Nice thread BT. Timely too.
I just spent the week revising a manuscript we submitted for publication this fall and was looking up a few references to cite. During this time, I found several other papers related to deer that I thought may be of interest to the members of the forum.
The first paper is a publication from the Journal of Wildlife Mgt. that Norton et al. (2011) published on PA buck harvest rates. There was a good statement made at the end of the paper that I think everyone needs to bear in mind when thinking about harvest estimates, deer population numbers, and any other hard and fast numbers. The paper suggests that managers need to put less emphasis on management decisions that don't acknowledge model uncertainty such as absolute population abundance but rather place more emphasis on population trends.
After doing quite a bit of population modeling over the past few years, I couldn't agree more. There is entirely too much emphasis and demand on exact numbers which are never exact if estimated due to random effects and other known effects.
This is largely the reason why I am not concerned if reporting rates for deer are low but more interested in being able to estimate the reporting rate to correct the estimated harvest from those that do report. This is also why I am not in favor of check stations or reporting at the time of buying a license. These methods may seem intuitive, but they could lead to even more variation in harvest reports which would affect the harvest trends.
The second paper I read dealt with genetic structures of wild and captive deer herds in response to selective harvest based on antler size. This paper Webb et al. (2012) simulated both changes in antler points and total B&C score of antlers in a captive herd and a free ranging herd.
In the captive herd, the simulation indicated about a 2-4 antler point increase and 2-4 inch increase in B&C score in an 8 year period when the largest bucks were selected for and spikes were selected against in both the bucks and does.
In the wild herd, the simulations assumed a closed population (no deer moving in or out of the study population) and tested a number of selection scenarios such as B

ratios, age structure, etc. The average number of points and B&C score did increase but not significantly after a simulation of 20 years. Since the simulations assumed a closed population which would not occur in a natural population, the chances of altering the genetics for antler potential through antler selection in a free ranging herd would be difficult if not impossible under normal circumstances.
Only a simulation but I found it interesting nonetheless.