1. Project Title: Simmons College Sea Squirt Population Modeling
2. Participants (students and faculty)
The student participants in this research project are Nicole Richards and Elisabete Baker. Nanette Veilleux, Computer Science. and Mary Owens, Biology, will act as faculty advisors..
3. General Project Description
Computer Scientists, mathematicians and biologists have made significant scientific progress in working collaboratively. Many of the most exciting research of this new millennium, such as mapping the human genome, has come from applying computer algorithms and mathematical models to biological systems. In this project, we extend this approach by bringing computational models to the problem of modeling the complex parameters around what makes an invasive species successful.
In January of 2008, a group of scientists from Woods Hole Oceanographic Institute (WHOI) will be traveling to the Pacific coast of Panama in search of a group of creatures known as Sea Squirts. Sea Squirts, also known as Ascidians or Tunicates are filter feeders that have both an incurrent and an excurrent siphon and have small barrel shaped bodies which are covered in a tough layer called a tunic. In New England, Sea Squirts have been gaining notoriety as a recent invasive species which has begun to spread almost uncontrollably. This particular species, Didemnum sp. forms vast colonies resembling large rubber mats that can stretch for over 100 square miles. These colonies smother other filter feeders, disrupting both ecology and the economy of the industries that depend on the local shellfish. What is most concerning is that perhaps due to its highly acidic ph balance of 2, at the moment Didemnum does not seem to have a natural predator.
Until recently, Sea Squirts have received little attention. Woods Hole is hoping to change this with the recent research it has begun at the WHOI Ocean Life Institute – Liquid Jungle Lab (LJL) on Isla Canales de Tierra, located near the Pacific coast of Veraguas, Panama. The location of this lab was chosen for its pristine environment that results from its isolated and remote location. This location has allowed scientists to study many fauna in their natural and undisturbed state.
Dr. Mary Carman of WHOI and her team visited the LJL in January 2007 and surveyed the area around Isla Canales de Tierra and Isla Pacora, which are both located on the Pacific coast of Panama, in the southern Gulf of Chiriqui. This survey was the first step in creating a database that will eventually map the population of Sea Squirts along the entire Pacific Coast of Central America. Their return to Panama this winter will continue this recent effort, focusing on the area between the LJL and the North East coast of a neighboring island called the Isla de Coiba.
One member of this proposal’s research team, Elisabete Baker, has been invited to join Dr. Carman in Panama to search for, collect and document specimens of Sea Squirts. Elisabete is an undergraduate at Simmons College currently majoring in both Computer Science and Biology. Nicole Richards, the second team member, and also an undergraduate at Simmons College, will collaborate with Elisabete in developing a database of the information Elisabete collects. Nicole is currently majoring in Mathematics with a Computer Science minor and will use her experience in mathematical modeling to enable the team to develop computational models of the Sea Squirt population of southern Pacific Panama from this database. Using mathematical software and computer models such as decision trees, and their database to provide input parameters, the team will build computer models that predict the sea squirt population in a given area. The model will also make use of reproductive factors such as available food supply and pre-existence of other species in that region.
4. Specific Questions/Hypotheses (to be addressed)
WHOI has planned to focus this trip on answering the following questions:
1) How abundant and diverse are the Sea Squirts occurring in the area between LJL and Coiba?
2) Are there Sea Squirt larvae in the water column and are these larvae the same species that was previously documented?
3) How does the degree of development along the shoreline affect the current Sea Squirt population?
In addition to these questions, Elisabete and Nicole would also like to consider variables such as salinity, sampling methods, sub-strate, temperature, water depth, co-inhabitants of the areas where Sea Squirts are found and the presence (or lack) of natural known predators. They would also like to find out the carrying capacity of a sea squirt population under given conditions.
5. Methods
During the fall 2007 semester, Nicole and Elisabete will survey the research literature on known sea squirt populations and behavior. Using the parameters previously described, they will create potential models to describe the population behavior. During winter break, Elisabete will go to Panama for two weeks in January 2008 with WHOI to investigate sea squirt populations and collect data corresponding to the parameters. Through a combination of scuba, snorkeling and use of plankton tows, Sea Squirt data will be collected at various sites in the 10 mile target area between LJL and Coiba. Upon return from Panama, Elisabete and Nicole will input this data into a previously created database and consider modifications based on new information. During the spring semester, Elisabete and Nicole will analyze the data collected in January and identify which model(s) best describe the population behavior of the sea squirts. They will then create computer models that will predict the future behavior of Sea Squirt populations. Since Sea Squirt populations can increase either through migration or reproduction, both aspects will be modeled. The migration modeling aspect will consider changes in the input parameters and proximity to an existing population. The reproduction abilities of a population will consider current population density as well as availability of food resources.
For simplicity and to identify the most influential parameters, they plan to begin with a linear population model using least squares regression. Higher degree polynomials and exponential models will also be explored in this first attempt. A density-dependence model will be developed to address the question of ability to reproduce. Decision trees will be formulated to allow data with incomplete parameter sets to be modeled.
6. References:
http://en.wikipedia.org/wiki/Sea_squirt
Carman, MR, Molyneaux, SJ, Ji, R, Sievert, S, 2007, Ascidians of the southern Gulf of Chiriquí,Pacific-Panama: a native fauna at risk to bioinvasion: Marine Bioinvasions (submitted).
Tokioka, T, 1971, A new species of Rhopalaea from the Pacific coast of Costa Rica (Tunicata, Ascidiacea): Publ. Seto Mar. Biol. Lab. 19 (2/3), 119-121.
Tokioka, T, 1972, On a small collection of ascidians from the Pacific coast of Costa Rica: Publ. Seto Mar. Biol. Lab., 19 (6), 383-408.
Hickman,C.P, Animal diversity / Cleveland P. Hickman Jr., Larry S. Roberts, Allan Larson. – 3rd ed. p. cm.
Raloff, Janet. Dec. 24, 2005. Squirt Alert. Science News, Volume 168 (No. 26/27):p.411.
7. Impact on the goal of MRO-W
The participants are students at a women’s college and are deeply interested in the areas of biology, mathematics and computer science. One of the challenges in recruiting and retaining women in Computer Science is that women want to find meaningful applications for the theory. For Elisabete, that application is marine biology which she has independently pursued. For Nicole, that application is mathematical modeling of real world events. Both Nicole and Elisabete are considering graduate school in multi-disciplinary fields. This grant would allow them to pursue a self-initiated multidisciplinary CS/Math/Biology project in preparation for graduate school.
.
8. Student Activity and Responsibility
The student’s responsibilities will include finding resources about sea squirts and population modeling, developing population prediction models using decision trees, and writing and updating a weekly journal on a website documenting the student’s progress. After the Panama trip, the students will analyze the data pertaining to population size and adjust the most accurate model accordingly.
9. Faculty Activity and Responsibility
Professor Veilleux will meet with the research group on a weekly basis to set goals for the week, provide resources. In addition, she will be available to provide guidance on experimental and algorithmic development and troubleshoot on an ad hoc basis. Professor Veilleux will also help disseminate the results of the research project as well as the research experience with regards to undergraduate learning.
10. Timeline
Phase 1: Summer 2007 / Fall 2007
Phase 2: Spring 2008
Phase 3: Summer 2008
11. Budget
|
STUDENT EXPENSES |
|
|
per student |
3000 |
|
total, 2 students |
6000 |
|
summer travel (Panama), 1 student |
4000 |
|
MENTOR EXPENSES |
|
|
conference travel: sigse |
400 |
|
miscellaneous |
100 |
|
TOTAL |
13500 |
Justification:
Both of the research students expect to contribute about 12 hours each week on the project. Since these students are carrying a full course load, this is a reasonable time commitment. The stipend is based on an approximate hourly rate of $8.30, close to the prevailing wage for undergraduate research assistants in Boston. The summer stipend will allow Elisabete to travel to Panama with the WHOI team in Summer 2008. The faculty mentor plans to document the pedagogical as well as research experience of this project and present the pedagogically related findings at an appropriate conference ($400 support). In addition, $100 is budgeted for miscellaneous copies, supplies and computer software. Although the software resources we will need (database, decision tree) are generally much more expensive, we intend to leverage open source and Simmons College resources where possible.
12. This project is connected to the WHOI
efforts to study the Sea Squirt. Funding would allow Elisabete Baker to assist
the WHOI with their field collection since she is not otherwise funded for this
activity. Any results from this work will be shared with the WHOI investigators
and Dr. Carman has kindly agreed to provide input.