PS 32-196 - Confirming the insurance effect of rare species in ecological communities using an agent-based model

Abstract: The majority of species in ecological communities are rare. Environmental perturbations may favor rare taxa in communities where disturbance events decrease numerically abundant taxa, which ensures ecosystem stability. Rare species, thus, provide an insurance effect and contribute to ecosystem functioning when conditions do not favor common species, although mechanisms of the insurance effect of rare species remain unclear. This study aims to illustrate the insurance effect by means of an agent-based model that predicts the relationship between environmental variability, species richness and rarity, and stability of overall resource consumption by the community. Communities were simulated as sets of individuals belonging to different species that lived and reproduced over 500 time steps. Individuals of each species were assumed to have (i) the same half-life, after which the probability of death increased as a logistic function of lifespan; and (ii) the same critical biomass after which the probability of reproduction (i.e., death of the parent individual and birth of two individuals with half the body mass of the parent individual) increased as a logistic function of body mass. Resource consumption by individuals of a species was determined by whether the level of abiotic environmental factor was suitable for a species, where suitability was simulated as a Gaussian curve with random mean and variance for each species (i.e., resource consumption of a species was highest when environmental conditions approached the most suitable value for a species). Assuming constant values of critical body mass, half-life, initial resources, and resource income, I simulated 10,000 runs with species richness ranging from 2 to 20 and abiotic factors either remaining constant throughout the simulation or monotonically increasing or decreasing, with difference between starting and ending value of the abiotic factor denoting environmental variability. Model outputs showed that speciose communities with many rare taxa had consistently higher mean resource consumption, while the stability of resource consumption, as measured by a Dickey–Fuller time series stationarity test, was higher in species-rich communities at intermediate levels of environmental variability. This outcome suggests that rare species ensure ecosystem stability under environmental disturbance, confirming the insurance effect hypothesis.