Yellowstone Science a Quarterly Publication Devoted to the Natural and Cultural Resources

Home , Dale (crater), Eyes Wide Open (exhibit), Hydrothermal explosion, Lava Creek Tuff

Yellowstone Science A quarterly publication devoted to the natural and cultural resources

110˚35'W 110˚30'W 110˚25'W 110˚20'W 110˚15'W Yellowstone Pelican River Valley Indian Pond Mary Bay Storm Point Turbid Lake Bridge Mary Bay Bay crater outlet inflated spire graben plain Elliott's Steamboat field crater Point Sedge Gull Pelican Bay Stevenson Roost Point Island Lake Butte the 44˚30'N 44˚30'N fissures northern Sand basin Point Rock Point

Elk Point Pumice Point Dot Island west West Thumb Breeze central Park basin Point basin east Point Evil Frank central Duck Island Lake Twin Wolf basin Point 44˚25'N 44˚25'N crater

Delusion Lake Eagle Bay Biscuit Plover Basin Point Flat Mountain Arm

The Promontory Southeast South Arm Arm Alder

Lake 44˚20'N 44˚20'N

Yellowstone River

110˚35'W 110˚30'W 110˚25'W 110˚20'W 110˚15'W

The YS Interview: Lisa Morgan Mapping Yellowstone Lake Predator and Prey at Fishing Bridge

Volume 11 Number 2 “In War or in Peace”

Now, as these words are written, with prospects of a third world war looming up, with the need all the greater for a haven from the tensions of modern life, for an environment of quiet and peace and serenity, a book like Tilden's leads people's thoughts into channels upon which proper mental balance and perhaps even national sanity may depend. So much the more important, therefore, to cherish these crown jewels among the lands of the nation, to keep them unsullied and intact, to conserve them, not for commercial use of their resources but because of their value in ministering to the human mind and spirit. In war or in peace the national parks have their proper and proportionate place in the life of America. These lands are less than one percent of our area. Surely we are not so poor that we need to destroy them, or so rich that we can afford to lose them.

As Director of the National Park Service from 1940–1951, Newton B. Drury spent a great deal of his tenure fending off a constant flow of demands that the national parks be plundered for the resources they could contribute to wartime and post-war necessity. In that last year of his directorship, in the midst of the Korean War and a growing with the Cold War, he penned these words as an introduction to Freeman Tilden’s The National Parks: What They Mean to You and Me. Yellowstone Science A quarterly publication devoted to the natural and cultural resources Volume 11 Number 2 Spring 2003

Contents

Science with Eyes Wide Open 2 YS talks with USGS geologist Lisa Morgan about science, discovery, and the joys of life in the caldera The Floor of Yellowstone Lake is Anything but Quiet! 14 New discoveries from high-resolution sonar imaging, seismic reflection profiling, and submersible studies show there’s a lot more going on down there than we may have thought.

PAUL SCHULLERY PAUL by Lisa A. Morgan, Pat Shanks, Dave Lovalvo, Kenneth Pierce, Gregory Lee, Michael Webring, William Stephenson, Samuel Johnson, Carol Finn, Boris Schulze, and Stephen Harlan

Editor 7th Biennial Scientific Conference Announcement 31 Roger J. Anderson [email protected] Yellowstone Nature Notes: Assistant Editor and Design Predator and Prey at Fishing Bridge 32 Alice K. Wondrak In prose and on film, Paul Schullery captures a little-seen phenomenon: Assistant Editors the aesthetics of survival in the world of a trout. Tami Blackford by Paul Schullery Virginia Warner Technical Assistance News and Notes 40 and Printing Gray wolf downlisted • Bison operations commence outside North Entrance • Artcraft, Inc. Spring bear emergence reminder • Winter use FSEIS released • YCR 10th Bozeman, Montana anniversary Special Color Issue! Around the Park 43 Springtime in Yellowstone.

Cover: New high-resolution bathymetric Yellowstone Science is published quarterly. Submissions are welcome from all investigators con- relief map of Yellowstone Lake, acquired by ducting formal research in the Yellowstone area. To submit proposals for articles, to subscribe to multibeam sonar imaging and seismic map- Yellowstone Science, or to send a letter to the editor, please write to the following address: Editor, Yellowstone Science, P.O. Box 168, Yellowstone National Park, WY 82190. ping, surrounded by colored geologic map of You may also email: [email protected]. the area around Yellowstone Lake. Support for Yellowstone Science is provided by the Yellowstone Association, a Courtesy USGS. non-profit educational organization dedicated to serving the park and its visitors. For more information about the association, including membership, or to donate to Left: 1883 woodblock engraving of Yellow- the production of Yellowstone Science, write to: Yellowstone Association, P.O. Box stone Lake, probably produced for publica- 117, Yellowstone National Park, WY 82190. tion purposes and handcolored at a later The opinions expressed in Yellowstone Science are the authors’ and may not reflect time. either National Park Service policy or the views of the Yellowstone Center for Resources. Above: A trout rises to the surface of the Yel- Copyright © 2003, Yellowstone Association for Natural Science, History & Education. lowstone River. Yellowstone Science is printed on recycled paper with a linseed oil-based ink. Science with ‘Eyes Wide Open’ An interview with geologist Lisa Morgan ALICE WONDRAK

Dave Lovalvo, Lisa Morgan, and Pat Shanks launch a remotely-operated vehicle (ROV) into Yellowstone Lake. The ROV aids in ground-truthing recent bathymetric and aeromagnetic mapping of the lake floor conducted by the USGS.

A research geologist with the U.S. Geological Survey, Dr. Lisa Morgan has devoted 23 years to studying the geology and geophysics of volcanic terrains. Since 1999, she has been working in Yellowstone, mapping and interpreting the floor of Yellowstone Lake and its associated potential geologic hazards—an undertaking that was completed last summer. We felt that an achievement of this magnitude warranted extensive coverage in Yellowstone Science, and are excited to publish its results. In 2002, YS editor Roger Anderson had the opportunity to discuss this project and other issues with Lisa at her Boulder, Colorado, home. We are pleased to include this interview, which provides interesting insights into the mapping process as well as into her personal approach to the science of geology.

YS (Yellowstone Science): Lisa, when ory, which I always thought was pretty to draw something is going to be very dif- we first met, you described your back- interesting. My hope was that the course ferent than if you just took a photograph of ground as a little bit different than that of would enable me to have a better under- it, and you’re going to consider the rela- most geologists, in that you started out as standing of the color spectrum and how tionships somewhat differently when a fine arts major. Could you tell us a little light works. So I took that class and had to you’re drawing something than if you’re bit more about that, and how it affects how take prerequisites in physical and histori- just going to document it with a photo- you go about looking at your work today? cal geology and before I knew it, I was graph. So I think fine arts brings this abil- kind of hooked into geology. And I love ity to see or look. LM (Lisa Morgan): I did start out as a geology. One of the things I think fine arts I guess I would describe myself as a fine arts major. Along the way, I took a brings to geology is the ability or interest field geologist who studies the geology mineralogy and optical crystallography to look in detail at things, and to see things and geophysical characteristics of volcanic course in my pursuit of fine arts because I that you might not normally look for. Like terrains, and I think having a fine arts per- knew we’d be studying color and light the- when you’re drawing, how you’re going spective gives me another set of tools with

2 Yellowstone Science which to understand the Earth. My tant that we get out our products that we A great example from our 2001 field- approach is to do science “with eyes wide promised we’ll get out, but I think it’s also work is, we discovered charcoal and tree open.” And that’s where I see the connec- important, as natural scientists, to make molds in the Lava Creek Tuff. We have tion with art, because when you take on a sure we’re not missing something. So actual pieces of charcoal present in some canvas, you start with a specific drawing or while I have models in mind, I get really of the tree molds, which was surprising painting in mind but as you work on it, the frustrated when I’m in the field and some- because these pyroclastic flows are typi- painting begins to shape itself. You don’t body tells me, “well, this model tells me it cally erupted from very large calderas at start a painting saying, “I know exactly can’t be this.” I don’t care what your model pretty incredible speeds, and emplaced at what colors I’m going to use. I know tells you. Just look at the relationship. For- very high temperatures, probably on the exactly what I’m going to draw or paint get any model, and just look at that, study order of 800-850ºC. here.” You don’t know exactly what the what you’re seeing in that relationship, and final results will be. And I think you then see how that compares with your YS: And it’s unusual, because in that shouldn’t, either, with science. When you model. But sometimes, people go in there heat you would expect everything would start out on a proposal, certainly you have saying “well, it’s got to be something other be consumed. ideas of what you want to look at, how to than this.” proceed, certain goals, objectives; but you LM: That’s correct. At this location, need to make sure that we were close to an area you keep your options COURTESY LISA MORGAN interpreted as an eruptive open enough so that you vent for the Yellowstone don’t miss anything that caldera, and there’s a lot might be necessary in of evidence to suggest your final interpretation that there may have been of what you’ve seen, or some water involved in what you’ve recorded. So this particular part of the I think a lot of our work in emplacement of the Yellowstone Lake has deposit, which probably been a perfect example of decreased the tempera- going into an area using ture. techniques we really had- Tree molds are com- n’t used before, using a mon in basaltic lava broad group of different flows, such as those in individuals from different Hawaii and at Craters of disciplines all bringing the Moon, Idaho, but different skills to the same very little study has been table, allowing us to iden- done on the preservation tify things we didn’t of tree molds in rhyolite know were there. And Lisa points out a "twig mold" in the 0.64-Ma Lava Creek Tuff. A modern twig pyroclastic flow deposits, allowing us to come to has been placed above "twig mold" for comparison purposes. like the Lava Creek Tuff. conclusions that we did Tree molds and charcoal not know we were going to come to when YS: Is your approach unusual, do you are somewhat rare occurrences in these we started the original study. think, among scientists? types of environments. To find charcoal in this deposit, preserved charcoal, is an YS: Right. So if you had a pre-set par- LM: I don’t know; probably not. I interesting discovery, and contributes to adigm, and you found something else and think there are a lot who don’t come to the what we know about the climate 640,000 it didn’t fit into that, you don’t go with the table already working within a prescribed years ago when the Yellowstone caldera preconceived notion of what you’re going model. But I’m sure there’d be people who erupted. to find. That’s what you mean, “eyes wide would disagree with me on that, too. In open.” my experience, when I meet with people in YS: Where in the park did you find the field and we’re looking at different this? LM: Exactly. And a lot of times, you’ll things, some people already have kind of find, in science, and maybe in other things, an idea what it has to be. But I think it’s LM: It’s in the vicinity of Fern Lake, too, people want preferred outcomes. They okay to say “well, I don’t know what it is.” close to the topographic edge of the already know where they’re going to get I think nature is a continual puzzle for caldera. to, and they have their product that they’re most of us. And there are a lot of things we supposed to produce, and that’s what still don’t know or understand, which YS: Now, if I was out with you last they’re going to do. And I think it’s impor- keeps us going. summer on the trail, walking through that

Spring 2003 3 part of Yellowstone, what would you see a tree mold.” And I just happened to pick that everyone brings a somewhat different that I wouldn’t necessarily see that would this piece up, and there was all this char- perspective and set of experiences. It’s make you want to stop and take a closer coal, and also pine needles and impres- much better than just having your own look? What are you seeing in the land- sions. ideas and self to bounce concepts off. It’s scape that makes you want to investigate And Pat said, “well maybe that got in always good to have somebody who will this particular spot, and then how do you, there through some kind of later fluvial challenge one’s thinking. It keeps you in all of Yellowstone, get to that place and action, or maybe there was just natural honest and keeps you thinking. make that kind of find? plating for Later, when some reason, I told Ken LM: Here’s exactly what happened. It and you had I guess one of the things our project Pierce [of the had been raining on and off that day. Pat some flood- has exemplified is that not any one USGS] about Shanks and I were in one work group and ing, and you the tree molds Steve Harlan, Lydia Sanz, and Beth Erland got the pine person has all the answers. and pine needle were in another work group, and Pat and I needles in impressions were discussing where to go next. It was there,” and so then I went to another, and I found in the Lava Creek Tuff, his reaction starting to rain again, and we had to cross said, “what appears as the characteristic was, “Oh my gosh! That is so cool,” the creek. I had taken my backpack off and feature of this particular site and deposit is because in the field of paleoclimatology, a was putting my raingear on. I’m constant- the unusual nature of the platyness of the debate exists about whether the Yellow- ly, just always looking at everything. And unit and that’s a reflection of its content of stone caldera erupted during a glacial or as I was tying my shoelaces, I put my eyes organic matter.” I said to Pat, “I think the interglacial period. And he said, “I think on this little piece, that was just a surface platyness and the organic matter in the ign- you’ve got key evidence now for showing piece, probably no more than a couple cen- imibrite are part of the original deposit. I’ll the Yellowstone caldera erupted during an timeters long. It had a very fine rim, or bet you a beer that when I go over to that interglacial period. It has to be, to have all coating of silica on it, and then inside of platy zone and that platy zone and that one those pine needles and trees.” So that was that was this twig impression—this tree and all of these zones will be full of char- kind of cool. mold. And it was very, very tiny. So it was coal and have impressions of pine nee- a fluke. Just like a lot of things in science. dles.” He said ok to the bet. So I went and YS: It’s amazing. Without that collab- And so I saw that little thing, but then looked at all these different places in the oration, without people looking at the it started pouring. And so we skedaddled. rock exposure, and each one was full of resource from different perspectives, you We only had the next day left, and I just charcoal and pine needle impressions. So might not have made that really critical had this feeling that I had to go back to this Pat ended up buying me a beer after our connection. site and have a closer look at the impres- 13-mile trek out of the backcountry. But sion and site in general. Anyway, we went you see, the beauty of having people work LM: That’s right. So anyway, back to back up there, and I said, “look at this. It is with you with different backgrounds, is the eyes wide open, that allowed us to see that. A lot of the discoveries on Yellow- stone Lake have happened the same way. Before we started our West Thumb sur-

PAT SHANKS PAT vey, the current thinking was that most features in the lake are from the last glacial period since it’s pretty well established that over a kilometer of ice was over the lake 20,000 years ago. That certainly had to have had a pretty profound influence in shaping the lake. But what we’ve found is that it certainly is not the only, nor was it the most important, influence on shaping the floor of the lake. People had previously mapped Steven- son, Frank, and Dot Islands as being glacial remnants. And, certainly, if you went out there today, the rocks exposed on the islands are glacial tills. But what we’ve found with our high-resolution, aeromagnetic map, (based on the magnetic survey Lisa Morgan and fellow USGS geologist Ken Pierce kick back on the Mary Bay explosion done with an airplane), in tandem with our breccia deposit. sonar and seismic surveys of the lake, was

4 Yellowstone Science new bathymetric this particular survey, the plane flew lines maps of the lake 400 meters apart on an east-west orienta- floor show that tion, in a continuous pattern over the park.

ALICE WONDRAK many of the mag- The magnetometer measures the total netic anomalies magnetic intensity of the Earth’s field, and coincide with hum- that can be broken down into two main mocky areas of high components, the magnetic remanence and relief. We interpret magnetic susceptibility. Generally, the these as rhyolitic magnetic remanence records the signature lava flows, and sug- from the earth’s magnetic field that the gest that Frank, Dot, rock acquired at the time of its formation. and Stevenson Volcanic rocks are emplaced at tem- Islands are on these peratures above the Curie temperature, large lava flows. So which refers to the temperature below the glacial tills that which a mineral of a specific composition occur on Dot, becomes magnetic. Minerals in the vol- Frank, and Steven- canic deposit acquire the magnetization of son Islands are real- the Earth’s field at the time that that rock ly just mantling was emplaced and give the rock its specif- much larger fea- ic magnetic remanence direction. The tures; rhyolite lava earth’s magnetic field changes its polarity flows underlie the over time so that volcanic rocks erupted at islands and shape different times will have different and specific magnetic rema- ALICE WONDRAK nence directions. With magnetic intensity, we also measure the magnetic susceptibility, which is basically a measurement of how susceptible that rock is Dave Lovalvo at the controls of the ROV he to an ambient magnetic designed and built. What the ROV sees is field. visible on the computer monitor seen at the far right. A second monitor (not seen) dis- Susceptibility values plays water temperature readings and a sec- can vary depending on a ond picture of the lake floor. This photo was range of conditions. In taken inside the cabin of the NPS’s Cutthroat, the case of Yellowstone, which is dedicated for research purposes. what seems to be the major variable for sus- that the majority of the underwater topog- ceptibility is how raphy, or the bathymetry of the lake, is hydrothermally-altered really due to rhyolitic lava flows that were those rocks are. When emplaced some time after the Yellowstone your rock is extremely caldera formed. Now, in retrospect, I think, the lake bottom. And if we had held on to hydrothermally altered, the magnetic min- “Oh, well, that’s so obvious,” because the old model, we may have been blind to erals in that rock are also altered, and what do you see around there? It’s all lava seeing that those flows were there. become much less magnetic. A lot of times flows. And why would you think that all of what we’re seeing is titanomagnetites a sudden, just because you have a lake, the YS: Explain to the uninitiated about going to hematite or ilmenite. Hematite is lava flows wouldn’t be in there? But it was these aeromagnetic maps you’re talking nonmagnetic and ilmenite is weakly mag- a big, major discovery in mapping West about. What’s the process, and what does netic, so the magnetic susceptibility of the Thumb. it generate? rock goes to almost nothing. Most of the We’ve found that discontinuities, or rocks we’re looking at in much of the Yel- anomalies, on the aeromagnetic map coin- LM: Basically, we attach a magne- lowstone Lake area were erupted in the cide with the mapped extent of the rhy- tometer onto a fixed-wing airplane, and last 700,000 years, after the last big rever- olitic lava flows on land, and you can fol- then this airplane flies over the topography sal in the Earth’s magnetic field. So when low those out into the lake. Our detailed at a constant elevation above the terrain. In we’re looking at the total magnetic inten-

Spring 2003 5 sity of the rocks in Yellowstone Lake, the beam and seismic sonar systems. The able to measure temperatures and collect variable that is changing most is the mag- ROV is a one-meter-by-one-and-a-half solid and fluid samples of hydrothermal netic susceptibility, which in this case meter vehicle, built and piloted by Dave vents, lake water, and sinter. Later these reflects the amount of hydrothermal alter- Lovalvo of Eastern Oceanics. It’s attached can be taken to the laboratory and be ana- ation in the rock. to the boat by a 200-meter tether, which lyzed for mineralogy, chemical and iso- In many places in Yellowstone, allows continuous observation of the lake topic composition, and microscopic struc- hydrothermal alteration is associated with floor. At its front is a pan-and-tilt video tures. thermal springs. Hot waters come up along camera, which records images from the So the ROV allows us to observe and conduits and alter the rock sample what we have and magnetic minerals ALICE WONDRAK imaged bathymetrically and through which they’re flow- seismically. And that’s been ing. The hydrothermal alter- critical. The multi-beam ation of the rock lowers the mapping of Yellowstone magnetic susceptibility. Lake presented challenges When you look at the newly seldom found elsewhere. acquired total magnetic Thermal vents so dominant intensity map of Yellow- in different parts of the lake stone, you see a map that cause frequent changes in has areas with magnetic the temperature structure of highs and areas with mag- the lake, and therefore, the netic lows. In some areas, sound velocity profile. In we can use this map as a our first year (1999), when guide for where we might we mapped the northern expect hydrothermal alter- part of the lake, we identi- ation to be present; in other fied several features, which areas, we can use the map to turned out to be artifacts in identify faults and other the data. So we collected structures. more frequent sound velocity profiles than one would YS: So this is how you in a non-thermal environ- go about looking at the park, ment. Having the ROV as and looking at the rocks, and our eyes and hands on the trying to piece together all bottom of the lake allowed of the stories that they have us to confirm the bathymet- to tell? ric images. While we’re very confi- LM: I guess one of the dent of our imaged data, the things our project really has ability to sample fluids and exemplified is that not any solids, measure tempera- one person has all the tures, and photographically answers. I think if we went document the lake floor into Yellowstone Lake just adds an incredibly valuable doing a bathymetric map, component to our lake stud- we’d have a pretty appeal- ies. In 2000, we went with ing map, but the bathymetry Remotely-operated vehicle (ROV). The large orange balls are flotation the ROV to linear features combined with the total units. Water samples are drawn through the large tube mounted on the west of Stevenson Island magnetic intensity map and left. A thermometer/camera is visible in the mid-foreground, directly that were imaged in 1999. the seismic reflection pro- above the basket used for scooping up sediment samples from the lake As a result, we have photo- files enable our producing a floor. graphic evidence that the more powerful product with features are fissures with hot a higher level of confidence. Add to this floor of the lake. On the front is also water coming up along open cracks in soft the data we collect with the submersible mounted a 35-mm camera. The video mud and precipitating iron and manganese ROV (remotely operated vehicle), and the camera is on at all times so we can really oxides on the fissure walls. The fissures data set is pretty complete. The ROV is see the floor of the lake. That aids us in our are parallel to and part of the Eagle Bay wonderful, because it allows us to ground- sampling, and the still lifes are wonderful fault zone, which is a young fault system truth what we have imaged with the multi- to see. The ROV is great also because it’s mapped south of the lake at Eagle Bay.

6 Yellowstone Science This system probably continues northward had put the boundary based on the mag short time in a jewelry store with the inten- of the fissures to the young graben north of data. And so that was so cool, and then we tion of eventually becoming a gemologist, Stevenson Island. were able to take the ROV, and we could but then I got a job that paid twice as much see the caldera margin in the lake. It looks with an oil company. I stayed with the oil YS: Does that continue with the fault like a bunch of discontinuous, bathtub- company about 10 months but left to near the Lake Hotel? shaped troughs, kind of marching through return to the University to teach labs for the central basin. The multiple data sets introductory geology classes and work as LM: That’s right. The Lake Hotel is give the same conclusion, so that one can a technician in their analytical lab. I then near the fault. So moved to Colorado again, our mapping BETTY SKIPP and got my Mas- of the Lake has ter’s degree at the enabled us to look University of Col- at its geology in orado at Boulder, more detail and in focusing on a broader context. igneous petrology And Yellowstone and volcanology. Lake isn’t an easy I then had a lake at all to figure great opportunity out, or to work on. in 1980 to work at You think of most Mt. St. Helens, and lakes as being on August 7, I wit- quiet, calm, and nessed one of its good passive smaller pyroclas- recorders of the tic-flow-producing local climate and eruptions from a geologic process- plane about a kilo- es. But Yellow- meter or two away stone Lake is any- from the vent. That thing but quiet. eruption was sig- Lisa Morgan in 1980, doing field work for her master's degree. She is sitting next to the base nificant because it YS: Which of of the 6.65 million year old Blacktail Creek Tuff, the oldest caldera-forming ignimbrite from created quite a those technolo- the Heise volcanic field. The Heise volcanic field (4-7 Ma), on the eastern Snake River Plain, good eruptive gies—the aero- is similar in origin to the Yellowstone Plateau volcanic field, and immediately preceded its cloud, and in that magnetic survey, formation in space and time along the volcanic track of the Yellowstone hot spot. cloud you could the ROV, the see part of it col- bathymetry—played a role in determining say with much better confidence that this lapsing and forming pyroclastic flows on the caldera boundary under the water? is definitely where the caldera margin is. the flanks of the volcano. While the pyroclastic flow was moving, one could see LM: I would say the two most impor- YS: I’m going to just back up for a few how this flow concentrated in areas of tant ones were probably the recently minutes, and get us back to how we began lower topography, such as valleys coming acquired aeromagnetic data and the the discussion, from the fine arts to geolo- off of St. Helens. I could see fine ash being bathymetry. Much discussion has been had gy, to this philosophy of looking at your blown out of the front of the deposit. And about where to draw the caldera boundary work with your eyes wide open, and ask I just decided then I wanted to focus on through the lake. Using both of these data you to elaborate a little more about your how pyroclastic flows are emplaced. Being sets, we trace the topographic margin of background. Once you found geology, tell at St. Helens gave me a great opportunity the Yellowstone caldera right through us a bit about your schooling, where you to see what volcanologists do. And so in Frank Island. went, your degrees… the following year I decided to go for my Ph.D., and study with George P. L. Walk- YS: Weren’t you going back and forth? LM: I went to the University of Mis- er, at the University of Hawaii, whose pri- You were describing once about how you souri at Kansas City, and at that time it mary focus at the time was ash deposits, were out on the boat, and you were taking was just a small undergraduate geology their facies, and emplacement processes. measurements on what you thought was department, and I had great mentoring and inside the caldera, outside the caldera. opportunities there. That was key. I got a YS: When did you first work in Yel- job in the department, starting probably in lowstone on geology? LM: Yeah. Once we got the bathyme- my junior year. I was a lab technician try, it coincided perfectly with where we there. After I graduated, I worked for a LM: I started coming to Yellowstone

Spring 2003 7 probably 1979–1980, because I was work- molten, very hot material underneath Yel- LM: In 1977, when I moved to Col- ing on the Snake River Plain and there lowstone, and, in general, this mass of hot orado. were many similarities between the Qua- material causes the general Yellowstone ternary rhyolites in Yellowstone and the area to be topographically higher than the YS: What was your first job with slightly older rhyolites on the Snake River surrounding areas. Over time, if the Yel- them? Plain. I was working on my Master’s the- lowstone caldera is similar to earlier Qua- sis; its focus was a stratigraphic study of a ternary calderas in the Yellowstone Plateau LM: It was great. I made a Denver thick section of pyroclastic flow deposits volcanic field, which we have every reason dump map. My job was basically compil- exposed on the northern margin of the to believe is true, basaltic lavas will erupt, ing a map showing where all the landfills in the greater Denver area were. And that map transformed a lot of how I live my life today, and how I look at what our respon- sibilities are as citizens on Earth. The

ALICE WONDRAK USGS had been asked to do this because there had been a series of accidents associated with former landfills. Some accidents were due to spontaneous combus- tion of methane that caused some fires, some explosions. I think a couple of people were either seriously burnt or killed. Also, housing developments constructed on top of these landfills were developing cracked foundations and walls due to differential subsidence in the landfills. It was imperative that a comprehensive look be taken at where these landfills were located, so that city and county planners could make more informed choices of where to allow or deny development. I was blown Ground-truthing with the ROV means visiting several sites each day, requiring that the crew drop and haul anchor numerous times. Here, Lisa displays some hydrothermally-altered clay away by how many dumps there were, and that came up with the anchor. what went into these landfills. So much of it could be recycled, reused, and not put in there in the first place. And since then, eastern Snake River Plain. As we know eventually fill the caldera floor, and con- I’d say starting in like the mid-80s, our now, but didn’t know then, the Snake River ceal the Yellowstone caldera. What is now family has probably put out no more than Plain is a whole bunch of old Yellowstone- molten magma will eventually crystallize maybe three to four bags of garbage in a like calderas and volcanic fields. I was first and become denser, and thus less buoyant. year. formally assigned to Yellowstone in 1995, The overall topographic elevation will sub- but, over the previous 15 years, I used the side from today’s current elevation. With YS: Really. more complete exposures and caldera- continued southwest movement of the related features present in Yellowstone as North American plate over the thermal dis- LM: (Smiling) Yeah. We don’t sub- a way to better understand what I was turbance that causes Yellowstone today, an scribe to the landfill too much; they should looking at on the Snake River Plain, where area northeast of the present-day location be kept to a minimum. There’s a berm out exposures of rhyolites are mostly limited of the Yellowstone Plateau will become in our yard where we put all the inert to the margins of the Plain. Today, most elevated and rise above Yellowstone. In building material that would have gone to Yellowstone-like features in the Snake fact, we can already witness this. This the landfill, but that we took care of here. River Plain are covered by Quaternary or process of uplift followed by volcanism Right now I’m on the Boulder County very recent, young basalts. Eventually, like has been occurring for the past 16 million Recycling and Composting Authority, and the eastern Snake River Plain, Yellowstone years along the Snake River Plain starting our goal as a county is to divert, by 2005, will be much lower in elevation than it is in southwest Idaho. So today, Yellowstone our solid waste levels from 1994 by 50%. now and also will be covered by basalts. is anywhere from 1 to 2 kilometers above And I think we’re going to achieve that the Snake River Plain, depending on where goal. In the City of Boulder, our single- YS: Explain to me how in the future one takes measurements. family residential diversion rate is at 49%, Yellowstone will be much lower. so we have almost met our 2005 goal sev- YS: When did you begin to work with eral years ahead of schedule. However, in LM: Currently scientists can image the USGS? the arenas of commercial and industrial

8 Yellowstone Science waste and for multi-family units, our diver- stone to work on the Lava Creek Tuff, would have a very hummocky terrain. sion rates are only about 15-20% from which erupted from the Yellowstone You’d have a terrain very similar to what 1994 levels, so these are areas where we caldera, to better understand its formation. you see in the Central Plateau now, where still need to focus and significantly But I also came to do the ground-truth for there are a lot of very steep-sided, hum- increase our diversion rates. And we’re the aeromagnetic survey we flew in 1996. mocky terrain dominated by rhyolitic lava pushing the stakes up higher and trying to With Steve Harlan, I’ve collected oriented flows. And these lava flows would have a get to 80% diversion from 1994 levels. It’s core-samples from most of the Quaternary cap of glacial and lacustrine sediments. an informal goal for the City of Boulder. and Tertiary volcanic rocks in the park for Intermixed with this hummocky terrain Last year, the city council passed a new our magnetic studies. would be a whole series of hydrothermal ordinance, referred to as the “pay-as-you- As far as the lake, if you think of all of vent fields throughout the lake. The throw” ordinance, that really forced indi- the geologic maps in Yellowstone Nation- hydrothermal vents are associated with the viduals to pay the true cost of their trash. al Park, the one place that didn’t have a lava flows, generally near their edges. And This has resulted in major behavior geologic map was the Lake. What really so, one of the largest thermal fields in Yel- changes and a significant increase in our got me into Yellowstone Lake was my lowstone National Park is on the floor of level of recycling and reuse. People can interest in the hydrothermal explosion the lake. It’s pretty magical exploring these do it, but you have to have the infrastruc- deposits. We were already engaged in areas. On top of this very hot area, we’ve ture in place, like curbside pickup and detailed studies of the deposits from Mary seen a lot of fissures, which are linear mixed paper and commingled containers. Bay, Indian Pond, and Turbid Lake on cracks in the lake bottom. I can’t think of So that’s a long story, but that was my first land, and I was very interested in trying to an area on land in Yellowstone where you job with the USGS. understand the eruption of Mary Bay. In have big open fissures like these. Maybe in our 1999 survey, one of our big discover- some of the thermal fields, but some of the YS: What’s your current job? ies was what we are now calling Elliott’s lake-bottom fissures that were discovered crater, which is an 800-meter wide in 1999 and 2001, in the northern and cen- LM: Now I work at Yellowstone. This hydrothermal explosion crater complex on tral lake, extend for several kilometers. year I’m assigned to Yellowstone 100% of the floor of the lake in the northern basin. Another feature one would see are the very my time. We’ve finished the map of Yel- ALICE WONDRAK lowstone Lake, and are working on various aspects of the postglacial hydrothermal explosion craters and deposits that are probably the most immediate serious hazard in the park. The last very large hydrothermal explosion event that we know of was 3,000 years ago at Indian Pond. Of course, in recent years smaller hydrothermal explosion events have occurred in the Norris basin, Biscuit Basin (1915), West Thumb, and Potts thermal basins and elsewhere in the park. So they’re very much a current feature of activity that Yellowstone National Park has to deal with. I’ve also been working on the physical characteristics of the Lava Creek tuff and its emplacement, and how it relates to the formation of the Yellowstone caldera. And then there’s the mapping of Yellowstone Lake that’s basically con- The sun rises on Yellowstone Lake. The lake’s unpredictable summer weather makes it best sumed me for the past four years. to get an early start.

YS: Was your work in Yellowstone on YS: The work you and others have large lake-bottom explosion craters, simi- the caldera what ultimately brought you to done in recent years has really kind of rev- lar to Turbid Lake and Indian Pond on do the extensive work on Yellowstone olutionized the way we look at the lake. If land. Lake? What intrigued you about Yellow- we could look at the bottom of Yellow- stone Lake that has led you to do so much stone Lake, from the mapping you’ve YS: Duck Lake, too? work there? done, what would it look like? LM: Yes, Duck Lake is a large explo- LM: Yes, originally I came to Yellow- LM: If you took all the water out, you sion crater immediately west of West

Spring 2003 9 Thumb basin. You may have noticed a the west side, and about six meters of dis- responsible for the occurrence of hundreds steep slope west of the West Thumb placement on the east side. And we found of hot springs on the lake floor. The Geyser Basin. This is an apron of debris a lot more vents in West Thumb basin than hydrothermal fluids are very acidic and that was ejected during the hydrothermal we had previously thought. change the composition of the rocks explosion of Duck Lake. A morphological around them. And so in the lake, most of difference between the large explosion YS: Where else did you find them? the rock composition originally was rhyo- craters on land and those on the floor of the lite, which is mostly quartz, silica, lake is the radial apron of debris around the LM: They’re in the south-central part feldspar, and plagioclase. Feldspars and craters we see on land. In the lake, a well- of the West Thumb basin as well as in the plagioclase are altered easily by this acidic defined rim around the central crater is northern part of the basin, along the edges fluid and are changed into clays. And then absent. Most of this difference may have to of rhyolitic lava flows. these hydrothermal minerals precipitate do with the medium in which the eruption in this system forming a kind of imperme- occurred. YS: How about Mary Bay? able seal. At some point all the vents and fissures that were conduits for these fluids YS: And the vents, and the spires… LM: Mary Bay is a huge crater com- seal up. The acidic hydrothermal fluids ALICE WONDRAK

“We got bubbles!” Although GPS units are the crew’s primary mode of navigation, patches of bubbles rising to the lake’s surface can act as hydrothermal landmarks as those aboard the Cutthroat search for the exact spot to launch the ROV.

LM: And then you’d have spires, or plex. It’s a whole series of smaller craters and gases continue to do their work, which conical features, that are anywhere from inside a much larger main crater. One of is to alter the substrata, and at some time, one meter all the way up to about 8 meters the things that we need to get a better han- either these things explode and there’s a high, over in Bridge Bay. We think Monu- dle on is that not all of these craters are catastrophic failure of that sealant, or ment Geyser Basin, near the northwestern produced by explosions. We think some there’s collapse of all this material under- edge of the Yellowstone caldera, may be of these craters may also be produced by neath. Now, I don’t think we understand analogous in its origin to Bridge Bay. And dissolution collapse. As you know, the how we distinguish these two at this point, then just north of Stevenson Island, you’d lake has areas of very high heat flow, or how we can forecast what’s going to also see a large young graben, which is a which came out of research by previous happen. But I certainly hope some of our down-dropped block with bounding faults. workers such as Paul Morgan, Bob Smith, seismic profiles give us more insight into About two meters of displacement is on and Dave Blackwell. The high heat flow is our ability to look at the structural integri-

10 Yellowstone Science ty of the rocks underneath the lake. Or the ing a couple large detachment blocks, have Grand Teton National Parks. After melt- lake sediments. come off the eastern, and to a lesser extent, ing, voids left by the ice were later partial- western shores of Yellowstone Lake. ly filled with slumped sediment leaving YS: Of all the findings you’ve made These are kind of hummocky, but not as large, tens of meters wide, irregularly- on the lake, what surprised you the most, pronounced as the lava flows. The causes shaped depressions. would you say? of these landslides and their effects on Yel- lowstone Lake are an important topic for YS: Earlier, you talked a little bit about LM: I don’t know, I mean, the whole further study. the interplay between geology and bio- thing has been like a discovery a day. And logy. Could you elaborate? so it’s been really exciting, and has opened YS: How much of the lake bottom has new ways to examine multiple active been surveyed? LM: As you know, lake trout have processes, and has been quite fun along been discovered in Yellowstone Lake, and the way. You never are quite sure the native cutthroat trout is prey what you’re going to find. I COURTESY LISA MORGAN to the lake trout. Pat Shanks has guess I’d say the biggest surprise been working on the geochem- was either the fact that the rhy- istry of the sublacustrine olitic lava flows played such an hydrothermal fluids, looking at important role in shaping the toxic elements that we know floor of the lake and controlling exist in other hydrothermal sys- the location of the hydrothermal tems, including mercury, anti- vents, or that the caldera margin mony, and thallium. Crus- showed up so clearly in the taceans are a primary food bathymetry and coincided with source for cutthroat trout, so the areas of magnetic lows. That question arose, what kind of was really cool. transmission is there from the To see the caldera margin is vents to the lowest life forms really fascinating. And the that we could identify, on up hydrothermal craters are just through the food chain to the phenomenal. When you see cutthroat trout and up to the lake these large structures, you know trout? So he started looking at a very complex process is mercury content of fish muscle, involved, because not only do vital organs, and skin. And he you have 800- to 2000-meter found a higher than normal con- diameter structures, but through- centration in both the lake and out the floor of these structures cutthroat trout. hydrothermal processes are The park is interested in active and one can see and sam- identifying areas in the lake ple active hydrothermal vents. where lake trout spawn. Lake And so you think “well, where trout are anadromous, meaning can you go on land where you they stay within the lake their could see something like that?” Even graduate students need a break every now and then. entire lives. Cutthroat are poto- and so in the summer of 2001, modromous meaning they we started mapping Indian Pond spawn in the streams that feed and Duck Lake, and right now we only LM: We finished surveying the South into the lake during the early summer and have the seismic reflection profiles. But and Southeast Arms in 2002, so the bathy- later they come back and live in the lake. that should give us a lot of indication of metric mapping of the lake is completed! When they are spawning in the streams, what’s going on in those lakes. I think it’s About 75–80% of the lake is within the they become potential food sources for important for the park, from a public safe- Yellowstone caldera. Outside the caldera many species, some threatened or endan- ty perspective, to understand activity are the South and Southeast Arms, which gered such as grizzly bears, bald eagles, occurring in the hydrothermal explosion are fault-bounded valleys whose shape has otters, and osprey. If the cutthroat disap- crater lakes as well, because like geology, been enhanced significantly by glacial peared, the lake trout, which never leave nothing’s static. We know from our recon- processes. Much of the floor of the South- the lake, would not take their place in the naissance seismic surveys in Duck Lake east Arm is characterized with a hum- ecosystem. It’s a major resource issue for and Indian Pond that active hydrothermal mocky bathymetry with many depressions the park and understanding where lake vents are on their floors. reflective of kettle and glacial meltwater trout spawn is key to controlling their Also, large landslide deposits, includ- terrain seen elsewhere in Yellowstone and numbers and to the ultimate survival of the

Spring 2003 11 cutthroat trout. diatoms and silicified bacteria. And that LM: Well, for starters, I think the lake, Our understanding when we started was a big surprise. We had supposed that the park, and science would be well served this study is that lake trout like to spawn in the hot silica-enriched waters hitting the by doing a series of cores on selected sites gravelly areas. So we thought if we could cold lake water interface would precipi- in the lake. That would shed a lot of infor- identify gravelly areas in the lake with tate amorphous silica without biologic mation about the timing of different high-resolution bathymetry and seismic involvement. When we went and looked at events: timing of seismic events, of profiles, we could lead the biologists to the the spires under the SEM, sure enough, hydrothermal explosion events, of land- spawning areas for the lake trout. As it our compositions were for pure silica but slides. Cores collected in specific areas turns out, we’re finding would give information that the lake trout hang about what triggers the out in other areas in addi- landslides. Were they trig- tion to the deep gravelly gered seismically, or were areas. The cutthroat trout they triggered from the like to hang out in warm, hydrothermal explosion thermal shallow areas, events? Or from some- which have been called thing else? Potentially, “cutthroat jacuzzis.” The data from selected cores Park Service has found could tell us something lake trout coming into about evolution of the dif- some of these jacuzzi vent ferent hydrothermal sys- areas to prey on the cut- tems. Not just the throat trout. hydrothermal vents, but So biology has a also the large hydrother- major role in the Yellow- mal explosion complexes. stone Lake studies. For I would also like to know one, identifying what more about the climate of effects toxic metals pres- Yellowstone in the last ent in hydrothermal fluids 12,000 years. Certainly, have on lake water chem- coring into the lake would istry and how they affect give us a clearer idea of its ecosystem is impor- what the climate was like tant. Secondly, how those during these different effects are rippled up into events, and what kind of the larger animals outside influences there might the lake is equally impor- have been. tant. Chuck Schwartz, We need to have a bet- Charles Robbins, and the ter understanding of the Interagency Grizzly Bear issue between large-scale Study Team working with collapse of hydrothermal- Bob Rye and Pat Shanks ly altered features versus recently have analyzed large-scale hydrothermal hair from four bears in the Lisa Morgan’s field notebook. explosions and associated park. Two of those bears hazards. We need to come from areas very improve our understand- close to the lake, two are from farther the material was primarily silicified bacte- ing of doming activity on the lake floor. away. The two bears close to the lake have ria with diatoms. And so there’s something Do these doming events always end up in elevated levels of mercury in their hair happening on the floor of the lake that is an explosion, or do they end up in a col- whereas those bears not living near the very much involved in some of the very lapse, or do some of them not do anything? lake do not. So in this example, it seems a basic life forms that operate very closely I think that’s important. Are the domes that strong relationship exists between the with development of these hydrothermal have been identified in our surveys poten- geology of the lake and grizzly bear and vents. So it’s kind of interesting to see the tial precursors to hydrothermal explo- cutthroat trout ecology. full circle come back. sions? If so, how do we monitor these fea- That’s one issue. Another is the spires tures? Also the young and active graben and how they formed. Scanning electron YS: What work remains to be done? north of Stevenson Island should be mon- microscopic (SEM) images show that the What questions still need to be asked when itored, especially given this structure’s spires are composed of a variety of you look at the lake? proximity to the Lake Hotel. Putting CO2

12 Yellowstone Science and SO2 sensors next to vents in the domes on a daily basis, it’s very much an urban the University of Utah, and the National is important. I would like to see more work population. If you took the number of vis- Park Service (Yellowstone National Park). done using LIDAR (light detection and itors that you have coming to Yellowstone A lot of work remains that will contin- ranging) outside the lake. Ken Pierce and on an annual basis and divided it by the ue to build on previous investigators’ Ray Watts’s initial work with this tech- days, you basically have the city of Boul- research and findings. We still have a far nique in association with some other peo- der, Colorado in Yellowstone every day. way to go in improving our understanding ple shows that Storm Point may actually The problem with your population is it’s of the connections between geology and be an inflated structure. biology, and how the

And so we might want to LISA MORGAN biota react to different examine that pretty close- geologic events in the ly. Much work remains park. and, to summarize, this would include mapping, YS: Finally, please coring and associated describe some of your studies, assessing the memorable moments potential hazards, and working in the park. identifying the instru- mentation needed for LM: It’s been a chal- monitoring. lenging and rewarding research experience to YS: How about in the work in Yellowstone. It’s broader context of the been so much fun to caldera outside the lake? work in Yellowstone. And it’s just been kind of LM: Well, I would a dream, like the summer hope that we get a better when we did our back- understanding of the packing trip up to Fern hydrothermal explosion Lake, it was like, “I can’t potential inside the handle any more discov- caldera. Most, if not all of eries!” (Laughing) Just the hydrothermal explo- the number of discover- sions that we’ve identi- ies we’ve been able to fied occur inside the make through the course caldera. So that needs to of our research has been be assessed. I think it’s phenomenal, so I feel also very important to very lucky to have had understand the “heavy this opportunity. It’s also breathing” aspect of the been pretty awesome to Yellowstone caldera. work in this environment And again, Ken Pierce where the sight of a griz- has shown a lot of inter- zly makes one realize esting data that looks at Yellowstone Lake. what a unique, special, the coincidence between and still wild place Yel- uplift and subsidence of lowstone is. To under- the Yellowstone caldera with relationship moving all the time. But the park pretty stand the geologic framework in which to timing of some of these hydrothermal much controls where it moves. And so it’s bears and other species inhabit allows us a explosion events. So I think that’s very important that the park have a better under- more comprehensive understanding of important. From my perspective, probably standing of where these hazards may why certain species live where they do and one of the greatest and most likely poten- occur. the challenges they face in their environ- tial hazards in the park is the potential for Clearly, assessment of other potential ments in order to survive and what we a hydrothermal explosion. In terms of hazards, such as volcanic and seismic might do to enable their survival. For sev- scale, it’s not going to affect North Amer- events, are big items and will be included eral of these species, Yellowstone is their ica, but it potentially could affect the in the ongoing hazard assessment con- last outpost, so it’s up to those of us who park’s facilities, infrastructure, and visi- ducted under the auspices of the recently work in the park to make sure that they’re tors. And when you think of the transient established Yellowstone Volcano Obser- protected. population that goes through Yellowstone vatory, a joint effort between the USGS,

Spring 2003 13 “The lake was very rough. The waves coming in were equal to waves on the sea coast. Elliott says they were able to take but three soundings, it being rough all the time. The wind once was so strong that the mast was broken off and carried away. The boat rode splendidly." Albert Peale, mineralogist, US Geological Survey Hayden survey, August 14, 1871

Evolution of mapping Yellowstone Lake, 1871-2002 The Anna.

Henry W. Elliott, 1871 Hayden survey Hague report, 1896

Kaplinski, 1991 U.S. Geological Survey National Park Service 1999-2002

14 Yellowstone Science The Floor of Yellowstone Lake is Anything but Quiet! New Discoveries in Lake Mapping

by Lisa A. Morgan, Pat Shanks, Dave Lovalvo, Kenneth Pierce, Gregory Lee, Michael Webring, William Stephenson, Samuel Johnson, Carol Finn, Boris Schulze, and Stephen Harlan

HISTORY OF MAPPING YELLOWSTONE LAKE Lake. Mapping took 24 days and included crossing the central basin. Plate 1 shows approximately 300 lead-sink soundings. the map of Yellowstone Lake as drawn by Yellowstone Lake is the largest high- Navigation was carried out using a pris- Henry Elliott of the Hayden survey. The altitude lake in North America, with an matic compass. Albert Peale, the survey’s map not only shows a detailed topograph- elevation of 2357 m (7731 feet) and a sur- mineralogist, described the process in his ical sketch of the Yellowstone Lake shore- face area of 341 km2 (Plate 1, inset). Over journal (see box). line but many of the points where sound- 141 rivers and streams flow into the lake. The survey mapped a shoreline of 130 ings were taken for the survey. The Yellowstone River, which enters at the miles; the most recently mapped shoreline A second map of Yellowstone Lake, south end of the Southeast Arm, dominates gives the perimeter of Yellowstone Lake to published in 1896, incorporated elements the inflow of water and sediment. The only outlet from the lake is at Fishing Bridge, where the Yellowstone River flows north “A man stands on the shore with a compass and takes a bearing to the man in the and discharges 2000–9000 cubic feet/sec- Boat as he drops the lead, giving a signal at the time. Then the man in the Boat ond. The earliest attempt to produce a takes a bearing to the fixed point on the shore where the first man is located and detailed map of the shoreline and bathym- thus the soundings will be located on the chart...[Elliott will] make a systematic etry of Yellowstone Lake occurred during sketch of the shore with all its indentations [from?] the banks down, indeed, mak- the 1871 U.S. Geological Survey expedi- ing a complete topographical as well as a pictorial sketch of the shores as seen from tion, when Ferdinand V. Hayden led 28 the water, for a circuit of at least 130 miles. He will also make soundings, at var- scientists, scouts, and cooks in a survey of ious points.” what is now Yellowstone National Park. The sheer effort expended by this group, under the most primitive of working con- be 141 miles (227 km). Over 40 sound- of the original 1871 Elliott map from the ditions, is impressive on its own, but espe- ings were taken along the north and west Hayden expedition. While no mention is cially when considered in tandem with the shores, the deepest being around 300 feet. made in the official USGS report of addi- many accomplishments of the survey. A The survey estimated the deepest part of tional mapping or modifications made to primary goal of the party was “mak(ing) a the lake would be farther east and no deep- the Elliott Yellowstone Lake map, or even most thorough survey of [Yellowstone er than 500 feet. This depth range is com- of any additional work on Yellowstone Lake],” reflecting Hayden’s general inter- parable to what we know today; the deep- Lake during the years of the Hague survey est in watersheds and river drainage est point in Yellowstone Lake is due east of (1883–89, 1890–91, 1893), the lake was basins. Stevenson Island (Plate 3B) at 131 m (430 clearly resurveyed and triangulated by A 4.5 × 11-foot oak boat with a woolen feet) deep. In addition, the Hayden survey H.S. Chase and others, as published in blanket sail was used to map Yellowstone identified the long NE/SW-trending trough maps in the Hague report and reflected in

Facing page: Plate 1. From top left: (A) Henry Elliott's 1871 map of Yellowstone Lake. The headwaters of the Snake River, Upper Valley of the Yellowstone River, and Pelican River are shown. The area now known as West Thumb is referred to as the South West Arm. (B) W.H. Jackson photo of the survey boat, The Anna, with James Stevenson (left) and Chester Dawes on July 28, 1871. (C) 1896 map of Yellowstone Lake and surrounding geology as mapped in the Hague survey. (D) 1992 Kaplinski map. (E) New high-resolution bathymetric map acquired by multibeam sonar imaging and seismic mapping. The area surrounding the lake is shown as a gray-shaded relief map.

Spring 2003 15 Plate 1. The 1896 map built upon the 1987 survey. The map identified many Acronyms used in figures Elliott map and refined areas on the shore- thermal areas on the floor of the lake. The line, such as in the Delusion Lake area resulting bathymetric map has served as BFZ: Buffalo Fault Zone between Flat Mountain Arm and Breeze the most accurate lake map for Yellow- EBFZ: Elephant Back Fault Zone Point. Where the Elliott map of Yellow- stone National Park for over a decade, and EF: Eagle Bay Fault Zone stone Lake shows Delusion Lake as an arm has proven invaluable in addressing seri- HFZ: Hebgen Fault Zone of the lake, the Hague map delineates its ous resource management issues, specifi- IP: Indian Pond boundaries and identifies swampy areas cally monitoring and catching the aggres- LHR: LeHardy Rapids nearby. The maps from the Hague survey sive and piscivorous lake trout. LV: Lake Village also include a rather sophisticated geolog- Ten years after that bathymetric map, ic map of the subaerial portions of the park development of global positioning tech- MB: Mary Bay around the lake. nology and high-resolution, multi-beam PV: Pelican Valley The next significant attempt to map sonar imaging justified a new, high-reso- Qa: Quaternary alluvium (deltaic sedi- Yellowstone Lake came a hundred years lution mapping effort in the lake. Mapping ments) later and employed a single-channel echo and sampling conducted in 1999–2002 as Qg: Quaternary glacial deposits sounder and a mini-ranger for navigation, a collaborative effort between the USGS, Qh: Quaternary hydrothermal deposits requiring interpolation between track Eastern Oceanics, and the National Park Qhe: Quaternary hydrothermal explo- lines. Over 1475 km of sonar profiles were Service utilized state-of-the-art bathymet- sion deposits collected in 1987, using track lines spaced ric, seismic, and submersible remotely- Ql: Quaternary shallow lake sediments approximately 500 m apart and connected operated vehicle (ROV) equipment to col- (shallow water deposits and submerged by 1–2 km-spaced cross lines. An addi- lect data along 200-m track lines with later tional 1150 km of sonar profiles were col- infill, where necessary. The 1999–2002 Qld: Quaternary deep lake sediments lected in 1988 to fill in data gaps from the mapping of Yellowstone Lake took 62 (laminated deep-basin deposits) Qls: Quaternary land slide deposits Qpca: Quaternary Aster Creek flow 1A Qpcd: Quaternary Dry Creek flow Yellowstone National Park Qpce: Quarternary Elephant Back flow Norris- Qpch: Quaternary Hayden Valley flow Mammoth hot spring areas Qpcl: Quaternary tuff of Bluff Point Corridor postcollapse rhyolite lava flows

Qpcm: Quaternary Mary Lake flow resurgenti domes Yellowstone Qpcn: Quaternary Nez Perce flow 0.64-Ma Yellowstone caldera boundary Plateau X Qpcp: Quaternary Pitchstone Plateau N Lava Creek Tuff flow 2.05-Ma Huckleberry Ridge caldera boundary Qpcw: Quaternary West Thumb flow X Qpcz: Quaternary Pelican Creek flow Qps: Quaternary tuff of Bluff Point Yellowstone Qs: Quaternary sediments Lake Qt: Quaternary talus and slope deposits Qvl: Quaternary Lava Creek Tuff 0 10 20 30 40 50 KM Qy: Quaternary Yellowstone Group l lllll ignimbrites SI: Stevenson Island SP: Sand Point Figure 1. (A) Index map showing the 0.64-Ma Yellowstone caldera, the distribution of its SPt: Storm Point erupted ignimbrite (the Lava Creek Tuff, medium gray), post-caldera rhyolitic lava flows TFZ: Teton Fault Zone (light gray), subaerial hydrothermal areas (red), and the two resurgent domes (shown as Tl: Tertiary Langford Formation vol- ovals with faults). The inferred margin of the 2.05-Ma Huckleberry Ridge caldera is also canics shown. (B) (facing page) Geologic shaded relief map of the area surrounding Yellowstone TL: Turbid Lake Lake. Yellow markers in West Thumb basin and the northern basin are locations of active or Tli: Tertiary Langford Formation intru- inactive hydrothermal vents mapped by seismic reflection and multibeam sonar. (C) (facing sives page) Color shaded-relief image of high-resolution, reduced-to-the-pole aeromagnetic map. Sources of the magnetic anomalies are shallow and include the post-caldera rhyolite lava Tv: Tertiary volcanic rocks flows (some outlined in white) that have partly filled in the Yellowstone caldera. Rhyolitic YR: Yellowstone River lava flows (outlined in white) underlying Yellowstone Lake are shown clearly in this map.

16 Yellowstone Science Glossary of terms amphipods: crustaceans of small size and laterally-compressed body anastomozing: joining of the parts of branched systems bathymetric: relating to the measurement of depth and floor contour of bodies of water breccia: sharp fragments of rock embedded in a fine-grained matrix (as sand or clay) brittle-ductile transition zone: area where brittle and malleable rock meet beneath the earth’s surface dB: decibel diatomaceous: consisting of or abounding in diatoms (unicellular or colonial algae having silicified cell walls) en echelon: referring to an overlapped or staggered arrangement of geologic features fathometer: tool used to measure fathoms (6-foot units used to measure water depth) graben: a depressed segment of the earth’s crust bounded on at least two sides by faults and generally longer than it is wide

H2S: hydrogen sulfide ka: thousand years ago lacustrine: of, relating to, formed, or growing in lakes laminated: composed of layers of firmly united material lobate: having lobes Ma: million years ago mW/m2: milliWatt per square meter potamodromous: migratory in fresh water reduced-to-the-pole map: aeromagnetic map designed to account for the inclination of Earth’s magnetic field. Princi- pal effect is to shift magnetic anomalies to positions directly above their sources. seismic reflection profile: a continuous record of sound waves reflected by a density interface silicic: of, related to, or derived from silica or silicon strike-slip displacement: displacement whose direction of movement is parallel to the direction of its associated fault U-series disequilibrium dating: a method of determining the age of a desposit by analyzing the isotopes produced by radioactive decay of uranium isotopes Most definitions from Webster’s Third New International Dictionary (1981)

1B 1C º Qpcn Qpcn Qpcu 44 36' l l 44º36' PVIP Qpc? Spruce Creek Qpc? l º Qpcu Qpcw TL 44 35' flow LV Elephant West SPt Back flow Qpce MB Thumb Qpc? Qy Qpcw flow SI LB

l 44º30' Qpce northern West SP Qpcw basin Tv? Thumb 105.4 basin central -47.1 Dot Is. basin Qpca Qpce Qps Tv Park Absaroka Point -134.8 Qpcd l 44º25' Qpca Eagle Frank Is. Range Aster Bay Qpcd Tv Duck Creek Tv(?) fault Tv(?) -203.5 Lake Tv Qpca Plover flow Qps IN Point Plover -257.3 G Tv Point AR Flat M Tv -310.1 A Mtn. The

R l º E Qmf 44 20'

Arm ll l l l ll D Promontory -413.6 L Qpca A Southeast C Qy nT NE Qy Arm TO WS South LO EL Arm Qpcp Y Qy Tv Tv l l 44º15' 44º15' Qy l l l l l º l º º º -110 40' -110 40' -110 20' -110 10' -110º40' -110º10'

l l 0 10 km

Spring 2003 17 days over a 4-year period, compared to lowstone National Park have contributed development of a series of postglacial Hayden’s survey of 24 days in 1871. It to the unusual shape of Yellowstone Lake, shoreline terraces, and postglacial (<12- began in 1999 with mapping the northern which straddles the southeast margin of 15 ka) hydrothermal-explosion events, basin and continued in 2000 in West the Yellowstone caldera (Figure 1A), one which created the Mary Bay crater com- Thumb basin, in 2001 in the central basin, of the world’s largest active silicic volca- plex and other craters. and in 2002 in the southern lake including noes. Volcanic forces contributing to the The objective of the present work is to the Flat Mountain, South, and Southeast lake’s form include the explosive, caldera- understand the geologic processes that Arms (see Plate 1E). Unlike any of the pre- forming, 2.05-Ma eruption of the Huckle- shape the lake floor. Our three-pronged vious mapping efforts, the 1999-2002 berry Ridge Tuff, followed by eruption of approach to mapping the floor of Yellow- swath multi-beam survey produced con- the 0.64-Ma Lava Creek Tuff. Following stone Lake located, imaged, and sampled tinuous overlapping coverage, collecting explosive, pyroclastic-dominated activity, bottom features such as sublacustrine hot- more than 220,000,000 soundings and pro- large-volume rhyolitic lava flows were spring vents and fluids, hydrothermal ducing high-resolution bathymetric emplaced along the caldera margin, infill- deposits, hydrothermal-explosion craters, images. Seismic reflection records of the ing much of the caldera (Figures 1A, B). A rock outcrops, glacial features, slump

West Thumb Basin Northern Basin

o submerged o o o 110 26' 110 24 ' Qyl o ' Qpce lakeshore Qpce Qs 110 20 Qyl110 18 110 16 Qpcw Qci terraces Qpcw Qyl 44o35' Qpcw Yellowstone Pelican River Valley ' ' Qs Qs A Qs Qpce o o 44 28' 44 34' Qpcd hydrothermal vents Indian Pond Turbid Qpcd Qpcw Lake Qhe Lake Village Storm West Qpcw Qhe o Point Mary Bay lava Thumb 44 33' Qps flows outlet to Yellowstone B Qhe graben Qps Lake Bridge Bay lava Elliot's A' o 44 32' spire flows Crater explosion Steamboat field craters Point o 44 26' ejecta Qh explosion Gull deposits Qyl Qpcd crater lava Pt. flows o Stevenson 44 31' Qpcw B' Island lava hydrothermal Tl hydrothermal N flows vents Qpce Pelican Qhe vents submerged Roost Duck fissures Tl Lake lakeshore terraces hydrothermal vents Qh 44o30' Qs landslide Qyl deposits Qpce Sand o Pt. 1 km 44 24' Qs 1 km Qs Tl submerged lakeshore terraces Qs o o o 110 34'110 32' 110 30'

Figure 2. (A) New high-resolution bathymetric map of the West Thumb basin of Yellowstone Lake, acquired by multibeam sonar imaging and seismic mapping in 2000, showing a previously unknown ~500-m-wide hydrothermal explosion crater (east of Duck Lake), numerous hydrothermal vents, submerged lakeshore terraces, and inferred rhyolitic lava flows that underlie 7- to 10-m of post-glacial sediments. (B) High-resolution bathymetric map of the northern basin of Yellowstone Lake, acquired in 1999, showing large hydrothermal explosion craters in Mary Bay and south-southeast of Storm Point, numerous smaller craters related to hydrothermal vents, and landslide deposits along the eastern margin of the lake near the caldera margin. Post-caldera rhyolitic lava flows underlie much of the northern basin. Fissures west of Stevenson Island and the graben north of it may be related to the young Eagle Bay fault (see Fig. 1B). upper 25 m of the lake bottom were smaller caldera-forming event about 140 blocks, faults, fissures, and submerged obtained along with the bathymetry in the ka, comparable in size to Crater Lake, Ore- shorelines. entire lake excluding the South and South- gon, created the West Thumb basin. Sev- east Arms. This effort has produced a map eral significant glacial advances and reces- RESULTS AND DISCOVERIES OF HIGH- that is accurate to the <1-m scale in most sions continued to shape the lake and over- RESOLUTION MAPPING areas. The following report focuses on lapped the volcanic events. Glacial scour Topographic margin of the caldera. results of this mapping effort and the inter- deepened the central basin of the lake and pretation of the newly discovered features. the faulted South and Southeast Arms Geologic maps show the topographic (Figure 1B). More recent dynamic margin of the Yellowstone caldera as run- GEOLOGIC SETTING processes shaping Yellowstone Lake ning below lake level in Yellowstone Lake Powerful geologic processes in Yel- include currently active fault systems, between the western entrance to Flat

18 Yellowstone Science Central Basin

Qpce Qs hydrothermal Figure 2C. High-resolution bathymetric map central basin vents Tv? of the central lake basin, acquired by multi- Qpcw Qs Elk beam sonar imaging and seismic mapping in Point 2001, showing the Yellowstone caldera topo- lava flows Qpca graphic margin, a large hydrothermal explo- 44o30' Dot lava sion crater south of Frank Island, and numer- Island flow ous faults, fissures, and hydrothermal vents as indicated. Qpca Qy Tv? hydrothermal or faulting or fracturing along which vents caldera Park margin Point hydrothermal alteration has occurred. We Qs Frank Qs Island believe the lava flows are key to control-

o Tv 44 25' ling many morphologic and hydrothermal features in the lake. Areas of the lake bottom around the Qpca ll Eagle 1 km perimeter of West Thumb basin (Figures Bay explosion 2A, 2B) have steep, nearly vertical mar- Delusion Lake fault crater Tv gins, bulbous edges, and irregular hum- zone Qs mocky surfaces, similar to postcollapse -110o27'30" -110o20' Qs Tv rhyolitic lava flows of the Yellowstone Plateau. Seismic reflection profiles in the Mountain Arm and north of Lake Butte hummocky or ridged tops, and strongly near-shore areas of West Thumb basin (Figure 1B). Our mapping of the central jointed interiors. Stream drainages tend to show high-amplitude reflectors (indicat- basin of Yellowstone Lake in 2001 identi- occur along flow boundaries, rather than ing low magnetic intensity) beneath about fied the topographic margin of the Yel- within flow interiors. 7–10 m of layered lacustrine sediments lowstone caldera as a series of elongated A major discovery of the lake surveys (Figure 3A). troughs northeast from Frank Island across is the presence of previously unrecognized Areas such as the West Thumb and the deep basin of the lake. Based on our rhyolitic lava flows underlying much of Potts geyser basins in West Thumb basin, new data and high-resolution aeromagnet- the lake floor. Field examination of rhyo- and Mary Bay in the northern basin, cur- ic data, we infer the topographic margin of lite flows shows that many areas identified rently have extremely high heat flow val- the Yellowstone caldera to pass through through the aeromagnetic mapping as hav- ues (1650–15,600 mW/m2). Current heat the southern part of Frank Island. ing low magnetic intensity values corre- flow values in Bridge Bay (580 mW/m2) spond to areas with hydrothermal activity, are relatively low compared to Mary Bay, Rhyolitic lava flows. Large-volume, subaerial rhyolitic lava South, Southeast, and Flat Mountain Arms flows on the Yellowstone Plateau control much of the local topography and hydrology. Characteristic lava-flow morpholo- gies include near-vertical margins (some as high as 700 m), rubbly flow carapaces,

Figure 2D. High-resolution bathymetric map of the South, Southeast, and Flat Mountain Arms, acquired by multibeam "Potholes sonar imaging in 2002, showing the glaciat- of the Southeast Arm" ed landscape of the lake floor in the south- ernmost part of Yellowstone Lake and several faults. The bathymetry in the Southeast Arm contains many glacial meltwater and stagnant ice block features; the area is infor- 1 km mally referred to as the "Potholes of the Southeast Arm," and resembles much of the kettle dominated topography mapped by Ken Pierce and others in Jackson Hole (inset image). 1 km

Spring 2003 19 3A A' A 0.00 water bottom

draped lacustrine sediments high-amplitude reflector

domed r (rhyolite flow) ll 36.25 sediments fault domed sediments depth (m)

gas pockets

l 72.50

1 km

B' B

l 0.00 uplifted, tilted block water bottom domal lacustrine with deformed vent lacustrineB. feature sediments sediments sediments gas gas 36.25 pockets pocket domal vent structures vent ) m ( ll 72.50 th p de

108.75l

145.00l 1 km

Figure 3. (A) High-resolution seismic reflection image from northwestern West ization of the rocks there. South of Bridge east quadrant of West Thumb basin, along Thumb basin showing high-amplitude Bay and west of Stevenson Island, low the southern half of the West Thumb chan- (red) reflector interpreted as a sub-bottom magnetic intensity values reflect active nelway, and over the central basin of the rhyolitic lava flow. Glacial and lacustrine hydrothermal venting and relatively high lake well past Dot and Frank Islands (Fig- sediments, marked in blue, overlie this heat flow values. Low magnetic intensity ures 1, 2). The Aster Creek flow has few unit. (B) High-resolution seismic reflec- values in the northern West Thumb basin mapped faults, and few areas that have tion image across part of Elliott's explo- also may be due to past hydrothermal been hydrothermally-altered. Similarly, sion crater, showing small vents, gas pock- activity, as evidenced by vent structures the West Thumb flow (Qpcw) can be ets, and domed sediments in the lacustrine there. Comparison of geologic maps (Fig- traced into the lake in northeastern West sediments that overlie the crater flank. ure 1B) with the high-resolution aeromag- Thumb basin, along the northern half of Lacustrine sediment thickness in the main netic maps shows a crude relation of mag- West Thumb channelway, and into the crater indicates 5-7 thousand years of dep- netic anomalies to the mapped individual northern basin beneath Stevenson Island osition since the main explosion. More lava flows on land (Figure 1C). and Bridge Bay (Figure 2C). In contrast, recent explosions in the southern part of The magnetic signatures, combined the Elephant Back flow contains a well- the large crater ejected post-crater lacus- with the high-resolution bathymetric and developed system of northeast-trending trine sediments and created new, smaller seismic reflection data, allow identifica- faults or fissures that has been extensively craters and a possible hydrothermal siliceous spire. tion and correlation of sediment-covered altered so that the magnetic signature of rhyolitic lava flows far out into the lake this unit is fractured with a wide range of yet the Bridge Bay area has low magnetic (Figures 1, 2). For example, the Aster values in magnetic intensity (Figure 2D). intensity values. Evidence for past Creek flow (Qpca) southwest of the lake Field examination of subaerial rhy- hydrothermal activity is present as inac- (Figure 1C) is associated with a consis- olitic lava flows indicates that negative tive hydrothermal vents and structures, and tent, moderately positive, magnetic anom- magnetic anomalies, for the most part, are may have been responsible for demagnet- aly that extends over the lake in the south- associated with extensive hydrothermal

20 Yellowstone Science alteration or, in places, alteration Another newly-discov- due to emplacement of lava flows ered, large, subaqueous into water, such as ancestral Yel- hydrothermal explosion crater lowstone Lake. For example, the is the >600-m-wide elongate, West Thumb rhyolite flow due steep-walled, flat-floored west of the Yellowstone River is crater south of Frank Island glassy, flow-banded, and fresh; (Figure 2D). Muted topogra- the magnetic intensity values in phy suggests that this explo- this area generally are high (Fig- sion crater is one of the oldest ures 1B, 1C, 2C). In contrast, in still recognizable in Yellow- areas where flows were emplaced stone Lake. Further, this crater into water, such as the West occurs in an area where heat Thumb rhyolite flow exposed on flow values are at present rela- the northeast shore of West tively low. Submersible inves- Thumb basin (Figures 1B, 2B), tigations do not indicate magnetic intensity values are low hydrothermal activity within (Figure 1C). The low magnetic the crater. values of flows emplaced into In the northern basin of water may be primarily carried Yellowstone Lake, Mary Bay by the fine-grained and altered contains a roughly 1-km by 2- matrix in the massive rhyolitic breccias, of the offshore explosion crater. The 500- km area of coalesced explosion craters highly fractured perlitic vitrophyre, clastic m-wide West Thumb explosion crater is (Figures 2A, 2C), thus making it the dikes, and entrained stream, beach, and surrounded by 12– 20 m high, nearly ver- world’s largest known hydrothermal lake sediments in an altered matrix. tical walls, and has several smaller nested explosion system. Boiling temperature in

Plate 3. (A) High-resolution bathymetric image of hydrothermal siliceous spires on the lake floor. (B) High-resolution bathymetric image of hydrothermal vent craters along a northwest-trending fissure east of Stevenson Island. The deep hole at the southeastern end of the trend is one of the largest hydrothermal vent areas in the lake, and is also the deepest point in the lake at 133 m.

craters along its eastern edge. These nest- the deep part of Mary Bay is about 160ºC. Large hydrothermal explosion craters. ed craters are as deep as 40 m, and are Submersible investigations show that flu- Subaerial hydrothermal explosions younger than the main crater. Tempera- ids from a 35-m-deep hydrothermal vent have occurred repeatedly in YNP over the tures of hydrothermal fluids emanating have temperatures near the 120ºC limit of past 12 ka, and are confined primarily from the smaller northeast nested crater the temperature probes used, reflecting within the boundaries of the Yellowstone have been measured by ROV at 72°C. extremely high-heat flow values in this caldera (Figure 1). Large (>500 m), circular, steep-walled, flat-bottomed depressions are mapped at several sites in Yel- The Deepest Hydrothermal Vent on lowstone Lake in the West Thumb, cen- The Floor of Yellowstone Lake tral, and northern basins (Figure 2). These are interpreted as large composite _ 36.2 hydrothermal explosion craters similar in _ 43.8 origin to those on land, such as Duck _ 51.3

Lake, Pocket Basin, the Turbid Lake 44.5167o I crater, and the Indian Pond crater (Figures _ 58.9

1B, 2B, 2C). _ 66.4 A newly-discovered, 500-m-diameter, _ sublacustrine explosion crater in the west- 73.9 ern part of West Thumb basin, near the _ 81.5 currently active West Thumb Geyser _ 89.0 o 44.5111 I Basin, is only 300 m northeast of Duck _ 96.6 Lake (Figures 2A, 2B), a postglacial (<12 _ ka) hydrothermal explosion crater. Here, 104.1 _ 111.7 i i heat-flow values are as high as o o Depth -110.361 -110.356 1500mW/m2, reflecting the hydrothermal (m) activity that contributed to the formation

Spring 2003 21

4A 4B northeast linear trend

northeast linear trend Storm Z 105.4 F Point Qhe B -47.1 E "Inflated Plain" -134.8

Gull Point -203.5 Qs ek re Elliott's l C se ea crater -257.3 Qpce W -310.1 Sand Point -413.6 Rock Point nT

0 10 km

4C 4D

32'15" 32'15"

32'10" 32'10"

32'05" 32'05"

o o 44 32' 44 32'

o o 110 21'40" 21'30" 21'20" 21'10" 110 21'40" 21'30" 21'20" 21'10"

0 100 200 300 400 500 m 0 100 200 300 400 500 m

4E 4F

32'30"

5.7 m

5.56 N

11.07 32'20" 5.7 m 16.59

22.10

32'10" 27.62

33.13 46.6 m 44o32'18.78" 38.65 o 44 32' 44.16 46.6 m

49.67 o 110 21'07.52" o depth (m) 21'50" 21'40" 21'30" 21'20" 21'10" 110 21' 44o31'56.01" o 0 200 400 600 800 1000 m 110 21'46.01"

Spring 2003 23 area. Radiocarbon dates from charcoal in lapse of sealed cap rock, created the of opacity in the seismic data, and of low breccia deposits and underlying soils depressions. values of magnetic intensity in the aero- exposed in the wave-cut cliffs along the Following the initial major explosive magnetic data, represent larger zones of Mary Bay shore indicate that eruption of event of Elliott’s crater, lacustrine sedi- hydrothermal alteration than seen in the this crater occurred at 13.4 ka. Detailed ments accumulated in the floor of the surficial hydrothermal vents. stratigraphic measurements of the breccia crater and on its south flank. Based on sed- Seismic reflection profiles of the sur- deposit indicate that multiple explosions imentation rates in the lake, post-eruptive veyed areas in the northern, central, and and emplacements occurred during for- sediment thickness of ~8 m indicates the West Thumb basins of Yellowstone Lake mation of this large and complex feature. main hydrothermal explosion occurred reveal a lake floor covered with laminated, A dark, clean, well sorted, cross- to between 8 and 13 ka. Opaque zones with- diatomaceous, lacustrine muds, many of planar-bedded, generally fine-grained sand in the stratified sedimentary fill of the which are deformed, disturbed, and overlying varved lake sediments occurs as crater indicate the presence of hydrother- altered. High-resolution bathymetric map- a sedimentary interbed between breccia mal fluids and/or gases. The presence of ping reveals that many areas contain small deposits within the Mary Bay breccia two younger craters at the south end of the (<20 m) depressions pockmarking the lake deposit. These types of deposits are likely main crater floor further indicates more bottom (Plate 2, cover). ephemeral, and the likelihood of their recent hydrothermal activity, and possibly Many vent areas are associated with preservation in the stratigraphic record is younger explosions. A north-south seis- smaller domal structures in which the lam- slight. The sand unit below the Mary Bay mic profile across Elliott’s crater shows inated, diatomaceous, lacustrine sediments breccia is 1.5 to > 2 m thick and contains about 10 m of vertical difference in height have been domed upward as much as sev- numerous small en echelon faults. These between the rims. This difference may eral meters by underlying pockets of gas or deposits are similar to other paleoseis- result from doming associated with gas-charged fluids, presumably rich in mites. We conclude that this sand unit rep- hydrothermal activity prior to initial explo- steam and possibly CO2. Hydrothermal resents a deposit from a possible earth- sion. activity beneath the domes silicifies the quake-generated tsunami-like wave, which sediments causing them to become sealed, may be related to triggering the explosion Hydrothermal vents on the floor of impermeable, and weakly lithified so that of the Mary Bay crater complex. Yellowstone Lake. their resultant compaction is minimal. The One kilometer southwest of the Mary Geochemical studies of the vents indi- unaltered zones of muds surrounding these Bay crater complex is another newly-dis- cate that ~10% of the total deep thermal domes become more compacted over time covered, large (~800-m-diameter), com- water flux in Yellowstone National Park and contribute to the overall domal mor- posite depression informally referred to as occurs on the lake bottom. Hydrothermal phology. These domal structures may be Elliott’s crater (Figure 2C, Plate 3A), fluids containing potentially toxic ele- precursors to small hydrothermal explo- named after Henry Elliott who helped map ments (arsenic, antimony, mercury, molyb- sions, collapse zones, and areas where Yellowstone Lake in the 1871 Hayden sur- denum, tungsten, and thallium) signifi- active hydrothermal venting may develop vey. Development of Elliott’s hydrother- cantly influence lake chemistry, and pos- in the future. mal explosion crater is best illustrated in a sibly the lake ecosystem. ROV observa- An active domal structure informally north–south seismic reflection profile tions indicate that shallow hydrothermal referred to as the “Inflated Plain” was orig- (Figure 3B). Zones of non-reflectivity in vents are home to abundant bacteria and inally recognized in the 1999 bathymetric the seismic profile on the floor and flanks amphipods that form the base of the local survey of the northern basin as a relative- of the large crater probably represent food chain that includes indigenous cut- ly large “bulge-like feature”. The “Inflated hydrothermally altered, and possibly het- throat trout, grizzly bears, bald eagles, and Plain” covers a roughly circular area with erolithic explosion-breccia deposits, simi- otters that feed on the potamodromous cut- a diameter of ~1 km, and rises several 10s lar in character to those exposed on land throat trout during spawning in streams of meters above the surrounding lake floor. and associated with subaerial explosion around the lake. This area hosts numerous active and vig- craters. Seismic profiles in the hummocky In seismic reflection profiles (Figure orous hydrothermal vents, smaller domal area southeast of Elliott’s crater are also 3B), hydrothermal vent features are typi- structures, and vent deeps. non-reflective, and may represent a layer cally imaged as V-shaped structures asso- As seen in Figure 4A, the “Inflated of heterolithic and/or hydrothermally ciated with reflective layers that are Plain” lies along a northeast linear trend in altered material erupted from this crater. In deformed or have sediments draped across line with Storm Point and Indian Pond, contrast to the subaerial craters, which their edges. Areas of high opacity or no both areas of hydrothermal explosion ori- have radial aprons of explosion-breccia reflection occur directly beneath them, and gin, to the northeast; an unnamed trough to deposits that rim the crater, many of the are interpreted as gas pockets, gas-charged the southwest; and Weasel Creek farther sublacustrine circular depressions lack an fluids, or hydrothermally-altered zones. southwest, west of the lake (Figure 2B, obvious apron. This may indicate either Evidence for lateral movement of 4A). We informally refer to the northeast more widespread dispersal of ejection hydrothermal fluids is seen beneath and linear trend as the “Weasel Creek linea- deposits in the lake water or that some adjacent to hydrothermal vents identified ment.” Weasel Creek is an unusually other process, such as catastrophic col- in the seismic reflection profiles. The areas straight drainage, as are two smaller sub-

24 Yellowstone Science exterior interior parallel drainages due north of it, and may 5A 5B Depth 10 10 represent a linear zone of weakness. The (m) cm cm “Weasel Creek lineament” also is reflect- 13.66 l growth 14.13 fronts ed as a linear zone of low magnetic inten- 14.59 sity in the high-resolution magnetic map of 15.06 the area (Figure 4B), and may reflect a 15.53 15.99

zone of upwelling hydrothermal fluids that 16.46 llllll have contributed significantly to the 16.92 demagnetization of the rocks present. This 17.39 17.85 4 m

structure appears to left-laterally offset to 18.32 llll the “outlet graben” to the north from an cut interior incipient graben to the south associated section * with the “fissures.” In summer 2002, while traversing the “Inflated Plain” area in the boat, RV Cut- throat, we noted a strong scent of H2S, a

10–30-m diameter plume of fine sedi- 5C 5D ments, and large concentrations of bub-

10000 l bles, many of them quite vigorous, at the l l

ll samples from lll lake surface. The fine sediment plume was ll {oxidized exterior detected by the fathometer as a strong 1000 l l samples from l reflector, concentrated ~3 m below the lll {siliceous interior ll

lake surface. For Dave Lovalvo, this was 100 l l lll ll the first time in 18 years of working in this lll area that any of these phenomena have ll 10 l

µ l

10 m l been observed. The depth of the lake floor ll lll

ll ll ll ll ll lllll lllll lllll lllll lllll lllll here is ~28 m. lllll 1 l A close-up, bathymetric image of the l Asll Bal Mnl Mol Sbl Tll W “Inflated Plain” (Figure 4C) shows a bulging, domal structure, pockmarked Figure 5. (A) Bathymetric image of spires in Bridge Bay, showing roughly conical with numerous hydrothermal vents and shapes. About a dozen such siliceous sinter spires occur near Bridge Bay, some as tall as craters. Clear evidence of hydrothermal 8m. Many of the spires occupy lake-bottom depressions (possible former explosion or alteration is seen in the amplitude map collapse craters). (B) Photographs of the exterior and interior of a 1.4-m-tall spire sample (Figure 4D), where bright areas are reflec- recovered from Bridge Bay by NPS divers. The sediment-water interface of this spire is tive due to their relative hardness and apparent near the base of the exterior section, as seen in the dramatic change in color in degree of silicification. Figures 4E and F the outer rind of red-brown ferromanganese oxide to the light gray interior. (The red show the “Inflated Plain” in 2-dimension- asterisk on the photograph showing the exterior is on a natural external surface of the al and 3-dimensional perspectives, respec- spire below the sediment-water interface.) Former growth fronts on the spire can be seen tively, and plainly demonstrate how this as shown in the photograph of the interior section. (C) SEM image of diatoms, silicified feature rises as much as 30 m from the lake filamentous bacteria, and amorphous silica from a spire sample. (D) Summary bar graph floor. of chemical analyses of spire samples showing substantial concentrations of potentially toxic elements arsenic, barium, manganese, molybdenum, antimony, thallium, and Siliceous spires. tungsten. Siliceous spires in Bridge Bay (Figure 2B), in the northern basin of Yellowstone National Park Service in 1999 shows the interior of the collected spire is white, fine- Lake, were discovered by Dave Lovalvo in spire base to be shallow (~0.5 m below the ly porous, and has thin (from 0.3 cm to <3 1997, and are described here because they sediment-water interface), irregular, and cm diameter), anastomozing vertical chan- represent an end-member of hydrothermal rounded; spire material above the lake nels through which hydrothermal fluids deposit development in the lake, clearly floor constitutes about 75% of the entire flowed. Little oxide occurs in the interior imaged by multibeam sonar studies. structure. The lake floor level is recorded of the spire structure, but oxidation sur- Approximately 12–15 spires are identified on the spire as a zone of banded ferro- faces are present on former growth fronts in water depths of 15 m. These roughly manganese, oxide-stained, clay-rich, and (Figure 5B). Chemical and oxygen-isotope conical structures (Figure 5A) are up to 8 diatomaceous sediments. Below the lake analyses and scanning electron micro- m in height and up to 10 m wide at the floor, the spire is not oxidized, whereas scope (SEM) studies of spire samples base. A small, 1.4-m-tall spire collected above it, the spire has a dark, reddish- show them to be composed of silicified from Bridge Bay in cooperation with the brown, oxide coating (Figure 5B). The bacteria, diatom tests, and amorphous sil-

Spring 2003 25 6A 6C

6B Monument Geyser Basin 498.5 4947850 21.9oC SPIRES #4A &4B 2.3 6D 6E Bison & Shrub 3518.14.1 SPIRES #2 & 3 86.8oC Sphinx & Walrus SPIRE #5 4947800 Whale 3518.17.1 82oC 498.4 and 3518.15.1a (g) 1.6 87.7oC 86.8oC 2.5 498.3 and 3518.18.2 (g) 88.7oC 88.5oC 4947750 1.8 1.1 SPIRE #1 Monument 498.1 and 3518.19.3 (g) o o Geyser (steam) 83.7 C 88.7 C 1.1

4947700 528.1, 528.5 (duplicates) 71.1oC 498.2 1.1 23.4oC 1.2

4947650 SPIRE #6 495.4 90.6oC 1.1 495.2 4947600 88.4oC 6F 1.3

495.1 88.5oC 4947550 1.1

50 meters 4947500 519500 519550 519600 519650

ACID ALTERATION AREA UTM coordinates, Zone 12, NAD27 HOT SPRINGS SPIRES WATER SAMPLES Figure 6. (A) Monument Geyser Basin caps a larger hydrothermal system along a north–northwest-trending fissure. The area in white is composed of hydrothermally-altered Lava Creek Tuff. (B) Index map of Monument Geyser Basin showing the extent of hydrothermal alteration and distribution of the spire-like structures, hot springs, and water-sample localities. The values in the boxes represent individual sample numbers, temperatures, and pH. (C) Looking south into Monument Geyser Basin. Note that the basin has a central trough and contains as many as seven spire-like siliceous structures. (D) Spire-like structure on the northern edge of the basin, informally referred to as the Walrus. (E) Another spire- like structure actively venting steam and H2S. This structure is ~2m tall. (F) Underwater photograph of a large (~8 m) spire structure in Bridge Bay in the northern basin of Yellowstone Lake. The subaerial structures at Monument Geyser Basin are very similar to the spires in Bridge Bay (in terms of size, scale, distribution) and are irregular in form.

26 Yellowstone Science ica produced by sublacustrine hydrother- west of Stevenson Island extending south- and high-resolution seismic and bathy- mal vent processes (Figure 5C). ward toward Dot Island (Figure 2A); a metric data, provide the first accurate Geochemical studies of lake waters, series of en echelon, linear, northwest- information on location and displacement hydrothermal vent fluids, and waters in trending, fissure-controlled, small depres- of this important structure. Measured dis- tributary streams show that Yellowstone sions east and southeast of Stevenson placements along the two bounding faults Lake waters and vent fluids are enriched in Island; and a graben north of Stevenson are variable, but displacement along the As, Mo, Tl, Sb, and W. Similarly, the Island, nearly on strike with Lake Village western boundary is generally ~6 m, Bridge Bay spires are strongly enriched in (Figure 1B). whereas that along the eastern normal fault As, Ba, Mn, Mo, Tl, Sb, and W (Figure The subparallel fissures west of is ~2 m. The eastern bounding fault cuts 5D). Oxygen isotopic values suggest for- Stevenson Island (Figures 2A, C) cut as Holocene lake sediments, indicating recent mation of the spires at about 70–90°C. U- much as 10–20 m into the soft-sediment movement. Seismic profiles across the series disequilibrium dating of two sam- lake floor 0.5-km southeast of Sand Point. graben indicate that it projects (or strikes) ples from one spire yields dates of about These fissures represent extension frac- toward Lake Village (Figures 1B, 2C), 11 ka; thus, the spire analyzed is immedi- tures whose orientation is controlled by posing a potential seismic hazard in that ately postglacial. Spires may be analogous regional north–south structural trends, rec- area. in formation to “black-smoker chimneys;” ognized both north and south of Yellow- Another incipient graben may be offset well-documented hydrothermal features stone Lake. Active hydrothermal activity from and forming to the south-southwest associated with deep-seated hydrothermal is localized along the fissures as shown by of the Outlet graben, where the north- processes at oceanic plate boundaries. dark oxide precipitates and warm shim- northeast fissures are identified. This struc- They precipitate on the lake floor due to mering fluids upwelling from them. The ture is on trend with the Eagle Bay fault mixing between hydrothermal fluids and fissures, inspected with the submersible system. cold bottom waters. ROV for about 160 m along their NNE The sublacustrine fissures and faults Subaerial features analogous to the trend are narrow (<2 m wide), and cut ver- revealed by the high-resolution bathyme- spires in Bridge Bay may be found at tically into soft laminated sediments. No try are related to the regional tectonic Monument Geyser Basin, located along displacement is observed. A parallel set of framework of the northern Rocky Moun- the western edge of the Yellowstone N–S-trending fissures also occurs 1.3-km tains, variable depths to the brittle-ductile caldera (Figure 1A), on a ridgetop along a northeast of Sand Point (Figure 2C). Far- transition zone, and the subcaldera magma northwest-trending fault in altered Lava ther south along this trend, the fissures chamber and play important roles in shap- Creek Tuff (Figure 6A). As Figure 6B appear to have well-developed hydrother- ing the morphology of the floor of Yel- shows, the number and distribution of the mal vent craters, although investigations lowstone Lake. Many recently-identified siliceous, spire-like structures at Monu- with the submersible show only weak or features along the western margin of the ment Basin are similar to what is seen in inactive vent fields in the central basin. northern and central basins, such as the Bridge Bay. In Bridge Bay, the spires are Examination of the high-resolution mag- active fissures west of Stevenson Island cold and inactive. The structures at Monu- netic intensity map of this area shows a and the active graben north of it, are ori- ment, like the spires in Bridge Bay, have linear zone of relatively lower magnetic ented roughly north–south, and are proba- irregular forms, and similar dimensions intensities that spatially coincides with the bly related to a regional structural feature (Figures 6C, D, E, F). Currently, Monu- fissures and graben (Figures 1C, 2B, 2D). in western Yellowstone Lake on strike ment Geyser Basin sits about 250 m above Observation of the features east of with the Neogene Eagle Bay fault zone the water table, and emits highly acidic Stevenson Island (Figure 2C), using the (Figure 1B). Seismicity maps of the Yel- steam consistent with the intense alteration submersible ROV, indicates that small, lowstone region show concentrations of of the Lava Creek Tuff host rock. It is well-developed hydrothermal vents coa- epicenters along linear north–south trends unlikely that the monuments formed from lesce along northwest-trending fissures. A in the northwestern portion of the lake. an acid-steam system, because steam has a large hydrothermal vent at the south end of Landslide deposits. very limited carrying capacity for SiO2. the northernmost set of aligned vents, in We hypothesize these deposits also formed the deepest part of Yellowstone Lake, at Multibeam bathymetric data reveal from a hot water system in an aqueous 133 m (Plate 3B), emits hydrothermal flu- hummocky, lobate terrain at the base of environment, probably related to a glacial- ids as hot as 120°C. slopes along the margins, especially along ly-dammed lake during the waning stages Finally, east–west seismic reflection the northeast and east of the lake basin of the Pinedale glaciation about 12–15 ka. profiles across the down-dropped block (Figure 2A, Plate 2). Seismic reflection north of Stevenson Island reveal a north- data indicate that the deposits range in Fissures and faults. northwest-trending graben structure thickness from >10 m at the eastern edge Features identified in the western area bounded by normal faults. This graben, of the lake, and are recognizable as thin of the northern and central basins (Figures referred to as the Outlet graben, was iden- (<1m) units extending up to 500 m into 2A, C, D) include a set of sub-parallel, tified by previous investigations, but our the interior of the lake basin. We interpret elongate, north-northeast-trending fissures studies, using differential GPS navigation these as landslide deposits. The thickness

Spring 2003 27 7A lake level

o 120-150 rhyolitic lava flow large gas upper sheeting swath sonar shrinkage cavities vitrophyre hydrothermal joints joints explosion crater ROV hot springs hot springs

hot springs hot springs near-vent basal basal flow feeder pyroclastic vitrophyre breccia lake sediments dikes deposits focused fluid flow through basal breccias and fractures

7B 7C lake level lake level fractured sediments with no capping lava flow 0 1 1 0 lava flow 10 10 30

30 30 50 30

50 50 70 50 70 70 90 70 110 90 90

110 110 90

50 o 0.5 100 m 100 m Temperature, C Temperature, oC 0.5 50 1000 m 1000 m 1062.5 115 mm/yr mm/yr 1062.5 115

7D 7E impermeable hydrothermal vents in northern Yellowstone Lake lake level lava flow fractured basal 1 0 0 1 unit

50 90

130 90 90

Temperature, oC 50 1 km 100 m 50 1000 m 10 60 110 mm/yr of the lacustrine-sediment cap deposited events. The volume of material identified discontinuous, postglacial terraces that above the landslide deposits is variable, in these deposits would result in a signifi- record the history of former lake levels. and suggests that the landslides were gen- cant displacement of water in the lake, and Correlation of these submerged shoreline erated by multiple events. We suggest the may pose a potential hazard on shore. terraces around the lake is based primari- landslides were triggered by ground shak- ly on continuity inferred from multi-beam ing associated with earthquakes and (or) Submerged shorelines. bathymetric data and shore-parallel seis- hydrothermal explosions. The eastern Several submerged former lake shore- mic reflection profiles. These data indicate shore of Yellowstone Lake, near where lines form underwater benches in the West that lake levels were significantly lower in many of these landslide deposits occur, Thumb and northern basins of Yellow- the past. An extensive bench occurs south marks the margin of the Yellowstone stone Lake (Figures 2A, B, and C). The of Steamboat Point and along the western caldera and abuts steep terrain of the submerged, shallow margins (depth <15- shore of the northern basin south of Gull Absaroka Mountains to the east, both pos- 20 m) of the northern basin are generally Point (Figure 2C). In Bridge Bay, sub- sible factors contributing to landslide underlain by one-to-three relatively flat, merged-beach pebbly sand 5.5 m below

28 Yellowstone Science Figure 7. (facing page). (A) Schematic diagram showing physical features of a rhyolitic lava flow. (B) Two-dimensional fluid-flow model with simple glaciolacustrine sedimentary aquifer (no cap rock), which results in low flow velocities, recharge at the surface, and lateral flow out of both ends of the model aquifer. Subsurface temperatures never exceed 114°C, as indicated by contours and color map. Fluid flow rates are low (<0.7 mm/y), as indicated by velocity vectors. (C) Fluid-flow model with a fully-cooled rhyolitic lava flow acting as cap rock. The underlying sedimentary aquifer and heat flow are exactly the same as in the previous model. The addition of a 200-m-thick fractured crystalline rock cap strongly focuses the upward limb of an intense convection cell under the cap rock. In this model, fluid temperatures reach 140°C, and flow velocities are as high as 150 mm/yr. (D) Locations of hydrothermal vents on the lake floor mapped using seismic reflection. Lava flow boundaries are based on high-resolution bathymetry and aeromagnetic data. (E) Fluid flow model that includes a basal breccia zone beneath an impermeable lava flow. In this case, the lower sedimentary unit is overlain by a thin, fractured lava flow unit (20-m-thick) that extends the entire width of the sedimentary prism. Above the more permeable basal unit is a 170-m-thick, low-permeability, unfractured lava flow. Flow vectors indicate strong upflow under the lava flow, with maximum subsurface temperatures of ~150°C and flow rates up to 160 mm/y. Upflow is deflected laterally within the 20-m-thick "basal" fractured zone toward the flow edges, resulting in hydrothermal venting on the lake floor near the margins of lava flows. hydrothermal activity? the present lake level yielded a carbon-14 floor of Yellowstone Lake extend several date of 3,835 years. Well-developed sub- kilometers from their crater rims and One of the basic observations from our merged shoreline terraces are present in include rock fragments in excess of sever- surveys is that a close spatial relationship West Thumb basin, especially along its al meters in diameter. Deposits from the exists between the distribution of southern and northern edges. Indian Pond hydrothermal explosion event hydrothermal vents, explosion craters, and Relief on these terraces is as much as extend as much as 3 km from its crater and sublacustrine rhyolitic lava flows. Does the 2–3 m, a measure of post-depositional ver- are as thick as 3–4 m. In addition, the presence of fully-cooled lava flows in a tical deformation. Documentation of the threat of another large explosion event subaqueous environment affect the distri- submerged terraces adds to a database of may exist, as indicated by the abundance bution of hydrothermal vents? Could the as many as nine separate emergent terraces of hydrothermal venting and domal struc- identification of rhyolitic lava flows be around the lake. Changes in lake level over tures observed, especially in the northern used as a tool to help predict where some the last 9,500 radiocarbon years have basin, where heat-flow values and temper- hydrothermal activity may occur in the occurred primarily in response to episodic atures are extremely high. The area cov- future? uplift and subsidence (inflation and defla- ered by the “Inflated Plain” is very com- The relationship between sublacustrine tion) of the central part of the Yellowstone parable in scale to its neighboring feature hydrothermal features and the areas of caldera. Holocene changes in lake level to the east, the 800-m-diameter, 8.3-ka high relief, interpreted here as rhyolitic recorded by these terraces have been vari- Elliott’s hydrothermal explosion crater lava flows, can be seen in Figures 1B, 2A, ably attributed to intra-caldera magmatic (Figure 2B, 4A). and 7D. Based on our observations of the processes, hydrothermal processes, cli- The combination of active and vigor- abundant, present-day distribution of mate change, regional extension, and (or) ous hydrothermal vents, the plume of fine hydrothermal vents, we infer that fully- glacioisostatic rebound. sediments in the lake subsurface, the cooled rhyolitic lava flows exert a funda- strong, locally-sourced H2S scent, and the mental influence on subsurface hydrology DISCUSSION evidence for silicification of lake sedi- and hydrothermal vent locations. We spec- Do the newly discovered features in ments merit detailed monitoring of the ulate that upwelling hydrothermal fluids Yellowstone Lake pose potential geo- “Inflated Plain” as a potential and serious are focused preferentially through rhyolitic logic hazards? hazard and possible precursor to a large lava flows, whereas hydrothermal fluids hydrothermal explosion event. The “Inflat- conducted through lake and glacial sedi- The bathymetric, seismic, and sub- ed Plain” area was resurveyed in 2002 in ments tend to be more diffuse (Figure 7). mersible surveys of Yellowstone Lake order to compare any changes from the In addition, convective flow moves later- reveal significant potential hazards exist- 1999 survey; these analyses currently are ally away from thicker, more impermeable ing on the lake floor. Hazards range from under investigation. In addition to hazards segments of the rhyolite flow toward the potential seismic activity along the west- affecting humans, hydrothermal explo- fractured flow margin, where the majority ern edge of the lake, to hydrothermal sions are likely to be associated with the of hydrothermal activity is observed (Fig- explosions, to landsliding associated with rapid release into the lake of steam and hot ure 7E). explosion and seismic events, to sudden water, possibly affecting water chemistry collapse of the lake floor through frag- by the release of potentially toxic trace SUMMARY AND CONCLUSIONS mentation of hydrothermally-altered cap metals. Such changes could have signifi- This mapping of Yellowstone Lake rocks. Any of these events could result in cant impact on the fragile ecosystem of allows the lake basin to be understood in a sudden shift in lake level, generating Yellowstone Lake and vicinity. the geologic context of the rest of the Yel- large waves that could cause catastrophic lowstone region. Rhyolitic lava flows con- local flooding. Ejecta from past hydrother- Do rhyolitic lava flows control tribute greatly to the geology and mor- mal explosions that formed craters in the phology of Yellowstone Lake, as they do

Spring 2003 29 to the subaerial morphology of the Yellowstone Plateau. We infer from our high-resolution bathymetry and aeromagnetic data that Stevenson, Dot, and Frank Islands are underlain by large-volume rhyolitic lava flows (Figure 2A). Mapped late Pleistocene glaciolacustrine sediment deposits on these islands merely mantle or blanket the flows. Similarly, the hydrothermally-cemented beach deposits exposed on Pelican Roost, located ~1 km southwest of Steamboat Point (Figure 2C), blanket another submerged large-volume rhyolite flow. The margin of the Yellowstone caldera passes through the central part of the lake and northward along the lake’s eastern edge (Figure 1). Similar to most of the rest of the topographic margin of the Yellowstone caldera Lisa Morgan (foreground) is a research geologist for the USGS. Her focus for the past 23 years (Figure 1A), we suggest that post-col- has been the geology and geophysics of volcanic terrains, including that of the eastern Snake lapse rhyolitic lava flows are present River Plain and Yellowstone Plateau, Cascades, Hawaii, Japan, and other areas. With Ken Pierce, along much of the caldera margin Morgan developed major concepts and a model for the Yellowstone hot spot. Her current beneath Yellowstone Lake and con- research focuses on the Yellowstone Plateau, mapping and interpretation of the geology and trol much of the distribution of the potential geologic hazards in Yellowstone Lake, and physical processes associated with large-vol- sublacustrine hydrothermal vents. ume caldera-forming eruptions and hydrothermal explosions. Many potential hazards have been identified in our mapping effort. Next Pat Shanks (right) is a research geochemist with the USGS. He has extensively studied seafloor steps will include hazard assessments and sublacustrine hydrothermal vents and mineral deposits on the mid-ocean ridges and in Yel- and methodologies to be employed in lowstone Lake, using stable isotopes and aqueous geochemistry as primary tools. His current monitoring these potentially danger- research includes hot spring geochemistry, hydrothermal explosion deposits, hydrothermal alter- ous features under the aegis of the ation processes in volcanic rocks, and the geochemistry of metals in the environment related to sites of past mining activities. Yellowstone Volcano Observatory. Dave Lovalvo (left) is the founder and owner of Eastern Oceanics. For 30 years, He has filmed ACKNOWLEDGMENTS and supported underwater research projects in just about every major ocean and many of the world's major inland lakes. Dave has spent 15 years exploring, filming and now mapping Yel- We thank Kate Johnson, Ed lowstone Lake, and continues to support projects in Yellowstone and many other locations duBray, Geoff Plumlee, Pat Leahy, around the world. He has been a manned submersible pilot of the Deep Submergence Vehicle Steve Bohlen, Tom Casadevall, Linda (DSV), Alvin, and a pilot and member of the design team for the unmanned DSV, Jason 2. He Gundersen, Denny Fenn, Elliott Spik- currently operates Eastern's unmanned DSV Oceanic Explorer. His latest projects have been the er, Dick Jachowski, Mike Finley, John search and discovery of the PT-109 with Bob Ballard and the National Geographic Society, and Varley, Tom Olliff, and Paul Doss for the government-sponsored exploration of the new underwater volcano off Grenada called "Kick- supporting this work. We thank Dan em Jenny." Dave resides in Redding, Connecticut. Reinhart, Lloyd Kortge, Paul Doss, Rick Fey, John Lounsbury, Ann SOURCES Deutch, Jeff Alt, Julie Friedman, Brenda improved the manuscript. We are grateful Beitler, Charles Ginsburg, Pam Gemery, to Coleen Chaney, Debi Dale, Joan Luce, Space constraints prevent us from list- Rick Sanzolone, Dave Hill, Bree Burdick, Marlene Merrill, Mary Miller, Vicky ing the numerous citations and sources Eric White, Jim Bruckner, Jim Waples, Stricker, Sandie Williamson, and Robert included in this article as originally sub- Bob Evanoff, Wes Miles, Rick Mossman, Valdez for their skillful assistance with mitted. For a complete list, please contact Gary Nelson, Christie Hendrix, and Tim project logistics. This research was sup- [email protected]. Morzel and many others for assistance ported by the U.S. Geological Survey, the with field studies. We thank Bob Chris- National Park Service, and the Yellow- tiansen, Karl Kellogg, and Geoff Plumlee stone Foundation. for constructive reviews that substantially

30 Yellowstone Science Predators and Prey at Fishing Bridge

by Paul Schullery

For more than thirty years now, I’ve watched Yellowstone cutthroat trout rise in the slow waters at the outlet of Yellowstone Lake. Their reliable presence, their abundance, and their easy familiarity with gawkers like me, all made it seem like this was something I could get around to photographing someday, but didn’t have to do right now. After all, the only thing I really wanted was some beautiful overhead shots of these golden fish, sinuously distorted and glow- ing against the mottled greens of the river bottom. Last summer, I finally started taking the pictures. About noon on a very hot, bright early July day, I walked out on Fishing Bridge and discovered that quite a few fish were feeding steadily on small mayflies and stoneflies. Eagerly rising trout are as exciting to me as the sight of a grizzly bear, and I was immediately caught up in the scene. Rather than looking for a fish tastefully holding over just the right color of bottom so I could get my artful trout picture, I spent the next hour on the bridge or along the shore, banging away at these eager risers. Even as I was taking the pictures, I wondered if my autofocus camera and 300 mm lens were up to the challenge of stopping the action, and what I might find when I could finally examine the pictures. What I found was as exciting as watching the risers. In that first hundred or so images, a surprising number of which weren’t just blurry splashes, the camera stopped the action at many distinct stages of the rise and take. What was just a quick flash of action when I watched it was revealed as much, much more. The more I looked, the more I saw. The more I saw, the more I needed to go back and take more pictures. Each subsequent visit to the bridge led me back into the angling literature and (more fruitfully) into the scientific literature on the physiology of feeding fish. I would look through each new batch of pictures, notice something new, think about it until I wondered about something else, then look through the pictures again, and again, and again. I’m

Spring 2003 31 ALL PHOTOS PAUL SCHULLERY

1. A Yellowstone cutthroat trout with one of the many thousands of mayflies it will eat each summer. In all these photographs, wild Yellowstone cutthroat trout were photographed feeding naturally; neither trout nor flies were interfered with or manipulated in any way.

2. Something about an individual mayfly has sparked something in the brain of an individual trout, which turns to investigate. The fly has tipped over and a wing is pinned against the water surface. Perhaps the trout's interest was triggered by the panicky motion of the insect's struggles.

still not done looking, and the more I find the more I realize I’m a long way from being done taking pictures, too. If you’ve watched many nature films on television, there’s a powerful image you will almost certainly remember. The scene is a tropical reef—some colorful submerged landscape replete with coral forests, sponges, and other exotica. The whole thing is near enough to the surface for sunlight to dapple its happy, travel-poster community of plants and animals. But off to the side (sinister soundtrack here), you see the snout, or even the whole head, of some dark- ly porcine, heavy-jawed fish, shad- owed patiently amidst the undulat- ing vegetation.

32 Yellowstone Science Then a new camera angle reveals an innocent little creature—a tiny fish, a crusteacean, some other tidbit of biolog- ical mobility—going about its day (peppy, cheerful soundtrack here, to evoke additional sympathy). You know what’s going to happen, but it’s always startling anyway, because it happens so fast. The innocent little tid- dler comes doodling along until it’s directly in front of the big fish, then it’s suddenly gone and the fish, which hasn’t left its place, is closing its mouth (only the tackiest of producers put a small burp on the soundtrack at this point, but some do succumb to a little ascending pennywhistle toot, to signify the hasty sucking in of the prey). It’s a great nature film gimmick, always good for a startled chuckle. It’s also terrifically interesting predatory behavior. It’s evolution making the most

3. Even in the cleanest, clearest water, the trout must pick its food from a distracting assortment of flotsam in the surface film, caught here when the camera chose to focus on it rather than on the trout below.

4. The trout has a kind of visual lock on this drifting mayfly. The fly is now well within the range of the fish's suction. The tiny "lens" of distorted surface just above the trout's head indicates that the trout has already begun to create a "rise form," though whether or not the fish will take the fly is uncertain.

Spring 2003 33 of the animal’s tools and environment. It’s predation without the chase. It’s always dramatic, and for all its staginess and comic effect it’s also a little scary. It looks almost like magic. We don’t hear much about this sort of thing with trout, especially trout rising to feed near or on the surface. Their fastidious little “rise forms” (the spreading rings of rip- ples that follow each feeding episode) hint of a greater refine- ment, as if trout have better table manners than to go around acting like a starship with an overactive tractor beam. In fact, fishing writers have tended to describe the trout’s feeding behavior as quite passive, more or less like this: When the trout

5. The same stage of the process as the previous pictures, but with a different fish photographed from a different angle. The trout's mouth is slightly open, with the fly perfectly suspended across the gap. The lower jaw, seeming a bit underslung, appears to be filling out already as the fish begins to create the suction that will pull the fly in.

6. The decision to take the fly has been made. The fly, this time a mayfly "spinner," has barely begun to tip into the opening mouth of the trout (the spinner, with its wings extended flat across the surface, is the last life stage of the mayfly). Suction has also begun; the beginning of the suction trough is passing over the head of the trout.

34 Yellowstone Science sees a fly gliding toward it, the fish tigate. Forget for the moment that simply rises to the surface, opens in that one sentence is a world of its mouth and gills, and lets the river engaging wonders to do with the run through its head, carrying the trout’s visual acuity, its ability to fly in. The fish keeps the fly and identify prey, the refraction of light lets the extra water flow on, right in a stream and how that affects the out the gills. trout’s “window,” and a host of But trout use precisely the same other subjects that many writers suction forces as the big reef fish have capably explored. Right now described earlier. In a process that is we’re only concerned with the likewise too quick for us to observe challenges the fish faces in eating from the bank, or even from a few this fly. feet away, they take their food in by Anglers have spent centuries means of a complex watching fish feed. Vincent Mari- and forceful series naro’s beautiful book In the Ring of valve-like motions of surprising of the Rise (1976), with its series of power and elegant efficiency. photographs of trout rising, gave Let’s follow a mayfly to its anglers their first close look at the doom in a trout’s mouth, starting ways in which trout conduct their with the fly, poised on the surface, inspection of a prospective meal. riding the current downstream. The Water is a much thicker and poten- trout sees it, and moves in to inves- tially clumsier “atmosphere” than

7. The mouth is now as open as it gets. A mayfly is in it, and two more drift by to the left.

8. It isn't enough to get the fly over the lip. It must be pulled deep into the fish's mouth, and the powerful suction is now doing that. The suction trough is clearly visible around the fish's head as down-curving distortion lines. The trough is likewise revealed in its shadow on the river bottom—a twin-lobed circle encompassing the trout's head.

Spring 2003 35 air. A fish that simply charges up to its prey is likely to push it away with its own “bow wave.” But depending upon the speed of the trout as it approaches the fly, and the care it exercises, it may approach quite closely. Fish routinely get their little faces right up close to the insect, and seem to lock it in place right there in front of them as it drifts long. I wonder about this stage in the process. Marinaro showed us, in his pho- tographic series depicting what he called the “compound rise” and the “complex rise,” the way a trout noses right up to a fly, then drifts backwards along with the fly as it continues on its way downstream. The trout concentrates on the fly, and keeps it right there, just off the end of its snout.

9. The same process, with another fish viewed from another angle. Again the trough is revealed in the surface distortion around the mouth, and again the two-lobed shadow stands out on the river bottom. The twin-lobed shape is probably the result of the trout's "chin" dividing the suction trough. At this stage, though the gills may be partly open, they are not fully expelling water.

10. Though the fish is somewhat obscured by the distortion of the water, this is the busiest of the pictures in the sequence. The lower jaw is still distended; notice how the cutthroat markings stand out. Both gills are now open wide, and the trout's right gill is clearly expelling a strong current of water. Water is almost certainly also exiting the left gill, but the light is from the right (as the off-center shadow of the suction trough, on the river bottom, shows), and the distortion of light on that side is probably lost under the fish.

This behavior is certainly agonizing for the angler, and who knows what the fly must make of it? But here is what I find most curious about it. The whole time this is going on, often for several feet or even yards of drift, the fly is well within the suction range of the fish (rainbow trout in one study, feeding under the surface, rarely applied suction toward food that was much more than a head-length away). I wonder if while the trout is eyeing the fly from this close, if it isn’t also applying some subtle little out- ward or inward currents to the fly, testing it in some way? Animals take every evolutionary advantage that comes along. Maybe the trout is only toying with the fly a little (trout are known to “play” with their food more toward the end of an insect hatch, when they are presumably sated, than at the beginning). Or maybe such manipulation, jostling the fly around a little, would somehow help in the decision of whether or not to eat it. Imagine being the fly at this point. That’s all speculation, of course. What happens next is vividly real. If all goes well with the inspection, it’s time to feed.

36 Yellowstone Science The goal of the trout is to cre- structures that run the length of ate enough suction to ensure the bottom of the lower jaw (a that the fly is drawn well into cutthroat trout’s “cutthroat” the mouth. To increase the marks are partly hidden in force of that flow beyond its those pleats until they stretch own physical capacity to cre- open). ate suction, the trout will The photographs capture often move forward as it the effect of this suction clear- takes, its speed adding a little ly. The surface-feeding fish, in more umph to the current sucking down the fly, actually flow it is creating with suc- pulls a shallow hole in the tion. As it does that, it creates water surface—a little feeding suction with its mouth. There depression, or trough. The are now three distinct forces insightful British angling speeding the fly into the writer G.E.M. Skues recog- trout’s mouth: the down- nized the evidence of this stream flow of the current, process eighty years ago in The the upstream movement of Way of a Trout with a Fly. He the trout, and the suction of described the initial stage of the trout’s mouth. the take as “a faint hump on the The trout creates the suc- surface, often accompanied by tion by enlarging its mouth a tiny central eddy caused by capacity, which it does by the suction with which the opening and extending its trout has drawn in the fly.” jaws, and dropping the floor Now the fish’s mouth has of the lower jaw, deepening opened, the oral cavity has the mouth cavity. This is deepened, and the fly is either facilitated by those pleatlike in or on its way into the mouth.

11. In this revealing photograph, the trout has closed its mouth, and the suction trough is sliding back over its head. Most important, the rapid closing of the mouth and the contraction of the floor of the mouth is expelling water from the gills with such force that some of it also escapes in strong little spurts from the sides of the fish's mouth. This startling process occurred with more than one of the photographed trout.

12. The trout inadvertently inhales air along with any fly taken from the water's surface. This air is then expelled out the gills with the water. Here, the first bubble of air emerges from the trout's gill and reaches the surface as the fish turns down from its take.

Spring 2003 37 The gills are already in play, as some fly. It closes its mouth (a good bit water is moving out of them, but the more quickly than it opened it), fish is dealing with some involved flushes the water out the gills, and physics at this point. If it simply the fly is retained, presumably either drops the floor of its lower jaw and in the throat or against the gill rak- opens its mouth and gills all with ers—those hard arching structures to equal force and at the same time, which the gills are attached. there will still be a lot of suction, but There is one lovely lingering water may be pulled in from both aftereffect in the take of a trout, first ends—into the mouth and in through noted by the angling writer Skues. the gills (the latter, if it happens too Perplexed by rising trout whose prey dramatically, is apparently not an he could not see, Skues needed a especially pleasant experience for way to determine if a fish was rising the trout). This could defeat the real to floating flies, or feeding right goal of the suction. The trout needs under the surface. He reasoned that a to keep the suction going mostly one fish feeding beneath the surface way, into the mouth, to have the best would inhale only water when taking chance of capturing the fly. That a fly, but a fish feeding on the sur- said, even when the trout does this face, especially if taking an right, there may be a modest back- upwinged insect like an adult

13. More bubbles have appeared, and are drifting back over the trout as the fish settles back into its holding position. But one last fine stream of small bubbles can be seen, still underwater, as they emerge from the trout's right gill.

14. The serendipitous beauty of a complete rise: as the trout turns down from a successful take, another mayfly eases past, caught by the camera as it passes over the pectoral fin. wash into the gills, but not mayfly, would necessari- enough to interfere with ly engulf a fair amount of the capture of the fly. air with the water and the Now that the suction fly. That air would be has been successful, the expelled out the gills with trout has the fly in its the water, and would be mouth. But recall what evident as bubbles in the anglers dread at this stage: resultant rise form. A fish that the fish will reject, or feeding on insects that “spit out” the fly. They were under the surface of actually do this—ptui!— the water, such as mayfly and the reason they can nymphs or drowned do it so quickly and so adults, might cause a sur- forcefully is that they just face disturbance that reverse the process that looked like any other rise pulled the fly in. They can form, but it couldn’t have just contract that lower bubbles in it because the jaw expansion, collapse fish had no air to eject. the large oral cavity, and The photographs show the fly spurts back out. If this too. the trout was operating Setting aside what all only a passive, flow- this observation and pho- through system, it would tography and reading has have no capacity for such taught me about fishing, abrupt and decisive it has given me a deep- changes of plan, and we’d ened respect for trout— catch a lot more of them. creatures I already But let’s assume that thought I admired pretty the trout approves of the thoroughly. Perhaps most

38 Yellowstone Science important, I admire them much more as going through). Like its physiological ness or sophistication of their brains, I am individuals than I used to. The feeding attainments, which result from millions of infinitely more aware of their superiorities process is so full of opportunities for vari- years of evolutionary engineering, the than I am of their limitations. ation, not only in one fish but from fish to trout’s cognition seems to me a spectacu- fish, that I am much less likely than before larly successful tool. to make assumptions about one fish based Over the years, I’ve spent a huge on what the fish next to it has been doing. amount of time watching Yellowstone We fishermen have joked for so long about predators go about their work, making how we’re made fools of by these simple their assessments, passing their fateful little creatures that we have begun to judgments, making their perfect moves. believe not only that we’re fools but that Trout are unmistakably members of the trout really are simple. They may not be as same guild. Whatever rarified sphere of individual as humans, but I’m now con- consciousness or even wisdom these crea- vinced they’re a lot closer to it than I used tures may inhabit, and whatever we may to think. eventually conclude about the primitive- I also admire them more as predators. I don’t know what’s going on when a trout is nosing up against a fly, doing its equiv- Paul Schullery, a former editor of Yellowstone Science, is the author of many books, including alent of judging and deciding. But the

American Fly Fishing: A History (1987) and MARSHA KARLE more I stare at these pictures of fish staring Lewis and Clark Among the Grizzlies (2002). at insects, the more I respect whatever it is This essay appeared in different form in the that the trout is going through (so far, I try May 2003 issue of Fly Fisherman magazine. not to think much about what the insect is NEWS notes

USFWS Reclassifies Some Wolves from The threatened designation, which now population goals. Among those criteria are Endangered to Threatened applies to all gray wolves in the lower 48 requirements to ensure continued survival The U.S. Fish and Wildlife Service has states except for those in the Southwest, is of the gray wolf after delisting. This will changed the status of gray wolves in the accompanied by special rules to allow be accomplished through management western Great Lakes states and northern some take of wolves outside the experi- plans developed by the states and tribes. Rocky Mountains from “endangered” to mental population areas in the northern Once delisted, the species will no longer the less serious “threatened” designation Rocky Mountains. These rules provide be protected by the Endangered Species under the Endangered Species Act. options for removing wolves that cause Act. At that point, individual states and The reclassification rule also establish- problems for livestock owners and other tribes will resume management of gray es three “Distinct Population Segments” people affected by wolf populations. wolf populations, although the Service (DPS) for gray wolves under the Endan- Wolves in experimental population areas will conduct monitoring for five years after gered Species Act. The three DPSs in the northern Rocky Mountains are delisting to ensure that populations remain encompass the entire historic range of the already covered by similar rules that secure. gray wolf in the lower 48 states and Mex- remain in effect. The final rule reclassifying the gray ico, and correspond to the three areas of The USFWS will now begin the wolf will be published in the Federal Reg- the country where there are wolf popula- process of proposing to remove gray ister. For more information on the gray tions and ongoing recovery activities. wolves in the western and eastern United wolf, visit the Service’s wolf website at Wolf populations in the Eastern and States from the endangered and threatened http://midwest.fws.gov/wolf. Western DPSs have achieved population species list, once the agency has deter- goals for recovery, and Advance Notices mined that all recovery criteria for wolf Bison Capture Operations Outside of Proposed Rulemaking are being pub- populations in those areas have been met North Entrance lished concurrent with this reclassification and sufficient protections remain in place During the first week of March, bison rule to give the public notice that the Ser- to ensure sustainable populations. migrated near Stephens Creek along the vice will soon begin work to propose To delist the wolf, various recovery cri- park’s northern boundary, and capture delisting these populations. teria must be met, in addition to reaching operations began at the Stephens Creek

Spring 2003 39 capture facility outside the North Entrance and black bears. NPS PHOTO for the first time since 1996. Under the Visitors are asked final state and federal Records of Decision to be especially (ROD) for the Interagency Bison Manage- cautious of ment Plan (IBMP) that were signed in wildlife carcasses December 2000, and the December 2002 that may attract IBMP Operating Procedures, when the bears, and to take bison population in late winter/early spring the necessary pre- is over 3,000 animals, and they are moving cautions to avoid onto lands where cattle are being grazed an encounter. Do near the North Entrance, they will be cap- not approach a tured in the Stephens Creek facility and bear under any sent to slaughter facilities. The November circumstances. population estimate was approximately An encounter 3,800. About 25 bison have died this year with a bear feed- NPS and Montana Department of Livestock personnel meet at the either by management actions west of the ing on a carcass Stephens Creek capture facility. park, natural mortality, or motor vehicle increases the risk accidents. of personal injury. Winter Use FSEIS Released for Grand The IBMP and the IBMP Operating Bears will aggressively defend a food Teton and Yellowstone National Parks Procedures use a variety of methods along source, especially when surprised. the north and west boundaries of the park The National Park Service is continu- The Final Supplemental Environmen- to limit the distribution of bison and to ing the seasonal “Bear Management Area” tal Impact Statement for winter use was maintain separation of bison and cattle on closures in Yellowstone’s backcountry. made available to the public on February public and private lands. It also allows The program regulates human entry in 20, 2003. There will not be a public com- some bison on certain public lands where specific areas to prevent human/bear con- ment period. National Park Service and cattle are not grazed. flicts and to provide areas where bears can National Environmental Policy Act The first response to bison approaching range free from human disturbances. (NEPA) regulations call for a 30-day wait- the north boundary is to haze them to keep Visitors are asked to report any sight- ing period, but public comment is not cus- them inside the park. However, after ings or signs of bears to the nearest visitor tomary on a final environmental impact attempts at hazing the bison become inef- center or ranger station as soon as possible. statement. A Record of Decision was fective and unsafe, it may become neces- Permits for backcountry camping and expected to be signed near the end of sary to begin capturing the animals. Haz- information on day hikes are available at March 2003. ing occurred during the previous few visitor centers and ranger stations. Five alternatives for winter visitor use weeks on numerous occasions. For further information on spring con- in the three park units are evaluated in the A total of 231 bison were captured at ditions in Yellowstone National Park, call FSEIS. Three of the alternatives, including the Stephens Creek facility and sent to park headquarters at (307) 344-7381. the preferred alternative, are limited slaughter facilities. Meat, heads specifically to actions that allow snowmo- and hides will be donated to bile recreation to continue in the parks. Native American groups/individ- The other alternatives include a no action uals and other social service alternative that would implement the organizations. Recent aerial photos taken in Hayden and Pelican Val- Spring Bear Emergence leys (left and below, respectively) show that bears have Reminder begun their spring emergence. The park’s Bear Management Office has started receiving reports of bear activity within Yellowstone, indicating that bears are beginning to emerge from their winter dens. Soon after bears emerge from their dens, they search for winter- killed wildlife and winter-weak- ened elk and bison, the primary NPS PHOTOS sources of much-needed food during spring for both grizzlies

40 Yellowstone Science NEWS notes

November 2000 Record of Decision to ban program; initiation of thermophile sur- itage and Research Center to house the snowmobiles from the parks beginning the veys; successful bald eagle and peregrine park’s library, archives, and photo and 2003-2004 winter use season, and a sec- falcon recovery programs; meeting target museum collections; strengthening of trib- ond that would delay implementation of goals for grizzly bear recovery; six biennial relations through the consultation the November 2000 Record of Decision al science conferences (planning for the process; completion of the interagency until the 2004-2005 winter use season. seventh is underway!); the halting of the bison management plan and EIS; and The preferred alternative strikes a bal- New World Mine; initiation of a new Her- acquisition of the Jack and Susan Davis ance between phasing out all snowmobile collection. Yellowstone Science also cel- use—as required under the November ebrates 11 years this year, with a mailing 2000 Record of Decision—and allowing list that has grown to include more than for the unlimited snowmobile use of the 2,100 individuals interested in Yellow- past. Critical elements of the preferred stone’s research and resources. alternative include: reduced numbers of snowmobiles through daily limits; imple- Lake Conference Proceedings menting best available technology Available requirements for snowmobiles; imple- The proceedings from the Sixth Bien- mentation of an adaptive management pro- nial Scientific Conference on the Greater gram; guided access for both snowmobiles Yellowstone, Yellowstone Lake: Hotbed and snowcoaches; a reasonable phase-in of Chaos or Reservoir of Resilience? are period; a new generation of snowcoaches; now available. Conference participants and funding to effectively manage the win- will receive their copies in the near future. ter use program. Implementation of all the Others who would like a copy, please critical elements will address the adverse contact Virginia Warner at virginia_warn- impacts identified in the November 2000 [email protected] or (307) 344-2233. Record of Decision. Hard copies and CDs of the document are available by writing: FSEIS, Planning Office, P.O. Box 168, Yellowstone Nation- Above, YCR Director John Varley cuts the NPS arrowhead-shaped YCR birthday cake. al Park, Wyoming 82190. The document Below, Here from the start: original YCR employees Wayne Brewster, Kerry Gunther, Mary can also be found by accessing Hektner, Mark Biel, Jennifer Whipple, Ann Rodman, Paul Schullery, Sue Consolo Murphy, www.nps.gov/grte/winteruse/winteruse. John Varley, Joy Perius, and Melissa McAdam. htm. The FSEIS is loaded in two volumes. Volume 1 is the main document and the appendices. Volume 2 is the public com- ments and their responses.

Happy 10th Anniversary, YCR! The Yellowstone Center for Resources celebrated its 10th anniversary on March 13, 2003. Created with the goal of central- izing resource research and management, YCR now includes the park’s Branches of Natural and Cultural Resources, its Spatial Analysis Center (GIS lab), and YCR’s own support branch (including the Resource Information and Publications Team, AKA the people who bring you Yellowstone Science!). Although all major undertakings, such as wolf restoration, represent cooperative efforts among the park’s divisions, many such projects have been primarily directed out of the YCR. NPS PHOTOS Highlights of the past ten years include wolf restoration; the lake trout eradication

Spring 2003 41 Help keep Yellowstone Science coming! We depend on our readers’ kind support to help defray printing costs.

Please use the enclosed envelope to make your tax-deductible donation. Checks should be payable to the Yellowstone Association. Please indicate that your donation is for Yellowstone Science. NPS PHOTO

Yellowstone Science Yellowstone Center for Resources PRSRT STD AUTO P.O. Box 168 US POSTAGE PAID Yellowstone National Park, WY 82190 National Park Service Dept. of the Interior CHANGE SERVICE REQUESTED Permit No. G-83