Bipartite networks are topological descriptors of interactions between two trophic levels (e.g., interactions between plants and herbivores). The frequency of observed visits by animals to each plant species is used as a surrogate for the effect of the animal on the fitness of plant. This assumption is problematic, as interaction frequency is not synonymous with interaction strength. We suggest that to understand ecological and evolutionary processes, bipartite networks need to focus on more appropriate metrics of fitness effects.
We propose the term Demographic Networks: bipartite networks where links of at least one trophic level represent the actual fitness obtained through each interaction. One conceptual advance of this study will be the development of analytical tools to estimate mean fitness of species transitioning into different environments.
This study focuses on interactions between Zingiberales and their insect herbivores, rolled-leaf beetles (genus Cephaloleia) at La Selva Biological Station, a tropical rain forest in Costa Rica. We illustrate the concept by assembling a demographic network using life tables of this insect guild reared on multiple host plants. The quantitative link of each insect species to each host plant is represented by the instantaneous population growth rates (r) i.e., the average fitness of a cohort of an insect species reared on a particular species of host plant.
Results/Conclusions
We distributed individuals of four beetle species (Cephaloleia belti, N = 1108, C. dilaticollis, N = 1071, C. dorsalis, N = 697, and C. placida, N = 721) in cohorts raised on two or three different host plants. After two years, we estimated survivorship, fecundity and fitness (instantaneous population growth rate, r) on each host plant. After transforming life tables into Leslie matrices, we calculated the global fitness for individuals in each cohort experiencing multiple host plants. Models assuming no transitions of insects between host plants show that after multiple generations, the fitness of the population is an approximation to the fitness associated with the host plant that confers the highest instantaneous population growth rate. One key result is that insect species assumed to be generalists only attained population growth in one or a few environments. The extinction of a host plant may result in an increase in fitness if the instantaneous population growth rate on that host is < 0. We call this phenomenon cryptic specialization. This example illustrates the potential of demographic networks to integrate demography into network theory.