David S. Gilmer

Survival of wood duck and mallard broods in north-central Minnesota

Duckling survival in wood duck (Aix sponsa) and mallard (Anas platyrhynchos) broods was estimated from data obtained from 71 radio-marked brood hens on a study area in north-central Minnesota. Radio-marked hens produced 30 broods during the study, and 41 hens already leading broods were captured and radio-marked. Production estimates based on brood size counts were inflated by about 38 percent for wood ducks and 30 percent for mallards if total-brood losses were not taken into account. Mortality during the first 2 weeks of life was most severe, accounting for 86 percent of total recorded mortality in wood ducks and 70 percent in mallards. Rearing success was about 41 percent for wood ducks and 44 percent for mallards. Duckling survival was negatively correlated with distance of overland travel by young broods. Wood ducks had a shorter hen-brood bond than mallards, and presence of the hen appeared to affect duckling survival less in wood duck broods than in mallards.

Survival of radio-marked mallard ducklings in Northeastern California

Estimates of duckling survival are necessary to accurately assess recruitment of mallards (Anas platyrhynchos), yet few reliable estimates exist. During 1988-90, we estimated survival rates for 127 radio-marked mallard ducklings from 64 broods on Lower Klamath National Wildlife Refuge, California. In 1988, we restricted the survival estimate to the first 10 days post-hatch (S = 0.18, SE = 0.07). Survival from hatching to 50 days was 0.37 (SE = 0.09) in 1989 and 0.34 (SE = 0.07) in 1990. Total brood loss differed among years (P < 0.05); 81.2% in 1988 (n = 16), 36.8% in 1989 (n = 19), and 37.5% in 1990 (n = 24). Ninety-three percent of mortality occurred during the first 10 days of life. We detected no differences in the proportion of radio-marked ducklings fledged from early-hatched versus late-hatched nests (P = 0.74)

Feeding ecology of waterfowl wintering on evaporation ponds in California

We examined the feeding ecology of Northern Pintails (Anas acuta), Northern Shovelers (A. clypeata), and Ruddy Ducks (Oxyura jamaicensis) wintering on drainwater evaporation ponds in California from 1982 through 1984. Pintails primarily consumed midges (Chironomidae) (39.3%) and widegeongrass (Ruppia maritima) nutlets (34.6%). Shovelers and Ruddy Ducks consumed 92.5% and 90.1% animal matter, respectively. Water boatmen (Corixidae) (51.6%), rotifers (Rotatoria) (20.4%), and copepods (Copepoda) (15.2%) were the most important Shoveler foods, and midges (49.7%) and water boatmen (36.0%) were the most important foods of Ruddy Ducks. All three species were opportunistic foragers, shifting their diets seasonally to the most abundant foods given their behavioral and morphological attributes.

HABITAT USE AND HOME RANGE OF MALLARDS BREEDING IN MINNESOTA

Telemetry techniques were used to study habitat use and home range of 12 drake and 12 hen mallards (Anas platyrhynchos) in the forested region of north-central Minnesota during the 1968 to 72 breeding seasons. Circumneutral bogs and seasonal wetlands were the most frequently used communities; the lakeshore communities most used were sand-gravel, overhanging brush, and bog mat. Based on the availability of habitat, the highest preference shown was for the seasonal community and the least for the softwood swamp; however, all communities were used to some extent during the breeding season. Lakeshore was always among the habitats most frequently used by pairs. Twelve nests were found, nine in nonpermanent wetlands and three in upland forest sites. Mean home range was 210 ha for hens and 240 ha for drakes. The long axis of the home ranges of drakes and hens averaged 2.8 and 2.7 km, respectively. Hens had smaller home ranges during the laying period (anti x = 70 ha) than during prenesting (anti x = 135 ha).

Movements and habitat use of mallard broods in northeastern California

To increase recruitment of mallards (Anas platyrhynchos), wildlife managers must understand the habitat and space needs of mallard broods. During 1989-90, we examined the movements, home range, and habitat use of 27 radio-marked mallard broods on Lower Klamath National Wildlife Refuge, California. Twelve of the 27 broods made 22 relocation movements (>1,000 m in 24 hr) in the first week (n = 6) and after the fourth (n = 16) week of life. Mean home range size was 0.93 km 2 (SE = 0.25) and did not differ between years (P = 0.26). Brood-rearing females selected seasonally flooded wetlands with a cover component and avoided open or permanently flooded habitats

Distribution and movements of female northern pintails radiotagged in San Joaquin Valley, California

To improve understanding of northern pintail (Anas acuta) distribution in central California (CCA), we radiotagged 191 Hatch-Year (HY) and 228 After-Hatch-Year (AHY) female northern pintails during late August-early October, 1991-1993, in the San Joaquin Valley (SJV) and studied their movements through March each year. Nearly all (94.3%) wintered in CCA, but 5.7% went to southern California, Mexico, or unknown areas; all that went south left before hunting season. Of the 395 radiotagged pintails that wintered in CCA, 83% flew from the SJV north to other CCA areas (i.e., Sacramento Valley [SACV], Sacramento-San Joaquin River Delta [Delta], Suisun Marsh, San Francisco Bay) during September-January; most went during December. Movements coincided with start of hunting seasons and were related to pintail age, mass, capture location, study year, and weather. Among pintails with less than average mass, AHY individuals tended to leave the SJV earlier than HY individuals. Weekly distribution was similar among capture locations and years but a greater percentage of pintails radiotagged in Tulare Basin (south part of SJV) were known to have (10.3% vs. 0.9%) or probably (13.8% vs. 4.6%) wintered south of CCA than pintails radiotagged in northern SJV areas (i.e., Grassland Ecological Area [EA] and Mendota Wildlife Area [WA]). Also, a greater percentage of SJV pintails went to other CCA areas before hunting season in the drought year of 1991-1992 than later years (10% vs. 3-5%). The percent of radiotagged pintails from Grassland EA known to have gone south of CCA also was greater during 1991-1992 than later years (2% vs. 0%), but both the known (19% vs. 4%) and probable (23% vs. 12%) percent from Tulare Basin that went south was greatest during 1993-1994, when availability of flooded fields there was lowest. The probability of pintails leaving the SJV was 57% (95% CI = 8-127%) greater on days with than without rain, and more movements per bird out of SJV occurred in years with more rain and fog but fewer days with southerly winds. Movements by pintails and changes in pintail distributions, direct recovery distributions, and harvest rates suggest the disproportionate decline of pintails in Tulare Basin was due to a lower percentage of pintails moving there in fall and a greater percentage or earlier movements north and south out of Tulare Basin. With fewer in Tulare Basin to replace Grasslands EA pintails going north in December, pintail abundance in the northern SJV declined during late winter. Changes in movement patterns correspond to habitat loss in Tulare Basin and increased habitats in SACV and western mainland Mexico. Habitat improvements, especially in Tulare Basin, that increase food, sanctuary, and winter survival would probably help restore pintails throughout the SJV.

Swainson's Hawk Nesting Ecology in North Dakota

Swainsons Hawks (Buteo swainsoni) were studied at 270 occupied nest sites in south-central North Dakota on a 1,259-km2 intensive study block and on a surrounding study area (16,519 km2) during three breeding seasons. On the intensive study block the number of occupied nests ranged from 46 in 1977 to 100 in 1979. Average nest densities were highest on ground moraine (0.119 nest/km2) and on eolian sand deposit (0.102 nest/km2) landforms. Pasture and haylands made up 75% of the land-use within 1.0 km of a sample of 27 nests. Some pairs nested successfully in sites characterized by intensive agriculture and human activity; about 75% of all nests were in sites attributable to human activities. The most common nesting sites (43%) were in shelterbelts. Cottonwoods (Populus deltoides) were the most frequently used (44%) nest trees. At least 50% of the pairs constructed new nests each year. Mean nest success was 64% and mean number of young fledged per occupied nest was 1.5. Wind and hail caused over 30% of the nest failures each of the three years. Northern pocket gophers (Thomomys talpoides) were the most frequent prey, accounting for 44% of all animal remains found at nests. Man-made changes in central North Dakota during the last century have provided many nesting sites in areas previously sparsely populated by Swainsons Hawks. Swainsons Hawks (Buteo swainsoni) are a major component of the raptor fauna within their breeding range in the semiarid region of North America (Brown and Amadon 1968). Despite their wide distribution, few studies of their breeding ecology in the northern prairie region have been published. Major studies have been conducted in Colorado (Olendorff 1973), Wyoming (Dunkle 1977), Washington (Bechard 1980, Fitzner 1980), Saskatchewan (Schmutz et al. 1980), and California (Bloom 1981). Briefer but noteworthy studies include those of Cameron (1913), Bowles and Decker (1934), Bent (1937), Craighead and Craighead (1956), Platt (1971), Smith and Murphy (1973), Keir (1976), and Tyler (1981). Land use in North Dakota is changing, with increased emphasis being placed on intensive agriculture (Stewart 1975) and energy development (D. L. Evans, unpubl. rep., U.S. Fish and Wildlife Service, Jamestown, North Dakota, 1980). The effect of this trend on birds of prey breeding in the region is unknown. We studied the Swainsons Hawk in North Dakota in order to provide a basis for evaluating the future status of this species in the region. Our objectives were to evaluate population density, productivity, habitat, nest-site selection, and food habits of Swainsons Hawks in North Dakota. Our study was conducted from spring through the summers of 1977-1979. STUDY AREA The 16,519-km2 study area consists of Stutsman, Kidder, Logan, and Burleigh counties in south-central North Dakota. The area contains three biotic subregions (Stewart 1975): Drift Plain, 21%; Missouri Coteau, 62%; and Coteau Slope, 17% (Fig. 1). The Drift Plain is mostly gently rolling ground moraine and end moraine. The Missouri Coteau is mostly highrelief dead-ice moraine and gently rolling glacial outwash. The Drift Plain and Missouri Coteau are characterized by the eastern mixedgrass climax community. Glacial action only moderately affected the topography of the Coteau Slope subregion where principal landforms include sheet moraine and shallow ground moraine deposits. The climax vegetation is the western mixed-grass community. Major land uses in the Missouri Coteau and Coteau Slope subregions are haylands and pasture areas (mostly native prairie); croplands (mostly wheat and sunflower) dominate the Drift Plain (North Dakota Crop and Livestock Reporting Service 1980). Small stands of deciduous trees occur throughout the study area on river bluffs, on eastand north-facing slopes of high moraines, along margins of many permanent and semipermanent fresh-water lakes, in shelterbelts, and in groves associated with nearly all farmsteads.

Natural cavities used by wood ducks in north-central Minnesota

Radio telemetry was used to locate 31 wood duck (Aix sponsa) nest cavity sites in 16 forest stands. Stands were of 2 types: (1) mature (t = 107 years) northern hardwoods (10 nest sites), and (2) mature (i = 68 years) quaking aspen (Populus tremuloides) (21 nest sites). Aspen was the most important cavity-producing tree used by wood ducks and accounted for 57 percent of 28 cavities inspected. In stands used by wood ducks, the average density of suitable cavities was about 4 per hectare. Trees containing nests were closer to water areas (P < 0.05) and the nearest forest canopy openings (P < 0.01) than was a random sample of trees from the same stands. A significant (P < 0.005) relationship existed between the orientation of the cavity entrance and the nearest canopy opening. Potential wood duck cavities usually were clustered within a stand rather than randomly distributed. Selection of trees by woodpeckers for nest hole construction probably influenced the availability of cavities used by wood ducks. A plan for managing forests to benefit wood ducks and other wildlife dependent on old-growth timber is discussed. J. WILDL. MANAGE. 42(2):288-298 McCabe (1966) and Aultfather (1966) pointed out the lack of information needed to guide foresters in multiple-use programs beneficial to wood ducks. There is little information on the use of natural cavities by wood ducks, particularly in the northern forested regions of the Lake States. Increases in wood duck populations in certain areas have been attributed to nest boxes (Bellrose et al. 1964, Grice and Rogers 1965), but, in some forested areas, large scale nest box programs may not be practical. Furthermore, timber resources that can provide a sustained source of cavity sites may be available. The primary objectives of this paper are to describe: (1) the home range of wood duck hens in relation to the nest site, (2) features of the habitat used by the nesting pair and characteristics of the forest stands containing cavity trees, (3) characteristics of the cavity tree and its immediate surroundings, and (4) guidelines for foresters managing northern forests to consider in formulating multiple use programs that would benefit wood ducks and other cavity-dwelling wildlife species. We thank H. A. Doty for critically reviewing the manuscript and W. A. Aultfather and G. W. Gullion for providing suggestions in manuscript preparation. E. Johnson provided assistance in photo interpretation and in the description of forest site conditions. T. S. Klodfelter and R. S. Stott assisted with data collection. We were assisted in statistical analyses by D. H. Johnson and in computer processing by D. A. Davenport. Telemetry equipment was constructed and maintained by the engineering staff, Cedar Creek Bioelectronics Lab, University of Minnesota. 1This research was supported by the U.S. Fish and Wildlife Service, University of Minnesota, National Institute of Health Training Grant TO1-GM01779, and the U.S. Energy Research and Development Administration (C00-1332-126). 2 Present address: U.S. Fish and Wildlife Service, Northern Prairie Wildlife Research Center, Jamestown, North Dakota 58401. 3 Present address: Department of Zoology, Washington State University, Pullman 99163. 4 Present address: R.R. #1, Valmeyer, Illinois 62295. 288 J. Wildl. Manage. 42(2):1978 This content downloaded from 157.55.39.104 on Sat, 18 Jun 2016 06:48:44 UTC All use subject to http://about.jstor.org/terms WOOD DUCK USE OF NATURAL CAVITIES .Gilmer et al. 289

Survival of adult female northern pintails in Sacramento Valley, California

North American populations of northern pintails (Anas acuta) declined between 1979 and the early 1990s. To determine if low survival during winter contributed to declines, we estimated winter (last week of Aug-Feb 1987-90) survival for 190 adult (after hatching yr [AHY]) female radio-tagged pintails in late summer in Sacramento Valley (SACV), California. Survival rates did not vary by winter (P = 0.808), among preseason, hunting season, or postseason intervals (P = 0.579), or by body mass at time of capture (P = 0.127). Premolt (wing) pintails (n = 10) tended to survive at a lower rate (0.622, SE = 0.178) than pintails that had already replaced flight feathers (0.887, SE = 0.030) (P = 0.091). The pooled survival (all years) estimate for the 180-day winter was 0.874 (SE = 0.031). Hunting mortality rate (0.041-0.087) and nonhunting mortality rate (0.013-0.076) did not differ among years (P = 0.332) or within years (all P > 0.149). Legal hunting (n = 7), predation (n = 4), cholera (n = 2), illegal shooting (n = 2), botulism (n = 1), and unknown cause (n = 1) accounted for all mortality. Nonwintering survival (annu. survival/winter survival = 0.748) was lower than winter survival; thus, if gains in annual survival are desired for this population, managers should first examine the breeding-migration period for opportunities to achieve increases.

September-March survival of female northern pintails radiotagged in San Joaquin Valley, California

To improve understanding of pintail ecology, we radiotagged 191 hatch-year (HY) and 228 after-hatch-year (AHY) female northern pintails (Anas acuta) in the San Joaquin 7 Valley (SJV), and studied their survival throughout central California, USA, during September-March, 1991-1994. We used adjusted Akaike Information Criterion (AIC c ) values to contrast known-fate models and examine variation in survival rates relative to year, interval, wintering region (SJV, other central California), pintail age, body mass at capture, capture date, capture area, and radio type. The best-fitting model included only interval x year and age x body mass; the next 2 best-fitting models also included wintering region and capture date. Hunting caused 83% of the mortalities we observed, and survival was consistently lower during hunting than nonhunting intervals. Nonhunting and hunting mortality during early winter was highest during the 1991-1992 drought year. Early-winter survival improved during the study along with habitat conditions in the Grassland Ecological Area (EA), where most radiotagged pintails spent early winter. Survival was more closely related to body mass at capture for HY than AHY pintails, even after accounting for the later arrival (based on capture date) of HY pintails, suggesting HY pintails are less adept at improving their condition. Thus, productivity estimates based on harvest age ratios may be biased if relative vulnerability of HY and AHY pintails is assumed to be constant because fall body condition of pintails may vary greatly among years. Cumulative winter survival was 75.6% (95% CI = 68.3% to 81.7%) for AHY and 65.4% (56.7% to 73.1%) for HY female pintails. Daily odds of survival in the cotton-agriculture landscape of the SJV were -21.3% (-40.3% to +3.7%) lower than in the rice-agriculture landscape of the Sacramento Valley (SACV) and other central California areas. Higher hunting mortality may be 1 reason pintails have declined more in SJV than in SACV.