I am interested in elucidating general principles in evolutionary
biology through the study of hummingbirds. This reflects my
underlying belief that hummingbirds are exemplary model organisms
for many of the reasons prompting focus on yeast or Drosophila,
but with singular advantages that stem from their remarkable
taxonomic and morphologic diversity, their extensive use in
physiologic and ecologic studies, and their status as the extreme
for many vertebrate adaptations.
Molecular Systematics
At the time I began my studies, the main impediment to the
use of hummingbirds as a model system in evolutionary biology
was the lack of information on their evolutionary history.
Therefore, I undertook to generate phylogenies with various
molecular methods. Our first studies relied on the method
of DNA hybridization. More recently we have applied DNA sequencing
to obtain much more extensive sets of comparisons. To make
progress, however, considerable background work was required.
This included a decade-long effort to collect tissue samples
across the spectrum of hummingbird diversity (especially
in the wilds of South America), theoretical studies to assess
the statistical properties of the data generated, and consideration
of avian ordinal relationships to determine the nearest relatives
to hummingbirds. Substantial progress in all of these areas
culminated in the generation of the first extensive and well-resolved
phylogenies for the hummingbird family. The phylogenies confirmed
some traditional groupings (hermits versus non hermits) but
also revealed a host of previously unsuspected relationships,
especially among the species-rich non-hermits.
With the phylogenies already in hand, it has been possible
for me to achieve some of my longer-term goals to study how
history and ecology interact to shape patterns of adaptation,
community structure, and diversity. In addition, the DNA hybridization
studies have also been important to the burgeoning field of
molecular systematics. In particular, a vast “tapestry” of
DNA hybridization-based phylogenies by Sibley and Ahlquist
published in the 1970s and 1980s revolutionized long-held beliefs
about avian relationships. However, perceived short-comings
in the methods and analyses applied by these workers prevented
their work from gaining wide acceptance. By vindicating many
of Sibley's and Ahlquist's findings, our work has served to
promote the acceptance and use of their findings.
DNA Evolution And The Molecular Clock
As hummingbirds represent the vertebrate extreme for many factors
thought to effect rates of DNA change (small body size, short
generation time, high metabolic rate) they present an ideal
group for analyzing the biology of molecular evolution. Analysis
of our molecular data indicated relatively fast rates of
DNA change in hummingbirds, thereby confirming a negative
association of body mass with rate of DNA divergence that
had been observed in several other organisms. In addition,
I found evidence that rates of molecular evolution decline
in hummingbirds living at higher compared to lower elevations.
While the underlying mechanism responsible for this phenomenon
remains unclear (though evidence suggests that it has a physiological
basis independent of body mass), this novel result challenges
the view that molecular markers are insensitive to environmental
effects. These data also imply that the environmental context
can influence divergence dates estimated under the assumption
of a molecular clock. To investigate this hypothesis further,
I undertook studies of the fossil record to test the validity
of divergence times based on assumptions of a molecular clock;
such dates are often much older than those indicated by the
evidentiary fossil record. Using the statistical method of
gap analysis, I found that robust confidence intervals placed
the origin of selected avian orders much closer to their
first appearances in the fossil record than did prior estimates
based on a molecular clock. The practical fall-out from these
studies is that they have allowed me to place more realistic
time-scales on the phylogenies used in the comparative studies
of adaptation and evolutionary radiation.
We are currently generating DNA sequence data to extend these
studies. In addition to applying these new methods to hummingbirds,
we have initiated study of sequence evolution in large-bodied
raptors as a natural complement to our studies of the small-bodied
hummingbirds. Our results confirm the expectation for low rates
of genetic evolution and variation in raptors, even in geographically
widespread forms such as the Andean condor. The condor studies
also provide important baseline information for genetic studies
of endangered species, many of which are large-bodied, by demonstrating
that low genetic variability in such megafauna may exist for
reasons other than a severe population bottleneck. Thus, our
work has begun to reveal both general biological controls on
molecular evolution, and well as how these controls may bias
molecular characters as indicators of evolutionary history.
Plumage And Sexual Dimorphism
My interest in the evolution of plumage coloration and plumage
sexual dimorphism arose from my broader interest in relating
social behavior to evolution. It has taken some years for
these two lines of investigation to converge. My initial
aim was to test the validity of the conventional view that
plumage colors evolved principally through sexual selection
on males. For hummingbirds, I documented a number of patterns
that were at odds with this sexual selection hypothesis.
These included variable and often highly ornamented plumages
of females, the restriction in some species of bright female
plumage to the immatures, evolution of sexual dichromatism
through change in female plumage, and significant associations
between plumage and bill-length sexual dimorphism. These
studies pioneered the use of female plumage to address hypotheses
about the evolution of color and sexual dichromatism. Through
them, I also developed indices that allowed plumage to be
quantified easily and as a continuous trait, thereby permitting
application of rigorous phylogeny-based statistical methods
across many taxa. On the basis of these studies, I showed
that even birds with breeding systems thought to generate
the highest levels of sexual selection (lek-type mating systems)
nevertheless evolve a highly predictable set of plumages
that probably reflect differences in foraging ecology rather
than levels of mate competition. With hummingbirds, I was
then able to develop this idea further to suggest that sexual
dimorphism reflects priority of access to food resources
as determined by the combined influences of social dominance
and of constraints imposed on dominance by mating system.
Color Vision And Evolution
My interest in the role of perception on signal evolution is
an inevitable outcome of my interest in coloration, as birds
in general, and hummingbirds in particular, possess certain
perceptual abilities, such as sensitivity to ultraviolet
(UV), lacking in normal humans. Thus, it was necessary to
examine many of the hypotheses outlined above from the birds'
perspective. My two major lines of research in this area
concern the effects of discrimination (at long wavelengths)
and sensitivity (at short wavelengths). I developed a new
biological theory to explain why many hummingbird-pollinated
flowers are red based on hummingbirds' poor discrimination
of long (red) wavelengths. I also found that hummingbird
plumage reflects strongly in the UV, although I failed to
find evidence for ultraviolet patterns that were not also
represented in the range of wavelengths visible to humans.
These studies served to validate the indices developed above,
for example.
Future Directions
Although I have combined field and laboratory-based studies
throughout my career, my recent work has focused mainly on
the latter. In the past year, however, we constructed a research
station on the eastern slopes of the Ecuadorian Andes, where
hummingbirds attain their greatest diversity. This will allow
us to add field based studies into the overall research program.
In Madison, I continue to pursue the molecular and comparative
studies, which now also include major efforts to quantify
and study ecomorphology.
