Robot Dives Deep for Sinkhole Slime
In May, researchers successfully conducted the third and final field test of the autonomous underwater robot, DEPTHX. Their objective was to explore Cenote Zacatón, the world’s deepest water-filled sinkhole.
Zacatón lies near one end of a chain of sinkholes stretching nearly half a mile across Rancho La Azufroza (Sulfur Ranch), located in northeastern México, roughly 20 miles from the Gulf Coast. Even without the sinkholes, the biology of the region would make a fascinating subject of study. The landscape is dotted with a muddle of tropical deciduous trees and bromeliads growing side-by-side with agaves and cacti typical of desert climates. Each day, as dawn approaches, a flock of green parrots takes wing, shrieking and squawking as they circle the rim of Zacatón. Later in the day, the air grows thick with butterflies, more than a dozen different species, some with wingspans exceeding six inches. It is a languid, sun-drenched setting.
But what lies below ground, in the dark waters of Zacatón, where only microbial life can survive, is what has piqued the interest of scientists and engineers from Stone Aerospace, the University of Texas at Austin, Carnegie Mellon University’s Robotics Institute, the Colorado School of Mines, and other institutions, who make up the DEPTHX team.
The DEPTHX project was funded by NASA’s ASTEP (Astrobiology Science and Technology for Exploring Planets) program. ASTEP projects typically involve both technology and science components. The robot incorporates a number of innovative technologies. It is the first underwater vehicle that can be placed in an enclosed water-filled space and, without any previous knowledge, safely navigate its way around, build a three-dimensional map of its environment, and collect samples of scientific interest — all without human intervention.
Mapping and sampling Zacatón was DEPTHX’s most challenging task to date. In March of this year, the craft explored Poza La Pilita, a smaller sinkhole near Zacatón.
DEPTHX enabled investigators to explore an otherwise inaccessible ecosystem that extends far below the Earth’s surface. The robot had a mechanical arm that could be extended 2 to 3 meters (6.5 to 10 feet); at its end was a spring-loaded penetrator that could sense when it came within a few inches of the cenote’s wall. Once in position, it grabbed a gob of the microbial biofilm that coats the entire interior surface of the sinkhole, and brought it back to the surface for later laboratory analysis. Positioning the 1.5-ton robot precisely — not too far from the uneven surface of the cenote wall to obtain a sample, but not so close that the penetrator slams into rock and gets bent — was challenging, particularly when the robot was doing its own navigation. But DEPTHX successfully obtained half a dozen samples of microbial Zacatón wall slime. The deepest of these came from close to the bottom of the cenote, at a depth of 272 meters (892 feet).
Finding the bottom of Zacatón was another of DEPTHX’s accomplishments. Previously, no-one had been able to determine for certain how deep the cenote was. As it turns out, the bottom is sloped, ranging from 315 meters (1033 feet) at its high end down to 320 meters (1050 feet). And it may go even deeper. At the low end of the cenote, the robot found what appeared to be a narrow tunnel that extended outward, and perhaps farther downward. Because the research team was pressed for time, however, and because they wanted to make sure they could safely get the craft back to the surface, they told DEPTHX to come home without exploring the tunnel.
John Spear, the lead microbiologist on the DEPTHX team, speculates that this deep channel is connected to an underground system of thermally heated water. About one million years ago, geologists believe, the Zacatón region was a site of intense hydrothermal activity, not unlike Mammoth Hot Springs in present-day Yellowstone National Park. Although thermal activity around Zacatón has calmed down considerably since those fiery days, there are clear signs that something is still stirring underground: a pervasive scent of sulfur hovers around the cenote, and Zacatón’s water is a constant 30 degrees C (86 degrees F). In fact, says Spears, one of the surprising discoveries made by DEPTHX is that the temperature in Zacatón is constant all the way through its thousand-foot water column. He expected to find temperature variation with depth, a more common scenario.
“If we stuck DepthX in a place like Yellowstone Lake, for example, you would see gradients of change in temperature. It would probably be warm on the surface, cold in the middle, and then down at the thermal vents warm again,” Spear said. But something is keeping Zacatón unusually well-mixed. I asked Spear what caused the mixing. “Don’t know,” he replied, but in a later email he added that “there is a large amount of geothermally heated water flowing through the system.”
Whatever its cause, the unexpected uniformity put a kink in the research team’s sample-collection plans. They had hoped to use gradients in the water’s temperature, in its salinity and in its level of dissolved oxygen to guide DEPTHX toward the best sampling locations. Places where such changes occur are interesting because they are often accompanied by an ecological change. Different types of organisms thrive in cooler water than in warmer water, for example.
Often, there is also a visual indicator of such changes. In Yellowstone, where Spear has worked extensively, “I could walk up to a hot spring and say, ‘I want a sample right there,’ mainly because of my visual interpretation of it, what I see,” he says. Green, for example, indicates the presence of photosynthetic organisms, which can survive only in relatively cool water. Yellow-, orange- and red-hued organisms dominate in hotter waters.
“You’d like the robot to do the same thing, use a visual cue to understand a place,” Spear says. He was hoping to be able to use “color changes on the walls of the cenote,” which “might correlate with chemistry,” to guide the robot toward good places to collect samples. But, apart from a shallow oxygenated zone near Zacatón’s surface, the microbial life that clung to the cenote walls was visually uniform, from top to bottom.
Nevertheless, Spear expects the DNA analysis that his lab will perform over the next few months on the Zacatón samples to yield valuable results. Preliminary analysis of samples collected a couple of years ago by a diver, at a depth of 85 meters (280 feet), turned up “six new groups of bacteria.” And by “six new groups” Spear emphasized, he didn’t mean six new species. “The bacteria domain [one of the three main branches on the tree of life] has about 100 different divisions or phyla in it. So we found six new ones, from here,” he explained. “That’s kind of equivalent to walking out your door in the morning and finding plants for the first time.” To be fair, Spear points out that “you can often find new groups” even in places as pedestrian as common garden soil. “It could even be something that’s living between your teeth.” Still, six is a pretty good haul for one sinkhole. “And we think we can find more,” he adds.
The samples that Spear’s lab will analyze were collected while DEPTHX was under human control. Because the robot’s time in Zacatón was limited, DEPTHX engineers had to choose between pursuing science goals or technology goals, and they decided to tell the robot where to collect its samples, rather than to let the craft’s onboard computers make autonomous choices.
But the robot’s software-engineering team, which hails from Carnegie Mellon’s Robotics Institute, also got a chance to put the robot’s sophisticated technology to the test. In Poza Verde (Spanish for “Green Pool”), a wider but shallower cenote near Zacatón, DEPTHX was tested in “exploration mode." In this mode, the robot is not given any instructions about where to go or what to do. It’s dropped into the water and simply told to go find interesting stuff. It is responsible both for navigating its way around and for deciding what is interesting. The engineering team judged this test a success.
That is promising, because exploration mode will have to work well for the next phase of the robot’s life. Later this year, DEPTHX will morph into ENDURANCE, the same robot but with a slightly different configuration, and will transition from exploring balmy semitropical waters to swimming about in chilly ice-covered lakes.
The first such cold-water test will take place in February 2008, in the Midwest. That will be a trail run for an even more challenging mission late in 2008: autonomous exploration of the waters of Antarctica’s Lake Bonney, an ice-covered lake about 3 km (1.8 miles) long and 1.5 km (0.9 miles) wide.
To date, very little is known about Lake Bonney. Peter Doran, a University of Illinois at Chicago associate professor and the principal investigator for the ENDURANCE project, and his colleagues have been studying the lake for several years, measuring its temperature, salinity and a handful of other parameters. But those measurements have all been made “in the center of the lake,” Doran says. “We go back to the same spot every year.” ENDURANCE, he says, will enable researchers for the first time to develop a portrait of the lake – its temperature, its chemistry, and its microbial ecology – in three dimensions.