LETTER The dynamics of assembling food webs Ashkaan K. Fahimipour 1,2 * and Andrew M. Hein 2,3 1 Department of Biology, University of California, Riverside, CA, 92521, USA 2 Department of Biology, University of Florida, Gainesville, FL, 32611, USA 3 Department of Ecology & Evolutionary Biology, Princeton University, Princeton, NJ, 08544, USA *Correspondence: E-mail: ashkaan. fahimipour@gmail.com Abstract Community assembly is central to ecology, yet ecologists have amassed little quantitative informa- tion about how food webs assemble. Theory holds that colonisation rate is a primary driver of community assembly. We present new data from a mesocosm experiment to test the hypothesis that colonisation rate also determines the assembly dynamics of food webs. By manipulating colo- nisation rate and measuring webs through time, we show how colonisation rate governs structural changes during assembly. Webs experiencing different colonisation rates had stable topologies despite significant species turnover, suggesting that some features of network architecture emerge early and change little through assembly. But webs experiencing low colonisation rates showed less variation in the magnitudes of trophic fluxes, and were less likely to develop coupled fast and slow resource channels – a common feature of published webs. Our results reveal that food web structure develops according to repeatable trajectories that are strongly influenced by colonisation rate. Keywords Colonisation, community assembly, food webs, information theory, interaction strengths. Ecology Letters (2014) 17: 606–613 INTRODUCTION Food webs in nature develop through a dynamic assembly process; new trophic interactions begin as colonists enter com- munities, and the arrangement and magnitudes of energy and material fluxes change as populations grow and shrink. Understanding how biotic and abiotic conditions cause food webs to change through time is a goal of both fundamental and applied ecology (Dobson et al. 2009; May 2009). Yet, even the most basic empirical questions about temporal varia- tion in food web structure remain unanswered (May 2009). This is due to a paucity of quantitative food web data. Time series data of who eats whom and at what rates do not exist (Olesen et al. 2010; Emmerson 2011) and replicated food web experiments are scant. In contrast to empirical studies, theoretical work has dealt explicitly with food web dynamics, and the dynamics of web assembly in particular. Theory suggests that the emergence of structural features in assembling webs should be strongly dri- ven by colonisation rate – the rate at which new individuals enter a community (Yodzis 1981; Post & Pimm 1983). Subse- quent models have extended this prediction, positing that col- onisation rate influences the formation of web topology (Pillai et al. 2011), distributions of species’ per capita interaction strengths (Pawar 2009) and species’ abundance distributions (Sole et al. 2002). In conjunction with the arrangement of interactions (i.e. the way in which links of different magni- tudes are arranged in a web), these properties collectively determine the quantitative structure of a web at a particular point in time (Ulanowicz 2004). Unfortunately, because of a lack of data, it has been difficult to test existing theory (Emmerson 2011). Moreover, it is not clear what empirical patterns new models must explain. There is a pressing need for a more thorough dialogue between food web assembly theory and data (Dobson et al. 2009). Here, we describe results of an aquatic mesocosm experi- ment designed to determine how colonisation processes affect the trajectory of food web structure through assembly (data set provided in Appendix S1 in supporting information and as electronic supplementary data). Specifically, our goal was to manipulate colonisation rate in an array of field mesocosms, and to study the development of three major structural fea- tures of the webs that formed in these mesocosms: (1) web topology, (2) variation in the magnitudes of trophic fluxes, and (3) the arrangement of trophic fluxes. These features are discussed in detail in Materials and Methods and Appendices S2 and S3. We report that variation in colonisation rate among food webs causes consistent differences in the emer- gence of web structure during assembly. Our results reveal that food web structure develops according to repeatable tra- jectories that are determined by colonisation rate on timescales that are relevant to the dynamics of real ecological systems. MATERIALS AND METHODS In June 2008, Hein & Gillooly (2011) established an array of six 1-m diameter aquatic mesocosms around each of seven permanent lakes at the Ordway-Swisher Biological Station in Melrose, FL, USA (42 mesocosms total). Mesocosms were studied from 11 June 2008 to 6 August 2008. In this study system, natural temporary ponds are seasonally filled with water and colonised by aquatic species (e.g. wind and animal-dispersed phytoplankton and zooplankton, oviposited aquatic insect and amphibian larvae and flying aquatic insects) that immigrate from nearby permanent lakes. Manipulating colonisation Hein & Gillooly (2011) showed that increasing the distance between mesocosms and the nearest permanent lake reduces © 2014 John Wiley & Sons Ltd/CNRS Ecology Letters, (2014) 17: 606–613 doi: 10.1111/ele.12264