Translator Disclaimer
1 September 2015 Using Land-Based Surveys to Assess Sea Duck Abundance and Behavior in Nearshore Waters of Southern New England, USA
Author Affiliations +
Abstract

Nearshore waters provide very important habitat for sea ducks (Tribe Mergini) during migration and winter, but gathering information on sea duck use of shallow nearshore waters is challenging because traditional aerial and boat-based surveys are expensive, are usually conducted infrequently, and are often not feasible near the coast. The objective of this study was to use land-based surveys to characterize spatiotemporal variation in the abundance and behavior (e.g., foraging, flying) of Common Eider (Somateria mollissima) and scoters (Melanitta spp.) in nearshore waters of southern New England. Surveys (60–120 min per survey, n = 1,044 surveys) were conducted throughout the day from February 2009 to July 2010 to assess diurnal and seasonal variation in sea duck behavior and spatial distribution at nine sites in southern Rhode Island. The density of sea ducks resting or foraging on the water exhibited little diurnal variation, whereas flight activity dramatically increased nearer to sunrise. Sea duck densities and passage rates (individuals/km2/hr) peaked during migration periods from October through November and February through April, although there were important seasonal differences between sites. For example, the highest densities of Common Eider during fall were in a protected estuary, whereas abundance of scoters during fall was greater at a coastal headland. The relative activity of Common Eider on the water and in flight was similar among sites, whereas scoters exhibited highly variable activity among sites, particularly during winter and spring. The spatiotemporal patterns in abundance and behavior of sea ducks in nearshore waters that we detected using land-based surveys provides essential, complementary information to that available from other types of waterfowl and seabird surveys in southern New England.

There is considerable interest in sea duck (Tribe Mergini) use of nearshore waters along the Atlantic and Pacific Coasts of North America because these areas provide very important wintering habitat that is potentially vulnerable to anthropogenic disturbance (Austin et al. 2014; Baldassarre 2014; De La Cruz et al. 2014). Nearshore areas are important to sea ducks because most species typically forage in shallow waters less than 20 m that often occur close to shore (Guillemette et al. 1993; Fox 2003). Much more is known about the large-scale distribution and abundance of sea ducks in winter along the western Atlantic Ocean (Zipkin et al. 2010; Silverman et al. 2013) than their distribution and abundance at more regional and local scales where sea ducks demonstrate their preferences for certain habitats (Johnson 1980), and their response to many environmental factors is determined. An understanding of sea duck distribution and abundance in nearshore areas at these finer scales is required to evaluate potential anthropogenic effects on local sea duck populations and inform local planning efforts (Madsen 1998; Langston 2013; De La Cruz et al. 2014; Winiarski et al. 2014).

Several studies have used aerial or boat-based surveys to assess the winter distribution and abundance of sea ducks at finer spatial scales (e.g., White et al. 2009; Winiarski et al. 2014). However, aerial and boat-based surveys are expensive, tend to be conducted infrequently, generally only occur when weather conditions are favorable for observation, often are not conducted near the coast for safety or regulatory considerations, and can have low detection probabilities for sea ducks (Camphuysen et al. 2002; Winiarski et al. 2013, 2014). Land-based surveys represent an alternative approach to monitor sea duck use patterns at fine spatial scales (McKinney et al. 2006; Loring et al. 2013). For example, the Avalon Sea Watch detects over 800,000 marine birds annually as they migrate off the coast of New Jersey (New Jersey Audubon Society, unpubl. data), and sea duck migration has been monitored from 1996–2012 at Point Lepreau in the Bay of Fundy (Bond et al. 2007; Cameron 2014). Unlike aerial and boat-based surveys, land-based surveys potentially provide a cost-effective method to document inter- and intra-annual spatial shifts in sea duck abundance and site-specific variation in behaviors (e.g., season-specific movements and flight paths) that are essential for detecting the potential effects of anthropogenic disturbances in nearshore habitats (Perry and Deller 1996; McKinney et al. 2006; Merkel et al. 2009).

We used land-based surveys in Rhode Island's nearshore waters to characterize spatiotemporal variation in the abundance and behavior of the most common species of sea ducks in the region including Common Eider (Somateria mollissima) and Black (Melanitta americana), Surf (M. perspicillata), and White-winged (M. deglani) scoters. We assessed variation in sea duck abundance, distribution, and behavior (e.g., foraging, flying) on diurnal and intra-annual (seasonal) temporal scales.

Methods

Study Area

We surveyed for sea ducks along the southern Rhode Island coast at 11 land-based point count sites along four survey routes from Watch Hill (41° 18′ N, 71° 51′ W) to Goosewing Beach in Little Compton (41° 28′ N, 71° 09′ W; Fig. 1) from February 2009 through July 2010. Three survey routes contained three sites, while tire easternmost route comprised two sites. We estimated the area of visible water for each site using a viewshed analysis (Environmental Systems Research Institute 2011) on 1-m horizontal resolution LiDAR data (U.S. Geological Survey 2012). The viewsheds of adjacent count sites overlapped considerably at four sites; therefore, we classified Beavertail and Brenton Point as one site, and Sachuest Point and Sakonnet Point as another site; this reduced the number of distinct “sites” in our statistical models to nine (Fig. 1).

Figure 1.

Sites where land-based surveys for sea ducks were conducted in coastal Rhode Island from February 2009 through July 2010. Sites were segregated into the southern coast (South Coast), western Narragansett Bay (West Bay), and eastern Narragansett Bay (East Bay). Gray polygons represent the estimated viewshed surveyed from each site; dark gray areas indicate regions of survey overlap between adjacent sites.

f01_252.jpg

Surveys

We conducted surveys during mornings (AM: concluded before solar noon) and afternoons (PM: initiated after solar noon). During AM surveys, we surveyed the first site along each route for 120 min starting at dawn and subsequent sites for 60 min. For PM surveys, we surveyed the last site for 120 min and all preceding sites for 60 min. The surveys at dawn and dusk were longer to ensure adequate observation effort when activity of sea ducks was expected to be greatest. The exception to this pattern occurred in our first month of surveys due to logistical limitations — some February 2009 surveys were of intermediate length (typically 90 min) and most occurred during the morning. We varied the visitation order of sites along a route to better distribute the sampling times among sites. Survey effort varied from two to 10 monthly surveys at a given site, typically with two to four surveys per month in each AM/PM window (Fig. 2).

During each survey, a single observer recorded all sea ducks within 3 km of the count location (Fig. 1) and their location as on the ocean's surface (hereafter, “on the water”) or in flight. Observers surveyed the ocean surface and airspace with a 20–60× spotting scope and 10×42 binoculars and recorded tire number and species for all individuals or flocks during the survey period. We took care to avoid recounting sea ducks on the water during the observation period, and those in flight were visually tracked to confirm that they exited the viewshed. We recorded observations in the field using a handheld personal digital assistant.

Figure 2.

Summary of monthly survey effort (median and interquartile range) for land-based sea duck surveys by period (AM: morning; PM: afternoon) at 11 sites along the Rhode Island coast from February 2009 through July 2010.

f02_252.jpg

We considered two species groups: Common Eider (hereafter, eider) and Surf, Black, and White-winged scoters (50%, 42%, and 8% of individuals identified to species, respectively; hereafter, scoters). We analyzed scoters collectively as they were often difficult to distinguish at a distance (50% were not assignable to species) . We focused on eider and scoters as they represented nearly 75% of all sea ducks we observed. Long-tailed Ducks (Clangula hyemalis) and Red-breasted Mergansers (Mergus serrator) were infrequently observed during these surveys, and thus were excluded from analyses.

Adjusting for Survey Duration and Area Surveyed

We recorded most individuals observed initially on the water in the first 60 min of observation; whereas, previously uncounted flying individuals regularly entered the viewshed throughout a given survey. Thus, count duration was included as a fixed offset only in the models for flying sea ducks to account for “effort” effects on these counts. Because we counted few additional sea ducks on the water after the initial 60 min of a survey, we did not adjust for survey length in the models of sea ducks on the water. We included the area surveyed (i.e., viewshed area) as a covariate rather than a fixed offset because of the disproportionate addition of areas at longer ranges (i.e., 2–3 km), where fewer observations occurred.

Statistical Analysis

We used generalized additive models (GAMs; Hastie and Tibshirani 1990) to evaluate spatiotemporal variation in sea duck abundance and behavior. We created separate negative binomial GAMs for each species group and initial observation location (i.e., on the water or in flight). We allowed sea duck abundance to vary smoothly over daylight hours (i.e., proportion of daylight elapsed at the midpoint of each survey) using a penalized cubic regression spline. For seasonal patterns, we used a cyclic cubic regression spline that allowed sea duck abundance to vary smoothly over the course of the year. We preserved the association among consecutive fall migration, winter, and spring migration seasons by considering observations from 15 August to the subsequent 14 August as occurring in the same “year.” We allowed separate seasonal splines for each of the nine “sites.” We included a categorical covariate to accommodate potential differences in the two (partial) years that surveys occurred.

We could not incorporate temporal autocorrelation among counts in our GAMs due to convergence problems associated with fitting separate smoothing splines for each site. As countermeasures, however, we restricted the basis dimension of the spline terms and imposed an extra penalty to each term in the model (Marra and Wood 2011) and interpreted marginally important parametric effects with care. We fitted GAM models using the mgcv (Wood 2006, 2011) and nlme packages (generalized additive mixed models; Pinheiro et al. 2014) in statistical program R (R Development Core Team 2014).

We did not adjust statistically for differences in sea duck detection because we recorded only initial distances for sea ducks on the water or in flight. Conventional distance sampling methods typically fail for land-based surveys of the marine environment (Marques et al. 2010), and the techniques for separating the detection process from the expected non-uniformity in animal density (e.g., associated with bathymetry) additionally require the angle of detection (Cox et al. 2013). Thus, to compare abundances among sites, we necessarily assumed similar patterns of sea duck detection and density with distance from shore among sites. We varied the order of site visitations, viewsheds were similarly proportioned among sites, and the conditions most likely to impact detectability (e.g., wave height, atmospheric visibility, observers) varied similarly over time at each site. Nonetheless, we acknowledge that comparison of sea cluck abundances among sites represented the combined effects of the actual abundance of sea clucks as well as the imperfect detection process.

Results

Diurnal Variation

The density of eider and scoters on the water varied little during daylight hours (Fig. 3A). In contrast, eiders and scoters in flight were considerably more common during the first 25% of daylight hours compared to the rest of the day; sea duck flight activity varied little after mid-day (Fig. 3B).

Seasonal Variation

In general, we observed most flying eider and scoters in two distinct periods that likely represented fall and spring migratory movements through Rhode Island (Fig. 4). For both species groups, fall flight activity occurred primarily from October through November. During the fall migratory period, densities of flying eider were usually highest at sites near the western mouth of Narragansett Bay (Fig. 4A–C). Densities of scoters in flight during fall were greatest at Pt. Judith (Fig. 4E) and in late fall at East Beach (Fig. 4D). In the spring, flight activity occurred mostly from February through April, although we regularly detected scoters in flight into May. Estimated eider abundance in flight was lower during spring migration than fall at most sites, with peak abundances in spring at Beavertail/Brenton and Sachuest/Sakonnet (Fig. 4C). Scoter abundance during spring in flight was greatest at eastern Rhode Island at Sachuest/Sakonnet and Ruggles (Fig. 4F).

Figure 3.

Diurnal variation in the (A) estimated density (individuals/km2) of sea ducks on the water relative to that at solar noon and (B) estimated passage rates (individuals/km2/hr) of flying sea ducks relative to that at solar noon for Common Eider and Black, Surf, and White-winged scoters (Scoters) during land-based surveys from February 2009 through July 2010 based on generalized additive models. A value of 100% indicates an estimated sea duck density similar to that at solar noon. Gray polygons represent the point-wise standard error of the fitted relationship.

f03_252.jpg

Sea ducks in flight were generally less abundant in winter relative to the migratory periods. However, we observed unusually high densities of flying eider at Watch Hill (Fig. 4A) and flying scoters at East Beach (Fig. 4D); neither of these sites experienced corresponding increases in the respective species on the water.

Relative densities of eider on the water (resting or foraging; Fig. 5A–C) corresponded generally to relative densities of individuals in flight during the migratory periods (Fig. 4A–C). Sites near the western mouth of Narragansett Bay and at Deep Hole experienced the greatest densities of eiders on the water during fall. In spring, sites in eastern Rhode Island (eastward from the Beavertail/Brenton Point area) consistently hosted more foraging eider (Fig. 5A–C); East Beach was a notable exception to this spring pattern, hosting a relatively high density of foraging eider in late winter that persisted into the early spring migratory period. Foraging eider were consistently present in greater densities in the Beavertail/Brenton Point area during winter, although eider densities increased late in the winter period at East Beach and the Sachuest/Sakonnet Point area (Fig. 5A–C).

Figure 4.

Seasonal variation (August through June) in the expected passage rate (individuals/km2/hr) of flying (A–C) Common Eider and (D–F) Black, Surf, and White-winged scoters at nine sites in southern Rhode Island based on generalized additive models. Sites were grouped into three geographic regions (see Fig. 1): South Coast (A, D), West Bay (B, E), and East Bay (C, F); background shading indicates presumed fall and spring migratory periods.

f04_252.jpg

Densities of scoters on the water (Fig. 5D–F) corresponded less well to patterns of flying scoters (Fig. 4D–F). At a given site, densities of scoters on the water were strictly unimodal (Fig. 5D–F), failing to exhibit the scoters' bimodal pattern of density in flight (corresponding to migratory periods; Fig. 4D–F) or eider on the water (Fig. 5A–C). Densities of scoters on the water were uniformly low during fall migration, except at Pt. Judith, contrasting with highly variable abundance among sites during the winter and spring migration. The areas experiencing the highest winter densities of foraging scoter (Watch Hill, Beavertail/Brenton Point, and Ruggles) were scattered along the Rhode Island coast (Fig. 5A–C). During spring migration, Ruggles hosted consistently greater densities of scoters on the water, although Watch Hill and Beavertail/Brenton Point retained relatively high densities early into the spring migration period (Fig. 5A–C).

Discussion

The spatiotemporal patterns in abundance and behavior of sea ducks that we detected using land-based surveys provide essential information not otherwise available from other types of waterfowl and seabird surveys in southern New England. Prior to these land-based surveys, few quantitative studies existed on the spatial distribution and abundance of sea ducks in the region. In southern New England, the mid-winter waterfowl aerial survey, conducted once annually by Rhode Island Department of Environmental Management, does not survey nearshore areas where most sea ducks concentrate (J. Osenkowski, unpubl. data). Winiarski et al. (2014) conducted aerial and boat-based surveys throughout Rhode Island and Block Island Sounds, but nearshore habitats were not sampled due to Federal Aviation Administration restrictions and water depth constraints for larger survey ships. Therefore, the results from our land-based surveys provide valuable baseline information on fine-scale spatiotemporal dynamics of sea ducks of nearshore waters in the region. This type of information will be useful for planners interested in minimizing effects to local sea duck populations from anthropogenic disturbance, such as offshore wind energy facilities, by determining where the highest concentrations of sea ducks and other marine birds are located (Langston 2013; De La Cruz et al. 2014).

Figure 5.

Seasonal variation (August through June) in the expected density (individuals/km2) of (A–C) Common Eider and (D–F) Black, Surf, and White-winged scoters detected on the ocean surface at nine sites in southern Rhode Island based on generalized additive models. Sites were grouped into three geographic regions (see Figs. 1 and 4).

f05_252.jpg

We detected substantially greater numbers of birds in flight near sunrise, which concurs with other land-based surveys in the western Atlantic (Bond et al. 2007; Cameron 2014). Therefore, biologists interested in potential collision risk with wind turbines and other objects should conduct surveys of flight activity during the first 3 hours after sunrise. Increased flight activity near sunrise presumably includes movements from offshore nocturnal roosts to nearshore foraging areas (e.g., Lewis et al. 2005). However, we did not detect increased flight activity near sunset, which suggests that movements from nearshore foraging sites to offshore roosts in southern Rhode Island occurs after sunset or via swimming (Mudge and Allen 1980; Reed and Flint 2007).

Peak numbers of flying sea ducks during fall and spring corroborated the presumed migratory periods throughout the region (Veit and Petersen 1993; Beuth 2013; Baldassarre 2014; Loring et al. 2014), including the slightly extended spring migratory period for scoters relative to eider. Most sites exhibited a bimodal annual cycle, with peak numbers of eider and scoter's in flight in fall and spring. Due to the orientation of the coastline in southern Rhode Island, most sea ducks migrated east or west in this area. Therefore, we did not expect the abundance of sea ducks in flight to vary substantially among sites during migratory periods. The variation documented among sites suggests that monitoring a single site may be insufficient to capture variation in local (e.g., wintering) sea duck abundance and distribution, although a single site may be sufficient to monitor their general migratory movements (e.g., Cameron 2014).

Land-based surveys of nearshore environments complement existing standardized sea duck surveys along the Atlantic Coast, such as the aerial surveys conducted along much of the Atlantic Coast (Zipkin et al. 2010; Silverman el al. 2013) or at smaller spatial scales (Winiarski et al. 2014). We suggest that systematic, multi-site land-based surveys fill an important gap in aerial and boat-based surveys: they document the dynamics of sea duck abundance and distribution in important nearshore areas missed by most aerial surveys and single-site migration counts. This information can inform regional conservation planning (e.g., harvest management: Merkel 2004; placement of offshore wind turbines: Langston 2013). Furthermore, while satellite telemetry provides important movement and migratory connectivity details (Beuth 2013; Loring et al. 2014), inferences of resource selection from satellite telemetry typically require broad generalizations from relatively small samples of individuals that may not have been representatively sampled from the population. Land-based surveys could facilitate the evaluation of resource selection and anthropogenic disturbance at the population rather than individual level.

Acknowledgments

We thank John Veale and Alex Patterson for assistance collecting data in the field. CBC data were provided by the National Audubon Society and through the generous efforts of Bird Studies Canada and countless volunteers across the western hemisphere. We thank Tim Bowman and Ian Cameron for ahead-of-print access to manuscripts. This research was funded by grants from the U.S. Department of Energy the American Recovery and Reinvestment Act, Rhode Island Division of Fish and Wildlife, U.S. Department of the Interior, Fish and Wildlife Service Wildlife and Sport Fish Restoration Program, and the State of Rhode Island. We thank Keith Mueller for permission to modify his scoter and eider photographs for inclusion in our figures. Three reviewers substantially improved the manuscript. Data and R code necessary to reproduce the analysis are available from the corresponding author.

Literature Cited

1.

J. Austin , S. Slattery and R. G. Clarke . 2014. Waterfowl populations of conservation concern: learning from diverse challenges, models and conservation strategies. Wildfowl Special Issue 4: 470–497. Google Scholar

2.

G. Baldassarre 2014. Ducks, geese, and swans of North America. Johns Hopkins University Press, Baltimore, Maryland. Google Scholar

3.

J. Beuth 2013. Body composition, movement phenology and habitat use of Common Eider along the southern New England coast. M.S. Thesis, University of Rhode Island, Kingston. Google Scholar

4.

A. L. Bond , P. W. Hicklin and M. Evans . 2007. Daytime spring migrations of scoters (Melanittaspp.) in the Bay of Fundy. Waterbirds 30: 566–572. Google Scholar

5.

I. R. Cameron 2014. An analysis of spring migration rates of scoter species and Common Eider recorded at the Point Lepreau Bird Observatory on the Bay of Fundy over the period 2000 to 2012. New Brunswick Naturalist 41: 94–99. Google Scholar

6.

C. J. Camphuysen , C. M. Berrevoets , H. J. W. M. Creniers , A. Dekinga , R. Dekker , B. J. Ens , T. M. van der Have , R. K. H. Kats , T. Kuiken , M. F. Leopold and others . 2002. Mass mortality of Common Eiders (Somateria mollissima) in the Dutch Wadden Sea, winter 1999/2000: starvation in a commercially exploited wetland of international importance. Biological Conservation 106: 303–317. Google Scholar

7.

M. J. Cox , D. L. Borchers and M. Kelly . 2013. nupoint: an R package for density estimation from point transects in the presence of nonuniform animal density. Methods in Ecology and Evolution 4: 589–594. Google Scholar

8.

S. E. W. De La Cruz , J. M. Eadie , A. K. Miles , J. Yee , K. A. Spragens , E. G. Palm and J. Y. Takekawa . 2014. Resource selections and space use by sea ducks during the non-breeding season: implications for habitat conservation planning in urbanized estuaries. Biological Conservation 169: 68–78. Google Scholar

9.

Environmental Systems Research Institute (ESRI). 2011. ArcGIS Desktop v. 10. ESRI, Redlands, California. Google Scholar

10.

A. D. Fox 2003. Diet and habitat use of scoters Melanittain the western Palearctic. Wildfowl 54: 163–182. Google Scholar

11.

M. Guillemette , J. H. Himmelman , C. Barette and A. Reed . 1993. Habitat selection by Common Eiders in winter and its interaction with flock size. Canadian Journal of Zoology 71: 1259–1266. Google Scholar

12.

T. J. Hastie and R. J. Tibshirani . 1990. Generalized additive models. CRC Press, Boca Raton, Florida. Google Scholar

13.

D. H. Johnson 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61: 65–71. Google Scholar

14.

R. H. W. Langston 2013. Birds and wind projects across the pond: a UK perspective. Wildlife Society Bulletin 37: 5–18. Google Scholar

15.

T. L. Lewis , D. Esler , W. S. Boyd and R. Žydelis . 2005. Nocturnal foraging behavior of wintering Surf Scoters and White-winged Scoters. Condor 107: 637– 647. Google Scholar

16.

P. H. Loring , P. W. C. Paton , S. R. McWilliams , R. A. McKinney and C. A. Oviatt . 2013. Densities of wintering scoters in relation to benthic prey assemblages in a North Atlantic estuary. Waterbirds 36: 144–155. Google Scholar

17.

P. H. Loring , P. W. Paton , J. E. Osenkowski , S. G. Gilliland , J.-P. L. Savard and S. R. McWilliams . 2014. Habitat use and selection of Black Scoters in southern New England and siting of offshore wind energy facilities. Journal of Wildlife Management 78: 645–656. Google Scholar

18.

J. Madsen 1998. Experimental refuges for migratory waterfowl in Danish wetlands. I. Baseline assessment of the disturbance effects of recreational activities. Journal of Applied Ecology 35: 386–397. Google Scholar

19.

T. A. Marques , S. T. Buckland , D. L. Borchers , D. Tosh and R. A. McDonald . 2010. Point transect sampling along linear features. Biometrics 66: 1247–1255. Google Scholar

20.

G. Marra and S. N. Wood . 2011. Practical variable selection for generalized additive models. Computational Statistics and Data Analysis 55: 2372–2387. Google Scholar

21.

R. A. McKinney , S. R. McWilliams and M. A. Charpentier . 2006. Waterfowl—habitat associations during winter in an urban North Atlantic estuary. Biological Conservation 132: 239–249. Google Scholar

22.

F. R. Merkel 2004. Impact of hunting and gillnet fishery on wintering eiders in Nuuk, southwest Greenland. Waterbirds 27: 469–479. Google Scholar

23.

F. R. Merkel , A. Mosbech and F. Riget . 2009. Common Eider Somateria mollissimafeeding activity and the influence of human disturbances. Ardea 97: 99–107. Google Scholar

24.

G. P. Mudge and D. S. Allen . 1980. Wintering seaducks in the Moray and Dornoch Firths, Scotland. Wildfowl 31: 123–130. Google Scholar

25.

M. C. Perry and A. S. Deller . 1996. Review of factors affecting the distribution and abundance of waterfowl in shallow-water habitats of Chesapeake Bay. Estuaries 19: 272–278. Google Scholar

26.

J. Pinheiro , D. Bates , S. Debroy , D. Sarkar and R Development Core Team. 2014. nlme: linear and non-linear mixed effects models. R package v. 3.1-118. R Foundation for Statistical Computing, Vienna, Austria.  http://CRAN.R-project.org/package=nlme, accessed 12 October 2014. Google Scholar

27.

R Development Core Team. 2014. R: a language and environment for statistical computing v.3.1.2. R Foundation for Statistical Computing, Vienna, Austria.  http://www.R-project.org/, accessed 15 November 2014. Google Scholar

28.

J. A. Reed and P. L. Flint . 2007. Movements and foraging effort of Steller's Eiders and Harlequin Ducks wintering near Dutch Harbor, Alaska. Journal of Field Ornithology 78: 124–132. Google Scholar

29.

E. D. Silverman , D. T. Saalfeld , J. B. Leirness and M. D. Koneff . 2013. Wintering sea duck distribution along the Atlantic Coast of the United States. Journal of Fish and Wildlife Management 4: 178–198. Google Scholar

30.

U.S. Geological Survey. 2012. Spring 2011 Rhode Island Statewide LiDAR Data - Provisional Release. University of Rhode Island Environmental Data Center, Kingston, Rhode Island,  http://www.edc.uri.edu/rigis/data/download/lidar/2011USGS/, accessed 8 November 2014. Google Scholar

31.

R. R. Veit and W. Petersen . 1993. Birds of Massachusetts. Massachusetts Audubon Society, Lincoln, Massachusetts. Google Scholar

32.

T. P. White , R. R. Veit and M. C. Perry . 2009. Feeding ecology of Long-tailed Ducks Clangula hyemaliswintering on the Nantucket Shoals. Waterbirds 32: 293– 299. Google Scholar

33.

K.J. Winiarski , D. L. Miller , P. W. C. Paton and S. R. McWilliams . 2013. Spatially explicit model of wintering Common Loons: conservation implications. Marine Ecology Progress Series 492: 273–283. Google Scholar

34.

K. J. Winiarski , D. L. Miller , P. W. C. Paton and S. R. McWilliams . 2014. A spatial conservation prioritization approach for protecting marine birds given proposed offshore wind energy development. Biological Conservation 169: 79–88. Google Scholar

35.

S. N. Wood 2006. Generalized additive models: an introduction with R. Chapman & Hall/CRC, Boca Raton, Florida. Google Scholar

36.

S. N. Wood 2011. Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 73: 3–36. Google Scholar

37.

E. F. Zipkin , B. Gardner , A. T. Gilbert , A. F. O'Connell Jr , J. A. Royle and E. D. Silverman . 2010. Distribution patterns of wintering sea ducks in relation to the North Atlantic Oscillation and local environmental characteristics. Oecologia 163: 893–902. Google Scholar
Adam D. Smith, Scott R. McWilliams, Kristopher J. Winiarski, Carol L. Trocki, Brian Harris, Jason E. Osenkowski, and Peter W. C. Paton "Using Land-Based Surveys to Assess Sea Duck Abundance and Behavior in Nearshore Waters of Southern New England, USA," Waterbirds 38(3), 252-259, (1 September 2015). https://doi.org/10.1675/063.038.0305
Received: 26 January 2015; Accepted: 25 May 2015; Published: 1 September 2015
JOURNAL ARTICLE
8 PAGES


SHARE
ARTICLE IMPACT
Back to Top