Genetic and Evolutionary Basis of Behavioral Variability in Drosophila melanogaster

Abstract

Behavioral individuality is a ubiquitous phenomenon. Animals from genetically homogeneous populations reared in identical environments display persistent and variable behavioral phenotypes. This behavioral variability can be heritable and potentially adaptive through a bet-hedging strategy, wherein genotypes that produce a range of behavioral phenotypes enhance their long term fitness in unpredictably fluctuating environments by increasing the probability that some individuals will express traits well-suited to the current conditions. In this dissertation, I investigate the genetic and evolutionary basis of behavioral variability in Drosophila melanogaster, combining theoretical modeling with experimental approaches to test its evolvability and underlying mechanisms. In Chapter 1, I develop a computational model to examine how variability responds to artificial selection. I compare family-based and individual-based selection strategies and show that family-based selection, where variability is measured across related individuals, is more effective than individual-based selection when the trait under selection is variance-based but not when it is mean-based. I further explore how population size, genetic architecture, and selection strength shape evolutionary outcomes. In Chapter 2, I apply these insights in a 21-generation artificial selection in fruit flies experiment targeting increased variability in locomotor handedness. I observe a consistent increase in behavioral variability over selection without a shift in mean turning bias. This evolved variability is polygenic and potentially has some sex-linked basis. I also identify changes in central complex morphology and correlated tradeoffs in mating success. In Chapter 3, we shift focus to thermal preference, an ecologically relevant behavior. Using a panel of inbred lines, we show that the mean and variability of thermal preference are both heritable, and genetically independent. Genome-wide association implicates spag, a co-chaperone of Hsp90, as a regulator of thermal preference variability. In Chapter 4, we investigate the genetic basis of the offspring number–body weight tradeoff using a composite offspring index. We identify candidate genes influencing this tradeoff and demonstrate through functional experiments that specific mutations alter the balance between offspring size and number. In the concluding chapter, I outline future directions, including experimental evolution approaches for studying bet-hedging. Together, this work shows that behavioral variability is an evolvable trait shaped by genetic and neuroanatomical factors.