Scientists work on math model to better predict spread of CWD

A new, more accurate way to forecast the growth and expansion of chronic wasting disease (CWD) in whitetail deer also could help predict the spread of invasive species such as zebra mussels or the Ebola and Zika viruses, according to researchers.

The forecasting tool presents the probability of CWD infection in a heat map-like format, with darker "hot spots" showing where the disease is growing at a faster rate in a study area in southern Wisconsin, said Daniel Walsh, a U.S. Geological Survey scientist based in Madison, Wis., and an author of a paper describing the newly developed method.

In Minnesota, this type of modeling will be more valuable as the state collects more data, said Chris Jennelle, a research biologist with the Minnesota Department of Natural Resources. That may result from a special January 2018 hunt that the DNR has proposed in three southeastern deer permit areas and a special zone established last year after CWD was discovered in wild deer near Preston and Harmony, Minn.

In the forecast map, darker areas, which could indicate sources of disease, contrast with lighter "cool spots" where the disease is less likely to spread and where it then may die out, Walsh said. The heat map-style graphical representation offers more information about the growth and spread of disease over time in an area than more typical distribution maps, which simply pinpoint where the disease is present or absent.

"We can model the growth of the disease in an area and how it's spreading through an area," Walsh said. "That allows us to look at what's affecting growth in the disease in the area. The other nice thing is that we can look at past management actions to see which ones may be most efficacious or most effective at slowing growth or preventing spread. That's how we think it could be useful."

The new model outperformed traditional prediction methods, according to the USGS.

"The traditional techniques describe the pattern in the current data that you have in hand," Walsh said. "This model is better at forecasting, projecting out where we don't have data yet. Looking at the process of growth and spread allows us to do better at predicting what the prevalence rates, for example, might be in future years."

Walsh and a USGS colleague developed the forecasting model with researchers at the universities of Kansas State, Colorado State and Utah State. Their work, presented in a paper published earlier this year, focused on an area of southwestern Wisconsin.

The study found that the disease grew rapidly during the latter part of the 2002-2014 study period, especially among older males, and that the trend likely would continue in that area, according to a USGS statement. The model found that CWD could spread nearly twice as fast in the Wisconsin River corridor than outside of it, USGS reported, and likely will grow faster in highly forested areas.

The study involved nearly 2,600 positive cases of CWD found in close to 104,000 deer in southwestern Wisconsin tested over that period. CWD, a brain disease that's fatal to deer, elk and moose, isn't known to affect humans.

Walsh said researchers will test the model on 2015-16 CWD data recently received from the Wisconsin DNR. Where this study looked at proportion of hardwood forest and human development, whether the area was within or outside the Wisconsin River corridor, and the sex and age of the sampled deer as variables that could predict CWD prevalence, future work will include more detailed variables.

"There's a lot of work still to be done on getting a good handle on the disease and its ecology and how it operates in the landscape," Walsh said. "This is going to be a good tool to help us start to nail down some of the big questions."

It has applications beyond CWD. The Kansas State researcher has applied the model to the spread and growth of white-nose syndrome in bats, Walsh said, while the one at Colorado State has modeled the movement of an invasive species, Eurasian collared doves.

"It could be any kind of invasive species or disease, human or wildlife or agricultural," Walsh said. "Not knowing the data for sure, but something like Ebola or Zika cases where we have infection starting from a source and spreading out. Any type of diseases or invasive species that's introduced into an area and then spreads out would be a good case study for this type of tool or model."

The model offers a "solid foundation for characterizing disease growth in a location," said Minnesota DNR's Jennelle, who said he has been talking to the study's authors. Wisconsin, Jennelle noted, has a much larger data set to use.

"I'm definitely interested in using this approach if it might help us," Jennelle said. "I'm hoping that when we do surveillance again in that area this coming year we don't find any CWD. That would be the best thing. I hope our efforts to knock it down actually worked. We'll see this fall."

Jennelle said he also thought the forecasting model might be applicable to avian flu on a regional or national level.

Todd Nelson is a freelance writer from Woodbury.