All ecosystems are subjected to chronic disturbances, such as harvest, pollution, and climate change. The capacity to forecast how species respond to such press perturbations is limited by our imprecise knowledge of pairwise species interaction strengths and the many direct and indirect pathways along which perturbations can propagate between species. Network complexity (size and connectance) has thereby been seen to limit the predictability of ecological systems. Here we demonstrate a counteracting mechanism in which the influence of indirect effects declines with increasing network complexity when species interactions are governed by universal allometric constraints. With these constraints, network size and connectance interact to produce a skewed distribution of interaction strengths whose skew becomes more pronounced with increasing complexity. Together, the increased prevalence of weak interactions and the increased relative strength and rarity of strong interactions in complex networks limit disturbance propagation and preserve the qualitative predictability of net effects even when pairwise interaction strengths exhibit substantial variation or uncertainty.
Iles_Novak_2016_AmNat_ATN_Model_Simulation_Data
This is structured data file containing the ATN model simulation results described and analyzed in Iles & Novak 2016.
The file contains a cell array for each of the 5790 network simulations performed in Matlab.
Each cell array is named ‘ATN_#’, with the ‘#’ corresponding to the network number.
Each cell array for each network contains 9 variables, including:
1. the network number: a unique number for each network (1 to 5790)
2. the species richness: the number of species in the simulation; a network level property
3. trophic connectance: the proportion of all possible direct feeding links (L) between species (S) that are realized from the feeding matrix, C = L/S2; a network level property
4. dynamical connectance: a measure of connectance that includes all effects as encapsulated by the non-zero elements of the Community matrix. These include not only consumer-resource effects, but also intraspecific effects, apparent mutualisms and competition between producers; a network level property
5. the feeding matrix: a matrix of 1s and 0s indicating which species feed on which other species defined by the niche model; an interaction level property
6. species trophic levels: a vector of a trophic level values (one for each species) is calculated from the feeding matrix based on methods of Levine 1960; a species level property
7. species body mass: a vector of randomly drawn body masses (one for each species) based on the species trophic level; a species level property)
8. maximum eigenvalue of the Community matrix: a measure of asymptotic stability with which we can exclude unstable networks from the analysis; a network level property
9. Community matrix: a matrix of direct species interaction strengths from which the net effects can be calculated and predictability can be evaluated; the pairwise partial derivatives of the ith species’ growth rate with respect to the biomass of species j (a.k.a. a Jacobian matrix), evaluated at equilibrium biomass densities; an interaction level property
