By Brian Myers, Grant Recipient 2015

Historically, the evolution of novel morphological traits, such as the diversification of limb bones in mammals, has been subjected to rigorous study, while behavioral characters have lagged behind. In my dissertation, I am using recombination in a naturally occurring hybrid zone to study how differentiation and behavioral novelty arise between the recently diverged Allen’s (Selasphorus sasin) and Rufous (S. rufus) Hummingbirds. I will determine the genes responsible for a novel courtship behavior and, for the first time, directly test Tinbergen’s “Emancipation Hypothesis”, which states that behavioral novelty can occur from the duplication of one trait into two, where one character is the original trait, and the other is novel, to investigate how this behavior might have evolved. In addition, for the first time, I am investigating how genetics, behavior, and morphology vary across an avian hybrid zone and study the mechanisms of isolation at play within the area of contact.

Allen’s and Rufous Hummingbirds are phenotypically similar, but differ in display behavior. In both species, courtship involves a male performing an initial ascent and subsequent J-shaped dive. During the dive, a species-specific sound is emitted by the tail feathers. In addition, Allen’s Hummingbirds exhibit a novel behavior occurring immediately before the dive, the pendulum display, where the bird slowly flies back and forth in a shallow U-shape. Both displays can be broken into distinct elements, and some elements of the dive are analogous to those in the pendulum display. I hypothesize the dive and pendulum displays are behavioral duplications, where the pendulum display is a modified dive with its own character identity.

Because some populations appear to be declining, Allen’s Hummingbird is currently listed on the National Audubon Society’s Watchlist. Allen’s Hummingbird consists of two subspecies: Selasphorus sasin sasin and Selasphorus sasin sedentarius. S. s. sedentarius was previously endemic to the Channel Islands, but is now widespread throughout southern California. S. s. sasin is 15% smaller than S. s. sedentarius and is found from Oregon south to Santa Barbara County. In addition, the two subspecies differ in their migratory behavior, with S. s. sasin migrating annually to northern California and Oregon and wintering in central Mexico and S. s. sedentarius remaining in southern California as a year-round resident. My research will clarify which parts of the distribution are undergoing expansions versus declines, and I will compare the phylogeography of Allen’s Hummingbird to that of previous studies of co-distributed species in California.

Grant fund use
All of my funds were used towards the 2016 field season, towards camping fees, food, and transportation. With the help of this award, I was able to collect morphological, behavioral, and genetic data from over 100 individuals.

Research progress

Chapter 1: Describe the hybrid zone
Thus far, I have collected data from 104 males from the hybrid zone, 6 pure Allen’s Hummingbird males from their range, 15 pure Rufous Hummingbird males from their range, and 73 females from the hybrid zone, for a total of 198 birds sampled. Most females were measured, sampled for DNA (blood sample via toenail clipping), and released. Approximately half of the males were collected, and the other half released after data was taken.

To standardize measurements of pure parents, I measured specimens of pure Rufous and Allen’s Hummingbird at the San Diego Natural History Museum and the Los Angeles Museum of Natural History. After obtaining 30 specimen measurements each for pure Allen’s males, pure Allen’s females, pure Rufous males, and pure Rufous females, I will have a basis of comparison for hybrid individuals so that hybrids may be identified in the field based on both morphology and behavior.

Phenotypic analyses reveal a gradual north-south shift from a Rufous-like to an Allen’s-like phenotype. For example, the northernmost hybrid locality is Rufous-like, but individuals occasionally perform pendulum displays. Our southernmost hybrid population is Allen’s-like, but some individuals perform intermediate dive sounds as a result of mixed tail feather morphology (Fig. 1). I performed a linear discriminant function analysis (LDA) using Minitab to diagnose individuals sampled from the hybrid zone as pure Rufous, pure Allen’s, or hybrid (Fig. 2). I am currently working on the publication for this chapter, which will: 1) confirm the presence of the hybrid zone, 2) assess the degree of admixture of behavioral and morphological traits, and 3) provide the required framework needed for addressing variation in genetic data across the hybrid zone in Chapter 3.

1 Figure1 Image

Figure 1. Sum of four rectrix two tail feather measurements (a score of 0=Allen’s-like, 12=Rufous-like) versus fundamental dive frequency (Hz) produced by the tail feathers during courtship (1600-1800Hz=Allen’s-like dive sound, 700-800 Hz=Rufous-like dive sound). Intermediate tail morphology and dive sounds suggest hybridization

2 Figure2 Image

Figure 2. Hybrid and parent localities, showing the proportion of phenotype observed at each locality based on linear discriminant function analysis (LDA) of standard morphological characters and behavioral characters described in this proposal (green=Allen’s-like, orange=Rufous-like, blue=hybrid). Pure parent localities are in southern California and northern Oregon, far from the hybrid zone.

Chapter 2: Cline analysis of the hybrid zone using genetics, behavior, and morphology
In this chapter, the goal is to determine which evolutionary forces maintain the hybrid zone, and to investigate whether prezygotic or postzygotic isolation is a larger driver of divergence across the area of contact. To explore potential genetic markers, myself and collaborators sequenced the genomes of eight pure Allen’s and seven pure Rufous Hummingbirds far from the hybrid zone. We found our two focal species to be very similar, with only 82 species-diagnostic single nucleotide polymorphism (SNP) markers. The vast majority (61) of these markers are located on the Z chromosome (the sex chromosome in birds), and 11 others are located on a single autosome. We will genotype these 82 species-diagnostic SNP markers using PCR-restriction fragment length polymorphism (RFLP)-based analysis. We will also use ddRADSeq to discover additional SNP markers that are variable, but not species diagnostic, from throughout the genome. These data will also be valuable for Chapter 4.

In the 2015 and 2016 field seasons I have nearly completed a north-south transect along the southern Oregon coast. In the upcoming field season, I will finish this transect and begin an inland west-east transect in extreme northern California.

Chapter 3: Phylogeography of Allen’s Hummingbird
Various museum specimens and tissue samples are available from throughout the range of Allen’s Hummingbird. In addition, I have samples from Riverside, Santa Catalina Island, and Del Norte County from the previous field season. To fill-in the remaining gaps in sampling, I will sample the following counties in 2017: Ventura, Monterey, Humboldt, Mendocino, and south Del Norte.

Chapter 4: Admixture mapping to determine the genetic components responsible for the pendulum display
In this chapter, the goal is to determine the genes associated with the pendulum display in Allen’s Hummingbird, and to determine whether the dive and pendulum display are paramorphs, where the pendulum display is a modified dive that has attained its own character identity. Chapters 1-3 require completion and all required samples for this project must be obtained before analysis for Chapter 4 can be initiated.