|
By Mike Gannis
 |
 |
 |
 |
In addition to horns, these lizards have another
defensethey can actually squirt blood
from their eyes; coyotes and foxes apparently
dislike the taste and drop them.
Horned lizards are specialized for
eating ants. They are very difficult to feed
in captivity, and often die of starvation.
Being cold-blooded, horned lizards
bask in the sun to warm up and cover their
bodies with sand to avoid temperature extremes.
Females of some species lay eggs, while
females of others bear live young, which is
thought to be an adaptation for cold environments.
Several species exhibit rock
mimicry, in which they change the shape
of their bodies to mimic the stones in their
habitat.
They drink during rainstorms by rain
harvesting, standing with hind legs
up and head down, allowing water to drip into
their mouths.
Horned lizards tend to stay within
a moderate home range, but sometimes go on
long walkabouts, moving hundreds of meters
in a day. They have been known to move total
distances of up to a mile. |
|
 |
 |
 |
|
It looks like something from the Mesozoic Eraa squat
body covered with sharp spines, head encased in bony armor,
dark penetrating eyes, and horns worthy of a Triceratops
dinosaur. Phrynosoma
cornutum would indeed be a terrifying sight if you
were the size of an ant, its usual prey. But many humans find
the docile and harmless Texas Horned Lizard (sometimes called
a "horny toad") rather endearing. UC Riverside biologist
Wendy Hodges
certainly likes the creatures, and has studied horned
lizards for a decade. Her latest studies of the reptiles
involve an attempt to reconstruct the physical features of
the common ancestor of the 13 living species of North American
horned lizards.
Assisted by visualization
expert Reuben
Reyes at the Texas
Advanced Computing Center (TACC) at the University of
Texas at Austin, a member of the National Partnership for
Advanced Computational Infrastructure (NPACI), Hodges is applying
advanced computer analysis and graphics techniques to three-dimensional
data sets acquired through computed tomography (CT). The results
are both scientifically informative and visually striking.
Between 23 and 30 million years ago, the first North American
horned lizards branched off from sand lizards, becoming stockier
and spinier and evolving their distinctive crowns of horns.
A single common ancestor gave rise to the 13 species of North
American horned lizards, each developing a unique set of cranial
horns and spikes. But what did this common ancestor look like,
and how did the horns develop?
"CT reconstructions enable us to analyze and compare
the morphologies of different species, in this case the Texas
Horned Lizard, Phrynosoma
cornutum, and the Mexican Horned Lizard, Phrynosoma
taurus," Hodges said. "The goal of this
project is to visualize the evolution of horns in this group
of lizards and to determine how horn number increased through
evolutionary time."
In their initial
efforts, Hodges and Reyes are applying "morphing"
programs and algorithms for computing ancestral states in
combination with three-dimensional morphologies from CT scans
to visualize an intermediate form between two species.
The fossils of horned lizards and other small desert-dwelling
animals are small, fragile, and easily overlooked. The rarity
of fossils is one reason for Hodges and Reyes
efforts to reconstruct ancestral forms through computer-based
morphing.
PROJECT LEADERS
Wendy L. Hodges, University of California, Riverside
Reuben Reyes,
Texas Advanced Computing Center, University of Texas
at Austin |
|
Analysis of DNA sequences plays a key role in determining
evolutionary relationships. Over 2,500 base pairs representing
four genes from horned lizards were sequenced, aligned, combined
with morphological data used in previous studies, and used
in phylogenetic analyses to reconstruct the relationships
between horned lizard species. But genetic information alone
isnt everythingsmall differences in gene expression
during embryonic development can translate into significant
changes in adult morphology. Reconstructing and visualizing
the ancestral forms of organisms requires the full combination
of comparative genetics, phylogenetics, and 3-D morphology,
with a major assist from advanced computer modeling.
Structural information was acquired from preserved horned
lizard museum specimens using a high-resolution CT scanner
at the National Science Foundation (NSF)-supported University
of Texas High-Resolution
X-ray Computed Tomography Facility. The scanner is comparable
to a conventional medical diagnostic CT device, but was custom-built
to have greater resolution and penetrating power and was specifically
designed to explore the internal structures of natural objects
and materials at macro- and microscopic levels. In its five
years of operation, this instrument has scanned a variety
of specimens, from rocks, meteorites, and fossils to plant
and animal structures.
Eventually, images of each
horned lizard species head will be created from
the CT data using visualization software, to reveal both the
external configurations and the internal structures of the
specimens. One visualization application depicts
the external skin and internal skeleton using a color-mapping
scheme that represents the distance from the skin to the bone,
giving some of the lizard head reconstructions a polychromatic,
"tie-dyed" appearance.
The reconstruction and morphing software source code is written
in C++ using "Performer" graphics libraries. Computational
resources used in the effort included a Silicon Graphics Onyx2
computer with 24 processors, 25 gigabytes of main memory,
and six "Infinite Reality2" graphics pipelines,
and other NPACI computer resources.
"Were still working on the morphing reconstruction,"
said Reyes. "After we create visualizations from the
combination of data from the Mexican and Texas Horned Lizard
species as a test case, we intend to integrate the data for
all species. In the future we plan to include non-linear morphing
effects, and we hope to be able to visualize how the morphologies
varied over time as the various species branched off from
one another."
Hodges and Reyes efforts are conducted in coordination
with the University of Texas Digital Morphology Group, an
informal association of students, researchers, and educators
from many different universities that collaborate to develop
useful computer-based tools and exploit digital technologies
for research and education. The research is funded in part
by the NSF through a Postdoctoral Fellowship in Biological
Informatics to Hodges; Ted Garland of UC Riverside is the
primary sponsor and Tim Rowe of UT Austin is a co-sponsor.
Funding for CT sections is provided through NSF support to
Rowe. Additional support for Reyes is provided through TACC.
Mike Gannis is a senior science writer
at the San Diego Supercomputer Center.
|