Independent evolution of baleen whale gigantism linked to Plio-Pleistocene ocean dynamics

Abstract

Vertebrates have evolved to gigantic sizes repeatedly over the past 250 Myr, reaching their extreme in today's baleen whales (Mysticeti). Hypotheses for the evolution of exceptionally large size in mysticetes range from niche partitioning to predator avoidance, but there has been no quantitative examination of body size evolutionary dynamics in this clade and it remains unclear when, why or how gigantism evolved. By fitting phylogenetic macroevolutionary models to a dataset consisting of living and extinct species, we show that mysticetes underwent a clade-wide shift in their mode of body size evolution during the Plio-Pleistocene. This transition, from Brownian motion-like dynamics to a trended random walk towards larger size, is temporally linked to the onset of seasonally intensified upwelling along coastal ecosystems. High prey densities resulting from wind-driven upwelling, rather than abundant resources alone, are the primary determinant of efficient foraging in extant mysticetes and Late Pliocene changes in ocean dynamics may have provided an ecological pathway to gigantism in multiple independent lineages.

Keywords: body size; fossil; macroevolution; phylogeny; tempo and mode; trend.

Figure 1.

Mean body lengths for extant mysticetes and estimated length for fossil species (baleen-bearing mysticetes, circles; toothed mysticetes, triangles) are plotted according to their age as inferred from our phylogeny. Shaded areas correspond to 80 (white), 90 (grey) and 95% (black) quantiles of 1000 Brownian motion simulations on mysticete phylogeny and illustrate that the modern fauna is both lacking in small species (less than 5 m) and over-represented in large ones (more than 10 m), relative to the fossil record. To the right is a smooth-spline fitted to the Eocene–Present oxygen isotope curve [36], and used as a proxy in modelling temperature-dependent body size evolution. Higher δ¹⁸O values correspond to cooler temperatures.

Figure 2.

The phylogenetic distribution of large body size is young. A reconstruction of body length evolution under a simple Brownian motion model shows that large (more than 10 m) body size evolved independently in the bowhead (Balaena) and right (Eubalaena) whales, several lineages of Balaenoptera, and though not as pronounced, in the grey whale (Eschrichtius). The extant pygmy right whale Caperea is also large relative to its cetotheriid relatives.

Figure 3.

Disparity through time for mysticete body size is not entirely consistent with Brownian motion. Though the empirical curve (solid line) closely approximates the median DTT curve (dashed line) derived from Brownian motion simulation, the peak at 5.01 Ma is not consistent with such a process (MDI = 0.545, p = 0.064). (Online version in colour.)

Figure 4.

The support surface for the mode-shift model supports a young transition to gigantism. The correlation between shift-time and the trend parameter for the biased random walk phase is strong and the lack of Pleistocene fossil taxa in our phylogeny increases uncertainty in the estimation of both. Nonetheless, the 2-unit support region is sufficiently restricted to rule out traditional, pre-Pliocene explanations for gigantism. Contours show the 2, 3, 5 and 10 unit support regions.