Global urban signatures of phenotypic change in animal and plant populations

Abstract

Humans challenge the phenotypic, genetic, and cultural makeup of species by affecting the fitness landscapes on which they evolve. Recent studies show that cities might play a major role in contemporary evolution by accelerating phenotypic changes in wildlife, including animals, plants, fungi, and other organisms. Many studies of ecoevolutionary change have focused on anthropogenic drivers, but none of these studies has specifically examined the role that urbanization plays in ecoevolution or explicitly examined its mechanisms. This paper presents evidence on the mechanisms linking urban development patterns to rapid evolutionary changes for species that play important functional roles in communities and ecosystems. Through a metaanalysis of experimental and observational studies reporting more than 1,600 phenotypic changes in species across multiple regions, we ask whether we can discriminate an urban signature of phenotypic change beyond the established natural baselines and other anthropogenic signals. We then assess the relative impact of five types of urban disturbances including habitat modifications, biotic interactions, habitat heterogeneity, novel disturbances, and social interactions. Our study shows a clear urban signal; rates of phenotypic change are greater in urbanizing systems compared with natural and nonurban anthropogenic systems. By explicitly linking urban development to traits that affect ecosystem function, we can map potential ecoevolutionary implications of emerging patterns of urban agglomerations and uncover insights for maintaining key ecosystem functions upon which the sustainability of human well-being depends.

Keywords: anthropocene; ecoevolution; ecosystem function; sustainability; urbanization.

Fig. 1.

Global distribution of study systems of trait changes in wild populations. Symbols represent Urban Disturbances, wherein each study system is categorized according to its primary driver of phenotypic trait change. White regions represent City Lights. Background of the Earth in 2012 from NASA: earthobservatory.nasa.gov/Features/NightLights/page3.php.

Fig. 2.

Representation of the AICc model selection table. Rows represent models sorted by decreasing empirical support (row height represents model probability conditional on the full model set). Predictor variables were shaded if included in a model. Saturation corresponded to estimated variable relative importance. Note all high-ranked models contained urban (e.g., Urban Disturbance) and background variables (e.g., Generations). Models with little relative support were omitted for clarity (95% confidence set displayed).

Fig. 3.

Multimodel predictions for Urban Disturbance categories. Some categories were associated with relatively high phenotypic change (i.e., Social Interactions and Introduction of predator/prey/host/competitors), whereas others associated with background or even stable phenotypes (Habitat modification). Modeled, but not shown, variables were held constant at means (continuous variables) or reference values (categorical variables).