Journal of the Southeastern Association of Fish and Wildlife Agencies

A USDA program to eradicate the introduced sugarcane root weevil in the spring of 1969 was studied to assess effects on fish and wildlife. Two basic techniques were utilized: analyses of residue accumulations in selected species, and search for dead animals. Residues increased substantially in birds but not in earthworms, fishes, or aquatic invertebrates. Considerable bird mortality followed treatment and residues in specimens analyzed indicated death from heptachlor poisoning.

During 1961 and 1962, 311 wild-trapped black francolins (Francolinus francolinus asiae) from India were released in the Gum Cove area of Cameron Parish in southwest Louisiana. A substantial bobwhite quail (Colinus virginianus) population was present on the release site. In 1967, ten francolins and ten quail were collected for comparison of food habits and parasite fauna. Differences in food habits were noted but an occasional preference of francolins and of bobwhites for the same food items was revealed. This was not believed to be a limiting factor for either population. Parasite burdens were generally low, and the parasite fauna of each species was quite distinct.

Bobwhite age and weight data were collected from 5 study areas in North Central Texas to determine characteristics contributable to optimal harvest. A total of 24,324 bobwhites were aged and 12,735 were weighed and examined during December and January (hunting season), 1962-67. The_peak in nesting as shown by molt patterns occurred prior to July 15 each year. The mean weight of all birds was 176.2 - a standard error of .5 grams. Weights increased through December and began to decline during the final 3 weeks of January. Average weights of January-killed birds did not decline below the average weight for the study period. It is submitted that bobwhite weight and physical condition can be utilized as criteria for harvest and subsequent management of the species. The harvest beginning date should coincide with the average date on which 90 percent of juvenile birds reach the 150 gram acceptable size. This date during the period in North Texas was November 25.

The American woodcock is a neglected game bird in most southeastern states. Reasons for disregarding this species include: a belief that woodcock are sparse in areas other than Louisiana and southwestern Mississippi; little knowledge pertaining to woodcock habitat; the practice of only hunting woodcock incidentally to other game; and a consensus that quail dogs cannot be used for hunting woodcock. During field activities associated with a rangewide study of parasitism in woodcock, surprisingly high populations were revealed in South Carolina, Georgia, Florida, Alabama, Arkansas, and southeastern Mississippi. Woodcock habitat was characterized and suggestions were presented for alerting southeastern hunters to the presence and value of this game bird.

During spring and summer in 1969 and 1970 approximately 200 observations were made of wild turkey (Meleagris gallopavo) brood roosting sites, most of which were found by radio-telemetry. The age that broods began to roost in trees varied from 12 to 19 days for the 14 broods. Most of the ground roosts were located under forest canopies in sparse ground cover. Brood hens normally did not defecate in their ground roosts. After tree roosting began, broods utilized cypress (Taxodium ascendens and T. distichum) and pine (Pinus palustris and P. elliottii) more than all other trees combined. The first night off the ground was typically spent on a horizontal limb 2 to 3 inches in diameter about 22 feet above the ground. Within three days they began to roost higher in the trees, but roost limb diameters were about the same size. Most of the roost trees were over water.

Daily movements and brood rearing among wild turkey (Meleagris gallopavo silvestris) hens in the Piedmont of eastern Alabama was studied during 1968 and 1969. Twelve hens were tagged in the patagium and instrumented with radio transmitters. Daily movements did not cover the entire spring and summer ranges. Daily movements of non-nesting hens averaged 0.4 mile; those of nesting hens averaged 0.22 mile; and those of hens with broods averaged 0.32 mile. They were usually linear and were strongly influenced by feeding areas provided by permanent pastures. All brood rearing ranges were within grazed permanent pastures and adjacent grazed woodlands. Management suggestions for brood rearing habitat based on range and movement data are given.

Body weight and basic body measurements were recorded on 395 adult cottontails collected from three physiographic regions in Georgia. Eight different skull measurements were taken on 65 individuals. Little variation was noted in the percentage change of paunched weights as compared to the animals' total body weight. No significant seasonal variations were noted in total body weight. Coastal Plain adults exhibited significantly greater hind foot length, ear length, total length, and body weight than Piedmont or Mountain rabbits. Six of eight Coastal Plain skull measurements were significantly greater than measurements from either Piedmont or Mountain cottontails. These data are in opposition to Bergmann's Rule which states that mammals in general increase in size as one proceeds northward. Use of total body weight rather than paunched weight at any time of day or season was verified for cottontails in Georgia.

A field test to evaluate anticoagulant rodenticides as a method of controlling raccoon populations was carried out on the Cape Romain National Wildlife Refuge, South Carolina. Fumarin mixed with corn was dispensed at ten permanent feeding stations over a six-week period. A marked decline in the number of raccoons was noted on the study area as a result of the experiment. Feeders for the study were developed by refuge personnel and proved to be relatively bird and mouse proof, but easily accessible to raccoons. In addition to the field study, observations were made of six caged raccoons fed varying amounts of rodenticide to determine lethal dosage and the length of time required to bring death.

The cannon-net technique for trapping deer was used to capture 130 wild fallow deer (Dama dama) in western Kentucky. The net proved dependable as deer were captured on 25 of 27 times it was fired. The average catch was 4.8 per attempt and up to II animals were captured at one time. Mortality was 3.1%. Acepromazine maleate was used with satisfactory results for tranquilizing captured deer.

A study was conducted on the A.E.C. Savannah River Plant, Aiken, South Carolina, to determine if eye lens weight was a reliable indicator of age in feral swine. A strong correlation (r=0.95) was found between eye lens weight and tooth irruption age. In most cases, when body weight failed, eye lens weight remained a dependable indicator of age in feral hogs. Eye lens weight can therefore be used as a reliable age criterion in feral hogs. With the establishment of an eye lens weight - to - age curve based on known-age animals, eye lens weight may prove to be better than tooth irruption for indicating the age structure of a feral hog population.

Several Southeastern states using mail surveys currently apply an estimation procedure to adjust for non-response bias. Efficient planning of these surveys requires an idea of the precision obtainable for a specified number of questionnaires mailed, i.e. an idea of the sample size required. This paper graphically describes the empirical relation between sample size and relative precision of estimates of hunters, effort, and harvest for three game species, deer, turkey and dove. The data were selected from records on hand from six states and over the period from 1964-1969.

In November, 1969, four groups of channel catfish, Ictalurus punctatus (Rafinesque), with 20 fish per group were marked and put in cages in an earthen pond at the Auburn University Agricultural Experiment Station, Auburn, Alabama. Marking methods consisted of four dyes injected and applied with a tattoo machine, two fluorescent pigments and a hot and a cold brand. The fish were sampled weekly for a period of 28 weeks and each mark was evaluated on the basis of mark retention and ease of identification. Blue latex injections, hot brands and cold brands were the most effective marks used in this study.

A survey procedure was developed for judging the relative quality of trout waters and used in surveying streams in National Forests in Georgia, Virginia and South Carolina. The purpose of the survey was to provide land managers with information that would help in making decisions affecting trout streams. The procedure incorporates water quality analysis, fish population sampling, and a comparative evaluation of habitat conditions based on a stream's present and potential capability for producing wild or "native" trout fishing. Drainages surveyed ranged from one to over twenty miles in length.

In an effort to supplement the food of fishes in remote areas lacking power lines and also to short-circuit the natural food chain, an automatic selfcontained DC feeder was designed and built. The paper deals with the construction details of this feeder.

The results from gill nest samples, a series of cove rotenone samples, and a creel census were compared to the total fish population obtained through the drainage of Lake Russell, a 90-acre reservior in Northeast Georgia. An extended period of gill netting yielded the best representation of the various fish species present in the reservoir, but gave a poor estimate of the relative abundance of the fish species. Cove rotenone samples of known acreage gave a good representation of the total weight per acre of the reservoir. The three coves sampled yielded 45.3 pounds per acre, 48.5 pounds per acre, and 73.0 pounds per acre. The three cove samples combined yielded a value of 54.1 pounds per acre. Upon drainage of the reservoir the total weight per acre of all fish was 48.4 pounds per acre. The deepest of the three coves, lacking an abundance of aquatic vegetation, was overestimated in terms of the actual weight per acre.

The large number of meter net samples needed to determine the spawning success of various species of fish in Canton Reservoir, Oklahoma, made subsampling advantageous. The basic Folsom plankton splitter was enlarged and modified so that meter net samples with volumes up,to 4,000 m\\. could be split into IO approximately equal subsamples. The splitter was constructed from a 12 inch diameter Plexiglas2 cylinder. Construction was accomplished using common shop tools. Chi-square tests (0.05 level) showed that there were no significant differences between the observed subsample counts and the expected counts. A nonparametric sign test showed that each chamber did not consistently have higher or lower counts than any other chamber. The minimum total number of organisms per sample that could be subsampled yielding estimates of the total sample number with less than a IO percent error 95 percent of the time were determined for larval gizzard shad and larval Chaoborinae.

Ecological changes in plant communities in several Florida lakes have been noted by biologists in the past, but rarely have the parameters of these communities been documented for future reference. Soil moisture is an important factor in determining which plants will occur at particular elevations, but more important is the inundation tolerance of the various species. The purpose of this study was to compare the long range effects of water level fluctuation on the marsh vegetation of Lake Okeechobee. The Florida Game and Fresh Water Fish Commission, in 1956, conducted a vegetational analysis of the lake marsh in order to predict the probable results on vegetation from construction of a levee to contain the lake waters. A subsequent analysis of the same area was conducted by Commission personnel in 1969 to determine what vegetational changes had occurred.

Alimentary tracts were examined of 1010 carp collected with gill and trammel nets by commercial fishermen from four riverine reservoirs, and 211 adult and 45 young carp from Lake Carl Blackwell, a headwaters reservoir. The alimentary tract contents of carp in five Oklahoma reservoirs contained algae, plant fragments, seeds, entomostraca, chironomids, Chaoborus, pelecypoqs, caddisflies, Ceratopogonidae, animal fragments, and organic and inorganic matter. Terrestrial insects were of rare occurrence. The major volumetric constituents in the alimentary tracts of carp from the riverine reservoirs Grand, Fort Gibson, Eufaula, and Texoma Reservoirs were: unidentified organic matter (65.5%), animal fragments (12.5%), plant fragments (10.1%), entomostraca (3.8%), algae (1.9%), pupae (1.1%).

Food items were found in 47.0% of 1329 flathead catfish stomachs collected by gill and trammel nets from six Oklahoma reservoirs: Carl Blackwell, Eufaula, Fort Gibson, Grand, Hudson, and Texoma. The average number of food items was 1.6 per stomach and the average volume per stomach was 26.4 ml in stomachs with food. Fish comprised more than 95% of total food volume and total number of food items in all six reservoirs. Gizzard shad contributed from 49.5 to 91.7% of total stomach volumes. Freshwater drum were second in importance as forage, accounting for 3.3 to 38.2% of total volumes. Carp contributed 23.2 and 42.0% of total stomach volumes on lakes Eufaula and Hudson, respectively, but were not found elsewhere. Channel catfish comprised 13.8% of the stomach volume in flatheads from Fort Gibson Reservoir where they were very abundant. All species of centrarchids comprised only 5.4 to 10.0% of total stomach volumes in three reservoirs.

In many reservoirs predation on stocked fish has been considered as one of the major limiting factors in establishing a particular population. Following several years of water quality determinations it was found that Lake Ouachita maintained a sufficiently oxygenated hypolimnion to support trout. As a result several thousand catchable rainbow trout, Salmo gairdneri, were stocked into the lake. The results of this attempt to establish a trout fishery have been disappointing for a combination of reasons; however, while collectint for brookstock chain pickerel, Esox niger, and from fisherman reports, it became evident that predation on the stocked trout must be very high. Collections of largemouth bass, Micropterus salmoides, and chain pickerel were made with a boom-type electro-shocker during January and February, 1970. Sampling was done at night within an approximate 100-acre area adjacent to a trout stocking point.

Limnological data collected during a four and one-half year period at Lake Sinclair is reviewed. Comparison of temperature and oxygen profiles are made in areas of the lake receiving hot water discharge and in unaffected areas. Hot water discharge was found to increase temperature and dissolved oxygen substantially during both summer and winter. In addition, hot effluent eliminated or altered the thermocline in the affected area. These effects were evident in varying degrees for 2.3 miles below the source. Observations concerning the sport fishery within the area receiving hot water discharge are noted. The possible effects of hot water discharge on the habitat and an aquatic ecosystem is also discussed.

Stomachs from 285 longnose (Lepisosteus osseus) and 603 Florida gar (Lepisosteus platyrhincus) were collected from five lakes in central Florida. Ofall stomachs examined, 432 or 49 per cent contained food items. Stomach data grouped from the two species and five lakes indicated that gar in central Florida lakes feed principally on nongame forage fish and do not extensively utilize game fish species. Nongame forage fish comprised 57 per cent of the total number of food items consum~d, while only II per cent were game fish.