Project Summary

Within the last century, northern bobwhite quail (Colinus virginianus) populations have dropped substantially due to regional landscape changes, such as intensive agriculture and conversion of open pine forests to high density hardwood forests and dense pine siliviculture. However, quail hunting plantations throughout the southeast manage their lands specifically for quail through annual prescribed fire, periodic thinning of the timber overstory, and supplemental feeding. Even with these land management practices in place, bobwhite quail continue to face an upward struggle to persist and thrive due to high levels of predation. The primary source of nest loss and adult quail mortality is by mid-sized mammals (bobcats, coyotes, foxes, and raccoons), snakes, and birds of prey.

It was once a common belief that cotton rats were a regular predator of quail eggs and nests. Herbert Stoddard and others recorded many nests with eggshell fragments showing signs of chewing by small mammals. However, this fact has been recently disproved with the advent of nest monitoring with remote video cameras. Through video monitoring we have recorded that cotton rats are not a primary predator of the nest but are secondary visitors scavenging on only the remains. In less than 1% of all nests we have observed have cotton rats been the primary nest predator. Additionally, further support has come from laboratory research documenting that most cotton rats cannot open their mouth large enough to crack open a quail egg. However, concern regarding predation of quail nests by cotton rats still abounds in the southeast.

Hispid cotton rats (Sigmodon hispidus) are a primary prey species for many of the predators also feeding on bobwhite quail, and cotton rats and bobwhite quail often thrive in similar habitats. These habitats consist of early successional grasslands with sufficient cover from brush or shrubs, such as blackberry and Wax myrtle. Dr. Mike Conner at the Joseph W. Jones Ecological Research Center suggests that “predator communities, and food webs in general, of the South would look drastically different” in the absence of hispid cotton rats. Avian predators tend to feed on the most available prey species present in an area and will shift their attention to other species, when the most populous species declines. Cotton rat populations follow a regular cycle, the length of which varies with time and location, in which the populations increase to almost plague-like levels in some years and then crash in subsequent years. Predators most likely feed heavily on cotton rats in periods of very high population levels, but shift to other species, such as bobwhite quail, when their populations drastically drop. It has been suggested that quail predation may be minimized by maintaining a high level of cotton rats, which would absorb a large portion of avian predation.

Little is known about how land management practices that benefit quail affect populations of small mammals. Previous studies have demonstrated that food supplementation and fire may positively impact growth rate, population density and summer survival of hispid cotton rats. However, no one has investigated how the size of blocks to be burned or the “scale of burn” in combination with supplemental feeding affects cotton rat demographics. Therefore, Tall Timbers established a study to determine the combined effects of bobwhite quail land management practices such as, supplemental feeding, prescribed burning and scale of burn, on cotton rat abundance, survival, home range, mass, long term population cycles, and behavior.

Prescribed burning at Tall Timbers Research Station generally runs from March until May. Trapping and banding of hispid cotton rats began April, 2002, in order to obtain post-burn population levels, and continued monthly through August. In subsequent years, February, September and December were added to the months trapped. February was added to obtain pre-burn population levels; September allowed for early fall population estimates; December enabled early winter population estimates to be made. We used 8 trapping grids, each having 100 traps spaced 15 meters apart arranged into 10 parallel rows also 15 meters apart. Of these, 4 grids were burned, 4 were not burned, 4 were supplementally fed with grain sorghum, 4 were not supplementally fed, 4 were large scale, averaging 20 acres, and 4 were small scale, averaging 5 acres, to give a different burn, scale, and feeding regime per each grid. The grids were trapped four consecutive nights per month. Trapped animals were weighed, sexed, reproductive status noted, and banded with unique ear bands for recapture identification. Through this mark-recapture data, the combined effects of supplemental feeding, prescribed burning, and scale of burn on cotton rat population cycles, within and between year survival rates, abundance levels and growth rates will be determined.

In the spring of 2004 a radio telemetry study was initiated to assess behavioral and home range responses to prescribed burning. Four cotton rats from 20 different plots: 5 small scale burned, 5 small scale unburned, 5 large scale burned, and 5 large scale unburned were radio tagged with necklace style radio transmitters. Cotton rats were tracked 6 days per week with location and vegetation type noted. During days when treatments were to be burned cotton rat locations were taken just prior to burning and immediately after to assess fire avoidance movements and the time it took to move to areas with sufficiently dense vegetative cover. We observed that fire related mortality was minimal, but a few instances were noted, with one cotton rat running into a too shallow hole and another into a clump of grass, which inevitably burned as the fire passed over. Rats usually moved out of burned plots within 24 hours of the burn, but a few remained in plots that had sufficient unburned patches remaining. Vegetation surveys looking at plant density and composition and ground cover were completed every other week in areas where cotton rats had previously been located to estimate ground cover preferences and time of re-entry into a burned plot. Most of the radio-collared cotton rats took up permanent residence in their new unburned habitats, but a couple moved back into the burned areas a couple months after burning. At this time, regenerated ground vegetation was at least a foot high and dense enough to prevent overhead predators from locating the animals.

Survival analysis of banding data from 2002 depicts sex of the animal, followed by supplemental feeding to be the most important variables in determining cotton rat survival. Prescribed burning and scale of burn were far less important in determining cotton rat survival. Pooled survival estimates for males and females and for fed and unfed cotton rats were very similar with large confidence intervals suggesting that supplemental feeding, prescribed burning and scale of burn do not largely impact hispid cotton rat survival. Further analysis of data collected from 2002-2005 will take into account environmental and year specific population effects, abundance levels, and the effect of mass on survival. However, survival may not be the key demographic parameter driving populations in different treatment blocks. Preliminary results from the telemetry study and abundance measurements suggest immigration/emigration and reproductive output may be dramatically different between treatments, affecting populations within a single breeding season.

Future analysis will compare hispid cotton rat population levels to those of northern bobwhite quail to look for consistencies between the two cycles over an extended period of time.

Information

• MS candidate