…is explained in Miroslav Kummel et al.’s new online paper “How the aphids got their spots: predation drives self-organization of aphid colonies in a patchy habitat”.
A short summary is given here by Miroslav:
Spatial self-organization is the ability of a system to develop a spatially heterogeneous distribution of population sizes across otherwise identical locations This self-organization can result in static areas of high and low population density across otherwise homogeneous underlying conditions. Alternatively it can result in traveling population waves, or in spatial deterministic chaos. Self-organization has emerged as a key concept in population ecology, because it can fundamentally alter the outcome and stability of interspecific interactions. It has been studied extensively theoretically, but there are very few empirical studies that establish the presence and causality of spatial self-organization in the field.
In addition, the majority of current studies in spatial ecology (both empirical and theoretical) examine self-organization in laterally connected systems. These are systems where spatial effects are strongly determined by distance (e.g. the probability of colonization decreases with distance) and adjacency. However, space can be conceptualized in ways that are different. For example, space can be conceptualized as a network of connected patches, where the connections between patches are determined by other variables than pure distance. Human examples of such networks include a network of airports connected flights, natural examples include a collection of food patches connected by ant trails. The “network” conceptualization of space is very new to ecology and allows us to address previously intractable issues. Recent developments in network theory show that self-organization is possible in other network topologies such as random or scale-free networks.
In the field system that we study, foraging flights of coccinellid (ladybug) predators connect spatially discrete colonies of aphids into a network that has topology more complex than a laterally connected lattice. We show that predation by coccinellids induced self-organization in sessile aphid populations into small and large colony sizes on otherwise identical racemes of Yucca glauca that grew in close proximity to each other.
The self-organization was supported by a bi-modal frequency distribution of aphid colony sizes, and by the structure of density dependence that showed multiple attractors. The position of the attractors matched the position of the two modes in the bimodal distribution.
We demonstrated that predation was the key driver of self-organization both empirically, and through a simple field-parameterized mathematical model. In the empirical study we showed that the multiple-attractor nature of density dependence disappeared when coccinellids were experimentally excluded from the system. The simple field-parameterized mathematical model showed that the multiple attractor structure was likely a consequence of the distribution of coccinellids among the aphid colonies: coccinellid number increased as a power (less than one) of aphid colony size. Thus the self-organization likely originated from spatial foraging decisions of the coccinellids.
The nature of self-organization in our system resembles that which was found in mathematical models of more complex networks. Thus our study provides the first link between these recent theoretical developments and field ecology.