Hemorrhagic disease is a phantom enemy of deer. Known to hunters more commonly as EHD and bluetongue, this family of viruses strikes unpredictably and sporadically, hitting deer hard in some areas, lightly in others, or not at all. It often skips years or even decades before hitting the same area again. We know to look for it in late summer and early fall, ironically almost the same months as hurricane season, but we have no EHD radar that provides early warning or a likely landfall.
Well, we had no radar. That might be changing after a scientific study found a specific combination of weather conditions that aligned extremely well with past EHD outbreaks. That combination of warning signs could alert deer hunters and state wildlife agencies to increased probability of outbreaks before they begin.
Why Weather and EHD?
How can weather be related to a virus outbreak in deer? Epizootic hemorrhagic disease (EHD) and bluetongue viruses (BTV) are spread by a tiny, flying, blood-feeding insect called a midge. These midges in the family Culicoides become abundant in late summer. The photo at the top of this page by Scott Bauer of USDA-ARS shows a Culicoides sonorensis midge full of blood after feeding through an artificial membrane in a research lab. Midges breed in shallow, warm, silty, low-quality water sources – basically mud holes. Picture slick river banks, stranded pools of creek water, or trampled mud around livestock water tanks.
When this habitat is abundant, midge numbers explode. In droughts, deer are often visiting these mudholes to drink, where they encounter midges in the thousands. Deer can be exposed to thousands of “attacks” per hour at these sites, according to a study by the Southeastern Cooperative Wildlife Disease Study (SCWDS). It takes as few as four infected bites for deer to become sick. This is why outbreaks of EHD tend to occur in hot, dry weather. The return of cold weather drives midge populations down, and virus transmission fades.
But just because it’s dry and hot in your area doesn’t mean an outbreak of EHD is occurring. Plenty of hot, dry summers come and go without notable issues, particularly where deer have developed immunity to the virus, like in much of the Southeast. And outbreaks can occur in places when it’s not extremely hot or dry. Again, a phantom enemy.
With EHD and bluetongue outbreaks appearing in new states with increasing frequency and intensity over the last 20 years, a group of scientists organized a forensic investigation to look for predictable patterns in the phantom’s history. The team of investigators included Brett Skelly, State Deer Biologist for West Virginia DNR; Dr. Mark Ruder and Dr. David Stallknecht of the Southeast Cooperative Wildlife Disease Study at the University of Georgia; and Dr. Sonja Christensen of the Christensen Lab for Wildlife Population Health at Michigan State University. I had a chance to speak to Sonja, a wildlife biology and disease ecology assistant professor at Michigan State, who presented the team’s findings at the 2024 Southeast Deer Study Group meeting in West Virginia.
Round Up the Usual Suspects
EHD is nothing new in the Deep South states. The viruses have haunted those deer for so long and so frequently that southern deer built up herd immunity. Death rates are low when outbreaks happen, and the disease’s effects often remain unnoticed by deer hunters – except when they notice the cracked hooves of harvested deer that previously survived infection.
West Virginia is north of the historical immunity line. It’s one of the states where EHD and bluetongue appeared only in the last 20 years. Outbreaks are becoming more frequent and killing more deer because they lack natural immunity. The team of researchers examined 39 years (1981 to 2019) of EHD and BTV occurrences in West Virginia reported at the county level. They quantified several suspected variables and cross-referenced them against outbreaks, looking for fingerprint patterns. The first suspect to be exonerated was human population density.
“To be able to report a dead deer, you need a person to see it,” said Sonja. “So we thought maybe human population might influence reporting. But we’re picking up these outbreaks even in areas of low human population. The results didn’t really support our initial hypothesis.”
Deer density also had an airtight alibi. Disease locations did not line up with deer density in a predictable way, since deer don’t transmit this virus through contact with each other. The same was true for the percent of wetlands in the county. “Wetlands” as a habitat category don’t necessarily reflect quality midge breeding conditions in local areas.
The scientists also plugged in a number of weather variables, including precipitation and temperature. They also analyzed each of these factors seasonally and by month. This is when a clear suspect emerged, and it wasn’t just a lone actor. The fingerprints belonged to a gang working in cooperation.
A + B + C = EHD
“The best model was above-average spring temperatures plus increased July precipitation plus decreased August precipitation,” said Sonja. “These are stand-alone variables in the model, and when they were all together, that is what created the higher likelihood of disease outbreak. One alone might contribute, but it really wasn’t meaningful in the dataset unless you had all three players happening in a county at the same time.”
This combination makes sense given the reproductive ecology of Culicoides midges. Above-average spring temperatures mean insects can become active sooner and begin building populations early. High July precipitation fills lakes, swells creeks, and pushes rivers into floodplains. But if rain shuts off in August, those waters recede leaving abundant mudholes and silty breeding grounds for midges during the hottest part of summer when virus replication peaks inside the insects.
The result is a collision between clouds of virus-filled midges and hot, thirsty whitetails seeking a drink at shrinking pools of stagnant water. I asked Sonja if there’s a magic threshold of temperature or precipitation that triggers this scenario, but it’s not an on/off trigger. This is all about ranges of probability.
“The statistical association we found is that your probability increases as an increasing function of average spring temperature,” said Sonja, “but you can see the probability starts to increase quite a bit more. If the average spring temperature is higher, then your probability of disease presence is much higher, going from 10% probability to 30% with just a few degrees more on average.” (See graph below)
You can see the same is true of July and August precipitation when you look at additional graphs from the study, shown below. As July precipitation climbs, so does disease probability. But August is the reverse: Higher precipitation equals lower disease probability.
I asked: What happens if July is dry and August is wet? “If you flip any of them, like if you have lower than average July precipitation, it would reduce the probability of disease,” said Sonja. “Any one of these variables plays a role. For example, July plus August without higher spring temperature was the second best model for increasing disease probability. Add in the spring temperatures, and that’s what really helped increase the probability of occurrence.”
Where Would This Formula Apply?
Sonja told me she doesn’t think the team’s conclusions apply to deer in the Deep South that have some built-in immunity to EHD and BTV. She said, “This applies in places where it’s an emerging disease and deer are more susceptible.”
Sonja also said the relationships are strong, giving her enough confidence to say they have potential to hold up in other states from West Virginia north and westward.
“These are trends that could potentially be extrapolated out to other states. We haven’t verified that yet, and we need to do this analysis on a broader set of data to see if those trends hold. Forecasting would be the next step but a challenging one,” she said. “In Michigan, we really haven’t had a big outbreak since 2012. We’re kind of due, and I’m a little nervous about that because we’ve had a warm spring and a pretty wet summer so far.”
We Can All Expect Warmer Springs
Increased spring temperatures alone, without higher July precipitation and lower August precipitation, was the third best predictor of EHD probability. That’s particularly concerning since we can pretty much expect above-average spring temperatures almost every year as climate change continues. According to the National Oceanic and Atmospheric Administration, 2023 was the world’s warmest year on record, and it has been 47 years since Earth has had a colder-than-average year. This is likely the very reason EHD and BTV have moved northward into new territory in that same time period. Milder winters and warmer summers are inviting the Culicoides midge to spread northward into states where they couldn’t previously survive.
Take Wisconsin for example. Hemorrhagic disease crossed the southern border of Wisconsin and advanced into its southernmost counties around 2012.
A Wisconsin DNR report states: “Wisconsin is likely to become a much warmer state over the next few decades, with average temperatures closer to southern Illinois or Missouri. Results show that Wisconsin has warmed 2 to 3 degrees Fahrenheit since 1950, and it is projected that the state will warm an additional 2 to 8 degrees by 2050. Seasonal temperatures have also changed. Spring temperatures have increased by 1.7 degrees Fahrenheit with the onset of spring coming three to 10 days earlier since 1950.”
This means EHD and BTV are likely to continue a northward march, reaching more new areas of the north, with more intense outbreaks when favorable conditions align.
How To Use This Information
Let’s say you hunt deer in a state along the same latitude as West Virginia or further north. Let’s also say you witnessed a warmer-than-average spring, and now it’s raining a lot for July. If August turns dry, what can you do?
We can’t manage the insects that spread the virus, and we can’t vaccinate healthy deer or treat sick ones. Theoretically, providing clean water sources for deer in a late-summer drought could help reduce infection, but this is unproven.
Instead, the primary way deer hunters and state wildlife agencies can help is by avoiding over-harvest of deer immediately following a serious EHD or BTV outbreak that kills significant numbers of deer locally. This means early awareness of the situation, which gives hunters time to react and state wildlife agencies more time to make any emergency adjustments to bag limits or seasons.
So, use this new formula to remain alert. If conditions align to increase disease probability in your county, monitor deer where you hunt. When you encounter the odor of rotting carcass on the breeze, follow it up and investigate. Sick or dead deer found near water in late summer are a prime indicator of hemorrhagic disease. If you find deer that aren’t too decomposed yet, contact your state wildlife agency as soon as possible and report the sighting.
The good news is, deer populations are resilient and soon recover from sporadic losses to EHD, even in states where mortality rates run high – as long as deer harvest allows for recovery. Keep your eye on the weather forecast and watch out for a combination that predicts higher probability of disease this summer.