culture of earthworms, centipedes and Collembola
The earthworms used in this experiment were Amynthas agrestis, an invasive species that is wide-spread in the eastern part of the United States. We collected earthworms from the litter layer in a mixed hardwood forest at the Chattahoochee National Forest, in northern Georgia, USA (34.91°N, 83.62° W). Until they were placed into experimental mesocosms, earthworms were kept with sufficient litter, soil, and moisture in plastic containers at 20 °C.
Collembola used in this experiment were Sinella
Results
Results of the 3-way ANOVA for each response variable can be seen in Table 1. Abundances of Collembola were higher relative to starting conditions in all treatments on all sample dates (Fig. 1). In the factorial analysis, there were significant main effects of earthworm presence and time on collembolan abundance, as well as a significant interaction between earthworm presence and time (Fig. 2). Collembola abundance was significantly greater in mesocosms without earthworms early in the
Discussion
Collembolan abundance was depressed over time and by the presence of earthworms, with those factors interacting, such that after two and four weeks Collembola abundance was lower in mesocosms where earthworms were present, but after six weeks the effect of earthworm presence had diminished. This interactive effect was likely due to most earthworms losing biomass in all treatments as the incubation proceeded (Fig. 5), such that the large effect of earthworm presence early in the experiment was
Conclusions
Invasive earthworms influenced trophic dynamics in our experimental mesocosm, but not in the way we anticipated. Invasive earthworms directly depressed collembolan populations initially, and benefited centipedes throughout. Invasive earthworms may represent a novel food source for centipedes and other predators, and could potentially disrupt typical trophic relationships in ecosystems where they invade.
Acknowledgements
We thank the USDA Forest Service International Visitors Program and the China Scholarship Council (CSC) for the opportunity for this collaboration, our colleagues at the Southern Research Station and Harbin Normal University for providing valuable information and accommodations for this research. We thank D.A. Crossley, Jr. for providing the Collembola culture for this experiment, Roberto Carrera-Martínez for assistance in collecting centipedes, and two anonymous reviewers for constructive
References (39)
- et al.
Occurrence of an exotic earthworm (Amynthas agrestis) in undisturbed soils of the southern Appalachian Mountains, USA
Pedobiologia
(2003) - et al.
Unveiling soil food web links: new PCR assays for detection of prey DNA in the gut of soil arthropod predators
Soil Biology and Biochemistry
(2013) - et al.
Trophic interactions in centipedes (Chilopoda, Myriapoda) as indicated by fatty acid patterns: variations with life stage, forest age and season
Soil Biology and Biochemistry
(2012) - et al.
Temperature and moisture variables affecting the earthworms of genus Amynthas Kinberg, 1867 (Oligochaeta: megascolecidae) in a hardwood forest in the Champlain Valley, Vermont, USA
Applied Soil Ecology
(2016) - et al.
Can the invasive earthworm, Amynthas agrestis, be controlled with prescribed fire?
Soil Biology and Biochemistry
(2015) - et al.
Spider predation on forest-floor Collembola and evidence for indirect effects on decomposition
Pedobiologia
(2000) - et al.
Earthworms and collembola relationships: effects of predatory centipedes and humus forms
Soil Biology and Biochemistry
(2005) - et al.
Earthworm invasion in North America: food resource competition affects native millipede survival and invasive earthworm reproduction
Soil Biology and Biochemistry
(2013) - et al.
Les Arachnides et Myriapodes prédateurs en forêt tempérée
- et al.
Belowground competition among invading detritivores
Ecology
(2016)