Journal of the Southeastern Association of Fish and Wildlife Agencies

Food habits of sympatric and allopatric populations of adult brook trout (Salvelinus !ontinalis) and rainbow trout (Oncorhynchus mykiss) from 4 streams in the Great Smoky Mountains National Park were compared to determine if competition for food resources might explain the apparent exclusion of the former by the latter in Southern Appalachian streams. When diets of brook trout in sympatry with rainbow trout were compared with diets of brook trout in allopatry, there was little significant change in prey composition. Based on this information, exploitation competition for food resources does not seem to play a role in the invasion of rainbow trout into areas of Southern Appalachian streams occupied by brook trout.

We compared estimates of northern bobwhite (Colinus virginianus) density derived from line transect (LT) and capture-removal (CR) sampling with each other and with 2 independent indices of abundance (coveys flushed/party-hour of hunting and captures/l00 trap-sets). CR estimates were not correlated with either index (P ≥ 0.371), whereas LT estimates were correlated with captures/100 trap-sets (P = 0.009) and not with coveys/party-hour (P = 0.156). Although CR and LT estimates were not correlated (P = 0.288), the estimators gave similar (P > 0.05) estimates of density in each of 4 years of study, and estimates differed by <22% in 3 non-drought years.

Counts of whistling male northern bobwhites (Colinus virginianus) were conducted on standardized routes from 1981-86 at Fort Bragg Military Reservation, North Carolina. Harvest data were collected at check stations from 1967-86, and were used to monitor population fluctuations. A linear relationships (r2 = 0.89, P < 0.01) was observed between the mean number of quail heard per station and the subsequent fall harvest. The call index and hunter success also were related (r2 = 0.73, P < 0.03). A negative trend (b = -0.51, P < 0.001) in the mean number of calling males per station was recorded from 1981-86. Whistle-count surveys may provide a cost-effective method for predicting fall harvest levels for controlledaccess management areas.

A known population density was established for northern bobwhites (Colinus virginianus) on a lOO-ha pasture. Line transects were walked twice/day for 5 days to obtain a density estimate. Twelve density estimators using the linetransect data were compared to the known bobwhite density. On our study area, 6 estimators overestimated bobwhite density and 6 of them underestimated density. Because few coveys were flushed, individual transect replicates had to be pooled to provide estimates. This pooling of replicates did not enable us to calculate a mean density and specific variances for each estimator. Variances produced by a Jackknife method appeared to underestimate the true variances. During years of low bobwhite densities (which are common), use of line-transect methods may be inappropriate. Capture-recapture data, collected 1 week prior to running the line transects, underestimated quail density on the pasture.

We developed and evaluated 7 least-squares regression models to estimate body fat and protein reserves from external measurements of post-breeding male and female lesser prairie-chickens (Tympanuchus pallidicinctus). For males, protein was adequately modeled (R2 = 0.60), but attempts to model body fat were unsuccessful. For females useful fat (R2 = 0.92), log fat (R2 = 0.74), condition index (R2 = 0.58), and protein (R2 = 0.74) models were derived. These models can be used to compare post-breeding lesser prairie-chicken nutrient reserves among populations and habitats, without sacrificing the birds sampled.

Periodic observations from 1976 through 1989 on an unhunted whitetailed deer (Odocoileus virginianus) herd located in coastal South Carolina indicated that the population had likely incurred a major decline from a hemorrhagic disease (HD) outbreak, and another may be imminent. Weights of adult does were comparable to those of the hunted herd on the nearby Francis Marion National Forest. Conception rates were insensitive to the occurrence of HD or acorn mast failures. Among adult does collected in early February, femur marrow fat content was a better indicator of animal condition and the abundance of acorns during the previous fall than perirenal fat or serum cholesterol levels. Significant (P < 0.05) differences in blood urea nitrogen and albumin concentrations seemed to be negatively associated with prevalence of HD and positively associated with fall acorn abundance.

Ages of 76 known-age white-tailed deer (Odocoileus virginianus) were estimated by tooth wear and replacement (TWR) and by incisor cementum annuli (CA) determination. TWR and CA methods gave similar results (75% and 71% correct, respectively). However, accuracy was dependent on deer age, with TWR being more accurate than CA in age classes <3.5 years and CA being more accurate in age classes >3.5 years. Accuracy of 55 southeastern United States biologists in estimating age of 98 known-age deer jaws also was determined. Results indicated biologists were well-skilled in the TWR technique but generally underestimated actual age class >3.5 years. Using their mode response (N = 98), biologists correctly determined age for 71.4% of the jaws; however, when using their estimates given for all jaws (N = 5,390), 62.6% were aged correctly.

We evaluated how well observers in a helicopter classed male white-tailed deer (Odocoileus virginianus) into groups ≤3.5 or ≥4.5 years old. Resightings of individually marked anti previously aged deer on 2 southern Texas ranches were used for evaluation during repeated helicopter surveys from 1986 to 1988. Classifications of marked males were at least ≥90% correct on 11 of 14 flights on 1 ranch and 6 of 14 on the other. Although accuracy was acceptable on many flights, some flights gave inaccurate estimates (> 10% misclassified) of male composition. Because managers make only a single, annual helicopter flight, we recommend male classifications made during helicopter counts be used as trend data and be evaluated over a number of years. Managers should avoid using male composition data from individual flights in calculations of harvest rate for a particular age class.

We investigated 4 in vitro digestion procedures to estimate dry matter, energy, and apparent protein digestibilities of 2 southern Texas white-tailed deer (Odocoileus virginianus) forages. Standard 2-stage rumen inoculum technique consistently underestimated in vivo dry matter digestibility by 1% to 9%, but could be corrected to in vivo values by regression analyses (R2 = 0.89). This technique also predicted digestible energy (kcal/g) and digestible protein (g/100g feed) accurately (R2 = 0.84 and 0.71, respectively). Cellulase, pepsin, or multienzyme techniques did not predict dry matter, energy, or protein digestibility accurately or consistently.

We surveyed 611 rural landowners in the Arkansas Coastal Plain in 1987 to detennine perceptions of white-tailed deer (Odocoileus virginianus) damage to agricultural and forestry crops. Respondents (N = 231) owned an average of 148 ha, with the 2 greatest areas in row crops (98 ha), such as soybeans, cotton, and rice, and forests (38 ha). One-half (50%) of respondents had sustained deer damage, most (52%) of whom described it as minimal. Landowners who estimated financial losses from deer reported an average loss of $1,650: 61% lost <$1,000, and 1% lost >$10,000. Of those with damage, 23% said that damage was unreasonable and 46% had tried control. Respondents most often used scare devices, chemical repellents, and fences, but few respondents (<3%) felt that these devices worked. Most respondents (71%) wanted deer on their land, 15% wanted deer but they worried about damage, and 7% did not want deer.

An improvement cut that removed commercially low-quality trees from an unmanaged 20-ha, 60-year-old, mixed mesophytic forest in Kentucky reduced the availability of cavities, snags, and small seeds. Stand basal area was reduced from 21 to 17 m2/ha. The abundance of primary and secondary cavity-using birds, as well as neotropical, migrant songbirds, was not affected severely by the cut. Both winter and breeding populations of primary cavity-using birds, breeding great crested flycatchers (Myiarchus crinitus), tufted titmice (Parus hicolor), and Carolina wrens (Thryothorus ludovicianus) were unaffected for 3 years after the cutting. Winter abundance of tufted titmice decreased immediately after the cutting and remained low for at least 3 years. The abundance of breeding Carolina chickadees (Parus carolinensis) decreased for 2 years after the cutting. Of 25 migratory songbird species evaluated, only ovenbirds (Seiurus aurocapillus) decreased in number.

We documented red-cockaded woodpecker (Picoides borealis) occupation of cavity tree clusters on North Carolina study areas from 1980-88. Most occupied cavity tree clusters (93%) were used in consecutive years, but 6.8% were abandoned each year. Abandoned sites were captured or reoccupied at an annual rate of 8.7%, implying there is a 60% chance that an abandoned site will be reoccupied in a 10-year period. However, recently abandoned clusters were reoccupied at a much higher rate than long-abandoned clusters. Cluster occupation was dynamic with some being occupied, abandoned and reoccupied, or captured during our study. Cluster use by solitary males often was intermediate temporally between group occupation and abandonment. Habitat enhancement may improve reoccupation rates and could be an important short-term strategy to increase the number of social units of this endangered species.

We monitored 18 southern fox squirrels (Sciurus niger niger) by telemetry in coastal South Carolina to determine habitat use in relation to availability of habitat types within observed home-range areas. Fox squirrels used leaf nests primarily, although females preferred to nest in cavities during winter. Nests were concentrated in hardwoods with a preference for oak (Quercus spp.) and gum (Nyssa spp.) species with >20 cm DBH (diameter breast height). Fox squirrels preferred (P < 0.01) hardwood, mixed pine-hardwood, and pine with a hardwood midstory habitat types, and ecotone areas.

We used correlation analysis to test for relationships between habitat types and numbers of courting American woodcock (Scolopax minor) recorded along singing-ground survey routes in eastern Texas. Availability of 10 different habitat types was quantified at each survey route stop and compared to the number of woodcock recorded. Numbers of courting woodcock recorded were strongly correlated to pine (Pinus spp.) seedling and pine sapling habitats. Results may prove helpful in monitoring long-term trends in singing-ground habitat availability.

We studied habitat used for nesting by eastern wild turkey (Meleagris gallopavo silvestris) on intensively managed pine forest land in eastern Texas in 1980 and 1981. Most (89%) nests were within 100 m of a road or timber stand boundary. Nest sites had <60% canopy closure, low density of rnidstory trees (16 stems/0.04 ha), low basal area (9 m2/ha), and abundant herbaceous ground and shrub cover. Retaining forest openings resulting from logging activity likely will enhance nesting habitat for turkeys in eastern Texas.

A "home range" analysis using 23,285 observations of 6,036 neck-banded Canada geese (Branta canadensis) seen from 16 October to 28 February 1984-87, identified wintering concentration areas within the Atlantic Flyway. Biweekly sampling of 4 cohorts affiliated with these concentration areas during the last 2 weeks of January revealed that >90% of the geese wintering in the Chesapeake Bay and western Pennsylvania regions arrived in these areas by early October. Most geese (>90%) sampled in central New York appeared by early November, while 90% of the sample of birds wintering in North Carolina were not present until early December. Fifteen to 25% of the North Carolina sample was observed in the Chesapeake region prior to their arrival in North Carolina.

Visual observations of marked individuals in a resident flock of Canada geese (Branta canadensis) on the Eufaula National Wildlife Refuge, AlabamaGeorgia, were conducted from July 1980 through May 1982. Associations among geese were examined during 2 periods (breeding, February-May; and nonbreeding, June-January). Social groupings were determined for 231 geese in 1980-81 and 240 in 1981-82 based on 4,577 observations during the breeding season and 666 observations during the nonbreeding season. Adult pairs that nested successfully remained with their young-of-the-year throughout the nonbreeding season and composed families; and pairs that did not nest successfully usually joined pseudofamilies, social groups of adult and yearling geese. Pseudofamilies were the predominant social units (80%) in the flock. All social units except yearling groups dissociated during the breeding season, and most pairs established nesting territories or dispersed from the area.

We studied fall and winter food habits of Canada geese (Branta canadensis) in the Swan Lake Zone of north-central Missouri during the 1980-82 hunting seasons. Gizzards from hunter-killed geese were obtained during 5 periods of each hunting season. Average proportions of native foods, row crops, winter wheat, and other forage were highly variable among and within years. Row crops in the diet ranged from an average of 1.1%to 99.6% by volume, compared to 0.3% to 98.8% for native foods. In general, row crops increased and native foods decreased in average percent volume through the hunting season, except in 1982, when floods during the hunting season likely kept native foods available all season. Winter wheat (Triticum aestivum) comprised up to 26.2% average volume. Cold temperatures and snow and ice accumulations increased use of row crops and decreased use of native foods within a season.

Nest box use, nesting success, and nesting habitat characteristics were examined on 6 study areas in eastern Texas during 1984-85. Box use averaged 29% and nesting success decreased from 82% to 64% over the 2-year study. Nest box characteristics selected by wood ducks (Ab: sponsa) included a lower clear tree bole, orientation of the hole toward open or vegetated water, and proximity of 1 or more additional nest boxes. Habitat characteristics selected included high visibility of the box, closeness to open water, less canopy closure (%), presence of herbaceous marsh within 1 km, and presence of herbaceous wetlands (Le., floating vascular aquatic bed, rooted vascular aquatic bed, and open water). A discriminant function analysis (DFA) model was developed for classification of nesting habitats for wood ducks in eastern Texas.

Fifty of > 1,000 landowners participating in the South Carolina Wildlife and Marine Resources Department (SCWMRD) wood duck (Aix sponsa) nest box program were selected at random to evaluate the program's effectiveness. We inspected 311 of 8,561 boxes bi-weekly throughout the 1987 nesting season. Most (95%) of the boxes allocated had been installed, and 82% of those installed were considered acceptable for use by wood ducks. Most (90%) cooperators had at least 1 nest start, and 69% of the available boxes were used. There were an average of 6.5 ducklings produced per available box. Estimated duckling production in 1987 from all boxes in the SCWMRD program was 42,833 ducklings. Cooperator nesting activity reports provided an accurate estimate of duckling production.

Aerial survey data (1968-1989) and water gauge readings (1958-1989) were examined to detennine trends in and relationships between canvasback (Aythya valisineria) populations and water levels at Catahoula Lake, Louisiana. Wintering canvasback populations at Catahoula Lake have increased over the past 21 years. A peak population estimate of 78,000 canvasbacks was recorded in January 1988. There was a significant relationship between increases in wintering canvasback numbers and increased November water levels. No significant changes in December or January water levels nor relationships between wintering canvasback numbers and December or January water levels could be detected. In recent years, Catahou1a Lake has become one of the most important canvasback wintering areas in North America. We conclude that the increased availability of open water feeding habitat has been a major factor in the increased usage of Catahoula Lake by canvasbacks.

We examined the influence of age and sex on body composition and organs of lesser scaup (Aythya ajfinis) during mid-winter in southwestern Louisiana. Except for intestine weight, organs and body components were similar in size between adult and immature lesser scaup. Body components and organs of males generally were larger than those of females. Body weights of lesser scaup in our study were lower than those reported from the same area more than 20 years ago. Declines in condition could result in lowered survival or reproductive performance. Information on winter survival is not available, but harvest age-ratio data suggest that recruitment of lesser scaup has declined in the Mississippi Flyway.

Late winter dabbling duck use of experimentally manipulated saltmarsh bulrush (Scirpus robustus) was evaluated in managed brackish wetlands in the Santee River Delta, South Carolina, during 3-21 February 1989. Three 1-ha plots were established in each of 3 treatments: saltmarsh bulrush that was dewatered, burned, and reflooded (burned, flooded bulrush [BFB]); saltmarsh bulrush that was dewatered and then reflooded (unburned, flooded bulrush [UFB]); and low stemdensity saltmarsh bulrush that was flooded throughout winter (winter-flooded bulrush [WFB]). Species and numbers of dabbling ducks using treatments were identified and counted. Densities of dabbling ducks using BFB (x¯ = 100 ± 25.60 [SE], N = 12) and WFB (x¯ = 65 ± 8.28, N = 12) were similar (P > 0.05); densities of dabbling ducks using UFB (x¯ = 5 ± 1.82, N = 12) were significantly lower (P < 0.05) than for the other treatments.

We compared climatic and pond-management parameters with aerial census estimates of waterfowl and American coots (Fulica americana) on randomly selected clusters of channel catfish (lctalurus punctatus) impoundments in the Delta Region of Mississippi from November 1983 to March 1984 and from October 1984 to March 1985. Mean total waterfowl and coot estimates for 1983-84 and 1984-85 were 51,853 and 81,325, respectively. Numbers of lesser scaup (Aytha affinis) increased four-fold the second winter. Weekly waterfowl and coot populations fluctuated with temperature, rainfall, and ice over. Waterfowl and coot numbers were positively correlated with seining, distance of ponds from disturbance, and the presence of aquatic vegetation and trash fish, and negatively correlated with hunting pressure. More coots used recently drained ponds than ponds drained 2 to 4 years prior to surveys.

The Habitat Suitability Index (HSI) model (Lewis and Garrison 1984) for wintering American black ducks (Anas rubripes) was applied during 1985 and 1986 at Chincoteague, Virginia. HSI values of 0.66 and 0.56 were obtained during the 2 respective years. We suggested improvements in field methods for estimating biological variables. Modification of 1 variable (V6) and inclusion of a new variable (proportion of total land and water represented by saltmarsh, Vg) were proposed to improve model output.