Multi-objective optimization shapes ecological variation

Abstract

Ecological systems contain a huge amount of quantitative variation between and within species and locations, which makes it difficult to obtain unambiguous verification of theoretical predictions. Ordinary experiments consider just a few explanatory factors and are prone to providing oversimplified answers because they ignore the complexity of the factors that underlie variation. We used multi-objective optimization (MO) for a mechanistic analysis of the potential ecological and evolutionary causes and consequences of variation in the life-history traits of a species of moth. Optimal life-history solutions were sought for environmental conditions where different life stages of the moth were subject to predation and other known fitness-reducing factors in a manner that was dependent on the duration of these life stages and on variable mortality rates. We found that multi-objective optimal solutions to these conditions that the moths regularly experience explained most of the life-history variation within this species. Our results demonstrate that variation can have a causal interpretation even for organisms under steady conditions. The results suggest that weather and species interactions can act as underlying causes of variation, and MO acts as a corresponding adaptive mechanism that maintains variation in the traits of organisms.

Local adaptations and phenotypic plasticity may render gypsy moth and nun moth future pests in northern European boreal forests

“…into account in such studies, as geographically distant populations of the same species may have developed the ability to thrive in different conditions and even to cope with a continuously changing climate (Ammunét et al 2011, Valtonen et al 2011, Kaitaniemi et al 2012, Valladares et al 2014.…”

A guide to multi-objective optimization for ecological problems with an application to cackling goose management

“…We focused on examples related to natural resource management. Several other applications of multi-objective optimization relevant to ecology include model selection (Williams 2016), life history evolution (Kaitaniemi et al 2011), and behavior ecology (Rothley et al 1997, Schmitz et al 1998. Model selection is concerned with balancing the competing objectives of model fit and model complexity (Burnham andAnderson 2002, Williams 2016).…”

“…Life history evolution concerns balancing vital rates to maximize fitness (e.g., clutch size vs. survival probability ;Lack 1947). Kaitaniemi et al (2011) used multi-objective optimization as a mechanistic analysis of potential ecological an evolutionary causes and consequences of variation in life-history traits of a species of moth. Similarly, behavioral ecology concerns balancing competing interests such as finding food and avoiding predators (Mangel and Clark 1986).…”