Recent Research Projects
Evolving Life History Diversity in Chinook Salmon
Fall run Chinook salmon in the Snake River of Idaho are changing in response to human influences on their habitat. The population is restricted to 15% of their historic habitat, with 8 dams between their spawning area and the ocean. Dams upstream have changed water conditions, predation, and survival in spawning, rearing, and overwintering locations. While juveniles used to hatch in early Spring, grow quickly, and leave for the ocean by late Summer, many fish now leave for the ocean as late as the following Spring. Aspects of this delayed migration appear to be heritable and selected for, making this an example of recent life-history evolution.
Using the ongoing changes in Snake River fall Chinook salmon this project aims to investigate how the combination of location and environment create conditions that select for diverse life-histories. Using a dataset spanning over a decade I use otolith microchemistry, stage-structured and mixed-effects modeling, as well as paired growth and environmental data, to investigate the spatially explicit drivers of life-history strategy within the population.
This project has been very productive, resulting in four peer-reviewed journal articles in the Canadian Journal of Fisheries & Aquatic Sciences, Oecologia, Chemical Geology, and Heliyon investigating aspects of salmon life history, otolith chemistry, and analysis techniques for otolith data. Two additional manuscripts are in-review with Ecology of Freshwater Fish and Ecosphere. A third manuscript, my final PhD chapter, is in-prep.
Pictures from this Project
Migration of Amazonian "Goliath" Catfish
Several species of so-called "goliath" catfish in the Amazon basin make migrations from the mouth of the Amazon River to the foothills of the Andes Mountains. The longest of these travels 5,788 km, the longest freshwater migration on earth. To put this on a human scale, this is the same distance as driving a car across the United States from Seattle, to Chicago and then to Key West, Florida. Until recently this migration was only inferred from fisheries data, the rivers were too long, too big, and too muddy to collect detailed data about this migration. But, dams being built in the Amazon threaten this migration, making understanding this migration very important.
Working entirely by email and videochat I initiated a collaboration with Dr. Tommaso Giarrizzo, a professor in Belém, Brazil at the mouth of the Amazon River, to study these incredible fish. Using otoliths collected from fisherman in the central and eastern Amazon we were the first to show individual migratory movements in three species of "goliath" catfish, B. rousseauxii, B. vaillantii, and B. filamentosum. Subsequent studies have confirmed our finding that tmultiple life-histories exist, including a likely non-estuarian life history. This diversity may be important to the resilience of the population to fishing and anthropogenic change.
Our paper in PLOS One was one of the first to use strontium isotope ratios in Amazonian fish. It provided important information on the species as well as providing evidence that strontium prediction from geologic maps (see next section) could be useful in locations were water sampling was sparse. This collaboration also spurred a successful crowdfunding campaign to fund further collaboration with Dr. Giarrizzo, which has expanded into study of largetooth sawfish and Atlantic tarpon on the Brazilian Amazon coast (see below).
Pictures from this Project
Isoscapes & Predictions from Bedrock Geology
The isotopic ratio of strontium (87Sr/86Sr) has become a powerful tool for understanding movement and location across species. It's uses are broad, from determining immigration in human and animal populations in archeaology, to understanding the ecology of bird migration, to uncovering where heroin and marijuana were grown. But, these methods only work if you can compare the isotopic signature to known signatures on the landscape. Developing ways to predict strontium ratio across the landscape to create maps, called "isoscapes", is an important goal in the field to improve the accuracy and usefulness of strontium ratio methods.
This project used data from geologic maps to predict strontium isotope ratio within the range of Snake River fall Chinook Salmon in Idaho. These predictive methods are much simpler to implement, using easily available GIS methods and simple statistical regression models, than other methods. They provide a way for reseachers to easily extend the spatial resolution of strontium ratio studies when sampling is sparse, to predict the location of unknown populations, and to determine whether proposed studies will be possible given the geolgic variability in the study area.
Results of this work was first published in Chemical Geology in 2012. These methods were also used to extend the resolution of the strontium ratio study used to reconstruct Amazonian catfish migration.
Sonification as a Method for Data Exploration
When you think of data, most likely, you think of visual representations...graphs and infographics that we use to explain the meaning in data. But, our ability to visually understand large amounts of data at once is limited. Too many lines on a graph and things just get confusing. Our sense of hearing, on the other hand, does an excellent job sorting out important information from simultaneous sources. Think of our ability to converse in a crowded room, or our ability to hear a wrong note, or a drum out of rhythm, amongst all the sounds of a band or orchestra. This property of hearing makes it potentially powerful as a way to explore and analyze large datasets.
Our ears are particularly attuned to data that moves across time (time-series). Otolith chemistry data is, at it's core, a time-series showing the changes in chemistry through the life of a fish. Further, our current statistical methods have a hard time analyzing this data without chopping up the time series, in the process losing much of the information that the sequence of data holds.
Working with two musicians, Dr. Jonathan Middleton and Ben Luca Robertson, in a cross-disciplinary team, we translated my otolith data from Snake River fall Chinook salmon into meaningful sounds (sonification) and tested peoples ability to understand the underlying data. This sonification procedure. The sonification was built based on our knowledge of how humans interpret sounds, so that the timing fish movement through the Snake River could be explored for many fish at once, a process explained in a peer-reviewed proceeding for the 12th Annual Sound and Music Computing Conference in Maynooth, Ireland in 2015. We then tested peoples ability to identify fish movement from the sonification with a perceptual study using human volunteers.
The results of this study, published in Heliyon, were intriguing. Individual test-takers all identified movement differently, some identifying many subtle movements and others only a few. But, the data from all the test-takers together was very accurate...indicating that crowd-sourced explorations of big datasets might be useful. It also confirmed a prior study showing that people are less accurate when the sonification is accompanied by a visualization of the data. Our study provided evidence that this may be because it requires more mental processing to interpret both the visualization and the sound, while most of the necessary information is already contained in the sound alone.