The portfolio concept in ecology and evolution

© The Ecological Society of America www.frontiersinecology.org T greatest challenges in ecology and evolution are in understanding how physical and biological processes that play out over extensive spatial, temporal, and taxonomic scales interact to affect the dynamics of genes, phenotypes, populations, and ecosystems. The fact that many biological properties are “scale-dependent” has been recognized for decades; biological systems tend to be extremely variable when observed at some scales but relatively invariant at others (Levin 1992). Scale-dependencies in biological systems are produced by non-linear interactions within systems, and from weak or negative covariances in system components across space and time. While fascinating ecologists for decades in terms of quantifying patterns in nature, scale-dependencies have received limited attention in exhibiting portfolio effects, which have important implications for understanding ecological processes and evolutionary dynamics. Here, we show how the “portfolio” concept (WebPanel 1) – a construct that developed out of simple probability theory and has been widely adapted in financial investment theory – applies across ecology and evolution, and provides a framework for understanding how organisms and ecosystems achieve stability in their dynamics despite inherent volatility in their components. The portfolio concept was developed to recognize that diversified financial investment portfolios tended to produce more stable returns than simple portfolios. Thus, the dynamics of financial systems have scale-dependencies analogous to biological systems: dynamics may be extremely volatile at small scales but less variable at more aggregated scales. Modern portfolio theory was first described through the use of graphical and simple analytical models to show the value of diversification for reducing risks in investment strategies (Markowitz 1952). These models introduced the idea that the selection of an efficient investment portfolio should assess the trade-off between the expected return and the variance of alternative asset collections. The key insight was recognizing that the variability of an aggregate portfolio depends critically on the covariation among the component assets. We contend that the nearly ubiquitous scale-dependencies in biological systems (ie that systems are often less variable when viewed at coarse scales than at fine scales; this is the emergent property of interest, where the whole behaves differently than the individual components; Levin 1992), generated by weak and negative covariation within their components, stabilize the emergent properties of many ecological and evolutionary processes. Examples are evident in phenomena ranging from the behavior of individual organisms, to population and ecosystem dynamics, and even to evolutionary strategies. These portfolio effects produce reliable biological functions in a world characterized by stochasticity and unpredictability, and provide underappreciated options for considering risk in natural resource management and conservation.