Date of Award


Document Type


Degree Name

Master of Science (MS)


Forestry and Environmental Conservation

Committee Member

Dr. Greg Yarrow, Committee Chair

Committee Member

Dr. Richard Kaminski

Committee Member

Dr. Patrick Gerard


I conducted a landscape-scale survey of nest-structure use and production by wood ducks (Aix sponsa), black-bellied whistling ducks (Dendrocygna autumnalis), and hooded mergansers (Lophodytes cucullatus) across two riverine basins in coastal South Carolina during 2016–2017. For 364 and 354 nest boxes surveyed in each year (n = 718 box years), 61% were used by wood ducks, 15% by black-bellied whistling ducks, and < 1% by hooded mergansers. Across years, mean frequency of use of nest boxes by wood ducks and black-bellied whistling ducks did not differ between the Ashepoo, Combahee and Edisto Rivers (ACE) and Santee Rivers Delta and Winyah Bay (SRDW) Basins (F1,353 = 1.46, P = 0.2270, F1,353 = 1.48, P = 0.2247, respectively). Wood ducks nested from January–August (x̄ = 181-day nesting season, CV = 2.5%) with peak nesting in March–May. Black-bellied whistling ducks nested from May–September (x̄ = 116-day nesting season, CV = 8.6%) with peak nesting in June–July. Stepwise logistic, multiple regression revealed several correlations of physical and micro-habitat characteristics of nest boxes with their use by wood ducks and black-bellied whistling ducks. Wood ducks were 5.8% (β = −0.00006, P = 0.0035) more likely to select nest boxes for every 1,000 cm3 decrease in internal box volume. However, black-bellied whistling ducks were 19.6% (β = 0.00018, P < 0.0001) more likely to select nest boxes for every 1,000 cm3 increase in internal volume. Vegetative, canopy cover above boxes had a negative association with nest-box selection; boxes were 15.1% (β = −0.01642, P < 0.0001) and 11.1% (β = −0.01194, P = 0.0236) more likely to be selected by wood ducks and black-bellied whistling ducks for every 10% decrease in percent canopy cover. Additionally, nest boxes were 18.3% (β = 0.01677, P = 0.0053) and 9.8% (β = −0.01031, P = 0.0044) more likely to be selected by black-bellied whistling ducks for every 10 cm increase in distance from the base of boxes’ entrance vertically to ground or water surface and every 10 m decrease in distance between boxes. My data suggest the conventional nest box described by Bellrose (1980), with internal volume of 34,375 cm3 and internal dimensions of 25 × 25 × 55 cm, was usable for both species of ducks and therefore can be deployed where these species are sympatric. However, I suggest nest box entrances should have 12.7 cm diameters to facilitate use by larger black-bellied whistling ducks, as recommended by Bolen (1967a). Additionally, I observed both species of ducks selected boxes in open areas and often in ponds with predatory fish, which could create “ecological traps” for ducklings. I emphasize the importance of proper nest-box placement near suitable brood-rearing habitat (e.g., shoreline scrub-shrubs; Davis et al. 2007) to promote duckling survival and recruitment. For wood ducks, mean clutch size was 13 eggs, nest success was 65%, and hatching success was 49%. Based on egg-shell membranes found in nest boxes, mean number of wood duck ducklings exiting boxes did not vary between basins in 2016 (t1 = 1.04, P = 0.2966) and 2017 (t1 = 1.81, P = 0.0698). Pooled across years, an average of five wood duck ducklings exited structures. For black-bellied whistling ducks, mean clutch size was 11 eggs, nest success was 51%, and hatching success was 43%. Mean number of black-bellied whistling duck ducklings exiting boxes differed between basins in 2016 (t1 = 3.77, P = 0.0002) but not in 2017 (t1 = −0.88, P = 0.3776). In 2016, the arithmetic mean number of black-bellied whistling ducks ducklings exiting boxes in the two basins were 1.2 and 0.8 ducklings, indicating an average of one duckling of this species exited structures. Additionally, an estimated 3,378 wood duck, 531 black-bellied whistling duck, and 19 hooded merganser ducklings exited nest structures in 2016–2017 for an average of six ducklings across species/boxes/years (3,928 ducklings/718 boxes; x̄ = 5.5 ducklings). I also assessed costs of female wood duck recruits from nest boxes, based on contemporary costs to fabricate boxes, their annual maintenance, an assumed longevity of 20 years for an annually maintained bald-cypress (Taxodium distichum) box (total = $142.23/box [U.S.]), reproductive metrics from my study, and a wood duck female recruitment rate of 5.22% (Hepp et al. 1989). Calculated cost per female wood duck recruit for an assumed 20-year life of box was $65.65. This cost of female wood duck recruitment over 20 years was 2.2 time less than the cost of the box, mounting structure, predator shield, and maintenance over 20 years. Therefore, based on these data, nest boxes in my study seem cost-effective in recruiting female wood ducks. My study did not estimate brood survival and duckling recruitment into fall and breeding population. Therefore, I emphasize need to determine major influences of recruitment and other vital rates of box-nesting duck populations, levels of vital rates needed to stabilize and grow populations, habitat locations of nest boxes that promote duckling survival and ultimately recruitment into breeding populations, estimated proportion of box vs naturally cavity produced wood ducks in the harvest, and cost-efficiency of nest box programs at larger spatial scales in North America.