On a warm June day beneath the waters of the Choptank River on Maryland's Eastern Shore, oysters on one of the river's last remaining reefs begin to spawn. The male's shell parts slightly and a white thread of sperm issues forth from the gap in a steady stream. Nearby, a female oyster raises her shell and brings it down with a sudden clap, a pulse of whitish eggs puffing out. She claps again. Pretty soon, neighboring oysters join in, clapping their shells in unison, turning the water milky white with maybe billions of eggs and sperm. A single female may release as many as 25 million eggs during a single spawn.
When the clapping subsides, the clouds disperse. The now-fertilized eggs divide again and again. Soon they sprout hairlike cilia and begin a microscopic journey. If the larvae survive tides, currents, let alone a score of predators, they will change shape and begin to make active decisions about where to swim. After two weeks, these tiny animals will begin to scout out an oyster reef on which to attach permanently and transform into adults.
How far will the larvae travel? How many will find an oyster bar on which to settle and begin adult life? How many will die before reaching one? With millions of larvae no larger than a pencil dot, answers to these questions lie beyond the reach of the human eye.
So how can one follow larvae on this unseen journey, a task critical to predict whether oyster populations can once again thrive in the Chesapeake Bay? Mathematical models may be able to take over where the eye leaves off, translating years of laboratory and field observations into equations that account for the major forces at work — currents, tides, and larval behavior.
With a few deft keystrokes, scientist Elizabeth North calls up a schematic map of the Choptank River on her computer screen — now she fills it with clouds of blue dots, simulated oyster larvae spread throughout the river. Small irregular shapes on the map represent oyster reefs, settlement targets where larvae will begin life as adults.
North's fingers play over the keyboard and the virtual larvae lurch into motion. Blue dots slosh back and forth on the screen, subject as they are to forces that numerically mimic the tides. The clock at the top of the screen ticks forward rapidly — six hour tidal cycles advance in a matter of seconds. . .Day 1. . .Day 2. . . Still the blue dots slosh back and forth in this computerized Choptank. . .Day 9. . . Day 10. On Day 14, some dots suddenly turn green and stop moving. The larvae are now mature enough to settle if they encounter suitable habitat. The pressure is on. The larvae's genetic code dictates that after they become competent to settle — a life stage called pediveliger — they must find substrate within another 7 days. If they fail to find a place, they cannot metamorphose. They will die.
By Day 21, the sloshing stops. The larvae have met their fate. On North's screen, larvae that have successfully settled stay green, while the dead oyster larvae turn orange, rendering the virtual Choptank a patchwork of color.
This "settle or die" oyster drama will play out over and over again on her computer as North, a biologist and mathematical modeler, runs model simulation after simulation. From her quiet, uncluttered office on the shores of the real Choptank River, at the University of Maryland Center for Environmental Science (UMCES) Horn Point Laboratory (HPL), North first gives the blue dots the behavioral traits of the native oyster, Crassostrea virginica, derived from published data accumulated over years of scientific study and recent laboratory experiments. Then she will run the same scenarios again with the behaviors of the non-native Asian oyster, Crassostrea ariakensis.
As though comparing the performance of two cars, North's model serves as a tool to test drive the two species of oyster, projecting whether one species can reach the finish line (i.e., settlement on an oyster reef) more successfully than the other. Her work will help show which species might better repopulate the Chesapeake with sustainable oyster populations. During the half century from 1920-1970, oyster populations held steady while supporting a profitable — and sustainable — fishery for watermen and oyster farmers. A return to such levels is currently held as a restoration target for an ongoing Environmental Impact Statement (see Assessing a Potential Introduction).
North's steady presence in front of the flat computer screen reveals no hint of the contentious nature of the debate that swirls at the heart of her work. The outcomes of North's model — maps that predict where larvae of the two oyster species will settle — will likely play an important role in the decision about whether to introduce the fast-growing, non-native Asian oyster to Chesapeake Bay. On one hand, the Bay's oyster industry hangs by a thread — Maryland's harvests alone have declined by more than 90 percent from 1970s levels. Each year more watermen abandon their heritage for more economically sustainable work, while only a handful of shucking houses remain. Equally important, the filtering prowess of oyster populations could help reverse seasonal oxygen depletion and turbid waters, helping to renovate the Bay's damaged ecology. Many have argued that the Asian oyster might have the capability to do just that.
But a decision to introduce the Asian oyster may be a risky one. This species could not only invade areas outside of the Chesapeake Bay, it could also bring new shellfish diseases to the region. In addition, the Asian oyster could outcompete the native oyster for already-diminished reef habitat, which might deal a final deathblow to its restoration. Furthermore, there are no guarantees that such an introduction will even work to bring oysters back to the Bay or clear up its murky waters. Much of its promise is based on preliminary research and extrapolation from studies of its biology in other regions. North's computer simulations, which compare the settlement patterns of the native and non-native species, will help provide some of the first predictions of the potential for sustainable oyster populations in the Chesapeake Bay.
North works in the midst of this controversial spotlight. She recognizes that her research will provide tools to environmental managers directly, a rare and exciting opportunity to serve as a bridge between science and policy. But the glare can be intense. With policy makers in Maryland and Virginia awaiting the results of research from her and other oyster scientists in the region, she works rigorously and as quickly as possible. Her project is one of 12 funded by the Maryland Department of Natural Resources to address urgent scientific questions about the non-native oyster to help assess potential risks posed by an introduction. And whatever the final decision by the states about whether to introduce the Asian oyster to the Chesapeake Bay, North knows that the outcome of her modeling efforts could someday land at the center of a controversial debate.
Master of a virtual Choptank River, Elizabeth North turns hours into seconds and days into minutes, tracking the ebb and flow of oyster larvae as they search for a place to settle. North's models form part of an elaborate process aimed at predicting the survival of both native and non-native oysters in the Chesapeake. Photograph by Skip Brown.
As a young researcher, North is grateful for collaborations with colleagues Raleigh Hood, Ming Li, and Liejun Zhong of HPL, and Tom Gross of the National Oceanic and Atmospheric Administration & Chesapeake Research Consortium, and she welcomes the tough scrutiny of academic peer review every step along the way. Peer review will help ensure that her work is of the highest caliber and it will insulate her science against potential political jostling down the road. For now, North stays focused on providing decision makers with the best information possible.
The path North followed to her current place in the scientific high beams derives from a lifelong connection to the Bay. She grew up on the shores of the Severn River, "catching yellow perch and then not catching yellow perch" when major fish kills occurred in the Bay during her elementary school years. She attended ecology camps in the summer run by the Chesapeake Bay Foundation and interned at the National Aquarium in Baltimore. Her father, a physician, taught her to fish. Her mother, an artist, taught her to identify marsh plants.
As a college student at Swarthmore, North studied comparative religion, with some biology classes along the way. She wanted to learn about differences and commonality in the human experience. As it turned out, studying religion prepared her well for studying science. Both, she says, offer a framework, a structure for understanding the world.
After college, North started down a path of long-time interest. She took a job in Annapolis with the Chesapeake Bay Office of the National Oceanic and Atmospheric Administration. She went on to work for the Environmental Protection Agency Chesapeake Bay Program in Solomons and to pursue a master's degree in environmental policy at Johns Hopkins, becoming deeply aware of the need for good science to support sound resource management.
Her interest in applied research next led her to pursue a Ph.D. with UMCES fisheries biologist Ed Houde. At the Chesapeake Biological Laboratory, she focused on physical oceanography, studying how the Bay's complex water circulation affects the distribution of fish larvae in the Bay. She spent long hours on research vessels and peering through a microscope, honing her knowledge of the Bay's intricate biology. As she went on with her research, North realized that mathematical modeling would provide a valuable tool to help link her observations in the physical and biological domains, to "visualize the world" in a way that would be useful to fisheries managers and other decision makers.
North became fluent in the language of modeling through a post-doctoral fellowship with UMCES researcher Raleigh Hood, a biological oceanographer at Horn Point Laboratory who uses mathematical modeling to study algae and primary production in ecosystems around the world. She later accepted a faculty position at the Horn Point Laboratory — a rare occasion to remain at the institution that trained her. For North, this job was the ideal chance to still further strengthen her link to the Bay, an opportunity to continue crabbing and fishing on the Choptank with her husband Tim, a research vessel engineer who used to tong for oysters in the fishery's more prosperous days.
Through her research program, North tries to link the Bay's physical environment to its biological resources, combining modeling, field, and lab-based approaches to studies of blue crabs, underwater grasses, and oysters. Models, she knows, provide just one tool of many, an attempt to visualize the complex network of relationships in the Bay, making the real world easier to understand. But the right tool must match a specific problem, North is careful to point out. "If we had only one tool for every project," she says, "there wouldn't be Home Depot."
A few more keystrokes from North's slender fingers and a new screen pops up: a blank graph stares back, waiting for her to execute a subsection of code that accounts for the different larval behavior of the two species. As native larvae mature, they tend to cluster above the salt barrier (halocline) that cleaves the Bay in two layers: a buoyant, less salty layer of river water flowing seaward and a layer of dense, saltier ocean water flowing upriver. But non-native C. ariakensis larvae stay low and hover within one meter of the bottom, according to new experiments by oyster researchers Joan Manuel, Roger Newell and Vic Kennedy, also at Horn Point Laboratory.
North believes that these differences in behavior could strongly affect which parts of the Bay these two species will populate. For example, if non-native oyster larvae hang close to the bottom, they may ride bottom ocean currents up the estuary. If larvae hover in the surface water as do the native oyster, they might go down estuary, she explains. For her model to accurately simulate a virtual larval journey, she must capture these differences in behavior.
When North starts the model clock running, the two virtual oyster species behave as she expects. The simulated native oyster larvae float up above the salt barrier, five meters off the Bay's bottom. The non-native oyster larvae stay low. Day 1. . .Day 2. The model seems to be working well. Day 14. . . . The larvae are now biologically competent to settle.
On Day 15, however, North encounters a problem. All of the larvae of both species freeze in place. She recognizes this as an error, perhaps a bug in her computer code, perhaps a problem with the boundary conditions that keep particles from jumping out of the virtual water onto land. At this stage, the simulated larvae should have had at least another seven days to swim around looking for suitable habitat. She knows that she'll diagnose the problem, but needs to find it fast. She wants to present this portion of her model results at an upcoming scientific meeting.
With another series of rapid keystrokes, North calls up the screen that masterminds her model, filled with code that to the untrained eye might as well be hieroglyphics. Leaning forward, she scans the language intently, proofreading and editing in an attempt to pinpoint the source of the problem.
In many ways, North's work as a modeler is much like that of a writer. She writes in the language of mathematics, but the actual syntax is the computer code Fortran. She weaves together themes with a complicated architecture of concepts to recount a classic epic journey, a coming-of-age tale of sorts. Her model uses mathematics to reflect the story of an oyster's search for a place to start life. Not just one oyster protagonist but a cast of hundreds of thousands of each of the two species.
As complicated as the plot of a complex computer model can become, the larval transport model developed by North and her colleagues dramatically simplifies ecological reality. For example, the model cannot track more than 100,000 oyster larvae in one run, while in the real world, a single oyster in the Bay can release millions of eggs, explains North. The challenge of modeling is to represent reality as accurately as possible while dealing with necessary limitations of available data and computer power, she says. The model must be strategically simplified to maintain realism yet complete simulations within a reasonable time frame, she continues. "If we don't simplify, it will be 2050 before we have an answer."
If the craft of a modeler can be compared to that of a writer, then the model itself could be considered an elaborate work of literary nonfiction. In the case of the oyster model, this work would read like an epic in two parts. North's larval transport model would be Volume I, the story of oysters coming-of-age. Volume II would follow the oyster and its progeny and its progeny's progeny ten years into the future. Other scientists will write this tale, technically known as the juvenile-adult demographic model.
In the opening chapter of Volume I, hydrodynamics — the Chesapeake's currents and tides — drive the plot. These forces determine the large-scale movement of oyster larvae of two species (native and non-native) over a three-week period, from spawning to settlementÂ.
After hydrodynamics set the stage in Chapter 1, the oyster larva emerges as the central character of Chapter 2. Here a particle-tracking model takes information from the hydrodynamic chapter on currents and salinity and projects where in the Bay larvae will move during their journey. Though larvae in the wild begin to swim vertically, the comÂputer's particle-tracking routine at this stage treats them as passive particles, entirely at the mercy of water, wind and waves.
By Chapter 3, however, the model begins to account for oyster biology and the larvae develop depth and complexity. Mimicking real life, they are no longer passive particles, but acquire attributes of age, swimming speed and behavior. That is, virtual larvae are now able — like real larvae — to direct their movementsÂ.
Chapter 4, the denouement of the larval oyster's "settle or die" drama, brings all of the plot lines together to make predictions about the potential distribution of larvae, both the native and non-native species. The pieces (hydrodynamics, particle-tracking and behavior) link together mathematically to generate maps of the Bay that forecast the distribution of each species (see graphic on page 8).
Later, North's maps will feed into another model developed by her collaboratorsÂ, statistician Mary Christman of the University of Florida in Gainesville and quantitative ecologist Jon V¿lstad from Versar, a science and technology consulting company. Volume II is a sequel of sorts. This so-called juvenile/adult demographic model will make predictions about what will happen as the oysters grow, reproduce, and die over the next ten years — projecting populations of the two species into the year 2015.
The outputs of Volumes I and II of the epic — the larval transport and the juvenile/adult demographic model — will generate maps that predict the potential distribution and abundance of the native and non-native oyster in the year 2015. These results will feed directly into policy makers' evaluations of the different restoration scenarios, providing one tool of many to assist them in making a final decision (see When Science Meets Policy).
It's 5 p.m. on Monday. The day doesn't usually end so early for North, who puts her computer to sleep, picks up her gym bag, and leaves her office. This is the one day each week that she leaves work behind at a reasonable hour. She drives over to the Aquaculture and Restoration Ecology Laboratory, a new building on the Horn Point campus, to teach a class. She moves the tables out of the spacious lobby so that her students will have space to spread out across the floor.
Soon a small group of regulars arrive in the lobby and take off their shoes and socks. Several older women join a couple of graduate students at the lab for North's weekly Tai-chi instruction, their camaraderie evident as they fill each other in on the week gone by. Sitting in a circle on the floor, North begins to lead the group in a series of warm-up stretches, limbering up for the challenging poses to follow.
Class begins and North guides her students through a sequence of moves that they learned in the last session. She concentrates intently on her own form and balance, while providing tips to the class on how to improve technique. To become fluent in the full practice of Tai-chi can take years of study and practice and most of her students have only recently begun.
North's own practice of this ancient Chinese art form has evolved over the past 19 years, drawing from her skills in dance and her interest in Eastern philosophies, born of her studies of comparative religion. She takes Tai-chi, a powerful tool for mind-body relaxation, very seriously, participating in retreats and classes taught by masters of the art whenever possible.
When the class finishes the steps that they know, North goes through the complete sequence of 108 exercises on her own, while her students watch her form carefully. For this moment at least, North's mind and body are far away from models, oysters, and the pressures facing a young scientist in a political spotlight.
She anticipates the day when the spotlight may sharpen its glare in her direction, when the states of Maryland and Virginia issue the final decision on the oyster Environmental Impact Statement. She braces for the maelstrom of clashing worldviews that could hit, whatever the outcome. But for the most part, North works to make the larval transport model as iron clad as possible. She also reaches out to colleagues for advice and builds support in the academic community for her modeling efforts through seminars and presentations at national meetings.
Now that the Department of Natural Resources has provided updated maps of currently available oyster habitat, North can begin the final runs of her model. Her computer will run day and night to generate maps that show where the two species of oysters could distribute in the Bay to feed into the projections of the demographic model. Soon DNR will ask North to present her results at their headquarters in Annapolis in what will be the fourth in a series of public meetings, held to keep the EIS process transparent and open to all interested stakeholders.
North knows that communicating the idea of "model as tool" to the public could be challenging. "People get angry at weathermen when the 7-day forecast is wrong and this is a 10-year playing field. These are not predictions of what is going to happen," she says, only what could happen. She also realizes that her findings will likely face intense scientific, public, and possibly political scrutiny.
"It is scary and it is great. I like being involved. I like the idea that the tools I am developing are going to be useful," North says.
She has spoken with other scientists about how to insulate herself from the high profile nature of the EIS project. A respected colleague advised her first and foremost to publish her model expeditiously in the academic literature, to vet it through the peer review process. "If this ends up in court, which it very well could," North recounts, "published papers will be important for credibility."
North has already written the framework for the manuscript she wants to publish. An outline sits in a file folder on her desk. She needs to complete the final model runs before she can write the Results and Discussion sections. But on the same day that she delivers her final report to the Department of Natural Resources, North plans to drop the manuscript in the mail.
For More Information
Elizabeth North's Web Page
northweb.hpl.umces.edu/
Maryland DNR's Oyster InFocus
www.dnr.state.md.us/dnrnews/infocus/oysters.asp
Maryland Sea Grant Oyster Node
www.mdsg.umd.edu/oysters/
National Academy of Sciences Report on C. ariakensis
www.nap.edu/books/0309090520/html/
Chesapeake Bay Program Scientific and Technical Advisory Committee information on C. ariakensis
www.chesapeake.org/stac/ariakensis.html
www.chesapeake.org/stac/stacpubs.html