Plastic and evolutionary responses to climate change in fish

Abstract

The physical and ecological 'fingerprints' of anthropogenic climate change over the past century are now well documented in many environments and taxa. We reviewed the evidence for phenotypic responses to recent climate change in fish. Changes in the timing of migration and reproduction, age at maturity, age at juvenile migration, growth, survival and fecundity were associated primarily with changes in temperature. Although these traits can evolve rapidly, only two studies attributed phenotypic changes formally to evolutionary mechanisms. The correlation-based methods most frequently employed point largely to 'fine-grained' population responses to environmental variability (i.e. rapid phenotypic changes relative to generation time), consistent with plastic mechanisms. Ultimately, many species will likely adapt to long-term warming trends overlaid on natural climate oscillations. Considering the strong plasticity in all traits studied, we recommend development and expanded use of methods capable of detecting evolutionary change, such as the long term study of selection coefficients and temporal shifts in reaction norms, and increased attention to forecasting adaptive change in response to the synergistic interactions of the multiple selection pressures likely to be associated with climate change.

Keywords: adaptation; climate change; evolutionary theory; fisheries management; life-history evolution; phenotypic plasticity.

Figure 1

a Left: The North Atlantic Oscillation (NAO) station-based seasonal index for winter (includes December – February) since 1865 (Hurrell 1995): annual index (vertical lines) are shown in the top panel, and 5 year running mean (black line) is shown in all panels. The portion of the NAO time series in each lower panel show the years included in a selection of studies in Table 1: (i) Sundby and Nakken (2008); (ii) Rogers et al. (2011); (iii) Beaugrand et al. (2003); (iv) Kjesbu et al. (1998). Dotted lines are linear regression lines fit to the period of data shown. Right: The PDO index since 1900: annual index (vertical lines) are shown in the top panel, and 3 year running mean (black line) is shown in all panels. The portion of the PDO time series in each lower panel show the years included in selected studies from Table 1: (i) Quinn and Adams (1996); (ii) Kovach et al. (2012). Dotted lines are linear regression lines fit to the period of data shown. Note that local temperatures do not necessarily follow the PDO, but may have additional trend upon them, as in the Kovach study.

Figure 2

Frequency distribution of traits showing shifts that correlate with environmental drivers in recent decades. Migration timing is broken into adult migrations (A), which are usually spawning migrations (includes “appearance” in Table 1) or seawater to freshwater migrations, and juvenile (J) or freshwater to saltwater migrations.

Figure 3

Frequency distribution of environmental drivers correlated with phenotypic change. NAO, North Atlantic Oscillation, T, temperature, SST, sea surface temperature; ice out is the day when a lake is free of all ice. We grouped counts by the species within a reference.

Figure 4

Frequency distribution of the number of years analyzed in each study.