Evolutionary principles and their practical application

Abstract

Evolutionary principles are now routinely incorporated into medicine and agriculture. Examples include the design of treatments that slow the evolution of resistance by weeds, pests, and pathogens, and the design of breeding programs that maximize crop yield or quality. Evolutionary principles are also increasingly incorporated into conservation biology, natural resource management, and environmental science. Examples include the protection of small and isolated populations from inbreeding depression, the identification of key traits involved in adaptation to climate change, the design of harvesting regimes that minimize unwanted life-history evolution, and the setting of conservation priorities based on populations, species, or communities that harbor the greatest evolutionary diversity and potential. The adoption of evolutionary principles has proceeded somewhat independently in these different fields, even though the underlying fundamental concepts are the same. We explore these fundamental concepts under four main themes: variation, selection, connectivity, and eco-evolutionary dynamics. Within each theme, we present several key evolutionary principles and illustrate their use in addressing applied problems. We hope that the resulting primer of evolutionary concepts and their practical utility helps to advance a unified multidisciplinary field of applied evolutionary biology.

Keywords: adaptation; agriculture; climate change; conservation biology; contemporary evolution; evolutionary medicine; fisheries management; forest management.

Figure 1

Applied biology often considers mismatches between current phenotypes and those that would be best suited for a given environment. The graph shows the fitness of individuals with a given phenotype (fitness function: blue dashed line) and the distribution of phenotypes in a population under those conditions (numbers of individuals: black curve). The degree of the current mismatch is the distance between the peak of the fitness function and the peak of the frequency distribution. In Panel A, the mismatch is high and so average fitness in the population is low and the population size is small. In Panel B, the mismatch is small and so the average fitness is high and the population size is large. In some cases, we might wish a large mismatch to be smaller (e.g., conservation biology). In Panel A, then, we might manipulate phenotypes or the environment to decrease the mismatch (horizontal arrows). We might also find a way to increase fitness for a given phenotype (thin vertical arrow). The expected outcome is an increase in population size (thick vertical arrow). In other cases, we might wish the mismatch to be larger (e.g., pathogens or pests). In Panel B, then, we might manipulate phenotypes or the environment to increase the mismatch (horizontal arrows). We might also find a way to decrease fitness for a given phenotype (thin vertical arrow). The expected outcome is a decrease in population size (thick vertical arrow).

Figure 2

Phenotypic variation can be described by reaction norms. (A) Reaction norms depict the phenotypes a single genotype (or individual or population) expressed in different environments. Differences between reaction norms represent genetic differences. (B) Examples of reaction norms: shown are the adult male body sizes from different populations of male speckled wood butterflies (Pararge aegeria) when their larvae are raised at different temperatures (redrawn from Sibly et al. 1997). Temperature has plastic effects on body size in all populations, but the degree of its plasticity differs among populations. Genetic differences among the populations become more evident with decreasing temperature.

Figure 3

An example of how genetic differences and plasticity are jointly considered when evaluating potential responses to climate change. Panel A shows the mean breeding times of different UK populations of the common frog (Rana temporaria) in relation to the mean temperature experienced by those populations. Panel B shows how the mean breeding time within each of those populations varies among years with the mean temperature in those years. The lines thus represent adaptive phenotypic plasticity, and differences between the lines adaptive genetic differences among populations. Panel C shows the breeding time changes that each population is expected to undergo as a result of plasticity in response to projected warming between 2050 and 2070. Panel D shows the difference between these adaptive plastic responses and the changes in breeding time that would be necessary to keep pace with climate change if the trends on Panel A are fully adaptive. These differences thus represent the evolutionary change that will be necessary to maintain full adaptation. Adapted from Phillimore et al. (2010) with data provided by A. Phillimore.

Figure 4

Changes in protein content as a result of artificial selection in the Illinois maize (Zea mays) selection experiment. Initially, one line was selected for high protein (solid line trending up: IHP) and another for low protein (solid line trending down: ILP). Half way through the time series, new lines were established taking the high protein line and selecting for low protein (dashed line trending down: RHP) or taking the low protein line and selecting for high protein (dashed line trending up: RLP). Adapted from Moose et al. (2004).

Figure 5

The evolution of resistance to the cationic antimicrobial peptide pexiganan by Escherichia coli (Panel A) and Pseudomonas fluorescens (Panel B). Shown is growth rate (y-axis) in relation to the test concentration of pexiganan at the end of the selection experiment. The different colored lines in each panel represent different strains (each the average of multiple lines) selected for resistance (solid lines) and the same strains not selected for resistance (dashed lines). Adapted from Perron et al. (2005) with data provided by G. Perron.

Figure 6

Crop mimicry in barnyard grass. Panel A shows rice (Oryza sativa) on the left, a barnyard grass (Echinochloa crus-galli var. oryzicola) that mimics rice in the center, and a very closely related barnyard grass (Echinochloa crus-galli var. crus-galli) that does not mimic rice on the right. Panel B is a discriminant functions plot that shows the morphological similarity of multiple individuals (points) in these three groups. Centuries of hand weeding is thought to have led to the close similarity of rice and its barnyard grass mimic. Adapted from Barrett (1983) with a photograph and data provided by S. Barrett.

Figure 7

Examples of the community and ecosystem effects of phenotypic differences between fish populations. Panel A shows that mesocosms with guppies (Poecilia reticulata) from high-predation (HP) populations have more periphyton and fewer benthic macroinvertebrates than do mesocosms with guppies from low-predation (LP) populations. These data are adapted from Palkovacs and Post (2009)– see also Bassar et al. (2010). Panel B shows that mesocosms with alewife (Alosa pseudoharengus) from anadromous (ANAD) populations have fewer zooplankton and more phytoplankton than do mesocosms with alewife from resident freshwater (FW) populations. These data are for the first sampling date after fish were added to the mesocosms and are from Palkovacs et al. (2009) for zooplankton and from E. Palkovacs (unpublished) for phytoplankton. Panel C shows that mesocosms with benthic threespine stickleback (Gasterosteus aculeatus) have greater light extinction coefficients and greater UV absorption than do mesocosms with limnetic threespine stickleback. These data are adapted from Harmon et al. (2009) with data provided by L. Harmon. In all panels, the bars are standard errors around the mean value across replicate mesocosms.