In last year's August and December issues of Ag News and Views, I addressed some of the assumptions associated with infrared-triggered camera surveys. With another year of data collection under our belt, I thought I would update you on the findings. Please keep in mind that the study is ongoing and our data analyses are preliminary.
In order for infrared-triggered cameras to be reliable tools for surveying deer populations, their use should not negatively impact site visitation. Since the majority of the camera site visitations on the Noble Research Institute Wildlife Unit (NFWU) occur during nighttime hours, deer response to the camera flash associated with camera surveys is of significant concern. Many other users witness similar nocturnal visitation trends at other study sites. As reported last year, deer reactions to camera flash ranged from "no response" to "extremely startled." Some individuals had fairly consistent reactions to camera flash, while others exhibited different behaviors during different encounters. There really did not appear to be an "average deer" reaction.
Even though some deer appeared to be quite startled by the camera flash, our data indicate that camera flash did not preclude photographing most of the deer using the camera sites. We observed camera site visitation before and after camera flash initiation. Observations of identifiable deer (marked deer and antlered bucks) indicated that well over 90 percent of the deer using camera stations before the flash survey was initiated were photographed at least once at a camera station after the flash survey began. These observations also indicated that bait site use was quite dynamic, and it was not unusual for individual deer to switch from one camera site to another.
As stated in previous articles, one of the big questions concerning camera surveys is whether bucks, does and fawns have their pictures taken at the same rate. In most cases, the number of pictures taken per buck is used to extrapolate the number of does and fawns photographed during camera surveys. In our study, comparisons of the number of pictures taken per identifiable buck with those taken per marked doe continue to indicate that bucks and does do not necessarily have their pictures taken at the same rate. Not only can there be variation between bucks and does, but there is often tremendous variation among individuals of the same sex.
We have begun to look at how this affects population estimates derived using camera survey data. In 2002 and 2003, five camera sites were monitored concurrently with infrared-triggered fixed cameras and infrared-triggered video cameras. Video camera data is time consuming and labor intensive to analyze, but we believe we obtain a very accurate assessment of the animals using a particular camera site through video analysis. For our analysis, we regarded population estimates derived using video data from the five sites as our best estimate (BE). We are confident the estimates closely approximated reality. The BEs for 2002 and 2003 were compared with two other estimates. First, we compared the BEs with estimates derived using the number of pictures per buck to estimate doe and fawn numbers (traditional). Secondly, we compared the BEs with estimates in which the doe and fawn numbers were extrapolated using the number of pictures taken per marked doe (BPMD).
The results are depicted in the graph above. Buck estimates were similar for all methods in both years. While we did video a couple of bucks in both years that did not show up during the camera survey at one of the five sites, the data continue to indicate that infrared-triggered camera surveys yield very good approximations of the buck component of the herd.
The number of does estimated using the traditional method was quite a bit higher than with the other two methods, particularly in 2002. Estimates derived using pictures per marked doe and the best estimate were similar. These data reinforce the point that extrapolations made using only "pictures per buck" data are of questionable accuracy.
In both years, the traditional and the marked doe methods resulted in overestimating the number of fawns as compared to the BE. While using the number of pictures taken per marked doe to extrapolate fawn numbers resulted in estimates more in line with the BEs, these results indicate that fawns did not have their pictures taken at the same rate as bucks or does. Our data indicate that fawn estimates derived using these extrapolations are of questionable accuracy as well.