I designed a suite of studies in coordination with Utah Division of Wildlife Resources (UDWR) to evaluate waterfowl use of the GSL in winter and ecological aspects associated with GSL use. These studies provided insight into key information gaps previously identified by UDWR regarding management of GSL resources. Population surveys indicated total duck abundance was low when GSL surface elevations were low and wetland resources diminished because of persistent drought in the system. Also, ducks appear to use hypersaline parts of GSL more when freshwater habitats are limited from either drought or ice conditions. Common goldeneye, northern shoveler, and green-winged teal exhibited the most use of hypersaline areas. Dietary evaluations indicated all three species feed on hypersaline invertebrates from GSL to meet energetic and nutritional needs in winter. Brine shrimp cysts were important foods for northern shoveler and green-winged teal. Fat levels of ducks are important determinants of survival and fitness. Fat reserves of goldeneye were generally lower in the winter when both GSL and wetland habitat resources were lower. Results suggest brine fly larvae productivity, freshwater habitat availability, and temperature and wind speed likely play a more prominent role in goldeneye fat reserves than osmoregulation. Also, common goldeneye and northern shoveler using the GSL apparently accumulated biologically concerning amounts of mercury and selenium during winter. However, further research is needed to evaluate the effect of these elements on GSL ducks.

The purpose of this project was to evaluate effectiveness of aerial photography as a waterfowl census technique. An aerial photographic pattern was formulated and tested during the spring and fall months of 1971 and 1972 at Ogden Bay Waterfowl Management Area, Weber County, Utah. The reliability and feasibility of the technique for censusing waterfowl proved effective and practical. Eight flights per season, 4 routes per flight, and 30 photos per route taken over unit 1 of the Ogden Bay Waterfowl Management Area was the experimental design or pattern for the study. The level of accuracy and the cost involved with this design was: Spring 1971 27% sampling error cost $566.90 Spring 1972 33% sampling error cost $566.90 Fall 1971 30% sampling error cost $566.90 Fall 1972 17% sampling error cost $566.90 The effectiveness of other patterns were calculated using different intensities of flights, routes and photos. The patterns of flights, routes, photos, and costs calculated to achieve reliable and feasible estimates of waterfowl numbers at an acceptable sampling error of 21% or less at the 95% confidence level, calculated in tabular form, would have varied per season from 6 flights, 4 routes, and 5 photos to 16 flights, 6 routes, and 35 photos; costing from $220.00 to $1,550.00.

Results of past studies on the spring-fed salt marshes of Utah indicated that waterfowl production on these areas was much lower than on Utah's river-fed marshes. Waterfowl production on the newly-established Fish Springs National Wildlife Refuge, a spring-fed marsh, increased considerably, however, after its waters were impounded. It was believed that by determining what caused this increased production at Fish Springs it would be possible to obtain greater waterfowl production a t other spring-fed marshes. Consequently waterfowl populations and factors affecting waterfowl production on this marsh were studied during the summers of 1966, 1967, and 1968. Populations of breeding ducks averaged approximately 900 pairs a year in 1967 and 1968. Mallards, cinnamon teal, and redheads comprised about 80 percent of these birds. Nesting densities, on plots representing available habitat, averaged about one nest per acre during these two years. In 1967 and 1968 a total of 312 duck nests representing 10 species were studied and their fates determined. Overall nest success was 63 percent, and predators, principally coyotes and striped skunks, destroyed 25 percent of all nests. Calculated duckling mortality rates during 1967 and 1968, respectively, were 19 and 16 percent. Duck production at Fish Springs averaged about 3,000 birds or about 430 ducks per square mile of marsh habitat a year during this period. At present the sparsity of adequate nesting cover for gadwalls appears to be an important limitation to breeding by this species at Fish Springs. Nesting cover apparently has not limited the use of this marsh by other waterfowl species, however, as evidenced by a high incidence (76 percent) of nesting on dry ground by redheads and ruddy ducks. Aquatic insect sampling in waters inundating portions of the original Fish Springs marsh indicated that these waters produced large quantities of proteinaceous foods for ducklings. Populations of aquatic insects in the original marsh were probably low except in recently disturbed areas. It appears that limitations of brood-rearing habitat were primarily responsible for low waterfowl production in the past. Newly-created impoundments, rich in animal food, were probably the key to the increase in waterfowl production on this marsh.

Purpose and scope Each year thousands of western waterfowl succumb to disease, predators, mechanical injury and other decimating factors . Based on a review of records it is conservatively estimated that an average of 25,000 ducks have succumbed to botulism on western marsh areas annually. In a recent study, the United States Fish and Wildlife Service valued each duck and goose at $8.00 (McLeod, 1950). Applying this value to the estimated annual numerical loss, a total of $200,000 has been lost each season in mortality of western waterfowl from botulism. Control of this malady would reduce annual waterfowl and monetary losses. Prevention and cure are the only means of controlling botulism in wild ducks. At present, no economical preventative measure exists and control is based on curing stricken birds. The purpose of this study was to evaluate the cost of treatment and rate of recovery of birds stricken with botulism when treated by selected methods . The 4 treatments selected for evaluation were: (1) hospital inoculation, (2) fresh water, (3) field inoculation, and (4) no treatment or control . Research included a comprehensive evaluation of factors such as age, sex, species, body condition, degree of affliction, reaction to various amounts of antitoxin, and reaction t o selected treatment methods, thought to be pertinent in botulism control. This study was conducted during botulism outbreaks of 1953 and 1954, and was confined to state-owned marshlands of Utah. Study areas Three major areas provided sick birds used in this research: (1) Ogden Bay Refuge, (2) Farmington Bay Refuge, and (3) the Public Shooting Grounds . These man-made marsh areas are located on the saline flat lands adjacent to Great Salt Lake. The majority, 1,979 or 89.3 percent, of sick birds were taken from Ogden Bay Refuge on the Weber River Delta. This state-owned waterfowl refuge contains approximately 13,700 acres of diversified habitat . Excellent conditions for the production and existence of Clostridium botulinum, type c, the causative agent of botulism, were apparently present throughout the area. Farmington Bay Refuge, approximately 20 miles south of Ogden Bay Refuge in Davis County, Utah, was dried for improvements in 1953, but was traversed regularly during the 1954 season. In preparation for the hunting season, water was diverted into the north lake of Farmington Bay on October 1, 1953. A two-man crew picked up and disposed of approximately 2,000 dead ducks from approximately 20 acres of the reflooded marsh on October 6, 1953. This was the most serious outbreak of the study and indicated the rapidity with which sickness advanced. Few sick birds were noted, which indicated that the crisis had passed. Farmington Bay Refuge provided 121 of the 2,214 ducks treated during 1953 and 1954. Not more than 12 sick or dead birds were seen on the Public Shooting Grounds, 8 miles west of Corinne, Utah, in 1953. During the 1953 season, 3 sick birds were transported from the area to Ogden Bay Refuge for treatment. In 1954 this state-owned shooting area supplied lll sick birds for treatment. Other areas were observed during the study but did not provide sick birds for treatment. These areas consisted of : (1) State -owned lands beyond Bear River Migratory Bird Refuge, and (2) Smith and Utah Lakes west of Provo, Utah.

There are today many areas and types of water storage. The uses to which these are put are many, and the prospect for the future can only be that there will be an increase in the number of these reservoirs. These areas are usually operated with regard only for power production, irrigation, or storage, but there is a growing concern about the effects on wildlife which this type of operation may have. Fluctuating water levels, and falling water levels are not restricted to man-made impoundments tut are characteristic of many of then. Biologists have observed that these variations in water levels are often harmful to some forms of wildlife. Fluctuating water levels generally were found to cause damage to waterfowl nests. The amount of damage, the amount of water rise and the time involved are values which have generally remained unknown, largely perhaps because some other aspects was of greater importance in the nesting studies. There is a descending scale of values which have been placed on natural resources, and it is generally agreed that the waterfowl concerned would rank below the value derived from the impounded waters. Where it is practical, the management of these impoundments should take wildlife into consideration. To have intentions of good management is not sufficient to effect conservation; these intentions must be implemented with the proper tools. This study was designed to discover magnitudes of cause and effect, and perhaps it will point the way to a more efficient utilization of associated resources. It was possible that the results of this study would show that there was little damage to waterfowl. On the other hand, if significant damage were to be discovered, contributing causes could be more closely delimited and so point the way toward corrective measures.

The Mississippi Alluvial Valley (MAV) contained vast bottomland hardwood forests into the 20th century. Humans cleared forests, and altered hydrology, yet the MAV remains important for North American waterfowl and other wildlife. To estimate standing crops of aquatic invertebrates as food in hardwood bottomlands for wintering waterfowl, I quantified dry mass of invertebrates in naturally flooded forests (NFFs) and greentree reservoirs (GTRs) during winters 2008–2010. The MAV had greater invertebrate mass in NFFs (x̄ = 18.39 kg/ha; SE = 2.815 [CV = 15.3%]) than GTRs (x̄ = 5.16; SE = 0.492 [CV = 9.5%]), compared with lesser masses in Mississippi Interior Flatwoods’ GTR (x̄ = 2.26; SE = 0.320) and NFF (x̄= 1.45; SE = 1.305). Invertebrate diversity was greatest in NFFs and in depths from 10–40 cm. Flooding GTRs ≤ 40 cm and managing naturally dynamic hydrology may benefit invertebrates, ducks, and associated bottomland hardwood communities

Red oaks (Quercus spp.; Section Erythrobalanus) produce acorns which are valuable forage for wildlife especially mallards (Anas platyrhynchos) and wood ducks (Aix sponsa). Scientists have limited information on amount, timing, and persistence of these acorns in the Mississippi Alluvial Valley (MAV). Conservation planners rely on precise estimates of acorns and other forage to estimate habitat needed by waterfowl in the MAV and other regions. My study provided premiere landscape-scale, multi-year estimates of red oak acorn yield and on-ground abundance in the MAV. Mean yield of acorns was 534 kg(dry)/ha (42.3 acorns/ m2) across all sites, years (falls-winters 2009-2012), and oak species. Yield varied more within years (CV = 11 - 29%) than when data were combined across years (CV = 11%). Yield was not synchronized in any year among MAV sites. However, yield usually was synchronized among species within sites suggesting local factors influenced acorn yield more than landscape-scale factors. Among sites and years, acorn abundance generally was greatest in January (sample mean = 371 kg/ha) and least in November (198 kg/ha). Acorns persisted to February only in years of above-average yield. Except for Nuttall oak (Quercus texana), acorn persistence generally was stable regardless of yield from parent trees. Nuttall oak acorn persistence increased with yield perhaps revealing an evolutionary pressure that encourages masting. Red oak acorn abundance was linearly related to percentage of red oaks in the overstory, but this relationship differed in years of above- and below-average yield. Currently, conservation planners use 166 kg/ha as a forage estimate of red oak acorns, moist-soil seeds, and aquatic macro-invertebrates in bottomland hardwood forests with 100% red oak canopy. I sampled at 5 sites throughout the MAV over 3 years; therefore, I recommend conservation planners consider adopting my predicted estimate of 247 kg of acorns/ha of forest land with 100% red oak canopy. Because acorns persist through most winters and generally reach peak abundance in January, often concomitant with peak abundance of mallards and other ducks in the MAV, biologists and conservation planners may have undervalued the potential of bottomland hardwood forests to support ducks in mid-late winter.

Water resources for Bluff Lake in Mississippi are managed to achieve objectives related to waterfowl, waterbirds, fish, anglers, and Paddlefish (Polyodon spathula). Annually, the reservoir undergoes a nine-stage seasonal drawdown and re-inundation to improve waterfowl habitat. In addition, weekly discharges are released from the water control structure to encourage Paddlefish spawning and migration each spring. However, additional discharges throughout the year may provide additional passage opportunities. In this study, multiple discharge states were evaluated to identify optimal water releases during each drawdown period given reservoir objectives. First, I developed a hydrodynamic model to predict daily changes in lake volume. Second, I defined functional relationships between water surface elevation and management objectives. A structured decision-making framework was then applied to determine the optimum additional discharge strategy. This approach allowed trade-offs between management objectives to be evaluated and optimal water releases to be identified for this multiuse reservoir.

Waste rice in harvested fields and natural seeds in moist-soil wetlands are important foods for waterfowl in the Mississippi Alluvial Valley (MAV). Waste rice declines significantly between harvest and late autumn. I conducted experiments in 19 rice fields in Arkansas and Mississippi during autumns 2003 and 2004 to evaluate the ability of common post-harvest practices to conserve waste rice between harvest and late autumn. Greatest mean abundance of waste rice in late autumn occurred in standing stubble (x&95 = 105 kg/ha; CL = 72.84, 150.16 kg/ha) followed by burned (x&95 = 72 kg/ha; 49.57, 105.81 kg/ha), mowed (x&95 = 67 kg/ha; 46.65, 97.42 kg/ha), rolled (x&95 = 51 kg/ha; 35.54, 73.076 kg/ha), and disked stubble (x&95 = 48 kg/ha; 33.26, 68.41 kg/ha). I recommend leaving stubble or burning fields to create interspersion of stubble and water after flooding. Additionally, I estimated abundance of moist-soil seed in 72 management units on public lands throughout the MAV for autumns 2002-2004. Mean seed abundance was 496 kg/ha (SE = 62). I recommend increased active management of moist-soil wetlands to mitigate decreased availability of waste rice in the MAV.

The Laysan teal, an endangered species, is restricted to a single population on Laysan Island, a remote atoll of the Hawaiian archipelago. Little is known of the Laysan teal's ecology, therefore, I examined food habits, habitat use, and population dynamics. These aspects of its ecology are fundamental to the species management and conservation. I described diel and nocturnal habitat use, home range, and foraging with radio telemetry in 1998-2000. Most individuals showed strong site fidelity during the tracking period, but habitat selection varied between individuals. Mean home range size was 9.78 ha (SE 2.6) using the fixed kernel estimator (95% kernel; 15 birds with >25 locations). Foraging was strongly influenced by time of day: birds spent only 4% of their time foraging in the day, but spent 45% of their time foraging at night. Time activity budgets from the island's four habitat zones indicated that the coastal zone was rarely used for foraging. The birds foraged 42% of the time they spent in the terrestrial zone at night, but foraged only 4-6% of the time they spent there during other times. Fecal analysis and behavioral observations revealed that the Laysan teal is not a 100% macro-insectivore as previously reported, but consumed seeds, succulent leaves, and algae, in addition to adult diptera, diptera larvae and pupae, ants, seeds, lepidoptera, coleoptera, and Artemia. I concluded that this species exhibits high plasiticity in foraging behavior. Laysan teal appear to opportunistically select abundant, high energy prey for the breeding season, due to constrained resources on Laysan Island. I also studied the parameters influencing the Laysan teal's population dynamics. Adult survival is high, but duckling survival on Laysan is low, and is a primary demographic parameter limiting population growth. Estimates indicate the population density was high (between 546-827) from 1991 until August 1993, prior to a population crash that occurred between September and December 1993. The most current population estimate (Sept-Nov 2001) is 444 (SE 181) adults. Additional populations (translocation), along with control of non-native mammalian predators, are needed to reduce extinction risks to the Laysan teal. / Ph. D.