1998 Georgia Deer Special

Last deer season was the most phenomenal deer season that Georgia hunters have ever experienced, and we never saw it coming.

Compare the estimated deer harvest from last season to the population numbers that we reported in last year’s Deer Special, and you will notice that something doesn’t sound right: we killed more than half the deer in the state last year. The truth is, we had a far greater deer population, not just last year but for the past two years, than the Wildlife Resources Division (WRD) was estimating. In fact, we should have been reporting that in the season before last, we achieved a new record harvest of 459,135 deer, topping the previous record of 453,00 set in the ’93-’94 season. Instead, we reported that the ’96 harvest was about the same as the ’95 harvest: 396,000.

That change is big enough news by itself, but now we also get to report that last season deer hunters pushed the harvest record even higher to an incredible estimate of 509,533 deer. Of course we did not really harvest half the deer in the state, and WRD did not come up with this number by rolling a can full of deer bones onto the floor and watching how they fell. Here is the explanation for how the Great Curveball of ’98 occurred. 

Every year, when the deer season is over, WRD sends out survey questionnaires to a sample of deer hunters to find out how the season went. Most of the hunters who return their survey do so after the first mailing, but WRD sends a second wave of letters out to those who do not respond. A few more now fill out the questionnaire. Finally, WRD sends out a third and final wave of letters, and a small group finally gets a clue and responds. Still, each year, WRD gets an overall response rate of about 50 percent.

For years, there was a predictable pattern in the waves. The hunting success rate of those who responded with the second wave was lower than those in the first, and the hunters in the third wave had the lowest success of all, due to a non-response bias: folks believed that if they did not do well, WRD did not really need their response, so they took longer to finally catch on and respond. So, WRD could safely assume that the success rate of the third-wave respondents also spoke for the thou- sands who never responded at all. This is a proven and safe statistical assumption that is used frequently by survey researchers.

But in the summer of 1997, a curveball came from the mound: the third wave success rate from the ’96-’97 season was actually higher than the second wave. This had never happened before, and to count it as the truth and not a glitch would mean that a large proportion of hunters had done well, and harvest estimates would soar. WRD chose to ignore the “glitch” and transpose the drop in the success rate between wave one and two onto the third wave as well. The result was a modest harvest of 396,000 that all other field indicators seemed to agree with. 

Then, the bomb dropped. The “glitch” showed up again this summer in the survey results from last season, meaning that it was not a glitch. The high harvest that had been indicated in 1997 was not a fluke. Hunters in fact did have extremely high success last season, as well as the season before. WRD now knew that they had far underestimated the ’96-’97 harvest and the statewide deer population for at least two years. Those numbers have now been revised.

This means that for more than a year, WRD was working with harvest and population estimates that were off the mark, and deer hunters will be quick to point out that a lot of significant decisions were made by WRD in that year, like to end the December Break and continue increasing doe days. GON asked Todd Holbrook, WRD Game Management chief, if the more accurate numbers might have altered those decisions.

“These numbers don’t cause me any major heartburn from a regulatory standpoint, but it will occur to a lot of people that, oh no, we just upped the season three weeks and we had this huge harvest that we didn’t know about until now,” Todd said. “The big issue on the break is age structure on bucks, and regardless of what the harvest level is, we are still monitoring the age structure of bucks in the harvest with cooler-locker checks, and we are still seeing a much older age structure than we have seen in the past.”

Asked if the minor increases in either-sex days in some areas might have been scaled back, Todd said, “That would be in the ‘possible’ category.

“We didn’t have any kind of major adjustments, just some add-a-weekend kind of things, so where we are sitting on doe days right now is about where we would be sitting anyway, had we known. Now, in the long run, it might well make a difference. If we start seeing this harvest coming back again in 54 percent does, like we saw this past season, then we’ll know that maybe we can’t increase doe days anymore above what we’ve done.”

Now that the non-response bias problem has been overcome, the harvest numbers are back on track as one of the most reliable estimates WRD is able to calculate. The harvest numbers are determined by combining mail survey success results with data collected from cooler-locker checks, including sex and age data, date of kill and antler measurements from yearling bucks.

Interestingly, the only other threat to accuracy that still remains, although a small one now, is the presence of QDM counties in DMUs four and six. In those counties, cooler checks don’t produce any antler data on yearling bucks because yearling bucks aren’t legal, so the data for those DMUs is very slightly watered down. Scott said that in the future, if more counties are added, then they may have to be included in their own separate DMU to maintain accuracy.

For last season, the DMU data shows that the continuing harvest increase came mainly from the western Piedmont and the upper and lower Coastal Plain. The most significant jump in harvest came in DMU 4, where the total harvest went from 70,712 to 108,193 deer. Other DMUs either stayed the same or shrank slightly.

The chart on this page that shows the buck age distribution in the harvest also shows little or no change over the past two seasons by DMU. Button bucks continue to be a significant part of the harvest, due in part, most likely, to the “Oops” factor on either-sex hunting days. Meanwhile, 2 1/2-year-olds and older made up significant portions of the harvest in DMUs six and nine.

In DMU 3, it appears that hunters killed very few mature bucks in their harvest, but recall that this is the metro Atlanta DMU. Though plenty of hunting goes on in some of these counties, sample sizes from cooler checks are small, so error rates are significantly higher in this DMU.

Record Level of Does in the Georgia Deer Harvest
For the past several years, the percentage of does in the deer harvest has hovered around 50 percent, but last season it rose to the highest level ever documented by WRD, 54.1 percent. Essentially, the increase in the harvest from 459,000 in ’96- ’97 to last year’s harvest of 509,000 was made up mostly of does; the buck harvest remained relatively unchanged.

With doe days increasing slightly each year, it is not surprising that hunters are taking advantage of the opportunity and it is being reflected in the harvest. How does WRD feel about the percentage?

“I’m comfortable where we are right now, for a few years anyway,” said Todd Holbrook. “There’s still a big need in certain places for people to kill more does, and we feel like we need to give those people the flexibility to do that.

“If age structure on bucks continues to get older and older, we could live with 60 percent does in the harvest. If you started seeing the buck age structure get younger like it used to be, you wouldn’t want to hold with 60 percent does in the harvest for very long, because there’s no question you’re going to bring your herd down.”

As an indicator of when the doe harvest has gone too far, Scott McDonald said he would rather look at the age structure of the doe harvest. Last season, 33 percent of the adult does killed in Georgia were 1 1/2 years old, which is a good number. If that age group approached 40 percent of the harvest, it would most likely be a sign that does were being hit too hard, Scott said. The percentage of 1 1/2-year-old does in the harvest has been holding in the low 30s for several years.

Deer Hunter Success Rate Explodes In Georgia
Over the past two years we had two record harvests, while at the same time the number of hunters in this state has, overall, shrunk. That is going to lead to one inevitable conclusion: fewer hunters killed more deer.

The deer population is still hovering at an all-time modern high, so there are plenty of deer in most areas for hunters to kill, but at the same time, the hunters who remain in the woods are experienced. So, last season, roughly 60 percent of all hunters were successful, meaning that they killed at least one deer. Incredibly, of those successful hunters, the majority killed multiple deer and almost half, 45 percent, killed three deer last season.

“That’s some pretty phenomenal success rates when you think about it,” said Todd Holbrook. “Anytime you have over half your hunters kill a deer, you’ve got some fantastic deer hunting, and over two out of three of those that are successful are successful at killing multiple deer.”

The phenomenally high success rates of the past two seasons may be tied up in the trouble with the non-response bias in the mail survey results. In fact, that’s exactly what a small or non-existent bias would point to: very high success. Also, Todd believes that hunters may be growing desensitized to deer-hunting success.

“Apparently, killing one or two or three deer is not so much an accomplishment and not the kind of incentive to return a questionnaire that it used to be back in the old days,” Todd said. “That could be why there’s not much difference in the success rate of those who return the survey early and those who respond late.”

’97-’98 Hunter Success From the Mail Survey
• 60 Percent of Georgia hunters killed at least one deer. This is not an average based on the number of hunters divided into the number of deer killed. It is the actual percentage of hunters who said they had a successful season last year.
• 70 Percent of the successful hunters killed more than one deer! Of all hunters who reported that they were successful, a third of them, 31 percent, killed only one deer. The rest killed multiple deer, including: 24 percent killed two, 17 percent killed three, 12 percent killed four, 13 percent killed five, and 4 percent of the survey respondents killed six or more (some of these may have been illegal kills, but there are ways to kill more than five deer in one year legally, and that is through bonus WMA or refuge hunt tags.

Georgia Deer Hunting In 1968
According to the November 1969 issue of “Georgia Game & Fish,” an official publication of the Georgia Game & Fish Commission, the 1967-68 hunting season in Georgia recorded a “respectable success ratio” when 140,310 deer hunters killed an estimated 20,902 deer. On average, one out of every seven hunters managed to bring down a deer that year. Compare that to seven average hunters who participated in last year’s season: not only did five of those hunters kill a deer, but a couple of them may have killed two apiece, and one might have killed as many as four or more deer. That was accomplished with twice as many hunters in the woods as 30 years ago.

Of course there were also 12 times as many deer last year as 30 years ago. In 1968, biologists said they estimated the herd at around 100,000 animals, and in the understatement of the century, added that they believed “the state could easily support 400,000 deer with proper protection and management.”

Other things have changed in 30 years: those 140,310 folks who said they were deer hunters were outnumbered by, and put in fewer days of effort, than squirrel hunters. Then, deer hunters had only recently surpassed quail and rabbit hunters — in 1962, quail hunting was the most popular sport, followed by squirrel, rabbit and then deer hunting.

What To Expect For Georgia’s 1998 Deer Season
Two record harvests in a row and more than half of the deer were does in the past year — where do we go from here?

In trying to estimate deer populations, you always get more accurate results looking back at a season (with harvest numbers and other data to help out) than looking forward at a season. With the road bumps in the survey data behind them, WRD is now saying that last season there were 1.39 million deer in this state, which is 450,000 more than they estimated going into the season, and the highest estimated deer population ever. That was the curveball that we struck at: we never expected that last season would be so successful. The “correction” in the harvest numbers forced WRD to seriously reevaluate their population estimates for the past couple of years, because with the phenomenal harvest rates and hunter success indicating that half the deer herd landed on a charcoal grill, the population estimates had to be way low.

“We knew we weren’t harvesting even close to 50 percent of the deer,” Scott McDonald said. “If you’re harvesting 50 percent of your deer and 54 percent of the harvest is does, then we would essentially be wiping them off the planet.”

For certain, the population estimates in 1995 and 1996 were the most conservative of any in the last decade, which explains the big surprise from last year if you look at the chart on this page: the population was creeping back up faster than anyone realized (or either it had never dropped that much to start with) following the heavy harvests of the early 90s. This chart now reflects the revised population numbers, although the ’95 and ’96 estimates are probably still conservative.

“This doesn’t mean that our population has jumped up and there is going to be a whole bunch more deer out there for the hunters,” Scott said. “What it means is we’re becoming better at estimating what we do have, and in the past we’ve been real conservative in our estimates.”

All numbers aside, WRD believes that the deer herd has been stable or slightly on the increase for several seasons now.

“When you’re seeing these kind of harvest numbers two years in a row, either your population is falling out the bottom or it’s climbing slowly,” said Todd Holbrook. “All of our age structure data says that it ain’t falling out the bottom, so you can roll that option out of the picture.”

One fact, which may be relevant to the coming season, is certain: following the heavy-duty harvest of ’93-’94 there was a 20 percent decline in the harvest the following year.

“It would not surprise me to see that happen again,” Todd said. “I think that if we continue to hit does like we have, whether the harvest is 450,000 or 500,000 or more than that, yeah, you’re going to have some adjustments down. Is it going to happen next year? The year after? I don’t know, but it is going to happen.”

This season, chances are strong that hunter success will not be as high as it was last season, but then hunter success was through the roof last season. We could suffer a major decline in success and still find the majority of deer hunters dragging venison out of the woods. The computer models now point to a population of 1.28 million deer this season, down from last year, but Scott said that he still has a feeling that this number is on the conservative side.

“I don’t think we’re wiping them out. There may be some areas where the population is going down and some areas where the population is still high. Hopefully, some of these questions will be answered when we go to Point- of-Sale.”

Point-of-Sale licensing will begin next spring, when hunters go to buy their licenses for turkey season. Under this new computer system, information will be gathered from every single hunter that will allow WRD to conduct a much more thorough and truly random survey, which will render far more accurate and detailed data on hunters and the herd. The ’99-’00 deer season will be the first to be sized up under the new system.

“For now, the bottom line is that these numbers really are gravy,” Scott said. “We don’t have to have these numbers to manage our deer herd. The hunters are real interested in the numbers, and we owe it to them to get the best numbers we can, but it doesn’t put us in any dire straits as far as managing our deer herd when these numbers get convoluted.”

Now, let’s forget the estimates, the computer models and the statistics jargon, and think about what we know. We still have one of the strongest, most well-managed deer herds in the country that annually produces trophies for tens of thousands of hunters, whether it’s a veteran’s trophy Boone & Crockett or a trophy doe or spike for a beginner. This year, we again have a few more either-sex days giving us room to manage according to our individual desires. We have one of the longest deer seasons in the country and this season the Northern Zone hunters get three extra weeks and get to find out what it’s like to hunt on Christmas Eve. Regardless of what a computer says next spring, it’s going to be one heck of a season for those of us heading to the woods.