Modeling Around Disease

Simulation models can give managers the ability to assess the effects of different environments and different restoration strategies on the virulence of oyster disease

What affects an oyster's growth, its reproductive success, its chances for survival?

"You cannot set management strategies based simply on what you see this week or what you've done in the past."

The natural set of new oysters in Chesapeake Bay reached record numbers in 1997, surprising scientists and managers alike. That was the good news -- the bad news is that the outlook for these young oysters is not promising. Why? Because the oyster diseases Dermo (Perkinsus marinus) and MSX (Haplosporidium nelsoni) can be so virulent that young oysters may not survive the three years it generally takes to reach market size.

To help head off this bleak prognosis, both resource managers and commercial growers are looking for better forecasts of year-to-year variations in disease. This means new computer models -- specifically, a dual-disease simulation model that will mimic the dynamic ebb and flow of MSX and Dermo in the estuary and their impact on oyster populations. In the long run, the goal is to help rescue susceptible oysters that would otherwise fall victim to Dermo or MSX.

Disease Virulence and the Environment

Just how virulent disease is likely to be each year depends on a suite of factors, ranging from environmental and climatic conditions to direct human impacts -- commercial harvesting, for example, or large-scale movement of oysters and shell by resource management agencies.

Even oysters specially bred to tolerate both MSX and Dermo (see "Breeding Disease Resistance in the Hatchery") may be more vulnerable given certain environmental conditions. Whether the goal is commercial production or large-scale restoration for ecological purposes, the need to manage around disease virulence has become paramount. Simulation models that can give managers the ability to assess the effects of different environments and different restoration strategies on the virulence of oyster disease are likely to play a key role in restoring oysters to estuarine waters.

Already models developed by Eileen Hofmann and John Klinck at Old Dominion University, in collaboration with researchers Eric Powell and Susan Ford at Rutgers University and Steve Jordan at the Maryland Department of Natural Resources (DNR), are making it possible to assess the relative outcomes of disease virulence and pressure under varying conditions -- outcomes that are not necessarily intuitive, says Jordan, director of Maryland DNR's Cooperative Oxford Laboratory.

What lies behind these shifts in patterns of disease? What affects an oyster's growth, its reproductive success? Its chances for survival?

Planting Shells and Moving Oysters

In most estuaries there are regions that from year to year will receive heavier sets of new oysters than other areas. This occurs for a number of reasons, among them, salinity levels and circulation patterns that may entrain both newly-spawned oyster larvae and the algae they feed on. Resource managers and growers try to take advantage of these natural sets by placing large volumes of clean shell (called cultch) in these areas to provide more surface area for setting oysters.

In Maryland, for example, the Department of Natural Resources dredges thousands of tons of fossil oyster shell annually for planting. Once spawning season is over and the free-swimming oysters have attached themselves, metamorphosed and hardened (as spat), the seed oysters are hauled up and moved to public grounds throughout the Chesapeake for growout. Waterman can harvest oysters some three years later -- except that many disease-ridden oysters do not live that long. Since the early 1990s, Dermo has decimated entire populations and before that MSX sporadically entered the Bay, especially in areas of high salinity, beginning in the late 1950s in Virginia.

In Delaware Bay, the state resource agency works with private operators to move seed oysters from public grounds that receive good natural sets of new oysters to leased grounds in high salinity waters. As Eric Powell points out, "the oysters grow more rapidly to market size there because food supply averages are greater, and high salinities are biologically more favorable to growth."

Before the spread of MSX in the mid-1950s, Delaware growers left oysters on the high salinity leased grounds for more than a year. But MSX began killing oysters before they reached market size -- it forced the industry, says Powell, to move oysters from higher salinity leased grounds after only a few months. "Unfortunately," says Powell, "while this procedure limits mortality, it also limits growth."

Nevertheless, the approach was generally successful, and growers were careful to move oysters from low salinity grounds in May or June, and then to harvest them from September to December.

Beginning in the 1990s, however, Dermo invaded Delaware Bay and competed with MSX as a killer of adult oysters. Practices that had worked to combat MSX were no longer working for both diseases, and oysters were succumbing between the May-June transplanting and the fall harvest.

Were growers at a dead end? Or were there management alternatives? According to Powell, simulation modeling in Delaware Bay had shown that the spring bloom of algae was extremely important for the growth of oysters to reach market size. Powell notes that the degree to which a spring bloom can be used by oysters may be "crucial" in determining the success of transplanting "when survival is limited by disease." It is such factors that Hofmann and her colleagues wanted to capture in developing mathematical models.

Modeling the Development of Disease

Hofmann originally developed a model for Galveston Bay, Texas to help resource managers predict potential impacts from dredging operations and resulting changes in freshwater flow. The model projected the impact of salinity changes on oyster growth and on the prevalence and intensity of Dermo infection. It did not take into account the impact of MSX, which, though a scourge in the Mid-Atlantic, is of less concern in the Gulf of Mexico.

To adapt the Texas model to the Chesapeake and Delaware bays, accounting for the effects of both Dermo and MSX, Hofmann needed detailed field information such as the prevalence of MSX spores during the winter on different sized oysters under varying salinities. Support from the Oyster Disease Research Program helped researchers gather that data, and led to Hofmann's collaborations with scientists in Delaware and Maryland to take on the heady work of adapting the original model.

"We have one of the best data sets on the recruitment of new oysters and disease," says Steve Jordan. In place since 1990, the Maryland monitoring program includes measurements on MSX infection during different seasons and over a whole range of oyster sizes.

"We couldn't have done our model without that long time-series of data," says Hofmann. In fact, because of the information that the model requires, Maryland DNR has modified its monitoring program. Employing electronic positioning systems that make use of satellites to give highly accurate location information, Maryland has aligned its oyster monitoring with the Chesapeake Bay restoration program's water quality monitoring. "They were never coordinated because they were done for different reasons," says Jordan. "Our collaborative project spurred us to do this."

Handtonging for oysters has been a way of life in coastal waters on the Atlantic and Gulf coasts for more than a century.

Models and Management

While oystermen in Delaware Bay were already working with Powell and Susan Ford to identify the best period for moving spat from seed grounds to leaseholds, the model shed new light on the all-important element of timing. "With the model," says Powell, "we looked at whether it is better to transplant oysters in spring or fall, given different circumstances, including the presence of disease."

What would be the differences in growth and survival, the researchers asked, if oysters were transplanted in November, rather than the following spring? What if they were planted in December or January or February?

Powell and his colleagues ran a series of simulations to analyze the role of disease and predators in determining the success of transplanting, and compared them with simulation results in which oysters were harvested directly from the seed beds.

Those simulations showed that transplanting in May resulted in the lowest harvest yield, while transplanting in November led to a high harvest the following August. "The reason," says Powell, "is that oysters apparently get the benefit of the large spring bloom of algae in the higher salinity waters." If, however, oystermen delay harvesting until late autumn, when prices are higher, the model shows that they would also be faced with considerably higher mortalities.

Waiting for an autumn harvest is especially risky, Powell says, if the principal source of mortality is Dermo disease rather than predation. Ultimately, a decision about when to harvest cannot be made by the model alone. Powell points out that oyster growers must balance "the increased price in the autumn with the increased loss through predation and, particularly, disease." Simulation models will simply help them understand the odds.

Looking Forward

An unexpected outcome of Hofmann's simulation model is that it is also helping researchers track down how MSX is transmitted. Unlike Perkinsus which releases spores that can then be filtered by nearby oysters, the means by which MSX infects other oysters remains a mystery. Though they have not yet discovered it, scientists have long thought there is an MSX "carrier," an intermediary host that causes infection.

In fact, for the simulation model to work, says Hofmann, "we had to put one [an MSX carrier] in so that we could reproduce what we were seeing in the field," though she cannot say what that carrier is.

The model results suggest it is an organism that has a relatively short life span, one that responds to variations in salinity.

Using Hofmann's information, researchers at the Virginia Institute of Marine Science (VIMS) have been searching for secondary carriers, employing molecular techniques developed through the Oyster Disease Research Program that enable them to sift through many microscopic organisms in Chesapeake Bay (see "Diagnosing Dermo and MSX"). "We have already found positive samples" says Burreson, meaning that genetic material from MSX is present in the sediment and water column. "Whether these are free spores or developmental stages [of MSX] we don't yet know," he says. Burreson's team is proceeding with its analysis.

Beyond the capability for projecting outcomes of different transplanting and shell planting strategies, Hofmann's simulation models have much broader implications for resource management of shellfish. "They demonstrate how important climate is in regulating diseases such as Perkinsus and MSX," says Hofmann. With the apparent warming trends globally, Perkinsus marinus has been extending its range -- where its northern limit was once Chesapeake Bay, Dermo disease is now being detected as far north as New England.

"We have to manage the disease populations with a long-term climate perspective," Hofmann says, "which means that you have to be aware of such occurrences as an El Niño or other climatic effects -- you cannot set management strategies based simply on what you see this week or what you've done in the past." Factors such as changes in freshwater inflow to estuaries -- which control salinity and vary year to year – have a huge impact on oyster reproduction, on survival and on disease. Oyster bars that historically have shown large natural sets of new oysters may well be affected by shifting climatic conditions, which can affect water circulation patterns.

The future is challenging us to develop flexible management approaches in all of our fisheries," says Hofmann, approaches that can take into account year-to-year changes in environmental conditions and the effects they have in ecological food webs -- with such approaches as the dual-disease simulation model, she says, we are developing the kinds of tools we need to meet those challenges.