Guest Seminar (4/27): Brett McClintock

Next Thursday, April 27, from 3:30 to 4:30 p.m. in Anderson 223, you are invited to join a special guest seminar with Dr. Brett McClintock from the NOAA National Marine Mammal Laboratory: “Hidden Markov models of animal movement: integrating more ecological realism and common challenges associated with telemetry data.”

Brett McClintockBrett is a research statistician (biology) at the NOAA National Marine Mammal Laboratory in Seattle. He earned a Ph.D. in wildlife biology and an M.S. in statistics at Colorado State University. His research focuses on the development and application of statistical models for ecological data, with a primary focus on marine mammals. He is the creator and maintainer of the animal movement R package ‘momentuHMM’ and the capture-recapture R package ‘multimark’. He also recently co-authored the book, Animal Movement: Statistical Models for Telemetry Data (CRC Press).

About the Talk
Discrete-time hidden Markov models (HMMs) have become an immensely popular tool for inferring latent animal behaviors from telemetry data, largely because they are relatively fast and easy to implement when data streams are observed without error and at regular time intervals. While HMMs of animal movement typically rely solely on location data, auxiliary biotelemetry and environmental data are powerful and readily available resources for incorporating much more behavioral realism and inferring ecological relationships that would otherwise be difficult or impossible to infer from location data alone. However, there is a paucity of generalized user-friendly software available for implementing (multivariate) HMMs of animal movement. Furthermore, measurement error and temporally irregular or missing data are often pervasive in telemetry studies (particularly in marine systems), and the incorporation of uncertainty attributable to location measurement error, temporally irregular observations, or other forms of missing data typically requires fitting HMMs using custom and computationally demanding model-fitting techniques. This is unfortunate because complex analyses requiring novel statistical methods, and custom model-fitting algorithms are not practical for many of the biologists and ecologists who collect telemetry data. Using several real-world examples, including African elephant and marine mammal telemetry data, Brett will demonstrate how a recently developed R package (momentuHMM) addresses these challenges and facilitates hypothesis-driven analyses of animal movement by non-statisticians.

Brett’s talk is free and open to the public, and we hope you can join us!

We hope you can join us!

Photo © Brett McClintock.

Against the Current

If you take a stream ecology course, you are generally taught that as a stream winds down from its headwaters at higher elevations, the water temperature will increase fairly steeply at first, and then gradually—and predictably—approach air temperature as the stream levels off at lower elevations. But several researchers at the School of Environmental and Forest Sciences (SEFS)—including doctoral student Aimee Fullerton and Professors Christian Torgersen and Josh Lawler—have recently published new findings in Hydrological Processes that could change the way we think about stream ecology and temperature dynamics.

The paper, “Rethinking the longitudinal stream temperature paradigm: region-wide comparison of thermal infrared imagery reveals unexpected complexity of river temperatures,” is a meta-analysis of thermal data from 53 rivers across the Pacific Northwest. Torgersen, a research landscape ecologist with the U.S. Geological Survey and an affiliate professor with SEFS, started building this massive data set in 1994. He partnered with long-time colleague Russ Faux at Watershed Sciences, Inc. (now part of Quantum Spatial), to collect thermal information from Faux’s aerial surveys of hundreds of rivers in Oregon, California, Idaho, Washington and a few other states.

Aimee Fullerton

Doctoral student Aimee Fullerton, lead author on the paper, grew up in Ohio and works as a research scientist with NOAA.

Thermal infrared imaging is usually accurate to within a half degree, so these readings provided a trove of high-level, high-quality spatial data to explore. “This is the first time we’ve had the kind of spatial data over many, many rivers—and at a fine resolution—to even look at these patterns,” says Torgersen. “It was my dream project.”

Using these data, the researchers set out to map spatial patterns in river temperature during the summer, when fish are most stressed. They wanted to determine, among other information, whether they could predict the location of cold patches, which provide useful “cold refuges” for fish as they migrate up a stream. And though the authors expected to find geographical indicators for how a stream’s temperature would behave, the actual results surprised them.

Rather than finding predictable patterns, they discovered a great deal of variability and complexity in the streams. About half of the rivers behaved as expected, with temperatures steeply warming from the headwaters, and then gradually tapering off as the stream progresses. With other streams, though, the pattern was more gradual and linear, or the temperature stayed the same; and then in other cases, the temperature actually decreased or fluctuated over lengths of 50 kilometers or more—starting out cold, warming a bit, and then getting cold again.

“I think most people would say it’s not super surprising that there’s variability in these patterns,” says Torgersen. “But at this broad scale to see some of these odd-ball patterns was kind of a humdinger. We just know a lot less about river temperature than we do about air temperature.”

That’s why the mapping of water temperature in this study was so valuable. Most mathematical models of stream temperature, while largely accurate, aren’t able to account for fine-scale variations. Yet there are so many factors that can impact temperature variability, says Fullerton, such as tributaries, groundwater and nearby vegetation, or even coastal fog deflecting solar radiation. So this research provides a crucial perspective for what is actually happening in the water—and, ultimately, how those variations impact all of the species depending on the stream.

Fullerton says an important caveat with these findings is that the researchers only studied a snapshot in time. Their data came exclusively during the summer, so they weren’t capturing temperatures during different seasons, or overnight.

Christian Torgersen

For terrestrial ecology, the same paradigm would have you assume that as you go up a mountain, the air generally gets cooler. “If it got warmer as you went up, you’d know there was something up,” says Professor Christian Torgersen.

Still, these results can already reshape how researchers think about stream restoration projects, and how they determine the “natural” template for a certain section of river. It will be vital to examine the broader context of any stream segment—what’s happening directly up- and downstream, or along the riverbanks—to get the most comprehensive and accurate reading.

The diversity of thermal habitats in these streams, moreover, could be good news for the long-term survival of existing species, especially salmon. It appears that species may already be accustomed to navigating through a variety of conditions, and coping with a range of temperature tolerances, which could make them more resilient and less susceptible to future land use and climate impacts. “That’s going to help them with whatever comes next,” says Fullerton.

An important aspect of the climate analysis came from Professor Lawler, who is a coauthor on the paper. He played a key role in developing the approach for comparing patterns of water temperature among streams. “He was essential as a reality check to make sure our assertions were valid,” says Torgersen. “He also helped us couch these results in the context of climate change, and what the implications of this work are for understanding how species respond to a warming climate.”

Next Steps
Fullerton has worked as a research scientist with the National Oceanic and Atmospheric Administration (NOAA) since 2002. Now into her fifth year of doctoral study—working with Torgersen and Lawler—she can’t wait to dive back into the data and expand their analysis.

This first paper focused on a broad-scale perspective, and the next step is to key in on a finer scale and begin to look at how these spatial patterns might be affected by climate change, and therefore might affect the vulnerability of salmon. Specifically, the researchers will be quantifying the size, location and distance between cold water patches that salmon use, and considering how those patterns might change under future climate scenarios. After that, a third component of this research will be to look at drivers of these patterns, and whether we can predict where colder patches will occur in the landscape.

Which is to say, there’s much more to come from this exciting research, which has already challenged a number of long-held assumptions. “My hope is that stream ecologists will be reading this paper and then teach students that you can’t assume the temperature will increase,” says Torgersen. “It could change the way people think about basic stream ecology questions, and how to develop their models.”

Photos © Aimee Fullerton and Christian Torgersen.

Fullerton on a research trip to the Salmon River in Idaho in 2013.

Fullerton on a research trip to the Salmon River in Idaho in 2013.

ONRC Hosts Community Program on Tsunami Debris

Dock Removal

This dock, set adrift from Misawa, Japan, by the tsunami in March 2011, beached on a remote shore of the Olympic Peninsula this past December.

On Tuesday evening, March 19, the Olympic Natural Resources Center (ONRC) invited members of the Forks, Wash., community to a program about the marine debris washing up on nearby coastal beaches.

Some of the debris is a result of the devastating tsunami in Japan two years ago in March 2011, and speakers at the event addressed various angles of the disaster and its ongoing effects. Nir Barnea, regional lead for the National Oceanic and Atmospheric Administration’s (NOAA) Marine Debris Program, provided an overview of the tsunami’s physical impacts and efforts to track and respond to tsunami debris as it is dispersed across the Pacific Ocean. Coastal biologist Steve Fradkin from Olympic National Park, along with resource protection specialist Liam Antrim from NOAA’s Olympic Coast National Marine Sanctuary, then shared updates on the removal of a large dock that beached last December on a remote shore within the boundaries of both Olympic National Park and the sanctuary.

The dock—which measured 65’x20’x7.5’ and was kept afloat by 200 cubic yards of a Styrofoam-like material in its concrete holds—is currently being sawed up into manageable sections and removed by helicopter. It was one of three docks set adrift from Misawa, Japan, says Rainey McKenna, a public information officer with Olympic National Park.

Dock Removal

Crews work to saw the dock into smaller sections, which are then removed from the beach by helicopter.

The Olympic Coast National Marine Sanctuary is overseeing the removal project, and they are collaborating closely with Olympic National Park. A subcontractor, Undersea Company of Port Townsend, is handling the actual dismantling and removal of the dock.

Among those who attended the hour-long program were about 35 members of the Port Angeles and Forks communities, including Forks Mayor Bryon Monohon. In addition to learning more about the tsunami debris and removal efforts, attendees also got a chance to connect with the local work and research at ONRC.

Located on the Olympic Peninsula in Forks, ONRC is a research center with the School of Environmental and Forest Sciences at the University of Washington. The facility provides scientific information to address critical issues and solve problems concerning forestry and marine sciences in the region. It serves as a catalyst for interdisciplinary and collaborative work, bringing together expertise from forest resources and ocean and fishery sciences. By integrating research with education and outreach, it unites researchers, students, professionals and the public.

If you’d like to learn more about ONRC or the tsunamis debris event, please contact Ellen Matheny at 360.374.4556, or visit the ONRC site.

Photos of dock removal © John Gussman/National Park Service.