Ecological Entomology (2010), 35, 53–60 DOI: 10.1111/j.1365-2311.2009.01155.x Predation and facilitation determine chironomid emergence in a bromeliad-insect food web BRIAN M. STARZOMSKI,* DARYL SUEN and DIANE S. SRIVASTAVA Biodiversity Research Centre and Department of Zoology, The University of British Columbia, Vancouver, BC V6T 1Z4, Canada Abstract. 1. Ecological theory has focused on negative interactions, such as competition and predation, to explain species’ effects on one another. This study demonstrates the importance of considering both positive and negative interactions in explaining how species influence abundances at the local scale. 2. Two experiments were conducted using the aquatic insect food web in Costa Rican bromeliad phytotelmata. Manipulations contrasted the strength of predation between trophic levels versus facilitation within a trophic level on the emergence of detritivore chironomids. 3. Predation had a strong negative effect on chironomids, reducing emergences by 81% overall. Most predation was as a result of the top predator, the odonate Mecistogaster modesta ; the intermediate predator, a tanypodine chironomid, had little effect. In the absence of predators, shredder and scraper detritivores (tipulid and scirtid larvae) increased the emergence rate of chironomid larvae by 86%. The mechanism of facilitation was likely the processing, by tipulids and scirtids, of intact detritus into fine particles that the detritivore chironomids consume or use to build protective cases. 4. This study is among the first demonstrations of a processing chain in a multi- species context, and in bromeliad-insect food webs. Our finding that top-down effects are of similar magnitude to facilitative effects suggests that the relative importance of processing chains in nature will depend on food web context. Key words. Bromeliad, chironomid, Mecistogaster modesta, phytotelmata, processing chain. Introduction Ecological studies have documented a wide variety of inter- actions by which species in a food web affect other species, including vertical (between-trophic level) effects, such as pre- dation (e.g. Paine, 1966) and parasitism (e.g. Lafferty et al., 2006; 2008), and horizontal (within trophic-level) effects such as competition (e.g. Polis & Strong, 1996) and facilitation (e.g. reviewed in Bruno et al., 2003). However, few studies have examined the relative importance of vertical and hori- zontal processes on the fitness of individual species (Wootton, 1993; Bertness & Callaway, 1994; Heard, 1994a; Fox, 2004), Correspondence: Brian M. Starzomski, Biodiversity Research Centre and Department of Zoology, The University of British Columbia, 6270 University Blvd. Vancouver, BC V6T 1Z4, Canada. E-mail: starzom@uvic.ca ∗ Current address: School of Environmental Studies, University of Victoria, Victoria, British Columbia V8P 5C2, Canada although examples continue to accumulate. Even rarer are con- trasts of positive horizontal (within trophic level: facilitation) and negative vertical (between trophic level: predation) effects on species abundance (but see Stachowicz & Hay, 1999). More progress has been made comparing the effects of vertical and horizontal interactions on the biomass of entire trophic levels. The intriguing conclusion is that vertical interactions are at least as strong, if not stronger, than horizontal interactions (Duffy, 2003; Srivastava & Vellend, 2005). In this study, we compare the effects of predation and facil- itation on one fitness component of a detritivore insect. We examine in detail processing chains, a type of facilitation, in which at least one species conditions, or prepares, a resource for another species. The conditioning species has been termed the upstream species (it is at the beginning of the resource pro- cessing), and the receiving species is termed the downstream species (this species is at the receiving end of the flow of resources, in this case smaller particles of detritus – Heard, © 2010 The Authors Journal compilation © 2010 The Royal Entomological Society 53