Journal of the Southeastern Association of Fish and Wildlife Agencies

The food habits of common moorhens (Gallinula chloropus) and purple gallinules (Porphyrula martinica) on a north-central Florida lake were studied through monthly analysis of gizzard contents from May 1981 through April 1982. Common moorhens consumed an average volume of 93% plant and 7% animal food, whereas purple gallinules consumed an average volume of 71% plant and 29% animal food. The major foods of the common moorhen were leaves and stems of Hydrilla verticillata, seeds of Polygonum spp., and snails, Planorbella spp. The major foods of purple gallinules were seeds of Nuphar luteum and Polygonum spp. and insects of the family Chrysomelidae.

Mourning dove (Zenaida macroura) nests were located on the Texas A&M University Campus from February-October 1981. Mortality differed between the various stages of the nesting cycle, and in different months within each stage. The number of days in which wind speed exceeded 27.6 km/h (15 kts) each month explained 34% of the variability (P =0.10) associated with total nesting mortality. Mean monthly temperature was positively correlated (P < 0.10) with mortality occurring during the 1st nestling week and negatively correlated during the 2nd week after hatching. Loss of nestlings was correlated (P < 0.05) with an index to blue jay (Cyanocitta cristata) production and total monthly rainfall (P =0.10).

Seasonality of nesting by mourning doves (Zenaida macroura) in north-central Florida was investigated in 1979-80 by assessing changes in call (coo) counts, gonad size, nesting activity, and crop gland development. We assessed the incidence of nesting during the fall and possible impacts of hunting on dove productivity. Weekly call counts taken during this study indicated that incidence of cooing was highly variable; the highest levels of cooing occurred during February and March and they diminished as the spring and summer progressed. Very little cooing was recorded from October through December. Recrudescence of testes in adult males occurred in December, with some regression in size during September. Ovaries increased in size during February and March and diminished during August. Most adult ovaries appeared to be inactive in fall. Eggs and nests first appeared in late February of each year, peaked in both April and June, and were scarce by August and September.

Based on 162 European wild hogs (Sus scrota) collected from 1971 to 1973 in the Great Smoky Mountains National Park, the male: female ratio was 52: 48 and the age composition consisted of 52% of the hogs being <12 months, 21% 12-26 months, and 27% >26 months of age. No sex difference in collection by trapping and shooting occurred (0.1

Stomachs from 36 hunter-harvested black bears were collected over a 5-year period from October through January in Baker and Columbia Counties, Florida. Saw palmetto (Serenoa repens) was the most important item in the diet. Other important foods included black gum (Nyssa biflora), gallberry (Ilex glabra), yellow jackets (Vespula spp.), and armadillo (Dasypus novemcinctus). The practice of winter burning may have artificially increased the importance of saw palmetto to bears in the study area.

Depredating black bears (Ursus americanus) were trapped and tagged at beeyards in an attempt to create aversion to those beeyards. Of the 9 bears trapped and released, only 1 bear was known to return to a beeyard during the study period. Electric fences were also shown to be effective for protecting beeyards from depredating bears. There was a 70% greater probability of bear damage among unfenced beeyards than among fenced beeyards. A combined, electric fence, trap-and-release program is described.

Dispersal of 54 restocked eastern wild turkeys (Meleagris gallopavo silvestris) was studied using radio telemetry on 2 different areas in east Texas from February 1979 through July 1981. Twenty-five turkeys dispersed farther than 5 km from the release site. The maximum distance recorded from the release site for the first year following release varied from 1.3 km to 11.9 km and averaged 4.6 km. Dispersal increased through the spring and summer following release, then did not further increase until the next spring. The mean annual dispersal area was 1,688 ha. For the entire study period, the population dispersal areas were 19,300 ha on I study area and 10,545 ha on the other.

Gizzard fat (GF), wing fat (WF), eviscerated body weight (BW) spleen weight, and thyroid weights were analyzed in 84 ring-necked pheasants (Phasianus colchicus) to determine area, sex, and seasonal variations. Birds were collected from 2 study areas in the Texas Panhandle during 4 seasons. Area differences in major grain crops and food habits were not reflected in the nutritional indices examined. However, differences in pheasant densities between areas were detected. Gizzard fat and WF were greater in hens than in cocks. Highest values for GF and BW occurred in February; lowest values occurred in August. In May, a drastic decline in WF and GF occurred in cocks, and probably reflected reproductive/territorial activities. Spleen weights were greater in males than females, and were greater in August and November than in February. In juveniles, no differences between sexes in WF were detected, but GF was greater in hens than cocks.

Quail food plots established on regenerated pine sites in South Carolina were evaluated in relation to planting dates, seeding rates, and site preparation methods for pines. Stem count data indicated Piedmont food plot establishment should involve broadcasting a mixture of 0.14 kg of bicolor lespe. deza (Lespedeza bicolor) and 0.34 kg of Kobe lespedeza (Lespedeza striata) seeds per 0.04 ha plot in March on either disked or burned sites. Sandhill data indicated that broadcasting 0.14 kg of bicolor and 1.36 kg of Kobe seeds per 0.04 ha plot in March on disked soils would effectively establish food plots in that region.

Food plots for quail (Colinus virginianus) were established on 2 regenerated pine plantations in the Sandhills Region of South Carolina during 1976 and 1977. Bicolor (Lespedeza bicolor), Japonica (L. japonica), Thunberg (L. thunbergii), Kobe (L. striata), Korean (L. stipulaceae) lespedezas, and a reseeding variety of soybean (Glycine max) were evaluated along with various planting methods. The planting methods included combinations of broadcast seeding, dibbling seedlings, disking, inoculating seeds, covering seeds, and fertilizing. All plant materials tested were successfully established in persistent food plots except for reseeding soybean which failed to reestablish itself. A legume mixture of an annual and perenniallespedeza was recommended so that seed would be available to quail throughout the 5 to 7 years that newly regenerated pine stands could successfully support hunting. All planting techniques were successful in establishing food plots of the planted materials.

Home ranges and habitat preferences of radio-collared cottontail rabbits (Sylvilagus floridanus) on an intensive farm and a traditional farm were monitored during 1979-81 in eastern Wake County, North Carolina. On the traditional farm, mean home ranges of males (6.37 ha) were larger (P < 0.05) than those offemales (3.15 ha) . On the intensive farm, mean home ranges of males (2.60 ha) were not larger (P > 0.05) than those of females (2.11 ha). Home ranges of males on the traditional farm were larger (P <0.05) than those of males on the intensive farm; however, home ranges of females on the 2 farms were not different (P > 0.05). On both farms, rabbits preferred brush to woodlots with understories, woodlots with negligible understories, and edge, and preferred fields the least. These data indicate that habitat preferences did not change with agricultural practices. Diurnal and nocturnal habitat preferences were similar.

The effects of whole-tree harvesting upon white-tailed deer (Odocoileus virginianus) , snowshoe hare (Lepus americanus) , and ruffed grouse (Bonasa umbellus) were studied from 1978 through 1981 in southcentral West Virginia. Transect sampling by recording fecal pellets was conducted in mature forest and in clearcuts ranging from 0 to 8 years of age. Use of clearcuts created by whole-tree harvesting was compared to use of those created by conventional clearcutting. Deer used whole-tree clearcuts more than conventional clearcuts, but hare used conventional clearcuts more than whole-tree clear- cuts. Ruffed grouse data were too limited to indicate the influence of type or age of clearcut. Whole-tree harvest can be used to increase certain species of wildlife, but additional studies are needed to determine the long-range impacts.

The response of gopher tortoises (Gopherus polyphemus) to chopper-type forest site preparation was investigated. Of 5 burrows presumed to have been occupied prior to treatment, 2 were reopened by tortoises following chopping, and the other 3 showed various signs of tortoise emergence. Three additional radio-instrumented tortoises emerged from their collapsed burrows during the eighth week following site preparation. Also, several hatchling tortoises apparently survived the treatment.

Winter birds were censured on snag and snagless transects during 3 winters from 1977 to 1980 in a clearcut in east Texas to assess the impact of snag retention on the bird community. For the cavity user group there was greater bird abundance, species diversity, and number of species on snag than on snagless transects. The cavity user group included: eastern bluebird (Sialia sialis), loggerhead shrike (Lanius ludovicianus), northern flicker (Colaptes auratus), Carolina chickadee (Parus carolinensis), red-bellied woodpecker (Melanerpes carolinus), red-headed woodpecker (M. erythrocephalus), and red-tailed hawk (Buteo jamaicensis).

Wildlife mortality resulting from aerial ignition prescribed burning was evaluated in the Lower Coastal Plain of South Carolina. Observers took census of animals leaving the burn to confirm that representative species and densities were present in the study area. Based on the low mortality observed during a systematic post-fire survey, it was concluded that aerial ignition prescribed burning did not cause significant direct mortality to wildlife.

Sawtooth oak (Quercus acutissima) seedlings of unknown provenance were planted on International Paper Company's Southlands Experiment Forest, Bainbridge, Georgia, on 2 sites, an old-field and a wild-land site. Treatments consisted of all oak, oak and loblolly pine (Pinus taeda), all pine, cultivation, and fertilization. After 9 years, pine was significantly larger (P < 0.05) than oak on both sites. On the old-field site neither cultivation nor fertilization showed a significant influence on the growth of either species. However, on the wild-land site cultivation significantly increased diameter, height, and volume of both oak and pine and survival of oak. Fertilization significantly increased diameter and volume of both oak and pine. Acorn production at age 9 appeared to be influenced only by spacing. Fertilization increased acorn crude fat and protein and decreased crude fiber.

A Statistical Analysis System (SAS) program has been developed for analysis of data gathered from radio-telemetry studies of animal movement. The program requires as input data from each observation (1) a pair of directional readings, one from each of 2 known reference points to an animal location, (2) animal identification number and (3) time of observation. From these inputs, x-y coordinates of each animal location are computed relative to a specified origin. A number of animal movement statistics are computed and displayed graphically. Boundary points for the home range (convex polygon method) are determined and plots of home range with or without interior locations are generated as well as a measure of the enclosed area. Any of these outputs may be displayed for an individual animal or for a group of animals by sex, age, season, or other subgrouping.

Directions for construction and attachment of a device for weighing large animals in the field and cost estimations are presented. The device described was developed by the author so that it could be easily transported, mounted and a scale attached whenever a tree, or telephone pole was available.

Recaptures of mourning doves (Zenaida macroura) banded with spiral plastic leg bands revealed these bands were constricting and resulting in loss or severe damage to the legs of doves. Analysis of data from an experiment to determine the effects of color, environmental exposure, and treatment with acetone on the shrinkage of bands indicated that all 3 variables affected (P < 0.01) band shrinkage. Black bands experienced the greatest shrinkage of the 8 colors tested. Bands exposed to the environment shrank more (P <0.01) than bands kept at room temperature or in a freezer. Acetonetreated bands shrank less (P < 0.01) than those without acetone treatment. We recommend that all spiral plastic leg bands used on mourning doves be treated with acetone.

The age structure of 227 bobcats (Lynx rufus) collected from hunters and trappers in West Virginia during the 1977-78 season was dominated by young animals. Juveniles and yearlings constituted 46.7% and 23.3% of the sample respectively. All females> 1 year old had ovulated and had 4.0 ± 0.2 (X ± SE) current year luteal bodies. Yearlings had 1.75 ± 0.27 placental scars, which was significantly less (P <0.05) than the 3.4 ± 0.4 of 2-year-olds or the 2.63 ± 0.27 of bobcats ≥3 years old. White-tailed deer (Odocoileus virginianus) and lagomorphs comprised 45% and 19% of the stomach contents by weight, respectively.

Mourning dove (Zenaida macroura) behavior, habitat use, and food habits were studied in southwestern Oklahoma during faIl-winter. Birds present during the September-October hunting season were summer residents and migrants. They formed large feeding flocks containing up to 1,000 birds, flew 3-12 km from night roosts to feeding sites, and fed in morning and late afternoon. These doves fed in wheatfields when waste grain was available; otherwise, they fed on haygrazer (sorghum x sudan grass), sunflower (Hetianthus sp.), and weed seeds. Doves day-roosted in ravines or uplands near the stock ponds where they drank at midday. At evening they drank at other stock ponds before night-roosting in river bottom or upland trees and thickets. Winter residents in smaller flocks (15-300) roosted and watered close to their feeding fields, and at the roosts they perched low in the trees or sat on the ground.

Controlled hunting (9-night season) and livetrapping of raccoons (Procyon lotor) on the Chuck Swan Wildlife Management Area was assessed in relation to total harvest, age and sex characteristics, reproductive parameters, and density estimates. From 1976 to 1978 hunting pressure and hunting success remained relatively constant while the total harvest decreased at an average of 23% annually from 175 (1975) to 80 (1978). Seventy-five, 69 and 69% of the 1976, 1977, and 1978 harvests were juveniles suggesting a greater vulnerability compared to adults; this occurrence also may be related to dispersal or other factors. Reproductive data did not appear to compensate for high harvest and may be related to the quality of the upland hardwood habitat areas of the southern Appalachians. Tag return ratios in the harvest (Lincoln Index) yielded an estimated density of 1 raccoon per 23.1-27.5 ha. Thirty-four (30.4%) of the trapped raccoons (112) were harvested in 1976, 11 in 1977 and 4 in 1978.

This paper is an effort to demonstrate the importance of being prepared for retirement in the career planning of a wildlife officer. It also expresses the author's contention that the morale (hence productivity) of the enforcement officer is closely related to his perception of the concern of his administrators for his welfare, and it describes certain steps that a wildlife officer contemplating retirement should carefully consider.

Effectiveness of the routine, public cooperation, group, and airplane patrol methods in apprehending closed-season deer violators was determined through questionnaires. Missouri conservation agents submitted 3,551 monthly questionnaires during the 2-year investigation. The mean arrest rates for the public cooperation, group, and airplane patrol methods were not significantly different; however, the mean arrest rate for routine patrols was significantly (P ≤ 0.05) lower than those for the other methods. Public cooperation patrols produced the highest rates of arrests; nearly 55% of the deer violator arrests were the direct result of citizens' complaints about deer violations.

The conservation officer's job should be to mold favorable public opinion while enforcing the conservation and boating laws. Those laws that are biologically sound ensure the public safety, and applied uniformly form the core of a sound enforcement program. Laws for the convenience of the officer or the administration have no place in the program. The power to invoke the criminal process carries with it the power for leniency. If the objective of all laws is voluntary compliance, then reasonable laws applied with the discretion of good common sense will accomplish that objective.

The United States is an important producer nation of wildlife and is one of the world's largest consumers of wildlife. The United States wildlife trade is briefly reviewed, primarily live birds, mammals, fish, reptiles, amphibians, and their products. This multi-million dollar business is growing larger each year causing greater demand for native wildlife and creating the need for each state to more closely monitor these commercial activities.

The stressful nature of law enforcement requires specialized selection and training for those who seek to make it a career. In recent years, psychological testing has been used by psychologists and law enforcement administrators as an aid in selecting successful candidates. An additional use for such testing might be to improve training in social interaction through selfunderstanding.

Occupational stress exists in all jobs. Wildlife conservation officers are, by nature of their jobs, subjected to job-related stressors. In order to better understand occupational stress, it is necessary to identify those events that occur in the professional lives, or as a result of the professional lives, of conservation officers. In response to this, all 160 conservation agents of the Missouri Department of Conservation were questioned by 2 separate mail surveys. These questionnaires allowed Missouri's wildlife officers to tell what was stressfUl to them in their jobs and evaluate these stressful events. Missouri's conservation agents identified and rated 95 events which produced stress in their occupation. These represent the Occupation Critical Life Events Scale for conservation officers in Missouri.

During the study period of 1 May 1981 through 30 April 1982, compliance rates (game arrests! 100 hunters checked) for individual hunting cases were computed for 2 groups of Tennessee wildlife officers. One group of 35 officers was driving high visibility, state-owned vehicles. The other group of 35 officers was driving low visibility, personally-owned vehicles. No significant difference was found between the compliance rates of the 2 groups.

The various benefits from a correctly timed drawdown on reservoir fisheries have been well documented; however, little has been written on the public relations aspect. An extreme (2.7 m -9 foot) drawdown on the most heavily populated lake in Arkansas is discussed, focusing on the procedures and problems involved with "selling" such a management technique to the public. By discussing the public relations work involved with the project, the problems that can arise and recommendations based on this experience, it is hoped that other management biologists planning similar operations will receive some insight into situations they may encounter.

One hundred and seventeen solutions to endangered species problems were presented by gifted students participating in the Georgia Future Problem Solving Bowl, a competition which encourages students to develop solutions to problems they will encounter in the future. The Nongame/ Endangered Wildlife Staff subjectively evaluated the students' attiutdes on endangerment based on their responses to Bowl questions. Solution responses overemphasized the impact of hunting on endangerment while demonstrating little knowledge of the impact of habitat degradation. Attitudes expressed by students strongly suggest serious deficiencies in wildlife education in Georgia schools. A well-developed endangered wildlife educational program in schools would help students assume their future role as land stewards. If wildlife is to remain an important part of the environment, wildlife professionals must ensure that Georgians understand the conservation of our natural resources.

To acquaint young people with fishing, and to provide an initial fishing opportunity, the Missouri Department of Conservation promoted "Kids Fishing Day" and a series of statewide fishing clinics as an expansion of its urban fishing program. Special emphasis is given to selecting clinic participants with little or no fishing experience, and particularly to disadvantaged or handicapped youngsters. Clinics are limited to 25 children to ensure individualized attention. Community organizations are solicited as co-sponsors to foster local involvement and to provide instructional assistance. The Department stocks fish, provides all fishing equipment, and a variety of fish and fishing literature. Fishing tackle companies provided a variety of items as gifts for all participants. In 1982, more than 2,000 children attended Kids Fishing Day programs, and another 900 attendsd the 36 clinics.

Representing the land grant institutions, extension fisheries and wildlife specialists contribute to management of fish and wildlife resources through education. In a telephone survey, 17 state specialists ranked their audiences by time spent working with each and ranked 10 communication methods by frequency of use in reaching each audience. In order of priority, the audiences served were extension agents, commercial interests, private landowners, youth, general public, faculty and students, natural resource agencies, and conservation organizations. The most frequently used communication means was the telephone call. Other heavily-used methods were personal letters, extension publications, on-site visits, workshops, and conferences. Least used were magazine articles and research publications. Moderate use was reported for newsletters, and radio and television programs.

With the creation of a waterfowl stamp in 1981, Texas Parks and Wildlife Department opted for a bid procedure rather than an art contest and established stringent specifications to select a successful bidder. The procedure paid off, and Texas made $1.2 million from the sale of stamps and art prints with its first year stamp. All revenue from stamp sale receipts may be spent only for research, management and protection of waterfowl; for the acquisition, lease, or development of waterfowl habitats in the state; or for grants as provided by the enabling legislation.

Although Mississippi currently has a hunter education program, there is no mandatory requirement. Therefore questions evaluating attitudes toward mandatory hunter education and hunter safety training backgrounds were included in a mail survey to a sample of 1981-82 Mississippi hunters. The majority (65.7%) of hunters would favor a mandatory hunter education program for new hunters. Graduates of a safety training course were significantly more in favor of mandatory training than non-graduates. Hunters with children had the same attitudes toward mandatory hunter education training as hunters without children. Approximately 18% of the 1981-82 Mississippi hunters had received hunter safety training.

In 1982, the Missouri Department of Conservation surveyed 370 volunteers for their experiences, perceptions, and opinions related to use of "hunter orange" during spring hunting of eastern wild turkey (Meleagris gallopavo silvestris). Results showed that gobblers can indeed be called and taken while a hunter orange item is being displayed, but hunters were ambivalent as to the effects of the band on heightening the alertness of turkeys. Subjects generally agreed that hunter orange served to alert other hunters to their presence, but were hesitant to say they "liked" the idea of using an item of hunter orange during spring turkey season.

Microhabitat variables were measured at 79 locations used by migrant common snipe (Capella gallinago). The maximum water depth used was 27 mm. Snipes demonstrated diurnal rhythms of habitat selection, using wet sites with short vegetation during feeding periods, and dry sites with taller vegetation during nonfeeding periods. Lower vegetation densities were selected in late afternoon. During feeding periods, 67% of snipes were within 15 cm of a soil/water interface, but during midday, 69% were more than 5.0 m from water. Twenty-three plants were associated with snipe microhabitats, but snipes selected microhabitats because of water conditions and vegetation structure, not vegetation composition.

The accuracy of Lincoln Index estimates for a wild population of mourning doves (Zenaida macroura) of known size was examined. Data indicated there was a tendency to overestimate population size regardless of the percentage of the population that was marked. Learned trap-escape behavior apparently caused the observed overestimations. A similar bias probably exists for other studies on birds.

Nesting success and production of mourning doves (Zenaida macroura) was studied on the Texas A&M University Campus during 1978 and 1979. Data indicated that mourning dove nesting and production were highly variable between months within years and monthly between years. Nests initiated in August were the most important both in quantity (20% of total) and rate of success (32% of total).' Although September-initiated nests contributed only 5% of total nests and 6% of fledged young, August-initiated nests still active during early September contributed 20% of the total fledglings in 1978 and 26% in 1979.

Aerial surveys were used to determine numbers of wintering waterfowl and their habitat preferences on Sam Rayburn Reservoir in East Texas during the winters of 1979 - 81. The magnitude of relative preference for 6 wetland types was defined as the mean difference between the ranks of wetland use and availability. A total of 23,277 waterfowl representing 14 species was tallied from 12 surveys. Mallard (Anas platyrhynchos), gadwall (A. strepera), American widgeon (A. americana), and teal (Anas spp.) were the most abundant species and composed 89.0% of the total Ducks preferred scrub-shrub, emergent wetlands, and aquatic beds over unconsolidated bottom and shore, live forested wetlands, and dead forested wetlands (P < 0.05). Results suggest that waterfowl habitat management should be primarily considered in the design and construction phase of reservoir projects.

We banded 1,987 Canada geese (Branta canadensis) at Ballard Wildlife Management Area (WMA), Kentucky 1974 - 81. Additionally we observed 1,459 neckbanded individuals at Ballard WMA 1977 - 81 and obtained 11,649 observations of these individuals in the Mississippi Flyway. Analysis of these observations and 195 hunting recoveries indicated a strong association between Ballard WMA and Wisconsin. Average annual survival was estimated as 72% for geese banded at Ballard WMA 1974 - 79. Of 783 movements between refuges in southern Illinois and Ballard WMA, 73% were movements into Kentucky. Sixty individuals originally banded on the Hudson-James Bay coast were observed at Ballard WMA 1977 - 81. The origin of these individuals does not suggest a strong association between Ballard WMA and a particular segment of the breeding range.

Drive counts of pheasants (Phasianus colchicus subsp.) were evaluated on playa basins in the High Plains of Texas. The counts were 97-99% precise when the density of counters averaged 1/0.85 ha. The change-in-ratio estimator was modified to estimate pre-hunt population size when post-hunt population size and the proportion of cocks in the harvest are assumed known. Problems associated with drive counts of pheasants are discussed. Costs averaged $38/playa basin ($1.52/ha).

Arkansas residents and personnel of the U.S. Forest Service, U.S. Fish and Wildlife Service and the Arkansas Game and Fish Commission reported 1,253 observations of black bear (Ursus americanus) and signs of their activity from 1976 through 1980. Cub:sow ratios ranged from 1.85 to 2.18 and averaged 2.06. Percent cubs observed ranged from 19.6 to 23.6 and averaged 22.4. Nuisance bear complaints increased during this period. Survey techniques and implications for present and future black bear management are discussed.

Eighty-one river otter (Lutra canadensis) carcasses were collected from trappers and a fur dealer in Georgia during the 1979 - 80 and 1980 - 81 trapping seasons for environmental pollutant analysis. Mercury occurred in all muscle, liver, and hair samples analyzed with ranges of 0.2 to 13, 1.2 to 60, and 7.1 to 155 ppm, respectively. Detectable cesium-137 occurred in 77% of the samples ranging from less than 141 to 6,660 pCi/kg. DDT and its metabolites were detected in 97% of the otter fat samples at levels ranging from 0.08 to 91.90 ppm. Mirex, dieldrin, and PCBs were found in 68, 59, and 66% of the samples with ranges of 0.16 to 75.40, 0.03 to 1.26, and 0.57 to 84.20 ppm. Muscle, hair, and liver mercury levels, cesium-137 and mirex levels were all significantly higher in the coastal plain than in the piedmont region of Georgia.

Beaver pelts from the Mississippi Beaver Cooperative fur sale held during March 1980 were examined to determine both the extent and origin of the damage responsible for downgrading. Of 1,914 pelts in the sample, 1,255 (65.6%) were significantly damaged. Damage inflicted by beaver accounted for 27.5% of the total. Skinning and fleshing cuts accounted for 31.5% of the damage. All other types plus those of unknown origin totaled 41.0%. Methods of dealing with the damage problem are discussed.

Food habits of cottontail rabbits (Sylvilagus floridanus) were determined by microscopic analysis of stomach contents from 2 physiographic regions of Virginia during 4 seasons of the year. The area examined in the mountain region (Radford Army Ammunition Plant [RAAP)) contained a high population of rabbits whereas the area studied in the piedmont region (Fort Pickett) was an area with a history of relatively low rabbit populations for the past 2 decades. Food habits of rabbits from the 2 areas were different; rabbits from Fort Pickett subsisted on much greater quantities of forbs throughout the year than RAAP rabbits which consumed primarily grasses. Collectively, leaf and stem parts of grasses and forbs made up well over 90% of total food items found in stomachs throughout the year. At Fort Pickett grasses occurred in greater volumes in winter and spring, whereas forbs were more numerous in summer and fall. Grasses predominated in all seasons at RAAP.