UC Berkeley

Evolutionary novelty may be an important driver of diversification, yet theory does not agree on the circumstances or processes that may lead to novelty in the first place. Current theory may be insufficient for explaining observed natural patterns because it does not consider the multivariate nature of novelty and instead makes predictions assuming that evolution occurs along a single axis. Furthermore, it is challenging to measure variation across many biological levels in a single focal organism, and there is consequently a lack of empirical data to inform new hypotheses. Here, I investigate evolutionary novelty within an adaptive radiation of Cyprinodon pupfish endemic to San Salvador Island, Bahamas. This radiation contains two fantastic examples of novelty in the snail-eating (C. brontotheroides) and scale-eating pupfish (C. desquamator), and contains a third generalist pupfish (C. variegatus) similar to outgroup populations.

In chapter one I investigate if shifts in behavior are associated with novelty in pupfish. I specifically test the aggression hypothesis, which suggests that scale-eating arose due to the incidental ingestion of scales during aggressive interactions and predicts that scale-eating pupfish should have the highest levels of aggression compared to all other species. I empirically tested this prediction and found that both scale- and snail-eating pupfish are more behaviorally aggressive than algae-eating or outgroup species and show differential expression in the same aggression related pathways compared to the generalist pupfish species. These results suggest that behavioral shifts are indeed associated with the evolution of novelty and may be adaptive for dietary specialization.

In chapters two and three I investigate the contributions of morphological and behavioral adaptations to scale- and snail-feeding performance. I developed a new method for measuring bite size (a proxy for scale-feeding performance) across pupfish species and found that scale-eating pupfish have a unique, behaviorally mediated, kinematic profile adaptive for taking large bites. The kinematic profiles of F1 scale-eating hybrids from this study were dissimilar to those of purebred scale-eaters—even for traits that are behaviorally mediated—suggesting that shifts in morphology and behavior are likely necessary for high performance in this niche. I also measured snail-eating performance between pupfish species, to determine if the novel nasal protrusion (morphology) of the snail-eating pupfish is adaptive for snail consumption. I found that snail-eaters, snail-eating hybrids, and generalist pupfish all exhibit similar levels of performance and observed no relationship between nasal protrusion size and performance. These chapters suggest that shifts along a single axis (e.g., behavior) may be sufficient to produce some instances of novelty, but that others may require shifts along several axes in order to be successful.

In chapter four I investigated the genetic underpinnings of adaptive behavioral and morphological traits for both snail- and scale-feeding, with the goal of identifying if unique genomic regions, alleles, or SNPs were associated with each instance of novelty. I investigated the genetic basis of 31 adaptive traits for scale- and snail-feeding using a QTL mapping approach in two pupfish populations. I found that similar genomic regions were associated with craniofacial traits, but that many of these shared regions affected different, but highly correlated traits. I also found evidence of increased levels of adaptive introgression within shared QTL regions, suggesting that: 1) there is a surprising amount of genetic flexibility when adapting to a novel niche, and 2) introgression may be important for this transition.

Ultimately, this dissertation research provides a new framework for studying the multivariate nature of novelty. By quantifying variation across behavioral, morphological, and genomic axes I documented that scale-feeding may require shifts along multiple axes while snail-feeding may only require behavioral shifts. This is one of the first studies to empirically document variation across multiple biological levels in the context of novelty and the patterns described above further emphasizes that hypotheses considering shifts along only a single axis are not sufficient for fully understanding evolutionary novelty.