Search

Introduction The dynamical behavior of individuals in ecosystems involves a multifaceted set of interaction types and processes that take place at different hierarchical levels. We present an individual-based, stochastic model that considers species dynamics at three hierarchical levels: population, community, and metacommunity. We use an individual-based model to show how the consequences of mechanisms that are specific to each hierarchical level may interact with processes that belong to other hierarchical levels. The strength of these effects is quantified in terms of impacts on metapopulation sizes and spatial distribution of populations. Results indicate the following: (1) the cohesion of the social network structure among conspecific individuals heavily affects their feeding efficiency at food-web level; (2) more generalist feeding habits trigger homogeneous spatial distribution of species at the landscape scale; and (3) high frequency of migration movements limits the local success of a generalist species thus leading to small metapopulation sizes. We illustrate how such a hierarchical framework may contribute to understanding the emergence of macroscopic patterns (i.e., metapopulation size and spatial heterogeneity) starting from elementary, bottom–up rules defined at the individual level. Hierarchical Organization and Individual-Based Modeling in Ecology Concurrent processes and interactions occur at different hierarchical levels in ecosystems (i.e., individual, population, community, and metapopulation/metacommunity) and do often spread their effects beyond the levels in which they actually originate. Some studies describe how ecological dynamics involving two hierarchical layers may interplay with each other. Social interactions among conspecific individuals may be regulated by metapopulation and community dynamics, community composition may be molded by landscape fragmentation, and species coexistence in metacommunities may result from the trade-off betw

Individual-Based Modeling in Conservation Biology A major challenge for food-web research is studying diversity and variability more explicitly. This means a focus on individual-level variability in populations (Bolnick et al., 2011) that hopefully might help to better understand how structural properties predict dynamical behavior (Dunne, 2006). One reason why linking structure to dynamics is still a hard challenge can be that intrapopulation variability is relatively poorly considered in most models. Yet defining developmental stages as graph nodes is a step toward managing this challenge: for example, in many food-web models certain species are represented by separate graph nodes that include juveniles and adults. Trophic status and network dynamics can be quite sensitive to this kind of demographic aggregation (or resolution) of the web and are expected to be influenced by individual-level differences in terms of behavior and feeding habits. Individual-level variability includes genetic, demographic, and stochastic factors and to date it is not easy to incorporate all these in most modeling frameworks. Yet developing the methodological background of individual-based modeling seems to be very useful for future research and applications. Since individual-level differences are more important in smaller populations (Lande, 1988), studying their effects explicitly is relevant for conservation efforts. Using network metrics as proxies or predictors of food-web dynamics is an old but still open issue. For an example, network hubs are supposed to be species of key importance and hubbish networks are thought to be safe against errors but vulnerable against attacks (Montoya and Solé, 2002). We make structural predictions routinely but we are very poor in testing these on either real-time series or simulation models. One way to make our knowledge more robust here is by adopting a comparative approach: studying spatio-temporal food-web gradients can inform about possible re

Introduction Trophic interactions are one of the most important aspects shaping ecological communities, and the food-web paradigm has played a major role in the development of ecology as a science. Early food-web models attempted to simulate the flow of energy and biomass within local communities (Odum, 1956) or describe the structure of feeding relationships (Pimm, 1982). Species or “trophospecies” (i.e., groups of species that supposedly share the same sets of predators and prey; see Yodzis and Winemiller, 1999) have served as the building blocks of both types of trophic networks (usually referred to as flow webs and topological webs, respectively). Topological food webs are static “snapshots” lacking magnitudes (i.e., estimates of the rate of energy flow or the strength of trophic links) and therefore have limited utility for examination of ecological dynamics. The strength of trophic interactions can be estimated in different ways (e.g., observing magnitudes of biomass/energy transfers, modeling of consumer feeding preferences, functional responses or metabolic constraints, varying interaction coefficients in Lotka–Volterra multispecies competition models, and quantifying responses from manipulative field experiments; see Berlow et al., 2004, 2009). Food-web models have been used to predict risks of secondary extinction (Allesina and Bodini, 2004), examine consequences of biodiversity loss for ecosystem stability (McCann, 2000), and study direct and indirect effects of predators on prey populations (Bondavalli and Ulanowicz, 1999). Food-web models that lump individuals into species and trophospecies lose much valuable information concerning influential attributes associated with age, body size, foraging history, location, and reproductive tactics. Such lumping may inflate the number of trophic interactions associated with species or trophospecies and thus fails to take into account how variation in feeding preferences between individuals affects food-web structu