Incorporating nocturnal activity of wintering Atlantic brant (Branta bernicla hrota) into 24 hour time-activity and daily energy expenditure models
| Author(s) | Heise, Jeremiah | |
| Date Accessioned | 2012-11-27T14:48:21Z | |
| Date Available | 2012-11-27T14:48:21Z | |
| Publication Date | 2012 | |
| Abstract | The Arctic Goose Joint Venture has established numerous priority areas of research for Atlantic brant (Branta bernicla hrota). Several of these priorities focus on the wintering grounds, as the wintering period is often a limiting time for many species, especially migratory waterfowl. Addressing these areas of research often centers on behavioral observations to construct subsequent time-energy models. While waterfowl have qualitatively been shown to be active nocturnally, limited research has addressed incorporating this activity into time-activity and daily energy expenditure models. Thus, popular assumption promotes that observed diurnal behavior is representative of 24 hrs. As a result, my objectives were to 1) directly quantify the behavioral dynamics of wintering Atlantic brant within the 24 hr period inclusive of morning crepuscular, diurnal, evening crepuscular, and nocturnal periods along southern coastal New Jersey, USA, 2) determine if environmental variability (e.g. freezing temperatures, wind, snow cover, ice cover, etc.), or anthropogenic pressure (e.g. hunting disturbance) affect time spent in different behaviors across the 24 hr period, and 3) evaluate the effect of hunting and non-hunting areas on brant behavior both during and outside the Coastal Zone waterfowl hunting seasons. I observed brant during October-February 2009–2010 and 2010–2011, across their main wintering grounds along the Great Bay estuary of southeastern coastal New Jersey. I conducted behavioral observations using an instantaneous scan method across two different combinations of 6 hr observation periods, effectively covering the 24 hrs of a day. These observations were paired across either a diurnal and nocturnal period or morning and evening period inclusive the respective crepuscular periods, thus effectively being matched over a given day’s tide cycle. Explanatory environmental variables, as well as hunting location and season data, were also collected to model against the collected behavior data. During two years of observation, I collected observation data in a total of 5,682 instantaneous scans. Because of the close association of instantaneous scans (10 min), I tested for effects of pseudoreplication with the use of a semivariogram. I found that samples were not independent for up to four consecutive scans and thus averaged consecutive scans across 40 min. As a result, I ended up with a reduced (1,921 scans), but more robust sample for further analyses. I compared behaviors across periods against explanatory variables using an Akaiki Information Criterion (AIC) model selection approach. As a result of the number of models I constructed and the number of included explanatory variables (n = 8), I was inherently left with few models that contained a substantial level of empirical support. To address this, I calculated a 95% confidence set of models and performed model averaging to determine a given variable’s effect on behavior. This modeling resulted in my open/closed hunting season variable having a strong association with several behaviors across observation periods. Further analysis showed that behavior did indeed change between combinations of hunting and non-hunting locations across open and closed seasons between the diurnal and nocturnal period. I estimated energy expenditure across the landscape by applying behavior-specific multipliers across my observed behavior proportion data, accounting for costs of thermoregulation, which allowed me an hourly energy expenditure (HEE) estimate. Applying this HEE estimate across months, accounting for average amount of hours in each observation period, I was able to calculate the first true 24 hr daily energy expenditure (DEE) for wintering brant. Energy expenditure for brant averaged 1,594.98 ± 85.57 kJ/day across observation months. By comparing this value to estimates calculated 1) by scaling my diurnal data to be representative of 24 hrs (1,774.49 ± 189.17 kJ/day), 2) using a brant specific allometric method (1,294.56 kJ/day), and 3) a generalized allometric method (1,580.50 kJ/day), I found that values can differ significantly from one another (P = 0.03–0.90). As a result, inputs into these equations should be considered carefully, as they will greatly affect one’s estimates, which can in-turn have specific consequences for subsequent management decisions. | en_US |
| Advisor | Williams, Christopher | |
| Degree | M.S. | |
| Department | University of Delaware, Department of Wildlife Ecology | |
| URL | http://udspace.udel.edu/handle/19716/11791 | |
| Publisher | University of Delaware | en_US |
| dc.subject.lcsh | Brant -- Wintering -- New Jersey. | |
| Title | Incorporating nocturnal activity of wintering Atlantic brant (Branta bernicla hrota) into 24 hour time-activity and daily energy expenditure models | en_US |
| Type | Thesis | en_US |