Determinism vs. stochasticity in competitive flour beetle communities

TL;DR

This study compares deterministic and stochastic models in predicting competitive outcomes of flour beetle species, finding deterministic models reliably predict outcomes despite stochastic influences, and clarifies mechanisms behind priority effects.

Contribution

It demonstrates that deterministic population dynamics sufficiently explain beetle competition patterns, challenging the necessity of stochastic models for such predictions.

Findings

Deterministic models predict competitive outcomes with 94% accuracy.

Stochasticity influences the predictability of the winning species.

Egg predation is identified as the mechanism behind priority effects.

Abstract

As ecologists increasingly adopt stochastic models over deterministic ones, the question arises: when is this a positive development and when is this an unnecessary complication? While deterministic models -- like the Lotka-Volterra model -- provide straightforward predictions about competitive outcomes, they are often unrealistic. Stochastic models are more realistic, but their complexity can limit their usefulness in explaining coexistence. Here, we investigate the relative importance of deterministic and stochastic processes in competition between two flour beetle species, Tribolium castaneum and Tribolium confusum. Specifically, we use highly-replicated one-generation experiments (784 microcosms) to parameterize a mechanistic model. Both the full stochastic model and the underlying deterministic skeleton exhibit priority effects, where one species excludes the other, but the identity of the winning species depends on initial abundances. Stochasticity makes the identity of the winner less predictable, but deterministic dynamics still make reliable predictions (94% accuracy across a range of reasonable initial abundances). We conclude that deterministic population dynamics are sufficient to account for patterns of coexistence (or lack thereof), a potentially general finding that is supported by recent field studies. Additionally, we resolve longstanding issues in flour beetle research by identifying selective egg predation as the mechanism for priority effects, demonstrating the primacy of demographic stochasticity (compared to environmental stochasticity), and reinterpreting classic competition experiments to show that apparent coexistence often represents long-term transient dynamics.