Page 6

Ducks were marked throughout the study using nasal saddles
and leg bands. Ducklings were marked with web tags. Here a
female Blue-winged Teal has been marked with a nasal saddle.
  Nest success was calculated using the method devel-
oped by Mayfield (1961, 1975) with the 40% modifica-
tion of Johnson (1979). Nest success comparisons were
made using LIFETEST Proc. of the SAS Statistical
Package (SAS Institute 1990). Factors affecting nesting
success were determined through the use of exponential
regression analysis using 34 variables.
  Nest densities were estimated by dividing the num-
ber of successful nests found by the Mayfield nest suc-
cess estimate. These density estimates should be
considered minimum estimates since it is assumed that
not all successful nests were found.
  All nests, regardless of their fates, were used to deter-
mine habitat preferences of nesting ducks (Greenwood
et. al. 1995). All nests found were pooled for all years
by species.
  On the day prior to projected hatch, Mallard and
Blue-winged Teal nests were visited to capture the
females with hand nets and mist nets (Bacon and Evrard
1990). Captured females were marked with aluminum
leg bands and individually color-coded nasal saddles.
During the early morning of the projected hatch day,
nests were revisited to capture newly hatched ducklings
before they left the nest. Captured ducklings were
marked with numbered monel tags placed in the foot
webbing between their toes (Alliston 1975). Duckling
sexes were not determined but a 52 male:48 female sex
ratio was assumed for Mallards (Sowls 1955), and a 58
male:42 female sex ratio was assumed for Blue-winged
Teal (Bennett 1938).
  In July 1982-90, flightless young and adult ducks
were captured by drive trapping (Cooch 1953) and night
lighting (Cummings and Hewitt 1964) and marked with
leg bands. Flightless ducklings, estimated to be at least
4 weeks old (Evrard 1996a), were leg banded and nasal
saddled. Younger ducklings, having feet too small to
hold a standard leg band, were web tagged only. Adult
ducks were marked with individually color coded nasal
saddles while flightless leg banded ducklings were
marked with nasal saddles color coded to the marshes
in which they were captured.
Duck Broods
All complete duck broods were aged using the classifi-
cation system of Gollop and Marshall (1954) based upon
growth and plumage characteristics. Estimated brood
mortality was based upon attrition in brood size over
time from hatch through Class I, II, and III. Differences
were tested using the Student's Mtest.
  Counts of marked (nasal saddled) and unmarked
Class II and older Mallard and Blue-winged Teal duck-
lings provided mark/resight estimates (Otis et al. 1978)
of duckling production on WPA wetlands. Since most
duckling mortality takes place during the first few weeks
following hatch (Dzubin and Gollop 1972, Ball et al. 1975,
Ringleman and Longcore 1982, Talent et al. 1983,
Duebbert and Frank 1984, Fleskes 1986, Orthmeyer and
Ball 1990, Fleskes and Klaas 1991, Higgins et al. 1992,
Mauser et al. 1994), estimates of numbers of Class II and
older ducklings can approximate production of fledged
ducklings (Hestbek et al. 1989, Mauser and Jarvis 1994).
  In the morning following marking, ratios of marked
and unmarked Class II and older ducklings were
obtained using Bennett's (1967) point sampling method.
These efforts met all four assumptions of mark/resight
population estimates. First, no more than 8 hrs elapsed
from marking to resighting, so there was no natality and
very little mortality. Emigration or immigration from
the wetland on which they were marked was minimal
since the ducklings were flightless and little time had
elapsed from marking to resighting. Second, very few
ducklings lost their nasal saddles. Based upon recapture
of banded birds that had lost their markers, reported
nasal saddle loss for birds age >1 yr ranged from 0.2%
(Evrard 1986a, Evrard 1996b) to 2.3% in Saskatchewan
(Arnold and Clark 1996) to 19% in North Dakota
(Lokemoen et al. 1990). Third, errors in correctly identi-
fying, counting, and recording the ducklings during
marking and resighting were minimal (Evrard 1996a).
Finally, based on observed duckling behavior, it was
assumed that marked and unmarked birds had equal
chances of being observed during the point counts.
6
Mallard eggs hatching.
2
uco
:E