Sea ducks in the Atlantic flyway: population status and a review of special hunting seasons

              U.S. Fish & Wildlife Service
Sea Ducks in the Atlantic Flyway:
Population Status and a Review of
Special H'unting Seasons
David F. Caithamer
Mark Otto
Paul I. Padding
John R. Sauer
George H. Haas
~ ..... ............•....•..
: UOSODARThlllN..T.~.o ..F£ • $ lil~""'N.cw,.\_
~ -.,; .. ·.-
• ".~'- :'<," , .•.
• ' ..... ,.....':O:~ '_""I,~¥
• .,,'h-', '~
: ~1~~~'OC ~V '~~~
: MIGRATORY BIRD HUNTING srAMP
Sea Ducks in the Atlantic Flyway:
Population Status and a Review of Special Hunting Seasons
David F Caithamer
. Office of Migratory Bird Management
U.S. Fish and Wildlife Service
11500 American Holly Drive
Laurel, MD 20708-4016
Mark Otto
Office ofMigratory Bird Management
U.S. Fish and Wildlife Service
11500 American Holly Drive
Laurel, MD 20708-4016
Paul 1. Padding
Office of Migratory Bird Management
U.S. Fish and Wildlife Service
10815 Loblolly Pine Drive
Laurel, MD 20708-4028
John R. Sauer
Biological Resources Division
U.S. Geological Survey
11510 American Holly Drive
Laurel, MD 20708-4017
George H. Haas
U.S. Fish and Wildlife Service
300 Westgate Center Drive
Hadley, MA 01035-9589
February 2000
Sea Ducks in the Atlantic Flyway:
Population Status and a Review of Special Hunting Seasons
Executive Summary
Special seasons for sea ducks began in the Atlantic Flyway in 1938 when 5 northeastern
states were allowed 16-30 day extensions to their regular duck seasons. Initially, only scoters
were legal game during the extensions. Over time, the season became more liberal aniby 1972,
13 of the 17 states in the Flyway had special seasons that lasted 107 days and had bag limits of 7
eiders, oldsquaws, and scoters in the aggregate. The Special Sea Duck Season in the Atlantic
Flyway remained essentially unchanged until 1993, when a review ofthe status of sea ducks led
to a reduction in the scoter bag limit to 4. Growing concern for the status of sea ducks and the
need to evaluate the special season, especially the effects of restrictions on scoter bag limits,
prompted our investigations.
We assessed trends in harvest and 4 long-term measures of sea duck abundance: (1)
breeding population estimates from Canada and Alaska, (2) Christmas Bird Counts along the
Atlantic Coast, (3) Mid-winter Inventory estimates from the Atlantic Flyway, and (4)
Availability Indices (harvest/successful sea duck hunters) for the Flyway. We tested for changes
in trends and levels that coincided with major changes in hunting regulations in the Flyway. In
addition, we tested for annual changes in estimates of sea duck densities observed on the Atlantic
Flyway Sea Duck Survey during 1991-97.
Harvest ofoldsquaws, eiders, and scoters increased during 1963-71 as increasing
numbers ofstates adopted the special season. Oldsquaw harvest remained stable during 1972-96,
while harvest ofeiders continued to increase. Scoter harvest declined during 1972-93, but
appears to have stabilized since bag limits were restricted in 1993. We detected a decrease
(64%) in the harvest of white-winged scoters coincident with bag limit restrictions, but detected
no change in harvests of black scoters, surf scoters, and total scoters.
Numbers of common eider in the Atlantic Flyway appeared to have increased during
1972-97. Trends in indices ofoldsquaw abundance were inconsistent for this period, but the
index that we believe is most reliable reflected a stable population. Scoter population indices
2
were stable or declining during 1972-92. Since 1993, we can only weakly infer stable or
increasing populations for scoters.
We concluded that changes in hunting regulations in the Atlantic Flyway can produce
measurable changes in harvest. However, we found only weak evidence that changes in
population status coincided with changes in regulations. Nonetheless, we believe that mortality
of sea ducks from hunting is nearly completely additive to natural mortality, considering the life­history
characteristics ofsea ducks. Although sea ducks have smaller harvests than many other
waterfowl, conservative hunting regulations seem prudent considering our overall state of
knowledge of these birds. We recommend consideration of a regular sea duck season that
replaces the Special Sea Duck Season, and eliminating sea ducks from the regular duck season.
We suggest eliminating provisions for special sea duck zones from the Federal framework of
regulations.
We recommend that the Atlantic Flyway Council in cooperation with others develop a
management plan for sea ducks. The plan should not be restricted to harvest management, but
should address other issues and information needs. The plan should be developed in concert with
the Sea Duck Joint Venture ofthe North American Waterfowl Management Plan. Lastly, we
suggest a goal to maintain sea duck populations at or above levels observed during the 1970's.
3
Introduction
Concern over the status of sea ducks worldwide has increased in recent years. In North
America, this concern relates to the limited state of knowledge of this group compared to many
other waterfowl (Bellrose 1980, Goudie et aL 1994), reports of declining populations (Goudie
1989, Kertell 1991, Stehn et aL 1993, Ad Hoc Sea Duck Committee Atlantic Flyway Technical
Section 1994), the potential impact that hunting may have on their status (Reed and Erskine
1986, AdHoc Sea Duck Committee Atlantic Flyway Technical Section 1994, Krementz et aL
1996, Krementz et aL 1997), and the susceptibility ofthese birds to catastrophic and chrJlnic
environmental degradation (Di Giulio and Scanlon 1984, Ohlendorf et aL 1986, Ohlendorf and
Fleming 1988, Piatt et aL 1990, Franson et aL 1995). Our purpose was to assess population
status and trends of sea ducks commonly found in coastal areas of the eastern U.S. The species
we considered are oldsquaw (Clangula hyemalis), harlequin duck (Histrionicus histrionicus),
common eider (Somateria mollissima), black scoter (Melanitta nigra), white-winged scoter (M.
fusca), and surf scoter (M. perspicillata). We chose to not analyze data on king eider (S.
spectabilis) because their wintering range barely extends into the U.S. portion of the Atlantic
Flyway.
Sea ducks:are hunted in the Atlantic Flyway during regular duck and Special Sea Duck
Seasons. The Special Sea Duck Seasons have never been reviewed and evaluated since their
inception in 1938, although the status ofsea ducks in eastern North America was evaluated in
1993 and 1994 (Office of Migratory Bird Management 1993, Ad Hoc Sea Duck Committee
Atlantic Flyway Technical Section 1994). Those assessments revealed mostly decreasing trends
in indices of scoter populations, stable or increasing trends in indices ofcommon eider
populations, and stable or decreasing trends in indices of oldsquaw populations. Based largely
on these findings, the USFWS and Atlantic Flyway Council agreed in 1993 to restrict bag limits
ofscoters during Special Sea Duck Seasons in the Atlantic Flyway. USFWS policy states that
special seasons "may be re-evaluated for their effectiveness, appropriateness and necessity when
situations (are) warranted" (U.S. Fish and Wildlife Service 1988). In addition to assessing the
status and trends of sea duck populations, we also assess impacts of recent restrictions on scoter
bag limits and other impacts of special seasons on scoter, eider, and oldsquaw populations.
We retrospectively described and examined trends in estimates of harvest, the ratios of
young per adult duck in the harvest, and indices of breeding and wintering populations. We
4
searched for correlation between changes in hunting regulations and changes in estimates of
harvest and indices of population size. Our analyses focused on data that were most pertinent to
the U.S. portion ofthe Atlantic Flyway.
We thank G. T. Allen, A. W. Brackney, J. P. Bladen, R. E. Cummins, K. M. Dickson, B.
A. Hoover, P. D. Keywood, R. J. King, and J. R. Serie for assistance. Biologists from states and
provinces in the Atlantic Flyway provided information on hunting seasons within their
jurisdiction. The Wildlife Management Institute graciously allowed reproductions offigures
from Bellrose (1980). Numerous other persons and agencies supplied data via support for, or
participation in, various wildlife management activities and surveys.
History of sea duck hunting regulations in the Atlantic Flyway
Since adoption ofthe Migratory Bird Treaty Act in 1918, sea duck hunting regulations in
the Atlantic Flyway have become progressively more liberal and complex (Appendix 1). Sea
duck harvests were regulated through regular duck season limits from 1918 through 1937 in all
states ofthe Flyway. During most ofthis period (1918-31), the season was closed for eiders. In
1938, a Special Sea Duck Season was established for Maine, New Hampshire, Connecticut,
Massachusetts, and Rhode Island. This season, during which scoters could be taken "in open
coastal waters only, beyond outer harbor lines," was open from September 15 until opening day
ofthe regular duck season. The daily bag and possession limits during this special season were
the same as those for the regular duck season. Other than changes in the length of the special
season due to changes in opening dates of the regular duck seasons, those regulations remained
the same during 1938-46, except for the addition of a special season in New York, including
Long Island, in 1940. During 1938-46, the length of the sea duck season was 5-20 days in Maine
and New Hampshire, and 16-41 days in Connecticut, Massachusetts, New York, and Rhode
Island.
Regular duck seasons were reduced from 45 days in 1946 to 24 (Maine, New Hampshire,
and New York) or 30 (Connecticut, Massachusetts, and Rhode Island) days in 1947, and daily
bag limits were reduced from 7 ducks in 1946 to 4 in 1947. This appears to have led to an
expansion of sea duck seasons beginning in 1947; New Hampshire (36 days, opening on
September 1), Connecticut (63 days), Massachusetts (63 days), and Rhode, Island (77 days) had
special seasons prior to their regular seasons, whereas Maine (72 days) and New York (89 days)
5
had sea duck seasons that began before and extended through their regular seasons. Also, bag
limits for sea ducks were separated from limits for other ducks at this time; the daily bag and
possession limits for sea ducks seasons was 7 and 14 scoters, respectively. In 1948, eiders were
included in the legal bag for Special Sea Duck Seasons, and oldsquaw was added in 1950.
Connecticut was allowed an additional 6-day late season immediately after the end of its 1948
regular season, and in 1949 and 1950, all 6 states had 92-day special seasons from September 17
through December 17. During 1951-57, season dates and lengths were state-specific, with
opening dates of September 14 - October 17 and closing dates of December 29 - January 5.
Sea duck season dates remained standardized for all participating states during 1958-72.
The seasons ranged from 100-108 days in length, opened in late September or early October, and
closed in early or mid-January. A significant change in sea duck hunting regulations was made
in 1960; all states in the Atlantic Flyway were allowed "in addition to the bag limit on other
ducks, a daily bag limit of7 and a possession limit of 14 eider, old-squaw (sic), and scoter ducks,
singly or in the aggregate of these species" during the regular duck season.
During 1963-71, existing sea duck zones were expanded and the number of states
offering Special Sea Duck Seasons increased from 6 to 13. Sea duck zones in Connecticut,
Maine, Massach~setts, New Hampshire, and Rhode Island were redefined in 1963 as "all coastal
waters and all waters of rivers and streams lying seaward from the first upstream bridge," with
similar but more detailed area-specific zone descriptions for New York. Maryland, New Jersey,
and North Carolina were allowed Special Sea Duck Seasons beginning in 1966, "in any waters of
the Atlantic Ocean and/or in any tidal waters of any bay which are separated by at least 1 mile of
open water from any shore, island, and emergent vegetation: provided, that any such areas have
been described, delineated, and designated as special sea duck areas under the hunting
regulations adopted by the respective States." Georgia, Virginia, and South Carolina were
allowed Special Sea Duck Seasons under the same guidelines in 1968, as was Delaware in 1971.
The minimum allowable distance from any shore, island, or emergent vegetation was apparently
intended to protect riparian landowner rights, reduce disturbance of other waterfowl, and avoid
competition between sea duck hunters and waterfowl hunters using stationary blinds (Stotts
1966, L. Hindman, Md. Wildlife Heritage Div., pers. commun.). This distance was reduced to
1200 yards in Maryland in 1969 and 800 yards in 1970; by 1975, the prescribed distance for
Delaware, North Carolina and Virginia was also 800 yards (Fig. I).
6
From 1973 to 1997, more general frameworks have been used for Special Sea Duck
Seasons and regulations have remained relatively similar from year to year. In 1973, framework
dates were established at September 1 - January 20, season length at 107 days, and bag and
possession limits at 7 and 14, respectively. The opening framework date was changed from
September 1 to September 16-18 during 1976-1978, and to September 15 during 1979-1997.
The flyway-wide "bonus bag" of sea ducks during the regular season ended in 1987, when eiders,
oldsquaw, and scoters became part of the overall duck bag in all areas except designated sea duck
zones. In sea duck zones, however, hunters could still take a limit of sea ducks in addition to a
bag of other ducks during the regular season. The season on harlequin ducks was closed in the
Atlantic Flyway in 1989, and has remained closed since then. In 1993, the daily bag limit on
scoters was reduced to 4, while the aggregate bag limit on eiders, oldsquaws, and scoters
remained at 7. These bag limits have been used in Special Sea Duck Seasons of the Atlantic
Flyway since then.
In 1997, states that prohibited hunting on Sundays were allowed additional waterfowl
hunting days to compensate for Sunday closures. Sea duck seasons in Delaware, Maine, New
Jersey, and North Carolina encompassed 112-125 days, although the number ofhunting days
remained g 07 days. Other states eligible for "compensatory days" did not use them to extend
sea duck seasons.
Natural History
01dsquaw
No subspecies of oldsquaw are recognized (Sibley and Monroe 1990). They nest
circumpolarly in tundra habitat near coastlines, lakes, and ponds and winter along the Pacific and
Atlantic coasts and in the Great Lakes of North America (Fig. 2) (Johnsgard 1978, Bellrose
1980). Oldsquaws usually fonn their first pair bonds in their second winter and attempt nesting
the following spring (Johnsgard 1978, Bellrose 1980). Pair bonds generally last only through the
nesting period, although some females will pair with the same male in successive years (Alison
1975, Johnsgard 1978, Bellrose 1980, Oring and Sayler 1992). Philopatry to nesting areas can be
strong, as some nesting pairs return to the same pond in successive years(Alison 1975, Bellrose
1980). Clutch size averages 6-7 eggs and incubation lasts about 26 days (Alison 1975, Johnsgard
7
1978, Bellrose 1980). Nest success averaged 59% and no renesting was observed in one study
near Churchill, Manitoba (Alison 1975). Males leave their mates during incubation (Alison
1975, Johnsgard 1978, Bellrose 1980, Oring and Sayler 1992). Females occasionally abandon
their brood to begin the postnuptial molt, which can lead to amalgamations of several broods
without parents (Johnsgard 1978, Bellrose 1980). Most adult males migrate to molting areas for
the summer wing molt (Salomonsen 1968, Hohman et al. 1992). Young oldsquaws require only
35 days to attain flight (Johnsgard 1978, Bellrose 1980). Oldsquaws feed primarily on
crustaceans and mollusks (Cottam 1939, Stott and Olson 1973, Johnsgard 1978, Bellro1ie 1980).
Harlequin Duck
No subspecies of harlequin ducks are recognized (Sibley and Monroe 1990). In North
America, the range of harlequins appears discontinuous (Fig. 2) and is assumed to be comprised
of2 distinct populations (Bellrose 1980, Cassirer et al. 1991). In the east, birds nest from
Greenland south to the Gulf of St. Lawrence, and winter in coastal areas from Greenland
southward to the Chesapeake Bay (Bellrose 1980, Vickery 1988). Recent studies have
demonstrated that some harlequins move between Greenland and Quebec (P. Laporte, Can.
Wildlife Service; pers. commun.). In the west, harlequins range from Alaska southward to
California. The western population is much larger than the eastern population (Bellrose 1980).
Harlequin ducks were listed as endangered in eastern Canada in 1991. Harlequins typically nest
on rocky shorelines of turbulent mountainous rivers and spend the winter along rocky ocean
coastlines (Bellrose 1980). They probably form their first pair bonds late in their second winter
and are not known to nest until they are 2 years old (Johnsgard 1975, Bellrose 1980). Some pairs
may remain paired for more than one year or re-pair repeatedly (Oring and Sayler 1992). Clutch
size averages 5-6 eggs (Bengtson 1965, Jarvis and Bruner 1996) and incubation lasts 28-30 days
(Bengtson 1965, Johnstone 1970). Little information is available on nest success, time required
for young to attain flight, and post-breeding movements of adults (Bellrose 1980). Harlequin
ducks eat mostly crustaceans, mollusks, and insects (Cottam 1939).
Common Eider
Common eiders have a circumpolar distribution in the Northern Hemisphere (Fig. 3)
(Bellrose 1980). Sibley and Monroe (1990) recognize 2 subspecies ofcommon eider in North
8
America, while Bellrose (1980) and Johnsgard (1978) recognize 4; we accept the classification of
Bellrose and Johnsgard. The American eider (Somareria mollissima dresseri) is the only race
typically encountered in the eastern U.S. They breed in coastal areas from Massachusetts to
southern Labrador, and spend winters in coastal waters from the Gulf of St. Lawrence to New
Jersey. American eiders have recovered from extremely low numbers ofthe late 1800's (Krohn
et al. 1992). At various times, adult eiders and eggs have been transplanted to potential nesting
sites (Heusmann 1995). The other North American races are found in Hudson and James bays
(8. m. sedentaria), western Canada and Alaska (8. m. v-nigra), and northeastern Canada and
Greenland (S. m. borealis) (Bellrose 1980) (Fig. 3). All races utilize coastal marine habitats
extensively. Common eiders form pairs in their second or later winter (Spurr and Milne 1976).
Common eiders are seasonally monogamous and some may reestablish pair bonds with previous
mates (Spurr and Milne 1976, Bellrose 1980). American eiders do not breed until they are at
least 3 years old (Mendall 1968). Female eiders rely extensively on nutrient reserves acquired
before nesting to lay and incubate a clutch of eggs (Korschgen 1977). They commonly nest in
colonies on islands where many females return to their same nesting site in successive years
(Cooch 1965, Johnsgard 1975, Bellrose 1980). Clutch sizes are typically 4 or 5, and incubation
lasts about 26 days (Korschgen 1977, Bellrose 1980, van Dijk 1986). If their nest is destroyed,
some females will renest (Cooch 1965, Korschgen 1977). Multiple broods and females often
coalesce into creches (Munro and Bedard 1977, Prach et al. 1986). Predation rate of flightless
young can exceed 90% in some areas (Mendenhall and Milne 1985, Mawhinney and Diamond
1997). Young require about 60 days to be capable of flight (Cooch 1965). Some eiders migrate
to molting areas (Abraham and Finney 1986). Mollusks, especially blue mussels, crustaceans,
and other invertebrates are important foods to common eiders (Cottam 1939).
Black Scoter
Two subspecies of black scoters are recognized; the American black scoter (M. n.
americana) is found in North America (Sibley and Monroe 1990). Its nesting range includes
Alaska, the lowlands near Hudson Bay, and other areas across Canada (Fig. 3) (Johnsgard 1975,
Bellrose 1980, Savard and Lamothe 1991). Black scoters spend the winters in salty or brackish
waters along the Pacific and Atlantic coasts. Pairs are first formed during their second winter
and nesting is attempted the following spring. Little is known about nesting of black scoters in
9
North America (Johnsgard 1978, Bellrose 1980). Apparently, males leave their mates when
incubation begins (Johnsgard 1975, Bordage and Savard 1995). On the Yukon Delta of Alaska,
black scoter clutch sizes range from 5 to 8 eggs (Brandt 1943 cited by Bellrose 1980). In
Iceland, initial clutches average 9 eggs, and clutches of renesting scoters average 6 eggs
(Johnsgard 1978). However, Bellrose (1980) believes that renesting is uncommon among black
scoters in North America. Incubation lasts 27-28 days and brood mixing is not typical
(Johnsgard 1978). It probably takes 6-7 weeks for young to achieve flight (Johnsgard 1978,
Bellrose 1980). In North America and Europe, black scoters migrate to molting are$lS during the
summer (Salomonsen 1968, Bellrose 1980, Hohman et al. 1992, Bordage and Savard 1995).
Black scoters eat mostly mollusks, crustaceans, and other invertebrates (Cottam 1939, Stott and
Olson 1973, Johnsgard 1975).
White-winged Scoter
Three subspecies of white-winged scoter are recognized (Sibley and Monroe 1990). The
American white-winged scoter (M.f deglandi) is found in North America where it breeds
primarily in coniferous forest and parkland habitats of Alaska and western Canada (Fig. 4)
(Johnsgard 1975'; Bellrose 1980, although see Savard and Lamothe 1991). However, their
breeding range has apparently contracted northward since the late 1940's (Ad Hoc Sea Duck
Committee Atlantic Flyway Technical Section 1994). White-winged scoters winter in salty and
brackish habitats along the Pacific and Atlantic coasts (Johnsgard 1975, Bellrose 1980). Like
black scoters, white-winged scoters are seasonally monogamous and they do not pair or attempt
to breed until their second year (Johnsgard 1975, Bellrose 1980, Brown and Houston 1982).
Females often return to the same nesting site in successive years (Brown and Brown 1981). In
southern portions of their breeding range, their clutch size averages about 9 eggs and the nesting
success rate averages about 70% (Brown and Brown 1981, Brown and Fredrickson 1989).
Bellrose (1980) believes that renesting is uncommon in white-winged scoters. Many females
abandon their broods within the first few weeks after hatch and the ducklings aggregate into
creches that are accompanied by variable numbers of adult females (Brown and Brown 1981,
Kehoe 1989). Young scoters require 63-77 days to attain flight (Hochbaum 1944, Brown and
Fredrickson 1997). Adults commonly migrate to molting areas in the summer (Salomonsen
1968, Johnson and Richardson 1982, Brown and Fredrickson 1989). Young and adult scoters in
10
central Saskatchewan feed primarily on amphipods during the summer (Brown and Fredrickson
1986). On wintering areas, they eat mostly mollusks, crustaceans, and other invertebrates
(Cottam 1939, Stott and Olson 1973).
SurfScoter
No subspecies ofsurf scoter are recognized (Sibley and Monroe 1990). They breed in
boreal forests of Alaska and Canada and winter in brackish and salty waters along the Atlantic
and Pacific coasts (Fig. 4) (Bellrose 1980, also see Reed et al. 1994). Surfscoters are probably
seasonally monogamous and first pair and breed at the end oftheir second year (Johnsgard
1975). Average clutch size is probably 5-7 eggs (Bent 1925). Little is known about nesting
success of surf scoters and length of time required for young to be capable of flight (Bellrose
1980). In a southern portion of its breeding range, creching behavior seems common (Reed et al.
1994). Adults migrate to molting areas (Johnson and Richardson 1982, Salter et al. 1980).
Mollusks and crustaceans are important foods during winter while insects are more important to
juveniles in summer (Cottam 1939, Stott and Olson 1973).
METHODS
Data Collection
Harvest and Recruitment Index.- Recreational harvest of ducks in the U.S. is annually
estimated by the U.S. Fish and Wildlife Service through a questionnaire survey ofFederal duck
stamp purchasers (Martin and Carney 1977). Survey respondents report the number of days they
hunted waterfowl and the number of sea ducks, other ducks, geese, and coots they bagged in each
state in which they hunted waterfowl. Combined with a complete count of the number of Federal
duck stamps sold, results of this survey provide estimates of the total U.S. harvest of sea ducks
and other waterfowl. The survey also provides estimates of the number of active waterfowl
hunters in each state and the number of days they hunted waterfowl, but it does not provide any
estimate of hunter activity specific to sea duck hunting. The estimated number of successful sea
duck hunters is the only index of sea duck hunting effort that is available from this survey.
The U.S. Fish and Wildlife Service also conducts an annual waterfowl parts survey, the
sample for which consists of hunters who reported bagging 2:1 duck, goose, or coot during the
11
previous hunting season (Martin and Carney 1977). Respondents are asked to send a wing from
every duck and coot they bag and the tail feathers of each goose they bag, and to report the state,
county, and date of harvest for each bird. Biologists can determine the species, sex, and age
(immature or adult) of a duck from its wing plumage (Carney 1992). Thus, sea duck wings
received through this survey, combined with estimates of the total sea duck harvest, provide
estimates of the species, sex, and age composition as well as the geographic and temporal
distribution ofthe sea duck harvest in the U.S. The precision of these estimates is dependent on
the number of hunters responding to the questionnaire survey and the number of wings. received.
The U.S. Fish and Wildlife Service does not estimate the variances of its estimates, but they are
probably large for sea ducks harvest estimates, especially at the state level. Precision ofthe U.S.
Fish and Wildlife Service harvest estimates diminishes for species with small harvests, such as
sea duck species, and for smaller geographic areas (Geissler 1990).
The Canadian Wildlife Service similarly estimates sport harvest in Canada (Cooch et al.
1978). Canada's National Harvest Survey consists of a questionnaire, sent to a sample of current
and previous-year national migratory bird permit purchasers, that asks hunters to report how
many days they hunted migratory birds and how many ducks, geese, and other migratory game
birds they baggell. Responses to this survey coupled with counts of total migratory bird permits
sold provide estimates of the number of active migratory bird hunters, the number ofdays they
hunted, and the number of ducks, geese, and other migratory game birds harvested in Canada.
As in the United States, the National Harvest Survey includes an annual parts survey that enables
the Canadian Wildlife Service to estimate the species, sex, and age composition of Canada's
waterfowl harvest.
Breeding Population.-Annual indices to the size of scoter, eider, and oldsquaw breeding
populations are obtained from an aerial survey across much of Canada, Alaska, and the
northcentral U.S. (Appendix 2) (Canadian Wildlife Service and U.S. Fish and Wildlife Service
1987, Smith 1995). The Breeding Waterfowl and Habitat Survey (hereafter called Breeding
Waterfowl Survey) is directed primarily at mallards and does not provide complete coverage of
the breeding ranges of some sea ducks. The survey generally begins in early May in southern
strata and finishes by mid-June in northern areas. Protocol for this survey does not require
identifying species of eiders or scoters, except since 1998 in Alaska. Here, species ofscoters is
12
identified when they are located within 100 m ofthe transect center-line. Data on harlequin
ducks are pooled with several other species.
Aerial estimates are adjusted for visibility bias. Visibility adjustments for southern areas
are determined annually through concurrent ground counts, while those for northern strata were
determined through concurrent helicopter counts that were conducted in 1986-91. These
helicopter-based visibility adjustment rates for northern strata have been used for all years ofthe
survey. However, the visibility ofwaterfowl improved dramatically in Alaska and the Yukon
Territory (strata 1-12) beginning in 1977 due to a change in the type ofairplane used for surveys
(Hodges et al. 1996). Thus, visibility adjustments determined in 1986-91 may be biased low for
pre-1977 estimates in Alaska and the Yukon Territory.
This survey became operational in most strata in 1955. However, it was not operational
in Alaska and the Yukon Territory until 1957, and in eastern Canada (strata 51-57, and 62-69)
not until 1990 or later. Because a large proportion of the total sea duck population is found in
Alaska, and because data from only a few years are available from eastern Canada, we chose to
restrict our analyses to data from strata 1-50 and 75-77 (Traditional Survey Area) during the
years 1957-97.
Mid-winter Inventory.-Waterfowl populations in states of the Atlantic Flyway are
annually surveyed by the Mid-winter Inventory (MWI), which is a series of coordinated aerial
and ground counts conducted in early January (Martin et al. 1979, Eggeman and Johnson 1989).
Survey coverage of the MWI typically includes inland and near-shore habitats, but not deepwater
areas of the ocean that often harbor large numbers of sea ducks. This survey has been criticized
because of inconsistent methodology across regions and time (Montalbano et al. 1985, Eggeman
and Johnson 1989). Because this survey does not extend northward into Canada, variable
numbers of sea ducks wintering in Canada are uncounted. Data are not tabulated separately for
each species of scoter and eider. Despite these limitations, results from mid-winter surveys have
been found to reflect changes in the size of other duck populations (Conroy et al. 1988), and we
believe that they may also reveal large and long-term changes in numbers of sea ducks.
Christmas Bird Counts (CBC).- This annual survey of birds across North America is
coordinated by the National Audubon Society (Butcher 1990, Sauer et al. 1996). Counts of birds
are collected in sample units (circles) that are 15 miles in diameter. On a selected day within 2
weeks of 25 December, volunteers search the predefined area, and record all birds encountered.
l3
· Most birds are identified to species, but occasionally the species is not determined. For example,
some counts are recorded for unidentified scoters. The number of circles surveyed in North
America increased from 512 in 1955 to 1644 in 1995. The number ofparticipants varied greatly
among circles in any year and within circles over time. Survey effort varied from a mean of 40.8
party-hours per circle in 1955 to 70.4 hours per circle in 1995. Consequently, analyses of these
data must include some adjustment for varying effort, and the exact form of the adjustment is
often not evident from the data (Butcher 1990.) In earlier years, most counts were located in
coastal areas or near large cities, and circles were developed by local coordinators ratheUhan
placed within a sampling frame. Although some coastal areas are surveyed, there have been no
consistent surveys of areas offshore. The results therefore do not provide statistical samples of
absolute abundance. Despite these limitations, we believe that CBC may be useful as long-term
indices to changes in sizes of sea duck populations.
Sea Duck Survey.-Aerial surveys designed to estimate the density of sea ducks in coastal
habitats were conducted during late January and early February of 1991, 1992, 1994, 1995, and
1997-99. Sea ducks were counted from airplanes flying over predetermined transects centered
approximately 500 meters offshore and parallel to the coast. Transects were 500 meters wide
and divided into segments that were lO nautical miles long. Survey coverage extended from
southern Georgia northward to New Brunswick and Nova Scotia. Approximately 440 flight
segments were surveyed in each of the years by 2 crews. During 1991 and 1992, < 50% ofthe
scoters were identified to species; in other years> 98% of the scoter were identified to species.
Bandings and Band Recoveries.-We obtained electronic data files ofall sea duck
bandings and band recoveries from the Bird Banding Laboratory of the U.S. Geological Survey
in September, 1997. Total numbers of bandings and recoveries were tabulated, and banding and
recovery locations were plotted on maps. No maps were made for black and surf scoters because
there were fewer than 15 band recoveries for each of these species.
Analyses
Time Series Analyses ofBreeding Waterfowl Survey, MWI, Harvest, Age Composition,
and Availability Indices.-We modeled and tested for different linear trends in annual estimates
of breeding populations, mid-winter populations, harvest, and in an index of availability. The
Availability Index is the estimated harvest divided by the estimated number of successful sea
14
duck hunters. For age composition of the harvest, we had no hypotheses to test, so we only
plotted a locally weighted regression (lowess) curve (Cleveland 1979) of each time series. Age
composition of the harvest is the proportion ofimmatures (immature ducks harvested/total ducks
harvested). Lowess is a robust regression that uses nearby time points to calculate each
smoothed value or prediction. We used Auto-Regressive-Integrated-Moving-Average (ARIMA)
time series analytical methods (Box and Jenkins 1970) to test for lack of independence in the
regression errors and to model the correlation over time (Appendix 3). We obtained valid
statistical tests of selected hypotheses by jointly modeling the linear regression terms and the
ARIMA time series errors (Box and Jenkins 1970, Time Series Staff of Census Bureau Statistical
Research Division 1995). We used an alpha-level of 0.05, unless specified otherwise.
We tested hypotheses that changes in sea duck hunting regulations in the Atlantic Flyway
would effect trends in estimates of harvest and population indices (Table 1). These regulatory
periods varied some by species. For oldsquaw, the periods were 1956-62, with relatively stable
and conservative regulations; 1963-71, when regulations became increasingly liberal; and 1972­96,
with relatively stable and liberal regulations. The regulatory periods for scoters were similar
to those for oldsquaw, except that a restrictive period was implemented beginning in 1993. Thus
the periods for scoters were 1956-62 (relatively stable and conservative), 1963-71 (increasingly
liberal), 1972-1992 (stable and liberal), and 1993-96 (stable and moderate). Population
responses were expected to lag behind regulatory changes (e.g., a change in hunting regulations
in the fall of 1993 would change the 1994 breeding population estimate). We evaluated interval­specific
population changes using a procedure similar to piecewise linear regression.
Regulations for eider hunting have remained relatively stable in Maine, Massachusetts,
Connecticut, and New Hampshire. Since 1953, the eider season has been 100-108 days with a 7
bird daily limit in these 4 states. Because eider regulations were stable there, and because these
states account for >99% of the eider harvest in the Atlantic Flyway, we had no opportunity to
assess the impacts of regulatory changes on eider harvest or population status.
In our analyses of time series data, we: (I) checked for nonstationarity or changes in the
variance over time and against the size of the estimate; (2) checked for nonstationarity or changes
in the level ofthe series; (3) conducted stepwise backward elimination to choose the significant
regression effects; (4) modeled the time series structure of the regression errors; (5) identified
point and level-shift outliers; (6) verified that the final regression plus time series model was
15
appropriate with no systematic patterns in residuals and no large autocorrelations; and (7)
graphically compared regression predictions to estimates produced through lowess smoothing
techniques (Cleveland 1979).
To check for nonstationarity in variance, we plotted each series over time and looked for
systematic changes in variability. Estimates of breeding population, mid-winter population,
harvest, and age ratio were modeled using the log-transformed values because the variance
increased with the level of the estimates. When checking breeding population estimates, which
had associated estimates of sampling error, we also plotted the cube root ofthe samptingerror
variances against the survey estimates and again against the model predictions as suggested by
Carroll and Ruppert (1988). We used log-transformations ofthe breeding population estimates
because of the positive relation between sample errors to the original estimates. Next we plotted
the logged-transformed series against its sampling error, the relative variance (variance/mean2
).
These plots revealed no relation, suggesting this was a proper transformation. After transforming
the data, we verified that there were no patterns of variability over time.
We checked for nonstationarity in the level of the series by adjusting the series with all
the possible regression variables and checking autocorrelation statistics in the regression
residuals. If the autocorrelations diminished more slowly than the rate of exponential decay, then
the data were not stationary. No evidence of nonstationarity in the levels was found in any of the
series. Stationarity was important because reporting overall means is only meaningful if the data
are stationary. Also, we then could fit simple stationary autoregressive (AR) or moving-average
(MA) models.
Next, out of all the possible periods of population change hypothesized (modeled by
slope parameters called "ramps" and interventions called "level-shifts") (Table I), we used
stepwise backward elimination to evaluatemodels with different combinations ofslopes and
level-shifts. We retained models where all regression variables were significant (Draper and
Smith 1981). Estimating the time series and the regression parameters jointly required testing of
some non-nested models, so we used the difference in the bias-corrected version of the Akaike's
Information Criterion (AIC) (Hurvich and Tsai 1989) to test for significant differences between
models. A variable was also removed if the AIC did not increase. We conducted the backward
elimination process manually and checked for outliers and for changes in ARIMA time series
16
error structure at several points in the elimination process. Changes might have indicated that an
important feature in the regression was being removed. These rarely occurred.
We identified point and level-shift outliers with an automated procedure (Time Series
Staffof Census Bureau Statistical Research Division 1995). Because of the large number of
tests that were conducted during the process, we used a critical t-value of3.8 corresponding to an
experiment-wise p-value of 0.01 in the tests (R. Templeton, Statistics New Zealand, pers.
commun.). We investigated the validity of outliers and excluded from models those that we
believed were due to deviations from standard data collection procedures.
After we determined the final set of regression variables, we again checked plots ofthe
regression residuals for systematic differences over time and variability with the level of the
residuals. We also plotted the cube root of the survey sample variances against the regression or
model predictions. This is similar to the diagnostic methods we used earlier to look for
nonstationarity in variances of raw samples.
In addition to modeling linear changes over time, we used lowess regression with a high
smoothing value (j= 0.67) to describe each time series. We checked our modeling results against
lowess estimates, expecting similar slopes from both the linear and lowess regressions.
However, when analyzing age composition data we relied exclusively on lowess regressions,
since we had no a priori hypotheses regarding change in these data.
We evaluated possible visibility differences for breeding population estimates from
Alaska and the Yukon Territory (strata 1-12) prior to 1977 (Hodges et al. 1996). Since the
change in visibility was not estimated directly by comparing simultaneous counts from new and
old aircraft in 1977, we tested models that included a level shift in 1977 and assumed that
changes in levels ofbreeding populations between 1963-76 and 1977-96 were due to improved
visibility of waterfowl in the more recent period. We modeled the log-transformed population
estimates, and then used the regression coefficients as multiplicative correction factors for the
pre-I 977 estimates.
Christmas Bird Count.- Data from 1955-95 were provided by the Patuxent Wildlife
Research Center (B. A. Hoover, Patuxent Wildlife Research Center, Personal Communication).
We analyzed data from the 6 species of interest and for all scoters combined, which included
unidentified scoters. Data on unidentified eiders were excluded. We calculated population year
effects (composite yearly indices of abundance) and estimates of trend using the methods in Link
17
and Sauer (1998,1999). In this procedure, a generalized linear model is used with effort
adjustments of form (~ P), where ~ is the effort at a site. The size of the exponent p determines
the form ofthe effort adjustment. Link and Sauer (1999) developed a method for estimating p in
which the model is fit with alternative values of p, and the value of p that produces the model
with smallest deviance is used in future modeling. Once p is chosen (i.e., the appropriate form
of the effort adjustment is specified), the significance ofthe overall effort adjustment is accessed
by determining whether the coefficient of the effort adjustment is different from O.
Once the need for effort adjustments was assessed for each species, the generalized linear
model with year effects and appropriate effort adjustments was fit for 10 regions (groups ofstates
and provinces in Atlantic Flyway). Estimated regional abundances were also calculated using
the appropriate effort adjustments, and were standardized to a consistent year using the estimated
year effects. Year effects were combined among regions using empirical Bayes procedures as
suggested in Link and Sauer (1998). In these procedures, differences in precision ofregional
year effects are accommodated by replacing them with a weighted average of the original time
series and a composite time series, and weights are determiued by the relative precision of the
original time series. These averaged time series are then weighted by relative abundances and
areas within regions to estimate year effects for the entire population.
Population trends were estimated as a linear regression through the year effects at the
regional level, accommodating for covariances among year effects. These trends were then
averaged among regions using the empirical Bayes procedures and weightings described above.
We determined trends for the Atlantic Coast, which we defined to include all states of the
Atlantic Flyway, plus New Brunswick, Nova Scotia, Ontario, Prince Edward Island, and Quebec.
We evaluated population changes during several periods that corresponded with changes in
hunting regulations (Table I). For harlequin ducks we evaluated 2 different periods: 1955-89 when
hunting was permitted in the Atlantic Flyway, and 1990-95 when no hunting was allowed. We
used one-sided t-tests to evaluate null hypotheses ofno differences in population trends between
periods. Alternative hypotheses were that trends would be lower (less positive) in periods with
more liberal regulations. For example, alternative hypotheses for comparisons of trends from 1955­63
and 1964-72 were that trends would be lower in the later period because of liberalizations in
regulations. Alternative hypotheses for comparisons of 1964-72 and 1973-93 were that trends
would be lower in the later period when regulations were the most liberal. For the comparison of
18
scoter trends between 1973-93 and 1994-95, our alternative hypotheses were that trends would be
higher in the later period when scoter bag limits were restricted.
Sea Duck Survey.-We used a simpler approach for analyzing data from the sea duck
survey because data from only 7 years were available. We tested for linear trends in annual
totals of each species.
RESULTS
Bandings and Band Recoveries
Numbers of sea duck band recoveries generally are inadequate to test hypothesis on the
presence of regional populations using multiresponse permutation procedures (Mielke et aL
1981, Slauson et al. 1991, Krementz et al. 1996) (Table 2). However, recovery locations of
common eiders, oldsquaws, harlequins, and white-winged scoters (Fig. 5) do not refute the
possibility of regional populations (Reed and Erskine 1986, Cassirer et al. 1991, Canadian
Wildlife Service et al. 1997) or reference areas where members of a population share similar
population parameters (Krementz et al. 1996). Most band recoveries ofcommon eider in the
Atlantic Flyway are from birds banded in areas along the Atlantic Coast. Oldsquaw banded near
Cape Churchill, Manitoba have only been recovered in the Great Lakes, Chesapeake Bay, or near
the banding location. Oldsquaws banded in Alaska have never been recovered in the Atlantic
Flyway. RecovelY locations of harlequins in western North America tend to be near areas where
they were banded. Many recoveries of white-winged scoter bands along the Pacific Coast were
from those banded in Alberta; however, some white-winged scoters banded in Alberta also were
recovered along the Atlantic Coast.
Harvest Estimates
Successful Sea Duck Hunters.-The number of successful sea duck hunters increased
during 1965-1972 as more states were allowed sea duck seasons, but appears to have declined
since 1974 (Fig. 6). The decline coincided with a more rapid decrease in active waterfowl
hunters in the Atlantic Flyway during the same period (Martin and Padding 1997).
u.s. and Canadian Harvests.-On average, the number of sea ducks harvested in the
Eastern Provinces of Canada was similar to the number harvested in the states of the Atlantic
19
Flyway and has totaled about 154,500 per year (Appendix 4). On an annual basis, the
proportion of the total harvest that occurred in each country has varied considerably. For
example, only 35% of the total sea duck harvest occurred in the U.S. in 1980, while in 1991,
65% occurred there. This pattern of annual variation, but overall similar harvest between the
Canadian and U.S. regions was evident for all the commonly harvested species. Only king eiders
and harlequins tended to be harvested in greater numbers in Canada, but the harvest ofboth of
these species was relatively low «500 birds per year).
Spatial and Temporal Patterns.-From 1987-96, the mean annual harvestoFo1dsquaw
during the Special Sea Duck Season was 13,500 birds, most of which were bagged in Maryland
(50%), New York (18%), and Maine (17%) (Fig. 7). Almost all (98%) of the harvest occurred in
special sea duck zones except in New York, where 28% of the oldsquaw were harvested from
other areas (mostly Lake Erie and Lake Ontario). Oldsquaws were harvested primarily from
November through the end of the hunting season (Fig. 7).
Almost all of the mean annual harvest of25,500 common eiders occurred in Maine (51 %)
and Massachusetts (46%) (Fig. 7), and virtually all (99%) of the eiders harvested were bagged in
special sea duck zones. Across the Flyway, harvest increased gradually as the season progressed ,.
(Fig. 7).
About 6,000 black scoters were harvested annually, with Maryland (37%), North
Carolina (18%), and New York (13%) accounting for most of the birds bagged (Fig. 8). Only
12% ofNew York's harvest was in special sea duck zones, compared to 97% for the rest of the
Flyway. More than half of the black scoters harvested were bagged in October (Fig. 8).
Hunters in New York (31%), Massachusetts (24%), Maine (14%), and Maryland (14%)
bagged most of the 10,500 white-winged scoters harvested annually (Fig. 8). Almost all (99%)
of the harvest occurred in special sea duck zones, except in New York (79%). The temporal
distribution of the Flyway-wide harvest showed a peak in mid-October followed by a gradual
decline through the rest of the season (Fig. 8).
The mean annual harvest of 12,000 surf scoters was more evenly distributed among states
(Fig. 9). Maryland hunters bagged 35% of the total, while Massachusetts (16%), Maine (14%),
New York (9%), and North Carolina (9%) each accounted for> 1,000 surf scoters bagged
annually. About 75% ofNew York's harvest occurred in special sea duck zones, compared to
20
98% for the rest of the Flyway. Most of the surfscoters harvested were taken during October
and November (Fig. 9).
Maine, Massachusetts and Maryland together account for most of the Flyway's total sea
duck harvest (Fig. 9). Across the Flyway, there is no clear peak in timing of harvest (Fig. 9).
Trends in Harvest.-Harvest trends are shown in Fig. 10-12 and annual rates of change
are reported in Table 3. Trends that were not significant are reported as 0 to be consistent with
the graphs and models used. Harvest of eiders increased on average 7.5% per year during 1961­96
despite rather stable hunting regulations (Table 4). An alternative model that treated data
from 1961 and 1993 as outliers and had a level shift in 1970 revealed that harvest increased 3.2%
per year during 1961-96. However, we have no a-priori reason to include additional parameters
that account for outlier points or a level shift in the model. Inspection of the lowess estimate
reveals that the rate of increase probably has diminished since about 1975.
For oldsquaws and scoters, the harvest trend during 1963-71 was consistent with our
hypothesis for that period, with annual rates of increase in harvest ranging from 5-17%. The
trend in oldsquaw harvest during 1972-96 was also consistent with our hypothesis, showing no
change. In contrast, the harvest of all scoter species declined during 1972-92, at rates ranging
from 6-8% annually. The harvest of black and surfscoters during 1993-96 did not undergo the
level shift decrease (i.e., drop) from the previous period that we hypothesized, whereas there was
a significant drop in white-winged scoter harvest (64%) following the scoter bag limit reduction
in 1993.
Availability Index.-Estimates of successful sea duck hunters were not available for
1961-64, therefore our analyses were limited to 1965-96. Availability of oldsquaw increased
during both 1965-71 and 1972-96, and the rate may have diminished (P = 0.10), as we had
hypothesized, during the later period (Fig. 10). The availability index for common eider
increased on average 4.8% per year during this entire period (Fig. 10). Trends in availability
indices for all three scoter species in 1972-92 were not different from those in 1965-71 (Table 5)
and thus did not support our hypothesized effects of liberalized hunting seasons. Availability
indices of black and surf scoters declined during 1972-92 (Table 5, Fig. 11, 12). Trends of scoter
availability indices during 1993-96 were similar to those of 1972-92 (Table 5).
Index ofRecruitment.-The proportion of immature birds in the harvest was highly
variable for all species, but we believe that this was due more to sampling error than actual year
21
to year differences in population age structure. Sample sizes (number of wings received
annually) were small for all species, averaging 75 for oldsquaw, 174 for common eider, 66 for
black scoter, 148 for white-winged scoter, and 129 for surf scoter. Despite the high variability,
common eider, black scoter, and surf scoter trends all suggested long-term declines in
recruitment (Fig. 10-12). However, the apparent declines for black scoter and surf scoter are
small compared to the variability among data points. In contrast, the lower variability in
common eider age data provided stronger evidence of an actual long-term decline. Oldsquaw
and white-winged scoter age structures showed little change over time (Fig. 10, 11). '~ -,'
Breeding Populations
Oldsquaw.-We estimate that the visibility rate of oldsquaws increased 68.71% when
. survey airplanes used in Alaska changed in 1977. The trend in population estimates of oldsquaw
in Alaska was stable during 1957-63, then increased about 5% per year during 1964-72, and
finally decreased about 6% per year after) 972 (Fig. 13). In Canada, the population was stable
during 1957-63 and 1964-72, and then decreased about 6% per year in 1973-97. To determine
trends for the entire Traditional Survey Area, we modeled the sum ofthe Alaskan series (after
adjusting for chang"e in visibility) and the Canadian series. For the Traditional Survey Area, our
results were similar to those for Canada: trends during 1957-63 and 1964-72 were similar (Table
5) and stable. The population declined during 1973-97 at about 5% per year (Table 4), which
was a faster decline than occurred in the previous period (Table 5). In development of our final
models, we detected an approximately 65% drop in 1990 estimates from Canada and Traditional
Survey Area. If these level shifts were accounted for, the rates of decline were only about 2%
per year during 1973-97. We rejected these models, since we had no a-priori reason to
hypothesize a level shift in 1990.
Nearly as many oldsquaws inhabit the Arctic Coastal Plain of Alaska as the entire
Traditional Survey Area (Table 6) (King and Brackney 1997). No trend was observed in
oldsquaw population estimates from the Arctic Coastal Plain during 1986-96 (King and
Brackney 1997). Only a few thousand oldsquaws were found in eastern Canada and Maine
during transect and plot surveys (Table 6). The combined estimates of oldsquaws from the
Traditional Survey Area, the Arctic Coastal Plain, and eastern Canada and Maine averaged about
289,000 during the 1990's.
22
Scoters.-We estimated that the change of survey airplanes in 1977 increased the
visibility of scoters 43%. Our best model of Alaskan data revealed that the number of scoters
decreased at a rate of 1.7% per year during 1957-97 (Fig. 13). Our best models for the Canadian
area and the Traditional Survey Area were similar, except that rate of decline was 1.6% per year
in these areas. The rate of decline did not differ among periods in the Traditional Survey Area
(Table 5). We evaluated 2 alternative models for the Traditional Survey Area that had different
slopes for the period 1994-97. In one model the slope was increasing, and in the other model the
slope was leveL AIC statistics were similar for these 2 models and the one we plotted with a
decreasing slope (Fig. 13). Thus, our inference on population trend during 1994-97 is equivocal.
As with oldsquaws, we identified a 43% drop in 1990 in the Traditional Survey Area. Although
this model had a better AIC statistic compared to the model we retained, we rejected it because
we had no a-priori hypothesis for a level shift in 1990.
Breeding populations of scoters in eastern areas and the Arctic Coastal Plain have been
smaller than those of the Traditional Survey Area during the 1990's (Table 6). On the Arctic
Coastal Plain, there was no apparent trend in scoter populations during 1986-96 (King and
Brackney 1997). Plot surveys in eastern Canada and Maine (Dickson 1995) indicated fewer
scoters than transect surveys in a similar area. Transects surveys during the 1990's revealed
around I million total scoters in the Traditional Survey Area, eastern areas, and the Arctic
Coastal Plain (Table 6). Species composition of the scoter population is unknown on these areas.
However, on plot surveys in eastern Canada and Maine, composition of the scoter population
averaged about 52% surf scoters, 24% white-winged scoters, 21% black scoters, and 3%
unidentified scoters (calculated from data provided in Dickson 1995).
Mid-winter Inventory
Numbers of oldsquaws on the MWI declined 1.1% per year during 1976-97 (Fig. 14).
The total number of oldsquaws observed on the MWI (about 10,000) was much lower than the
total observed during CBC (about 100,000) (Table 6). We detected no trend in total counts of
eiders or scoters on the MWI during 1976-97 (Fig. 14). Total eider counts averaged about
104,500, while scoter counts averaged about 38,000. We detected no change in the trend of
scoter counts following bag limit reductions in 1993 (Table 5).
23
Christmas Bird Counts
Effort acljustments.-Analysis of effort adjustments lead to a range of optimal p values,
from -2.5 (all scoters) to 2.0 (black scoters). However, only common eider (p= -1.0, slope =
0.14, seeslope)=0.055) and surf scoter (p=-0.5, slope = 0.874, seeslope)=0.071) were significant.
Accordingly, effort adjustments were incorporated in the analysis only for common eiders and
surf scoters.
Oldsquaw.-Trends ofoldsquaw based on CBC on the Atlantic Coast did not differ from
zero during any interval or the entire period. Trends ofCBC were not smaller during intervals with
more liberal hunting regulations (Table 5).
Harlequin Duck.-We estimated trends of harlequin ducks in 2 periods, 1955-88 when
hunting of harlequins was permitted in the Atlantic Flyway, and 1989-95 when no hunting was
. permitted. The trend was not different from zero in either period. The rate of population growth
was also not significantly different between periods (P = 0.11). Compared to other sea ducks,
relatively few harlequin ducks are observed on CBC (Table 6), and their trends should be
interpreted cautiously. For example, the extreme estimate oftrend (23% per year) is not
significantly different from zero (P = 0.95), reflecting its imprecision.
<
Common Eider.-CBC ofcommon eider along the Atlantic Coast during 1955-95 provided
no evidence of a population trend.
Black Scoter.-Trends in CBC ofblack scoters along the Atlantic Coast were not different
from zero during any of the intervals and during the entire period (P> 0.9). The trend in 1994-95,
when bag limits were more restrictive, was greater than during 1973-93 (Table 5).
White-winged Scoter.-Based on CBC, the trend for white-winged scoters from 1994-95,
when bag limits were restricted, was greater than in 1973-93, when regulations were most liberal (P
< 0.0 I) (Table 5). Over the entire period, the trend in white-winged scoters along the Atlantic
Coast was not different from zero (P> 0.9).
SurfScoter.-Trends in CBC of surf scoters on the Atlantic Coast did not differ from zero
in any interval, or during the entire 1955-95 period (P < 0.9). The trend during 1994-95, when
scoter bag limits were restricted, was greater than during 1973-93, when seasons were most
liberal (Table 5).
All Scoters.-No trends in total scoter CBC were detected during the entire period (P > 0.8)
or during any interval, except possibly during 1994-95 when the trend was positive (P = 0.06). The
24
trend during 1994-95 was greater than the trend during 1973-93, when regulations were most
liberal (Table 5).
Sea Duck Survey
No trend was detected in counts of oldsquaws, harlequins, common eiders, or total scoters
during 1991-99 (Fig. 15). Similarly, no trend was detected during 1994-99 for black scoter,
white-winged scoter, and surf scoter (Fig. 15). Common eider and harlequin ducks were detected
mostly in northern portions of the survey area. Harlequins were difficult to detect and only small
numbers (8-54) were counted each year. Species ofscoter was generally not determined during
the 1991 and 1992 surveys, but beginning in 1994, species was determined for 2:85% ofthe
scoters. The change in identification rate of scoters was due to greater emphasis on this objective
and a change in survey protocol.
DISCUSSION
Limitations of Data
Bandings and Band Recoveries.-The small number of sea duck bandings and recoveries
generally prohibits detailed analyses of harvest and survival rates, population affiliations,
migration corridors, and other aspects of their ecology. The greatest numbers have been marked
in areas where the birds are accessible and concentrated, such as at nesting colonies of eiders and
white-winged scoters in southern portions of their ranges. For these birds, survival and recovery
rates have been estimated (Krementz et al. 1996, Krementz et al. 1997). Although these
estimates are useful, they apply only to one age-sex class and represent only a small portion of
each species' range. It seems unlikely that sufficient numbers of most sea ducks could be banded
and recovered in consecutive years to allow estimation of survival rates with available analytical
methods. However, additional bandings and recoveries, even iffrom non-consecutive years,
could be useful in identifying or confirming the existence of separate populations units.
Harvest, Recruitment, and Availability Indices.-Our inability to document associations
between changes in regulations and changes in harvest may be due to our poor understanding of
the precision of annual harvest estimates. Ifharvest estimates are imprecise, it is unlikely that
even a strong association would be detected. This problem was magnified when we used
25
estimates of harvest per successful sea duck hunter as an index to availability, because variance
estimates for successful hunters were also unavailable. Furthermore, our use of successful sea
duck hunters as a surrogate for hunter effort was based on several potentially tenuous
assumptions. These assumptions include: (I) that the proportion of active sea duck hunters who
were successful did not vary among years; (2) that the number of days hunted per active sea duck
hunter was constant over time; and (3) that the proportion of successful sea duck hunters who
were hunting specifically for sea ducks was constant over time.
Our index to recruitment, the proportion of young in the harvest, did not account.·for age
and sex related differences in vulnerability to hunting. Differences in vulnerability can be
estimated from band recovery data (Martin et al. 1979), but banding data were too limited to
estimate relative vulnerability of age and sex classes ofsea ducks. Another limitation of our
indices of recruitment is that they are based on few samples of wings. Sampling error is greatest
for those species with the smallest number of samples (i.e., oldsquaw and black scoter), and for
these species we attribute much of the variation among years to sampling error.
Breeding Population.-The Breeding Waterfowl Survey appears to cover most of the
scoters breeding ranges, but has 2 limitations. First, the timing of the survey usually is too early
in important regiO'ns of Alaska and this may bias scoter estimates (U.S. Fish and Wildlife Service
1999). Second, species of scoters mostly have not been identified in the past, because of
difficulty identifying them reliably (U.S. Fish and Wildlife Service 1999) and probably because
scoters were considered less important than most other species encountered (e.g., puddle ducks)
when the survey protocol was established (1. Goldsberry, U. S. Fish and Wildlife Service,
personal communication). Identification of scoters to species is possible, and survey protocol in
Alaska since 1998 requires this for scoters within the closest one-half of the survey transect (U.S.
Fish and Wildlife Service 1999). Elsewhere in the surveyed area, protocol was modified in 1999
to allow identification ofscoter species when possible. Oldsquaws can be readily identified
during the Breeding Waterfowl Survey, but the survey covers only a small portion of their
breeding range. Waterfowl surveys of the Arctic Coastal Plain have nearly doubled the number
of oldsquaws counted compared to the Breeding Waterfowl Survey, but much of the oldsquaw
breeding range still remains unsurveyed. Although identification of species of eider is feasible
during low-level aerial surveys (King and Brackney 1997), this usually is not done during the
Breeding Waterfowl Survey. However, even if eider species were identified, the Breeding
26
Waterfowl Survey still would provide only poor estimates of common eider populations, since
these birds nest in coastal marine habitats that are sparsely sampled.
Mid-winter lnventory.-This survey is not based on a sampling design and provides no
annual measure ofprecision., It still may be a useful index to a population, if it counts a
consistent proportion of that population over time. The survey probably inventories only a small
proportion of sea duck populations, since it covers mostly inland and near-shore habitats (Forsell
1999). The survey has been criticized for inconsistency in survey methods, especially before
1976 (Eggeman and Johnson 1989). We believe that using data collected since 1976 minimized
this potential problem. We acknowledge, however, that variation in methods still occurred and
this probably affected the proportions of the total populations that were counted each year.
Another limitation ofthis survey is that species ofscoters and eiders were not distinguished. Our
assumption that essentially all of the eiders counted in the Atlantic Flyway were common eiders
is reasonable because other species ofeiders (e.g., king eiders) are rarely seen during Christmas
Bird Counts along the Atlantic Coast. Also, other species of eiders are rarely harvested in the
Atlantic Flyway, even though they are legal game. The common eiders observed in the Atlantic
Flyway are probably mostly American eiders, although a few northern eiders and intergrades of
American and northern races undoubtedly also appear (Reed and Erskine 1986, Heusmann
1995). We have no estimate of the species composition of seaters.
Christmas Bird COlll1tS.- This survey is primarily conducted from land areas, and
probably only samples a small proportion of the habitats used by sea ducks. Sea duck habitats
that are farthest offshore are very poorly represented in samples. Also, CBC tend to be
concentrated near urban areas, and these areas may be over-represented in samples. Observer
experience and effort varies over time and space, although effort appeared to only influence
counts for 2 species. Results from the CBC may provide general information on sea duck
population change, but for many species the CBC clearly does not provide precise estimates of
population change.
Sea Duck Survey.-This survey has only been conducted 7 times and thus is not yet
suitable for evaluation of long-term trends. Its coverage ofcoastal habitats is more complete
than the MWI or CBC, but it too poorly represents habitats that are farther than about 0.5 miles
offshore. The current extent of this survey's coverage, from Georgia to Nova Scotia, is
inadequate for monitoring populations that winter farther north (e.g., king eider, northern race of
27
the common eider) or in the Great Lakes (e.g., oldsquaws). Our analytical methods were
simplified considering that data from only a few years were available. With additional data, this
survey may prove to be a valuable measure of sea duck populations.
Population Status and Impacts of Hunting
The data available for managing populations ofsea ducks are limited compared to many
other groups ofwaterfowl. We have most confidence in trends that were consistent among
several data sets. Trends that were found in only one measure of a species' population.status
should be considered cautiously.
Oldsquaw.-Indices of oldsquaw populations during 1972-96 are inconsistent; two
indicated a decreasing population, one a stable population, and one an increasing population.
The range in trends probably reflects the variable quality of the monitoring data. All surveys had
limitations, but we believe that the CBC may be the most accurate index to oldsquaw populations
wintering on the Atlantic Coast. The CBC appeals to us because of the relatively large number
of oldsquaws that typically are encountered during this survey. We detected no trend in
oldsquaw numbers on CBC during 1973-95. The Breeding Waterfowl Survey showed a large
decrease during ICf73-97, but this survey does not include eastern breeding areas and may not be
reflective of oldsquaws that winter on the Atlantic Coast. We also are skeptical of results from
the MWI, since this survey counts so few oldsquaws compared to the CBC.
Oldsquaw harvest increased as regulations became more liberal in the Atlantic Flyway,
and stabilized when regulations remained stable. However, we found little evidence (l or 2 cases
in 5 tests) that changes in hunting regulations would lead to predictable changes in population
indices.
Harlequin Duck.-Little information is available on the status of harlequin ducks in
eastern North America. There are 1,000-2,000 harlequins in this area (Vickery 1988, Myers et
al. 1996) and data from CBC suggest no change in trend when the hunting season was closed in
the Atlantic Flyway in 1989.
Common Eider.-Population indices ofcommon eiders reveal increasing or stable
numbers in the Atlantic Flyway during 1955-97. This is consistent with a pattern of population
growth that has occurred since 1907 for the eiders that nests in Maine (Krohn et al. 1992).
Decreasing recruitment rates during 1961-97 may be a response to increasing densities of nesting
28
eiders. When recruitment equals mortality, the size of the population should stabilize, assuming
that immigration and emigration are negligible. Annual mortality rates for adult female
American eiders in eastern North America averaged about 13% during 1977-92 (Krementz et al.
1996); mortality rates for other cohorts are unknown. We can not estimate recruitment rate
without information on relative vulnerability of each age-sex class. Ifimmatures were about 2
times as vulnerable to harvest as adults, then in 1997 about 10% of the fall population would
have been immatures.
Harvest ofcommon eiders in the Atlantic Flyway has increased despite relatively stable
hunting frameworks in important harvest states and variable numbers of successful sea duck
hunters across the flyway. In Maine and some other areas, guided hunts for eiders has increased
during recent years (B. Allen, Maine Department ofInland Fisheries and Wildlife, personal
communication), and this may at least partly account for the increasing harvest.
Black Scoter.- Overall, the black scoter population appears to be declining, but changes
in regulations may affect their numerical trend. One index to their numbers during 1972-93 was
decreasing, while the other index was stable. No trend was observed during 1994-96, but this
inference is weak because it is based on relatively little data. 'Harvest in the Atlantic Flyway
increased with liberalizations in regulations, but did not change when scoter bag limits were
restricted in 1993. We found weak evidence from CBC that population trend changed in the
direction we hypothesized when regulations were modified. Proportion ofimmatures in the
harvest may have decreased during 1961-96.
White-winged Scoter.-The size of the white-winged scoter population appears to have
been stable during 1972-93. The proportion of young in the harvest increased from low levels in
the early-1980's. Hunting regulations appear to be associated with harvest and possibly
population trends of white-winged scoters in the Atlantic Flyway. Harvest increased as more
states selected special sea duck hunting seasons during 1963-71. Bag limit restrictions in 1993
coincided with a 64% decline in harvest of white-winged scoters and a significant increase in the
trend from CBC. However, when regulations were modified at other times, trends changed in the
opposite direction compared to our hypotheses.
SwfScoter.-Numbers of surf scoters in the Atlantic Flyway may have declined during
1972-93. The proportion of immatures in the harvest also appeared to decline. Harvest of surf
scoters increased when more states selected special sea duck hunting regulations during 1963-71,
29
but did not change when bag limit of scoters were restricted in 1993. There was only weak
evidence (I case in 5 tests) that regulatory changes coincided with changes in population trends
of surf scoters.
Total Scoters.-Collectively, numbers ofscoters in eastem North America were either
declining or stable during 1972-93. One survey, the Breeding Waterfowl Survey, indicated
declining numbers, while 3 others indicated stable numbers. We consider the Breeding
Waterfowl Survey to be the most reliable survey of continental scoter populations. However, it
may not accurately reflect trends for scoters that winter in eastem North America.. Harvest of
scoters increased during 1963-71 when increasing numbers of states used Special Sea Duck
Seasons. We found little evidence (I case in 9 tests) that population trend changed in a
predictable way when regulations were changed in the Atlantic Flyway.
Management Needs and Recommendations
Delineation ofPopulations.--Pattems of geographic distribution from breeding to
wintering areas is only poorly understood for most sea ducks in North America. For example,
the proportions of oldsquaw breeding in Alaska and wintering on the Pacific coast, Atlantic
»;<
coast, Great Lakes, and elsewhere are unknown. Many species of sea ducks return to the same
nesting area in successive years, however their propensity to return to the same wintering area is
only poorly understood. Without this infornlation, it is difficult to determine the scale at which
management should be directed. Historically, harvest has been managed separately for the
Atlantic and Pacific flyways in the U.S. We have implicitly adopted this strategy by analyzing
data solely from the Atlantic Coast, when possible. Additional analyses of existing data (i.e.
comparing population trends from different areas) may aid in determining if Atlantic and Pacific
coast sea ducks share the same population dynamics, and therefore if they should be managed as
one or separate populations. Additional recovery data from banded individuals would aid in this
assessment. The most efficient method for assessing the amount of interchange among birds
from different areas would likely utilize satellite or traditional radio telemetry techniques.
Monitoring ofPopulation Parameters and Harvest.-Each data set that we analyzed has
limitations. If population estimates were provided by species for scoters, the utility of data from
the MWI and the Breeding Waterfowl Survey would increase. We recommend investigations
into the feasibility of this potential improvement and implementation of consistent survey
30
protocol across the entire area surveyed in the Breeding Waterfowl Survey area. The annual
waterfowl survey of the Arctic Coastal Plain (King and Brackney 1997) appears complimentary
to the Breeding Waterfowl Survey. We recommend investigations into the feasibility of
integrating results from both surveys. The Sea Duck Survey warrants a more thorough review
and summary than we have completed. Specifically, consideration should be given to its
geographic coverage, whether sampling intensity should be modified throughout the survey area,
and additional analyses that may provide more efficient estimates ofpopulation change.
Regarding CBC, we encourage research to improve the efficiency ofanalyses and to reduce the
time lag between collection of data and posting it into electronic files. We believe that the
precision of harvest estimates has improved with the full implementation ofthe Harvest
Information Program in 1998. However, estimates of the proportion of young in the harvest will
still be based on relatively few samples. Managers should consider increasing the sampling
intensity of successful sea duck hunters.
Harvest.-Compared to other waterfowl, sea ducks are k-strategists (Patterson 1979).
They utilize relatively stable habitats, have high annual survivorship, and have low reproductive
potential. Because of these characteristics, sea ducks have limited capacity to compensate for
hunting mortality through increased recruitment or increased survival outside of the hunting
season (Patterson 1979, Nichols et al. 1984, Krementz et al. 1996). For purposes of harvest
management, we suggest that harvest mortality should be considered completely additive to
natural mortality.
Our analyses demonstrated that changes in hunting regulations coincided with changes in
harvest, but in only a few cases did they coincide with changes in population trends. We believe
that changes in regulations can effect population growth rates, but that in most cases we were
unable to detect those changes. These changes went undetected for at least 3 possible reasons:
(I) variation in important environmental parameters (e.g., habitat conditions, weather) hide the
minor effects of regulatory changes, (2) imprecision of surveys masked true population changes,
and (3) regulatory actions outside of the Atlantic Flyway (U.S.) diluted the effects of regulatory
changes within the Flyway.
The state of knowledge on sea ducks is limited compared to many other hunted
waterfowl. Limitations of the data we summarized led to equivocal interpretations. Perhaps the
grealest inadequacies were that tolal sizes of most sea duck populations have not been reliably
31
estimated. Although sea ducks have smaller harvests than many other waterfowl, conservative
hunting regulations seem prudent considering our overall state of knowledge of these birds. We
recommend continued closure of hunting seasons for harlequin ducks in eastern North America
unless it is demonstrated that these birds are part of a larger population. Regulations for scoters
should be very conservative because of evidence of downward population trends during 1972-93,
and the extreme paucity of information on black and surf scoters.
Managing the harvest ofsea ducks can be contentious because some individuals and
groups question the ethics of providing hunting seasons on birds that are perceived to.have a high
rate of non-use due to crippling and wanton waste (Federal Register 1994:42475). We
recommend research to determine the contemporary rates ofcrippling loss, wanton waste, and
hunter utilization of harvested sea ducks. We also recommend research to determine hunter
. preferences for bag limits and season lengths in sea duck seasons. Although this information
likely would not resolve debate over the ethics of recreational hunting, it could at least eliminate
speculation on the desires of hunters, and the true extent that shot ducks are utilized.
Hunting Regulations.-Two administrative issues regarding Special Sea Duck Seasons in
the Atlantic Flyway warrant consideration. Specifically, these issues include: (1) is there a need
to define to specific sea duck zones by Federal frameworks; and (2) do Special Sea Duck
Seasons provide additional opportunity outside the regular duck season, or essentially the only
opportunity for sea duck hunting? To resolve these issues, we recommend consideration of a
regular sea duck season that replaces the Special Sea Duck Season, and eliminating sea ducks
from the regular duck season. By doing this, sea duck season lengths and bag limits will be
clearer and more specific to sea ducks, rather than confounded within seasons structured for other
ducks. Regarding the need for special zones, we suggest eliminating this provision from the
Federal framework. This would allow uniform seasons for sea ducks across all areas of the
flyway, including inland areas on the Great Lakes. Also, those states that had specific needs for
sea duck zones could define and enforce those zone restrictions through state regulations.
Management Plan and Sea Duck Joint Venture.-We recommend that the Atlantic
Flyway Council in cooperation with others develop a management plan for sea ducks. The plan
should not be restricted to harvest management, but should address other issues such as
population monitoring, habitat management, diseases and contaminants, and information needs.
32
The plan should be developed in concert with the Sea Duck Joint Venture ofthe North American
Waterfowl Management Plan.
Many information gaps impede the development of optimal harvest rates and regulations
for sea ducks (Johnson et al. 1993). One deficiency is consensus on management goals for sea
ducks. Development of management goals should involve managers from Canada and the U.S.,
and consider subsistence, recreational harvest, and non-consumptive needs. We suggest a goal to
maintain populations at or above levels observed during the 1970's (U.S. Fish and Wildlife
Service et al. 1994). An alternative goal might be to maintain populations at or above levels
observed during recent years (1990-97).
Efforts ofthe proposed Sea Duck Joint Venture will not be focused solely on the Atlantic
Flyway. We recommend that waterfowl managers in the Atlantic Flyway Council fully
participate in joint venture activities, so as to insure that their regional needs are addressed while
also promoting coordinated management and strategic research. The proposed Sea Duck Joint
Venture can also serve to improve communications among researchers and managers, and aid in
the administration of research and management activities.
LITERATURE CITED
Abraham, K. F., and G. H. Finney. 1986. Eiders of the eastern Canadian arctic. Pages 55-73 in
A. Reed, editor, Eider ducks in Canada. Canadian Wildlife Service Report Series 47.
Ad Hoc Sea Duck Committee Atlantic Flyway Technical Section. 1994. Status of sea ducks in
the Atlantic Flyway with strategies towards improved management. Unpublished report,
Technical Section Atlantic Flyway Council.
Alison, R. M. 1975. Breeding biology and behavior of the oldsquaw (Clangula hyemalis 1.).
Ornithological Monographs 18.
Barry, G. 1., P. 1. Flint, R. F. Rockwell, M. R. Petersen, and T. 1. Moran. 1997. Population
dynamics ofspectactled eiders on Yukon-Kuskokwim Delta, Alaska (abstract only).
Page 23 in Proceedings First North America Duck Symposium and Workshop, Baton
Rouge, Louisiana, USA.
Bellrose, F. C. 1980. Ducks, geese and swans of North America. Third edition Stackpole
Books, Harrisburg, Pennsylvania, USA.
Bengtson, S. 1965. Field studies on the harlequin duck in Iceland. Wildfowl Trust Annual
Report 17:79-94.
33
Bent, A. C. 1925. Life histories of North American wild fowl (part II). U.S. Natural Museum
Bulletin 130, Washington, D.C., USA.
Bordage, D., and J. L. Savard. 1995. Black scoter. Birds of North America 177.
Box, G.E.P., and G. M. Jenkins. 1970. Time series analysis: forecasting and control. Holden­Day,
San Francisco, USA.
Brandt, H. 1943. Alaska bird trails: an expedition by dog sled to the delta ofthe Yukon River at
Hooper Bay. The Bird Research Foundation, Cleveland, Ohio, USA.
Brown, P. W., and C. S. Houston. 1982. Longevity and age ofmaturity ofwhite-wlifged
scoters. Journal Field Ornithology 53:53-54.
__, and L. H. Fredrickson. 1986. Food habits of breeding white-winged scoters. Canadian
Journal of Zoology 64: 1652-1654.
__'__. 1989. White-winged scoter, Melanittafusca, populations and nesting on Redberry
Lake, Saskatchewan. Canadian Field-Naturalist 103:240-247.
__, and __. 1997. White-winged scoter. Birds of North America 274.
__, and M. A. Brown. 1981. Nesting biology of the white-winged scoter. Journal of Wildlife
Management 45:38-45. ,"
Butcher, G. S. 1990. The Audubon Christmas Bird Count. Pages 5-13 in J. R. Sauer and S.
Droege, editors. Survey designs and statistical methods for the estimation of avian
population trends. U. S. Fish and Wildlife Service Biological Report 90(1).
Canadian Wildlife Service and U.S. Fish and Wildlife Service. 1987. Standard Operation
Procedures for aerial waterfowl breeding ground population and habitat surveys in North
America, unpublished report, U.S. Fish and Wildlife Service, Arlington, Virginia, USA.
__'__, and U.S. Geological Survey. 1997. Conservation issues for North American sea
ducks. 27 November draft of unpublished report.
Carney, S. M. 1992. Species, age and sex identification of ducks using wing plumage. U.S.
Fish and Wildlife Service, Washington, D. C., USA.
Carroll, R.I. and D. Ruppert. 1988. Transformation and weighting in regression. Chapman &
Hall, New York, New York, USA.
Cassirer, E. F., C. R. Groves, and R. L. Wallen. 1991. Distribution and population status of
harlequin ducks in Idaho. Wilson Bulletin 103:723-725.
34
Cleveland, W.S. 1979. Robust locally weighted regression and smoothing scatterplots. Journal
of American Statistical Association 79:368, 829-836.
Conroy, M. J., J. R. Goldsberry, J. E. Hines, and D. B. Stotts. 1988. Evaluation of aerial transect
surveys for wintering American black ducks. Journal of Wildlife Management 52:694­703.
Cooch, F. G. 1965. The breeding biology and management of the northern eider (Somateria
mollissima borealis) in the Cape Dorset area, Northwest Territories. Canadian Wildlife
Service Wildlife Management Bulletin (Series 2) 10.
Cooch, F. G., S. Wendt, G. E. J. Smith, and G. Butler. 1978. The Canadian Migratory Game
Bird Hunting Permit and associated surveys. Pages 8-39 in H. Boyd and G. H. Finney,
editors. Migratory game bird hunters and hunting in Canada. Canadian Wildlife Service
Report Series 43.
Cottam, C 1939. Food habits of North American diving ducks. U.S. Department of Agriculture
Technical Bulletin 643.
Dickson, K. 1995. Breeding waterfowl survey in Eastern Canada and the state of Maine.
Canadian Wildlife Service unpublished report.
Di Giulio, R. T., and P. F. Scanlon. 1984. Heavy metals in tissues ofwaterfowl from the
Chesapeake Bay, USA. Environmental Pollution Series A35:29-48.
Draper, N. and H. Smith. 1981. Applied regression analysis. Second edition. John Wiley &
Sons, New York, New York, USA.
Dunn, E. H., and J. R. Sauer. 1997. Monitoring Canadian bird populations with winter counts.
Pages 49-55 in Dunn, E. H., M. D. Cadman, and J. Bruce Falls, editors, Monitoring bird
populations: the Canadian experience. Canadian Wildlife Service, Occasional Paper 95.
Eggeman, D. R., and F. A. Johnson. 1989. Variation in effort and methodology for the
midwinter waterfowl inventory in the Atlantic Flyway. Wildlife Society Bulletin 17:227­233.
Federal Register. 1994. 50 CFR Part 20 Migratory bird hunting; final frameworks for early­season
migratory bird hunting regulations; final rule. 59(158):42475.
Forsell, D. J. 1999. Distribution and abundance of sea ducks wintering in Chesapeake Bay.
Page 81 in Goudie, R. I., M. R. Petersen, and G. J. Robertson, editors, Behaviour and
ecology of sea ducks. Canadian Wildlife Service, Occasional Paper 100.
Franson, J. C, M. R. Petersen, CU. Meteyer, and M. R. Smith. 1995. Lead poisoning of spectacled
eiders (Somateria flscheri) and of a common eider (Somateria mollissima) in Alaska. Journal of
Wildlife Disease 31 :268-271.
35
Geissler, P. H. 1990, Estimation of confidence intervals for federal waterfowl harvest surveys. Journal
of Wildlife Management 54:201-205.
, and J. R. Sauer. 1990. Topics in route regression analysis. Pgs 54-57 in 1. R. Sauer and S.
Droege, eds. Survey designs and statistical methods for the estimation of avian population
trends. U. S. Fish and Wildlife Service, Biological Report 90(1).
R. I. Goudie. 1989. Historical status ofharlequin ducks wintering in eastern North America-a
reappraisal. Wilson Bulletin 101:112-114.
Goudie, R. 1., A. V. Kondratyev, S. Brault, M. R. Petersen, B. Conant, K. Vermeer. 1994. The status of
sea ducks in the North Pacific Rim: toward their conservation and management. Transactions
North American Wildlife Natural Resource Conference 59:27-45.
Hochbaum, H. A. 1944. The canvasback on a prairie marsh. American Wildlife Institute, Washington,
D. C., USA.
Hodges, 1. I., 1. G. King, B. Conant, and H. A. Hanson. 1996. Aerial surveys of waterbirds in Alaska
1957-94: Population trends and observer variability. Information Technology Report 4.
National Biological Service, U.S. Department oflnterior, Washington, D. c., USA.
Hohman, W. L., C. D. Ankney, and D. H. Gordon. 1992. Ecology and management of
Postbreeding waterfowl. Pages 128-189 in B. D. J. Batt, A. D. Afton, M. G. Anderson,
C. D. Ank"ney, D. H. Johnson, 1. A. Kadlec, and G. L. Krapu, editors. Ecology and
management ofbreeding waterfowl. University of Minnesota Press, Minneapolis,
Minnesota, USA.
Heusmann, H W. 1995. The eider duck. Massachusetts Wildlife 45(1):31-37.
Hurvich, C. M., and C. Tsai. 1989. Regression and time series model selection in small
samples. Biometrika 76:297-307.
Jarvis, R. L., and H. Bruner. 1996. Characterization of habitat used by breeding harlequin ducks
in Oregon. Oregon State University unpublished Progress Report, Corvailles, Oregon,
USA.
Johnsgard, P. A. 1975. Waterfowl of North America. Indiana University Press, Bloomington, Indiana,
USA.
1978. Ducks, geese, and swans of the world. University of Nebraska Press, Lincoln, Nebraska,
USA.
36
Johnson, F. A., B. K. Williams, J. D. Nichols, 1. E. Hines, W. 1. Kendall, G. W. Smith, and D. F.
Caithamer. 1993. Developing an adaptive management strategy for harvesting waterfowl in
North America. Transactions North American Wildlife and Natural Resource Conference
58:565-583.
Johnson, S. R., and W. J. Richardson. 1982. Waterbird migration near the Yukon and Alaskan coast of
the Beaufort Sea: 2. Moult migration ofseaducks in summer. Arctic 35:291-301.
Johnstone, S. T. 1970. Waterfowl eggs. Avicultural Magazine 76:52-55.
Kehoe, F. P. 1989. The adaptive significance of creching behaviour in the white-winged scoter
(Melanittafusca deglandi). Canadian Journal of Zoology 67:406-411.
Kertell, K. 1991. Disappearance of the Steller's eider from the Yukon-Kuskokwim Delta, Alaska.
Arctic 44: 177-187.
King, R. 1., and A.W. Brackney. 1997. Aerial breeding pair surveys of the Arctic Coastal Plain of
Alaska - 1996. Unpublished report, U.S. Fish and Wildlife Service, Anchorage, Alaska, USA.
Korschgen, C. E. 1977. Breeding stress of female eiders in Maine. Journal of Wildlife Management
41:360-373.
Krementz, D. G., 1. E. Hines, and D. F. Caithamer. 1996. Survival and recovery rates of
American eiders in eastern North America. Journal of Wildlife Management 60:855-862.
__, P. W. Brown, F. P. Kehoe, and C. S. Houston. 1997. Population dynamics ofwhite­winged
scoters. Journal of Wildlife Management 61 :222-227.
Kr{)hn, W. B., P. O. Corr, and A. E. Hutchinson. 1992. Status of the American eider with
special reference to northern New England. U.S. Fish and Wildlife Report. 12.
Link, W. A., and 1. R. Sauer. 1994. Estimating equations estimates oftrends. Bird Populations
2:23-32.
__, and__. 1999. Controlling for varying effort in count surveys - an analysis of
Christmas Bird Count data. Joumal of Agricultural, Biological, and Environmental
Statistics 4:116-125.
Martin, E. M., and S. M. Carney. 1977. Population ecology of the mallard part IV. A review of
duck hunting regulations, activity, and success, with special reference to the mallard.
Resource Publication 130, U.S. Fish and Wildlife Service, Washington, D.C., USA.
__, and P. 1. Padding. 1997. Preliminary estimates of waterfowl harvest and hunter activity in
the United States during the 1996 hunting season. U.S. Fish and Wildlife Service, Laurel,
Maryland, USA.
37
Martin, F. W., R. S. Pospahala, and J. D. Nichols. 1979. Assessment and popnlation
management ofNorth American migratory birds. Pages 187-239 in J. Cairns, Jr., G. P.
Pati!, and W. E. Waters, editors Environmental biomonitoring, assessment, prediction,
and management - certain case studies and related quantitative issues. Statistical
Ecology, Volume 11. International Cooperative Publication House, Fairland, Maryland,
USA.
Mawhinney, K., and A. W. Diamond. 1997. Status ofthe common eider (Somateria mollissima)
in New Brunswick (abstract only). Page 26 in Proceedings of First North America Duck
Symposium and Workshop, Baton Rouge, Louisiana, USA.
Mendall, H. L. 1968. An inventory of Maine's breeding eider ducks. Transactions Northeastern
Section Wildlife Society Conference 25:95-104.
Mendenhall, V. M., and H. Milne. 1985. Factors affecting duckling survival of eiders Somateria
mollissima in northeast Scotland. Ibis 127:148-158.
Mielke, P. W., K. J. Berry, P. J. Brockwell, and J. S. Williams. 1981. A class of nonparametric
tests based on multiresponse permutation procedures. Biometrica 68:720-724.
Montalbano, F., III, F. A. Johnson, and M. J. Comoy. 1985. Status of wintering ring-necked
ducks in the southern Atlantic Flyway. Journal of Wildlife Management 49:543-546.
Mumo, J., and J. Bedard. 1977. Creche formation in the common eider. Auk 94:759-771.
<::;
Myers, J. E., G. H. Haas, and P. Kehoe. 1996. Population status of harlequin ducks in the north
Atlantic region. Unpublished report, Rhode Island Division of Fish and Wildlife, West
Kingston, Rhode Island, USA.
Nichols, J. D., M. J. Comoy, D. R. Anderson, and K. P. Burnham. 1984. Compensatory
mortality in waterfowl populations: a review of the evidence and implications for
research and management. Transactions North American Wildlife and Natural Resource
Conferance 54:535-554.
Ohlendorf, H. M., R. W. Olwe, P. R. Kelly, and T. E. Harvey. 1986. Selenium and heavy metals
in San Francisco Bay diving ducks. Journal of Wildlife Management 50:64-71.
__, and W. J. Fleming. 1988. Birds and environmental contaminants in San Francisco and
Chesapeake bays. Marine Pollution Bulletin 19:487-495.
Office of Migratory Bird Management. 1993. Status of sea ducks in eastern North Amcrica.
U.S. Fish and Wildlife Service unpublished report, Laurel, Maryland, USA.
38
Gring, L. W., and R. D. Sayler. 1992. The mating systems of waterfowl. Pages 190-213 in B.
D. J. Batt, A. D. Afton, M. G. Anderson, C. D. Ankney, D. H. Johnson, J. A. Kadlec, and
G. L. Krapu, editors. Ecology and management of breeding waterfowl. University of
Minnesota Press, Minneapolis, Minnesota, USA.
Patterson, J. H. 1979. Can ducks be managed by regulation? Transactions North American
Wildlife and Natural Resources Conferance 44:130-139.
Piatt, 1. F., C. J. Lensink, W. Butler, M. Kendziorek, and D. R. Nysewander. 1990. Immediate
impact of the 'Exxon Valdez' oil spill on marine birds. Auk 107:387-397.
Prach, R. W., A. R. Smith, and A. Dzubin. 1986. Nesting of the common eider near the Hell
Gate - Cardigan Strait polynya, 1980-81. Pages 127-135 in A. Reed, editor, Eider ducks
in Canada. Canadian Wildlife Service Report Series 47.
Reed, A., and A. 1. Erskine. 1986. Populations ofthe common eider in eastern North America:
their size and status. Pages 156-162 in A. Reed, editor, Eider ducks in Canada. Canadian
Wildlife Service Report Series 47.
__, Y. Aubry, and E. Reed. 1994. Surf scoter, Melanitta perspicillata, nesting in southern
Quebec. Canadian Field-Naturalist 108:364-365.
Salomonsen, F. 1968. The moult migration. Wildfowl 19:5-24.
Salter, R., M. A. Gollop, S. R. Johnson, W. R. Koski, and C. E. Tull. 1980. Distribution and
abundance of birds on the Arctic Coastal Plain of northern Yukon and adjacent Northwest
Territories. Canadian Field-Naturalist 94:219-238.
Saner, 1. R., and 1. B. Bortner. 1991. Population trends from the American woodcock
singing-ground survey, 1970-88. Journal of Wildlife Management 55:300-312.
__, and P. H. Geissler. 1990. Annual indices from route regression analyses. Pgs 58-62 in 1. R.
Sauer and S. Droege, editors. Survey designs and statistical methods for the estimation of
avian population trends. U. S. Fish and Wildlife Service, Biological Report 90(1).
__, S. Schwartz, and B. Hoover. 1996. The Christmas Bird Count Home Page
(http://www.mbr.nbs.gov/bbs/cbc.html). Version 95.1. Patuxent Wildlife Research
Center, Laurel, Maryland, USA.
Savard, J. L., and P. Lamothe. 1991. Distribution, abundance, and aspects of breeding ecology
of black scoters, Melanitta nigra, and surfscoters, M. perspicillata, in northern Quebec.
Canadian Field-Naturalist 105:489-496.
Sibley, C. G., and B. L. Monroe, Jr. 1990. Distribution and taxonomy of birds of the world.
Yale University Press, New Haven, Connecticut, USA.
39
Slauson, W. L., B. S. Cade, and J. D. Richards. 1991. User manual for BLOSSOM statistical
software. National Ecological Research Center, U.S. Fish and Wildlife Service, Fort
Collins, Colorado, USA.
Smith, G. W. 1995. A critical review of the aerial and ground surveys of breeding waterfowl in
North America. National Biological Service, Biological Science Report 5.
Spurr, E., and H. Milne. 1976. Adaptive significance of autumn pair formation in the common
eider Somateria mollissima (L.). Omis Scandinavica 7:85-89.
Stehn, R. A, C. P. Dau, B. Conant, and W. 1. Butler. 1993. Decline of spectacled eider~ nesting
in western Alaska. Arctic 46: 264-277.
Stott, R. S., and D. P. Olson. 1973. Food-habitat relationships ofsea ducks on the New
Hampshire coastline. Ecology 54:996-1007.
Stotts, V. D. 1966. Observation of waterfowl in deep-water areas from mid-October to mid­January,
1965-66. Unpublished report, Maryland Department of Game and Inland
Fisheries, Annapolis, Maryland, USA.
Time Series Staff of Census Bureau Statistical Research Division. 1995. REGARIMA reference
manual on the Census Bureau Home Page (http:/www.Census.Gov/pub/ts/regarima).
Washington, D. c., USA.
U.S. Fish and wilalife Service. 1988. Supplemental environmental impact statement: issuance
of annual regulations permitting the sport hunting of migratory birds. U.S. Department of
Interior, Washington, D. C., USA.
__. 1999. Population status and trends of sea ducks in Alaska. Unpublished report,
Migratory Bird Management, Waterfowl Management Branch, Anchorage, Alaska, USA
, Environment Canada, and Desarrollo Social Mexico. 1994. 1994 Update to the North
American Waterfowl Management Plan. U.S. Fish and Wildlife Service, Washington,
D.C.
Van Dijk, B. 1986. The breeding biology of eiders at lie aux Pommes, Quebec. Pages 119-126
ill A Reed, editor, Eider ducks in Canada. Canadian Wildlife Service Report Series 47.
Vickery, P. D. 1988. Distribution and population status of Harlequin ducks (Histriollicus
histriollicus) wintering in eastern North America. Wilson Bulletin 100: 119-126.
40
Table 1. Periods of sea duck hunting regulations in the Atlantic Flyway (U. S.) and hypotheses (alternative)
of period-effects on harvests and populations indices of oldsquaws and scoters.
Regulations
Hypothesized
effects on
oldsquaws
Period
1955-62' 1963-71' 1972-96' 1972-924 1993-96'
Increasing number
of states use Stable and liberal,
Stable and special sea duck bag limits for Stable, bag limits
conservative seasons Stable and liberal seoters:= 7 for seoters == 4
Halvest
Index of
availability
Breeding
Population
Mid-winter
Index
Christmas
Bird Count
Hypothesized
effects on
seoters
Harvest
Index of
availability
Breeding
Population
Mid-winter
Index
Not evaluated
Not evaluated
Stable
No data
Stable
Not evaluated
Not evaluated
Stable
No data
Increasing
Lower rate of
growth than in
previous period
Lower rate of
growth than in
previous period
No data
Lower ratc of
growth than in
previous period
Increasing
Lower ratc of
growth than in
previous period
Lower rate of
growth than in
previous period
No datJ
Stable
Lower rate of
growth than in
previous period
Lower ratc of
growth than in
previous period
Stable
Lowcr rate 0 f
growth than in
previous period
Stable
Lowcr rate of
growth than in
prcvious period
Lowcr rate of
growth than in
previous period
Stable
Stable, but at
lower level than
previous period
Grcater ratc of
growth than in
prcvious period
Greater rate of
growth than in
previous pcriod
Greater rate of
growth than in
previous period
Lower rate of Lower rate of Greater rate of
Christmas growth than in growth than in growth than in
Bird Count Stable previous period previous period previous pcriod
'Periods used were 1957-63 for breeding population estimates and 1955-63 for Christmas Bird Counts (CBC's).
'Periods used were 1965-71 for availability indices, and 1964-72 for breeding population estimates and CBC's.
'Periods used were 1973-97 for breeding population estimates and CBC's, and 1976-97 for mid-winter indices.
'Periods used were 1973-93 for breeding population estimates and CBC's, and 1976-93 for mid-winter indices.
'Periods used were 1994-97 for breeding population estimates and CBC's, and 1994-97 for mid-winter indices.
Table 2. Total numbers of bandings and band recoveries. of select species of sea ducks in North
America. Data were retrieved in September 1997 from records of the Bird Banding Laboratory, U.S.
Geological Survey, Laurel, Maryland.
Banded Recovered
Oldsquaw 2569 58
Harlequin 3765 231
Common eider 20425 2747
Black scoter 114 4
~-="
White-winged scoter 2950 136
Surf scoter 395 14
Table 3. Estimated changes in harvests of sea ducks in the Atlantic Flyway (U.S.) during different
regulatory periods, 1963-96.
Regulatory period
1963-71 1972-92(6)'
Species
Slope Slope
(% P(slope caeff::s; (% P(slope caefr 1:-
change/year) 0) change/year) 0)
Oldsquaw 17 <0.001 0 >0.05
Black seater 13
<0.001 -8 <0.001
White-winged
scoter 5 0.02 -6 0.001
Surf sceter 13 <0.001 -7 <0.001
All seoters II <0.001 -8 <0.001
1993-96'
Level shift(%)
from previous
period P (level shift ~ 0)
o >0.05
-64 0.002
o >0.05
o >0.05
'Periods used were 1972-92 for sCoters and 1972-96 for oldsquaws.
'Level shifts estimated for scoters only.
Table 4. Rates of changes (percent annual changes) in harvest and population estimates of sea ducks, during recent time periods. Breeding
popuiation estimates and mid-winter counts were not avaiiable for individual species of scoters or eiders. Time periods evaluated vary due to
limitations in data and variations in hypotheses.
Estimate and region
Oldsquaw
Rate Period
Common eider!
Rate Period
Black scoter
Rate Period
White-winged scoter
Rate Period
Surf scoter
Rate Period
All scoters
Rate Period
Harvest estimates for
Atlantic Flyway
A"ailability index for
Atlantic Flyway
Breeding population
estimates of traditional
survey area3
Mid*winter Inventories
of Atlantic Flyway
NS'
2.8
-5.3
-1.1
1972-96
1972-96
1973-97
1976-97
7.5 1961-96
4.8 1965-96
NS 1976c97
-4.1 1972-92
-1.9 1972-92
-3.2
NS
1972-92
1972-92
-3.7 1972-92
-1.7 1972-92
-4.2 1972-92
NS 1972-92
-1.6 1973-93
NS 1976-93
Christmas Bird Counts
from Atlantic Coast NS 1973-95 NS 1955-95
lMid-winter inventory includes all species of eiders.
'Not significantly different (P>0.05) from zero.
'The traditional survey area is strata 1-50 and 75-77.
NS 1973-93 NS 1973-93 NS 1973-93 NS 1973-93
Table 5. Comparisons, between regulation periods, of rates of change in population estimates of sea
ducks. Periods for breeding population estimates, mid-winter inventories, and Christmas Bird Counts lag
one year behind those of availability indices.
Species
Oldsquaw
Black scater
White-winged scater
Surf scater
All scaters
Regulatory periods compared
1955~621 vs 1963-71 1 1963-71' vs 1972-92(6)' 1972_923 vs 1993-964
EstimateS tf' P (d? 0) d p(d? 0) d P (d;5, 0)
Availability index NC' ·6 0.10 NC
Breeding population
estimate 4 0.78 -6 <0.01 NC
Christmas Bird
Count -10 0.39 -12 0.33 NC
Availability index NC -I 0.13 2 0.44
Christmas Bird
Count 13 0.36 -II 0.32 28 <0.01
Availability index NC 9 0.98 8 0.25
Christmas Bird
Count -19 0.23 -21 0.11 47 <0.01
Availability index NC 3 0.72 -2 0.55
Christmas Bird
Count -I 0.48 -II 0.24 69 <0.01
Availability index NC 6 0.96 2 0.42
Breeding population
estimate -3 0.13 3 0.94 -5 0.87
Mid-winter inventory NC NC 4 0.12
Christmas Bird
Count 12 0.25 -4 0.35 348 <0.01
'Periods used were 1957-63 for breeding population estimates and 1955-63 for Christmas Bird Counts.
'Periods used were 1965-7 t for availability indices, and 1964-72 for breeding population estimates and Christmas
Bird Counts.
'Periods used were 1972-93 for seater availability indices; 1972-96 for oldsquaw availability indices; 1973-97 for
oldsquaw breeding populations; 1976-97 for oldsquaw mid-winter inventories; 1973-93 for scoter breeding
populations and Christmas Bird Counts; and 1976-93 for scoter mid-winter inventories.
'Periods used were 1994-97 for scoter breeding populations and mid-winter inventories, and 1994-95 for scoter
Christmas Bird Counts.
'Availability index and mid-winter inventory were from the Atlantic Flyway, breeding population estimate was
from the Traditional Survey Area, and Chrislmas Bird COUll! was from the Atlantic coast.
"(slope coeflicient of second period)-(slope coefficient of first period); positive number means Ihat slope coefficieni
is larger in the second period.
7No comparison was possible.
Table 6. Means of selected sea duck population indices (thousands) in North America during the 1990·s.
Survey and region Years 01dsquaw Harlequin duck Common eider All eiders Black seoter White-winged seoter Surf seoter All scaters
Breeding Population Survey of ~
Traditional Area (strata \-50,75-77) 1990-97 1699 9.0 953.1
Breeding Population Survey of
Eastern Area (strata 51-68)1 1990-97 3.1 112.4 68.1
Breeding Population Survey of Arctic
Coastal Plain 2 1990-96 1164 1.9 19.5 12.9
Breeding Population Survey of all
areas! 1990;97 289.4 1.9 140.9 1034.1
Eastern Plot Survey of breeding
populations> 1990-95 31 0.1 6.2 4.8 5.5 12.2 23.2
Mid-winter Inventory of Atlantic
Flyway 1990-97 10.7 133.8 56.4
Sea Duck Survey of Atlantic Coast 1991-97 8.9 <0.1 33.1 8.34 3.1 4 10.34 25.7
Christmas Bird Count of Atlantic
Coast 1990-95 111.5 0.1 70.3 6.6 27.3 11.5 58.4
I All strata not surveyed in each year.
20ata from King and Brackney (1997).
>Data from Dickson (1995).
4Means from years 1994, 1995, and 1997~99 only.
Zone Restrictions
• Coast to first upstream bridge
~ L800 yards offshore
III L1 mile offshore
Fig. 1. Special sea duck hunting zones in the Atlantic Flyway during the 1997-98 hunting season.
"- / '~""'''/-''--
I
/ -\ ft~~~~s1\r~~~~ I ..J ••
I .,. , ..
~., Breeding Area '. ~ .
l;Bi" Winter Area '"~<--<---"-Ull
• Minor Breeding Area '
i
Fig. 2. Distribution of oldsquaw and harlequin ducks in North America (from Bellrose 1980). Reproduced with permission of Wildlife
Management Institute.
.~,
!~", Breeding Area "
.~ ,'t Winter Area
.:" Minor Breeding Area "
Fig. 3. Distribution of common eider and black scoter in North America (from Bellrose 1980). Reproduced with permission of Wildlife
Management Institute.
.......... i
".1.,
.j
/
,/
/
I'~ !.
Breeding Area
Winter Area
Fig. 4. Distribution of white~winged and surf seaters in North America (from Bellrose 1980). Reproduced with permission of Wildlife Management
Institute. ,I
Oldsquaw Harlequin
Common Eider White-winged Seater
Fig. 5. Banding and recovery locations (connected with a line) of all bands recovered from selected species of sea
ducks. No line is shown where the banding location was the same as the recovery location. Data were obtained in
September 1997.
18
16
14
(/)
"0 c 12 ro
(/)
:::l
.s0:: 10
I-
8
6
4
- 0
- 0 0
9. __
0
0 0
- o /'" ./ ~
00 ,
/ o '.Q
/ --- / o ----_Q.Q 0
- / "0-- 0
/ 0 o -- __
~o 9/ 0
-", __ 0
/ 0 0 "0-
-
/ 0
/ 0
- 0
I I J I I I I I I I I I I ! I I I I I I , , , , I I I I
1965 1970 1975 1980 1985 1990 1995
I0 Raw Estimate Lowess Estimate I
Fig. 6. Estimated numbers of successful sea duck hunters in the Atlantic Flyway, 1965"96. Vertical
lines mark periods with major differences in hunting regulations.
·A. Oldsquaw B. Common Eider
8 15
7
--6 -- rJ)
-0
rJ)
C
-0 10
co 5
c
rJ) co
::::l
rJ)
0
::::l
£;4
0 .c
'--' .....
+-' -- C/) .....
CD 3 rJ)
CD
C: co C: 5
I
co
2 I
1
0 Il!I!II!ilI .. 0
I-w<{wo<{I-'>-0C20<{ I-w<{wo<{I-'>--0C20<{
000222ZZZZ (J» 000:22:2ZZZZ (J»
State State
12 12
10 10
..... .....
rJ)
CD
rJ)
CD
C: 8
C
ro ro 8
I I
..c..o. ro
0
.....
6 0 6 l- I-
'<- '+-
0 0
+-' .....
c c
CD 4 CD 4 () ()
L.-
CD
L-CD
0.. 0..
2 2
0 0
Sept Oct Nov Dec Jan
5-day Period
Fig. 7. Average spatial and temporal distribution of the harvest of oldsquaws (A) and commoneiders (8)
in states of the Atlantic Flyway, 1987-1996.
1000
2500
2000 -(/)
-0
Cm1500
:::l o
..c
~
~ 1000
Cm
I
500
o
A. Black Seater
f-W«WO«I'Z>-°o.::°
State
B. White-winged Seater
3500 ,.......---------~
3000
--2 2500
c
m~
2000 o
..c
~
00 1500
0)
C
m
I
500
o
f-W «wo «I'Z >-00.::0 «
00<.9222Z zz w>
State
35 12
30 10
...... .....
(/) (/)
~ 25 0)
C
m co 8
I I
..m... 20 ..c.o..
0 0 6 l- I-
"'"- 15 4-
0 0 ...... ..... c c
- () a: () «
oO<.?:2:2:2z zz (f»
State
o
5.-----------~
4
......... ......... 15 (J) (J)
"'0 "'0
C c
ill 3 co
(J)
~ ~
0 0 .....c..:. ..c: 10 ........ ..............
...... ......
(J) 2 (J)
CD CD c: c: co co
I I 5
1
9
8
...... 7 (J)
CD
~ 6
I
..c..o.. 5
0
f-
4- 4 0 ..c....
CD 3 .(..).
CD
0.. 2
1
0
Sept Oct Nov Dec Jan
5-Day Period
Sept Oct Nov Dec Jan
5-Day Period
o
20 ,-----------------,
......
(J) ill 15
C
ro
I
..r..o..
0 10 f-
4-
0 ...... c
CD e
ill 5 0..
Fig. 9. Average spatial and temporal distribution of the harvest of surf scoter (A) and all sea ducks (8) in
states of the Atlantic Flyway, 1987-1996. .
A. Oldsquaw B. Common Eider
100000 100000
I °
..... ~ 1 0 0 00 ..... 10000 0
(1) k1r0
:- r!i/S~o (1)
(l) (l)
i:: 10000 i:: co co
I I lOy! Cbo~ 0
~ ! 1000 0
i0 I
G j
: 1
1000 1 1 100
1961 1968 1975 1982 1989 1996 1961 1968 1975 1982 1989 1996
10 10
x x
(l) ~ (l) "'0 ; "'0 0
C
I 0 c
>. >.
:t= .....
15 1
, :0
.§! 1 m
m 1 m 0'00 «> !0 «>
m ! «;. m LL LL
j
0.1 :
0.1
1961 1968 1975 1982 1989 1996 1961 1968 1975 1982 1989 1996
0'--------------'
(1)
~
::::i 0.8 ..... co
EE
0.6
~i
1 .:.1 0 0
1 0
1 01
~ol~ 0 C}J .
: - ..... 0
1 d>-- er .DQ.-c§20
1 0 1 0 00 0
; ; 0 0
1 f 0
1, 0
1 : : :
o L-..'..__...i.; --.l
(1)
:(sl) 0.8 ..... m
EE
0.6
Co
OA t
o
0.. o 02
'-
0...
1961 1968 1975 1982 1989 1996 1961 1968 1975 1982 1989 1996
o Raw estimate -- Regression estimate - - - - Lowess estimate
Fig. 10. Estimates of harvest, proportion of young in the harvest, and fall availability indices for oldsquaw
(A) and common eider (B) in the Atlantic Flyway, 1961-1996. Verticalljnes mark periods with major
differences in hunting regulations.
A. Blaek Seater B. White-winged Seater
9,' ;,1 ! 0
~, ,~
~ ~,.
~ 1000 1.---'-'__'-- ......:.'_-J
10000
100000 .--...........---,-----~--.
;
'-flA.~L...\-f!-~.c90 0 !,
1000
100 I..-"""--_---'- =--......I
10000
100000 .::---.----;------~---.
1961 1968 1975 1982 1989 1996 1961 1968 1975 1982 1989 1996
10 10
x ~ ~ x ~ 0 I
. , lo ; >,
~ ~ , 0 ; :t:::: 1 0 oC6 01
:.0 ~&..0- - "6-; ..0 1 f- ,
~~ , 0 0 Q m i m l 00 0 sa
f
m o ; 'ro 0 0 .%
I 1
«> ! > 0
;; Q) 0,0 «
: :
m l i I ro
LL l LL
0.1 : :
0.1
1961 1968 1975 1982 1989 1996 1961 1968 1975 1982 1989 1996
(J)
~
::J 0.8
+-' m
EE
0.8
Co
0.4 t
o
Q. e 0.2
0..
o
o
; 06
~c§)o j
''''' 0,0
i'''',~'c9 0 0
0
, 0 ..... _J)_0 0 P:,0
6 f 0 ---.:
! 1 0 ~ib
100000010 i 0 i 0d
i! o~
:: ~
O'-"";"--_:---' .....:.-C_--'
(J)
Q)
:; 0.8
+-' m
EE 0.8
Co
0.4 :eo
eQ. 0.2
a..
1961 1968 1975 1982 1989 1996 1961 1968 1975 1982 1989 1996
o Raw estimate -- Regression estimate - - - - Lowess estimate
Fig. 11, Estimates of harvest, proportion of young in the harvest, and fall availability indices for black
seater (A) and white-winged scoter (8) in the Atlantic Flyway, 1961-1996. Vertical lines mark periods with
major differences in hunting regulations.
A. Surf Seater
100000
10000 +-'
(f)
CD
2:
ill
I
1000
100
1961 1968 1975 1982 1989 1996
B. Total Seaters
100000
10 i;
10 I 0
01
00
+-'
(f)
CD 2: 01!
<0 ,
I 0 !
0:
0 P
d ! 0
i
10000
j i
1961 1968 1975 1982 1989 1996
10
x
CD
"'0 c
>.
.-t=
..D
ill
'co
>«
ill u..
0.1
(f)
CD
~ 0.8
::J
+-'
ill
E 0.6 E
C
0 0.4 t
0
0- e 0.2
£l.
0
I~.:!:, OQO,.\UfJ8P'L',',~O
i TU~
, i A.l 0 !, j ~
! < i i
1961 1968 1975 1982 1989 1996
1961 1968 1975 1982 1989 1996
10
x
CD
"'0 c
>.
.-t=
:0 1 <0
<0 >«
mu..
0.1
1961 1968 1975 1982 1989 1996
o Raw estimate -- Regression estimate - - - - Lowess estimate
Fig. 12. Estimates of harvest, proportion of young in the harvest, and fall availability indices for surf
seater (A) and total seaters (B) in the Atlantic Flyway, 1961-1996. Vertical lines mark periods with major
differences in hunting regulations.
Oldsquaws in Alaska
1000000..,---.,--:...-,--,-------,
Seaters in Alaska
1000000..,---,----,--,-------;--,
x x:
x
100000
o
o
x
o
10000 ...L-..,-_.i....,,--...L-....:,-__,--_-,--l 100000 ...L-,-_L-,--...L"":'---r---l.-r-l
1957 1967 1977 1987 1997
Seaters in Canada
1957 1967 1977 1987 1997
Seaters in Alaska and Canada
:
QCOO ) Orn c90-cYoOo
- ~ uC?o-o ® \~-
0 ;:a9J
100000 I I I I T
1000000
10000000
10000000
Q)
0
1000000 0
"I
O~
o 0
0
100000 100000
1957 1967 1977 1987 1997 1957 1967 1977 1987 1997
10000 ...L-...,-_.i....,,--...L-,-__,--_-,--l
0 0 0
(jiB
0
0 0000
0
100000 0°
1957 1967 1977 1987 1997
Oldsquaws in Alaska and Canada
1000000 ..,--!..-,.------.:...----~
x : 0 o
o
1957 1967 1977 1987 1997
Oldsquaws in Canada
1000000 ..,---.,-~-,--------,
Fig. 13. Breeding population estimates of oldsquaws and scoters, 1957-1997 (X =estimate adjusted for
change in aircraft type, 0 = unadjusted estimate, - = regression estimate,- - = lowess estimate).
Solid vertical lines mark periods with major differences in hunting regulations in the Atlantic Flyway;
dashed vertical lines mark change in aircraft type.
100000 r----:----------------------------,
Oldsquaw
o
o
=0. 0
o
o
D--- 0 0
o
o 0
o
o 0
o
o
o
0=O-Q
o
10000
1976 1981 1986 1991 1996
Common Eider
1000000 .....-----------------------------,
o 0
o 0
..9-fr - -G--- o
o 0 o
o
o
:Q}-
_--0- o 0
100000
1976 1981 1986 1991 1996
Scoters
1000000 .....-----------------------;--------,
o
100000 o 0
G"­lJ~-;::'-_
JJ
o
o
o 0
--~ 0 'J
-o-------~----=v
00 0
o
1976 1981 1986 1991 1996
oRaw Estimate - Regression Estimate •••• Lowess Estimate
Fig. 14. Mid-winter inventory estimates of sea ducks in the Atlantic Flyway, 1976-97. Vertical line
separates periods with differences in seater bag limits.
·Oldsquaw Common Eider
16 55
0
14 50
0
0
12
"' "' 45 "0 "0 0
c 10 c
ro "::>' ill 40
0 8
::>
0 0 0 .c; 0 .c; 0
l- I- 35
6
0 0 0
0
4 0 30
0
2 25
1991 1993 1995 1997 1999 1991 1993 1995 1997 1999
Bla.ck Scoter White-winged Scoter
16 6
0
14 5
0
0
12
"' "' 4 "0 0 "0
C 10 c 0
ro ro
"' 3 3 ::>
0 8 0
.c; 0
.c;
l- I- 2
6 0 0
4 0 0
2 0
1991 1993 1995 1997 1999 1991 1993 1995 1997 1999
Surf Scoter Total Scoters
25 60
0 0
20
50
"' "' 40 0
"0 15 "0
c c
ro ffi 30 "::>' ::>
0 10 0 0
.c 0 .c 0
l- I- 20 0 0
5
0
0 0 10 0
0 0
1991 1993 1995 1997 1999 1991 1993 1995 1997 1999
Fig. 15. Counts of sea ducks in the Sea Duck Survey of the Atlantic coast, 1991-1999. Most scoters
were not identified to species in the 1991 and 1992 survey.
Appendix 1. Summary of sea duck hunting regulations in Atlantic Flyway states with special sea duck seasons (Connecticut, Delaware, Georgia, Maine,
Maryland, Massachusetts, New Hampshire, New Jersey, New York, North Carolina, Rhode Island, South Carolina, and Virginia), 1938-97.
Sea duck season Regular season Total sea
Year(s) States(s) Opening date Closing date Days Bag Eligible species Zones Days Bag duck days
1938-39 ME,NH 9/15 9/30 16 10 Seaters Beyond outer harbor lines I 45 10 61
CT,MA,RI, 9115 10114 30 10 45 10 75
Others 45 10 45
1940-41 ME,NH 9/15 9/30 16 10 60 10 76
CT, MA, NY, RI 9/15 10115 31 10 60 10 91
Others 60 10 60
1942-43 ME,NH 9/15 9/24-25 10-11 10 70. 10 80-81
CT, MA, NY, Rl 9/15 10/14 30 10 70 10 100
Others 70 10 70
1944-45 ME,NH 9115 9/19 5 10 80 10 85
CT, MA, NY, RI 9115 9/30-10/12 16-28 10 80 10 96-108
Others 80 10 80
1946 ME,NH 9115 10/4 20 7 45 7 65
CT, MA, NY, RI 9/15 10/25 41 7 45 7 86
Others 45 7 45
1947 ME 10/6 12/16 72 7 24 4 72
NH 911 10/6 36 7 24 4 66
NY 9116 12/13 89 7 24 4 89
CT,MA 9116 11117 63 7 30 4 93
Rl 9116 1211 77 7 30 4 107
Others 24-30 4 24-30
1948 MA, NY, CT, Rl 9118 12117 91 7 Seaters, eiders 24-30 4 91
ME 10/6 12/16 72 7 24 4 72
NH 911 1017 37 7 24 4 61
Others 24-30 4 24-30
1949 New England2, NY 9/17 12117 92 7 32-40 4 92
Others 32-40 4 32-40
1950 New England', NY 9/17 12117 92 7 Seaters, eiders, oldsquaw 32-40 4 92
Others 32-40 4 32-40
1951 CT, ME, MA, NH, NY 9/28 12/31 95 7 36-45 4 36-45
Rl 9/28 1/5 100 7 45 4 45
Others 45 4 45
Zones
3.4, (;
3,4. (;
3,4,6
•
3,4,6
lIn coastal waters only, beyond outer harbor lines.
2CT, ME, MA, NH, Rl.
3All coastal waters and all waters of rivers and streams lying seaward from the first upstream bridge.
4Any waters of the Atlantic Ocean, and/or any tidal waters of any bay, that are separated by ;<; 1 mile of open water from any shore, island, or emergent vegetation.
SAny waters of the Atlantic Ocean, and/or any tidal waters of any bay, that are separated by ;<; 1200 yards of open water from any shore, island, or emergent vegetation.
6Any waters of the Atlantic Ocean, and/or any tidal waters of any bay, that are separated by ~ 800 yards of open water from any shore, island, or emergent vegetation.
7States were allowed to select a sea duck season of up to 107 consecutive days during the period 911-1/20, inclUSive.
8States were allowed to select a sea duck season of up to 107 consecutive days, during the periods 9/18-1/20 (1976-77), 9/16-1/20 (1978), or 9/15-1/20 (1979-97).
9No substantive changes have been made to special sea duck zones since 1975.
IOlncludes compensatory days for states in which Sunday hunting is prohibited.
Regular season Total sea
Days Bag duck days
30 3 39-46
30 3 100-104
30 3 107
30 3 53
30 3 105
30 3 107
40 3 59
40 3 107
40 J 107
50 5 55-60
50 4 106-107
50 3-5 107
60 6 60
60 4 107
60 4-6 107
Appendix 2. Transects and strata of the Breeding Waterfowl and Habitat Survey. Some strata were not surveyed in all years.
Appendix 3. Autoregressive Moving Average (ARMAl time series models.
We report on these parts of the models separately since the time series errors are
asymptotically independent from the regression. The series are short, so we could only fit simple
time series models to the regression residuals. Most series showed no time series error structure;
simple one parameter first order autoregressive or moving average structure could explain those
exhibiting error structure. The following table shows the ARMA structure where $1 is the first
order autoregressive parameter, and 81 is the first order moving average parameter. The variance
column shows the residual or innovation variance; i.e., the variance after the regression after
accounted for the time series error structure. The variance and AlC are not comparable between
series, e.g., between oldsquaw and common eider harvests. They are listed for reference
purposes.
In the table below, "AlC" is a statistic describing each chosen model. Lower AIC values
indicate better model performance. "AlC difference" (AIC of chosen model minus AlC of
alteruative model) represents the results of a comparison ofmodels with and without an ARMA
time series error structure. Negative differences indicate that the chosen model performed better
than the alternative model. Differences:2:2 are significant, so differences greater than +2 would
indicate choosing a significantly worse model. In some cases, a model with a higher AIC was
chosen, but the difference was never >2. In other cases the autocorre1ations did not indicate that
the residuals were anything but independent. In these cases, no test was done, and we relied on
the Autocorrelation Function (ACF) in place of the test.
Series ARMAModel Variance AIC AIC Difference
Hanrest
01dsquaw e,~0.52 0.12 678.8 -2.0
Common Eider e,~0.62 0.18 734.7 ACF
Black Scoter None 0.16 679.2 ACF
White-Winged Seoter None 0.12 723.9 ACF
SurfScoter None 0.13 719.3 ACF
All Scoters None 0.09 776.7 0.2
Availability Index
Oldsquaw None 0.12 30.3 0.3
Eider None 0.14 66.0 ACF
Black Scoter None 0.11 8.4 ACF
White~Winged Scoter None 0.10 47.8 ACF
SurfScotcr None 0.10 45.7 ACF
All Scoters None 0.04 82.8 ACF
Breeding Waterfowl Survey
Alaska Oldsquaw None 0.03 958.9 -1.8
US+Canada Oldsquaw None 0.28 1074.2 ACF
North America Oldsquaw None 0-07 1074.8 0.7
Alaska Seoters ~ 1~-0.35 0.02 1026.7 -2.2
US+Canada Seaters None 0-07 1151.6 ACF
North America Scoters ~,~0.45 0.03 1151.2 -6.8
Mid-winter Inventory
Oldsquaw None 0.13 442.9 ACF
Common Eider None 0.14 533.4 0.7
All Scoters None 0.36 516.7 ACF
Appendix 4. Estimated total harvest of sea ducks in eastern Canada1 and the Atlantic Flyway of the U.S., and the percent of the harvest occurring
in each area, 1974-1997.
01dsquaw Harlequin duck Conunon eider King eider
Year Harvest % Canada %U.S. Harvest % Canada %U.S. Harvest % Canada %U.S. Harvest % Canada %U.S.
1974 25,500 50 50 100 100 0 33,200 33 67 100 100 0
1975 44,000 48 52 0 36,700 60 40 0
1976 36,000 57 43 0 54,500 67 33 0
1977 17,000 51 49 100 0 100 52,400 72 28 200 100 0
1978 17,100 60 40 0 48,100 68 32 500 20 80
1979 37,000 53 47 200 100 0 40,300 56 44 400 100 0
1980 24,700 71 29 200 0 100 45,900 61 39 0
1981 33,200 44 56 0 47,200 53 47 400 100 0
1982 27,200 65 35 0 46,700 49 51 900 100 0
1983 26,500 75 25 0 81,800 61 39 0
1984 60,000 48 52 500 100 0 51,500 66 34 0
1985 25,400 45 55 0 45,100 44 56 300 100 0
1986 30,500 52 48 1,900 100 0 61,500 48 52 1,600 94 6
1987 24,000 47 53 1,300 100 0 48,600 52 48 1,300 . 100 0
1988 26,400 41 59 1,200 100 0 42,000 52 48 100 100 0
1989 16,500 56 44 300 100 0 38,200 62 38 200 100 0
1990 25,400 24 76 200 100 0 47,600 56 44 600 100 0
1991 17,500 29 71 200 100 0 63,900 36 64 500 80 20
1992 25,900 25 75 0 61,500 61 39 600 100 0
1993 19,600 31 69 100 100 0 33,300 73 27 1,000 100 0
1994 19,100 38 62 300 100 0 43,100 42 58 100 100 0
1995 16,700 36 64 400 100 0 54,300 40 60 700 100 0
1996 24,800 40 60 0 63,500 30 70 300 100 0
1997 21,700 25 75 0 55,200 36 64 200 0 100
Min 16,500 24 25 0 0 0 33,200 30 27 0 0 0
Max 60,000 75 76 1,900 100 100 81,800 73 70 1,600 100 100
Mean 26,738 46 54 292 86 14 49,838 53 47 417 89 11
Appendix 4. Continued.
Black scoter White-winged scoter Surf scoter Total sea ducks
Year Harvest % Canada %U.S. Harvest % Canada %U.S. Harvest % Canada %U.S. Harvest % Canada %D.S.
1974 33,700 38 62 42,200 37 ~ 63 38,800 43 57 173,600 40 60
1975 41,800 61 39 44,200 25 75 58,300 48 52 225,000 48 52
1976 41,300 78 22 33,000 45 55 58,800 72 28 223,600 65 35
1977 52,400 71 29 27,500 56 44 56,000 59 41 205,600 64 36
1978 18,000 56 44 23,000 47 53 31,100 52 48 137,800 58 42
1979 28,400 58 42 25,000 60 40 33,200 70 30 164,500 59 41
1980 20,900 73 27 37,200 57 43 31,600 69 31 160,500 65 35
1981 31,700 73 27 24,300 51 49 44,900 49 51 181,700 54 46
1982 20,500 80 20 25,900 47 53 35,200 83 17 156,400 63 37
1983 14,600 77 23 22,200 57 43 16,600 65 35 161,700 65 35
1984 23,900 56 44 46,600 40 60 36,500 50 50 219,000 52 48
1985 26,100 49 51 31,900 39 61 29,800 41 59 158,600 44 56
1986 15,600 56 44 14,700 31 69 28,500 32 68 154,300 46 54
1987 20,400 52 48 32,200 37 63 31,000 45 55 158,800 48 52
1988 11,300 51 49 30,000 42 58 17,800 65 35 128,800 50 50
1989 12,100 56 44 18,800 63 37 30,600 48 52 116,700 57 43
1990 19,100 37 63 22,000 43 57 32,600 55 45 147,500 46 54
1991 13,900 53 47 20,500 20 80 17,900 37 63 134,400 35 65
1992 8,800 50 50 16,200 45 55 17,400 35 65 130,400 48 52
1993 10,100 69 31 14,300 55 45 19,600 57 43 98,000 59 41
1994 11,900 53 47 13,200 72 28 34,500 54 46 122,200 49 51
1995 7,800 63 37 10,400 41 59 19,100 66 34 109,400 46 54
1996 8,300 43 57 12,700 38 62 16,900 33 67 126,500 34 66
1997 8,400 45 55 10,900 37 63 16,400 41 59 112,800 35 65
Min 7,800 37 20 10,400 20 28 16,400 32 17 98,000 34 35
Max 52,400 80 63 46,600 72 80 58,800 83 68 225,000 65 66
Mean 20,875 58 42 24,954 45 55 31,379 53 47 154,492 51 49
'Includes New Brunswick, Newfoundland, Nova Scotia, Prince Edward Island, and Quebec.