The Pennsylvania State University The Graduate School HUMAN BEHAVIOR AS A DRIVER OF NON-HUMAN MORPHOLOGICAL AND GENOMIC EVOLUTION A Dissertation in Biology By Alexis P. Sullivan © 2020 Alexis P. Sullivan Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy December 2020 ii The dissertation of Alexis P. Sullivan was reviewed and approved by the following: George H. Perry Chair of the Bioinformatics and Genomics Interdisciplinary Graduate Program Associate Professor of the Departments of Anthropology and Biology Huck Institutes of the Life Sciences and Center for Infectious Disease Dynamics Dissertation Advisor Tracy Langkilde Professor of the Department of Biology Dean of the Eberly College of Science Chair of Committee Jesse R. Lasky Professor of the Department of Biology Timothy M. Ryan Professor and Head of the Department of Anthropology Matthew Reimherr Professor of the Department of Statistics Stephen W. Schaeffer Professor and Head of the Department of Biology iii ABSTRACT The earliest fossil evidence of Homo sapiens in Africa dates to ~300,000 years before present, with signs of behavioral modernity appearing ~200,000 years later. If we condense the Earth’s entire timeline to just 24 hours, modern human behaviors only exist in the final 2 seconds. Yet in this very short time, humans have become highly adept at changing the world and the non-human lives around them. Plant and animal domestication are the obvious examples of the effects that human behavior can have on non-human morphological evolution, but we are only just realizing how many other non-human organisms have evolved due to indirect human influences. Modern studies have recorded evidence of rapid morphological trait change in a variety of non-human, non-domesticated species due to anthropogenic influences, including harvesting and predation pressures, landscape modification activities, and genetic modification. There is also long-term evidence of such processes from the archaeological record, extending as far back as ~50,000 years before present. We review these human behaviors in Chapter 1, and document that the most well-studied of these behaviors is size- or trait-selective hunting and harvesting. Consequently, most of the examples we include in the introductory chapter and two of the data chapters fall into this category. Another human behavioral model that we’ll be discussing in the third chapter is translocation, specifically how the introduction of a novel species can affect non-human morphological evolution in native species. We used an integrative suite of tools to study the ways that human harvesting and translocation behaviors can directly or indirectly effect morphological adaptation in non- model species. Our approach begins by quantifying a morphological change that has likely been precipitated by some anthropogenic influence. Once a morphological change (phenotype) is characterized in an organism, we sequence whole genomes from many individuals to identify genomic regions in the nuclear DNA associated with that phenotype with a genome-wide association study (GWAS). We also use population genetics techniques to detect any signatures of recent positive natural selection on alleles associated with that phenotype. These techniques can also be applied to examine allele frequency changes over time in ancient DNA recovered from well-preserved organisms, though extraction protocols may need to be modified for recovery from non-standard materials. Chapters 2, 3 and 4 detail the results of applying this approach to three different non-model study systems, one mammal, one reptile, and one invertebrate. In Chapter 2 we present the first formal comparative morphological analysis of non- human primate subfossil and modern skeletal remains to evaluate the hypothesis of recent dwarfism among the still-living lemurs of Madagascar. We investigated a purported morphological change due to size-selective human hunting pressures in a population of Propithecus verreauxi (Verreaux’s sifakas). We collected both traditional caliper measurements and high-resolution 3D surface scan data from 106 P. verreauxi sifaka long bones to test the hypothesis that sifaka body size decreased following the earliest appearance of cut-marked bones in the area. After comparing the body size-associated measurements of subfossil bones from an archaeological site at Taolambiby to those from modern osteological collection at the nearby Beza Mahafaly Special Reserve, we found that the subfossil sifaka bones were indeed larger than those of the same species currently occupying the area. Although we cannot determine the ultimate cause of this size difference on the basis of our limited data, we believe that the findings from our analysis and our publicly available 3D data repository on MorphoSource make an important contribution to several fields of study, from the potential evolutionary consequences of human-environment interactions to zooarchaeology and primatology. Expanded archaeological sampling could corroborate our findings, and evolutionary genomic analyses may support the adaptive hypothesis versus an assemblage or taphonomic scenario where only the largest of the subfossil bones were represented. Chapter 3 details our preliminary efforts in correlating Sceloporus undulatus (eastern fence lizard) phenotypic adaptations to predation by invasive Solenopsis invicta (red iv imported fire ants) in the southeast United States. Humans accidentally introduced the fire ants through the port in Mobile, Alabama in the 1930s, and the venomous ants have been known to swarm, attack, and prey upon fence lizards. Some fence lizard populations have been exposed to fire ant predation for more than 70 years. Adult fence lizards at sites with the longest times since ant invasion have 3.4% longer relative hind limb lengths and increased twitching and fleeing behavioral responses compared to lizards at not-yet invaded sites. These traits are heritable from mother to offspring and the relative hind limb length differences between sites are inconsistent with expected patterns of ecogeographic variation based on museum specimens collected prior to fire ant invasion. To identify the genetic basis of S. unudulatus morphological traits hypothesized to represent adaptations to predation by red imported fire ants, we extracted and sequenced DNA from 381 individual fence lizards and conducted an evolutionary analysis of the identified genetic regions using methods from genome-wide association studies. Heritable components that are associated with phenotypes such as limb length are often polygenic, an expectation borne out by our finding of several thousand genetic variants/regions that appear to underlie fence lizard hind limb length variation. Our preliminary genome-wide association study (GWAS) results indicate that the top 17,172 (0.1%) of all single-nucleotide polymorphisms are near genes significantly enriched for biological processes that could be related to both the increased hind limb length and behavioral twitching phenotypes, such as nervous system and cellular development and neural activity. We also generated deeper-coverage sequence data for 20 of the Alabama individuals, as well as 20 individuals each from a site with no fire ant exposure in Arkansas and Tennessee. These data were used to test both within and between these three populations whether the patterns of limb trait variation have been driven by positive natural selection. Initial population differentiation (Fst) results indicate that the Alabama and Arkansas populations differ in genetic regions that are associated with synaptic signaling and transmission. Our nucleotide diversity (π) and LASSI (T) scans for selective sweeps indicated a strong signal in chromosome 10 that overlaps with our GWAS results and suggests that the MYO9B gene could be of functional importance for muscular motor activity in the Alabama lizards. We are currently verifying the quality of the genome-wide variants and their corresponding signals in each of our association and selection tests. In Chapter 4 we describe our endeavors in protocol development for harnessing both modern and ancient DNA from challenging sources. Based on comparisons from a mid- Holocene fringing reef, late Holocene excavations, and live-caught specimens, Strombus pugilis (West Indian fighting conchs) overall body size has decreased by up to 40% due to thousands of years of low-level human harvesting. We collected both living conch and recently eaten and discarded shells, as well as archaeological and paleontological materials, from several sites in the Bocas del Toro archipelago. We assembled both a nuclear and mitochondrial reference genome for S. pugilis from a freshly preserved tissue sample, and evaluated several DNA extraction techniques that were created for mineralized substrates like ancient bone and bivalve shell materials. We developed a modified method to extract both modern and ancient DNA from the robust, crystalline calcium structures of the discarded shells and report our assessment of the mapping rates and DNA damage from modern live-caught, recently discarded, late Holocene, and mid-Holocene shell materials. We were able to successfully sequence nuclear DNA and proportionate amounts of mitochondrial DNA from all of the shell materials. We also confirmed the authenticity of ancient DNA recovered from the archaeological and paleontological shells based on the presence of typical molecular signatures of post-mortem DNA damage and fragment