Killfish evolution of brain cell proliferation: More plasticity according to a higher degree of predation

Predation drives the evolution of brain cell proliferation and brain allometry in male Trinidadian killifish, Rivulus hartii. Kent D. Dunlap, Joshua H. Corbo, Margarita M. Vergara, Shannon M. Beston and Matthew R. Walsh. Proceedings of the Royal Society B, Volume 286, Issue 1917, December 18 2019. https://doi.org/10.1098/rspb.2019.1485

Abstract: The external environment influences brain cell proliferation, and this might contribute to brain plasticity underlying adaptive behavioural changes. Additionally, internal genetic factors influence the brain cell proliferation rate. However, to date, researchers have not examined the importance of environmental versus genetic factors in causing natural variation in brain cell proliferation. Here, we examine brain cell proliferation and brain growth trajectories in free-living populations of Trinidadian killifish, Rivulus hartii, exposed to contrasting predation environments. Compared to populations without predators, populations in high predation (HP) environments exhibited higher rates of brain cell proliferation and a steeper brain growth trajectory (relative to body size). To test whether these differences in the wild persist in a common garden environment, we reared first-generation fish originating from both predation environments in uniform laboratory conditions. Just as in the wild, brain cell proliferation and brain growth in the common garden were greater in HP populations than in no predation populations. The differences in cell proliferation observed across the brain in both the field and common garden studies indicate that the differences are probably genetically based and are mediated by evolutionary shifts in overall brain growth and life-history traits.

1. Introduction
Researchers have devoted much attention to assessing whether changes in adult neurogenesis in response to the environment might be a mechanism of adaptive brain plasticity [1,2]. While the precise functional significance of adult neurogenesis is still debated, there is substantial evidence from many model systems that environmental stimuli alter neurogenic rates in specific brain regions, and that such neurogenic changes have behavioural consequences [3]. For example, complex odour environments increase neurogenesis in the olfactory bulb of rodents [4], and these new neurons enhance odour discrimination abilities [5]. Similarly, seasonal changes in day length promote neurogenesis in the song nuclei in the brains of several bird species, and these neurons are linked to seasonal song production [6].

Most of our understanding of environmental influences on adult neurogenesis comes from laboratory studies in which researchers manipulate environmental stimuli and document effects over the timescale of days to months. That is, they demonstrate an external factor driving phenotypic plasticity in the neurogenic rate. However, the neurogenic rate can also be influenced by intrinsic genetic factors [7–9], and thus, over evolutionary timescales, the environment can modify the neurogenic rate via natural selection acting within populations. Selection could act directly on the neurogenic rate if enhanced (or reduced) brain plasticity confers an advantage in responding to environmental change. Additionally, in species with indeterminate growth, such as most fishes, the brain grows in tandem with the body throughout adulthood [10,11], and selection on body growth trajectories could indirectly affect brain growth and the underlying cellular processes of brain growth [12]. Thus, population variation in the neurogenic rate could arise from phenotypic responses to different environments, or from evolved genetic divergence owing to direct selection on brain growth rate or as an indirect, correlated response to selection on body growth (figure 1). We evaluated these alternative explanations by examining one stage of adult neurogenesis, brain cell proliferation, in killifish (Rivulus hartii) populations from different predator environments. By measuring brain cell proliferation rates in populations exposed to differential predation pressure in the field as well as those same populations reared in a common laboratory environment, we assessed whether population variation in brain cell proliferation is attributable to natural environmental differences versus intrinsic population differences. Finally, we evaluated the predator effects on brain cell proliferation within the context of lifetime growth trajectories of the brain and body in populations [13–16] to assess how population variation in brain cell proliferation fits into the overall evolved difference in life history.

[Figure 1. Three alternative causal chains linking the environment with variation in brain cell proliferation.]

In Trinidad, R. hartii are found in sites where they are the only species present (Rivulus only (RO) sites) and lack predators as well as in sites where they are exposed to predatory fish such as Hoplias malabaricus and Crenicichla frenata (high predation (HP) sites). RO sites are typically located upstream from HP sites above barrier waterfalls that truncate the distribution of large piscivores [13,14,16,17]. These sites are located near each other and thus do not differ in physical habitat and environmental variables (i.e. water temperature and dissolved oxygen) [14]. In HP sites, Rivulus suffer increased mortality, are found at lower densities, and, in turn, they exhibit faster rates of individual growth (HP sites also have a more open canopy) [13,18]. Rivulus can also be bred and reared in the laboratory, allowing us to identify intrinsic (probably genetic) differences between populations that are independent of the environment, and many previous studies have indeed demonstrated that increased predation pressure is associated with evolutionary changes in life-history traits [13,15,19].

Recent work on Rivulus showed that divergent patterns of predation lead to evolutionary shifts in brain size [20]. Increased predation rates in HP sites are associated with the evolution of smaller brains in male (but not in female) Rivulus. Given this negative association between predator environment and brain size in male Rivulus and the negative effect of predators on brain cell proliferation in another freshwater teleost [21,22], we predicted that Rivulus from HP populations would have lower rates of brain cell proliferation than those from RO populations. In fact, we found the opposite: brain cell proliferation was higher in HP populations than in RO populations. These differences were maintained in first-generation laboratory-reared fish, indicating that they probably arise from evolved genetic divergence rather than through phenotypic plasticity. Population differences in cell proliferation were found across all sampled brain regions and correlated with population differences in overall brain allometry, suggesting that they evolved as part of broader evolutionary changes in overall brain growth rather than as a mechanism serving a specific behavioural adaptation.