


a ee 
16 
Table 23. Number and success of canvasback nests. by 
pond classes, in the Minnedosa study area, 1963-72. 










Nests 
Mean 
Habitat success 
classification Total No./ha (%) 
Wetland type 
] l T 0) 
3 74 0.12 40 
4 653 0.64 AQ 
5 456 0.12 4] 
Pond size (ha) 
O.1-0.5 487 0.62 43 
).6-1.0 281 (),42 37 
1,1-2.0 188 0.18 43 
2.1-5.0 206 0.10 44 
5.1-10 22 0.02 32 
> 10 0) 0.00 0) 
Land use type 
Cultivated land 727 0.17 40 
Ungrazed grassland 155 0.25 40 
Grassland pasture 143 0.20 43 
Wooded pasture 60 0.52 45 
Ungrazed woodland 68 2.32 53 
Hayland 31 0.64 32 
Nest cover type 
Cattail 542 0.02 43 
Bulrush 332 0.03 42 
Whitetop 180 0.01 34 
Willows 29 0.04 38 
Mixed 75 0.01 44 
Carex 21 0.01 33 
Reeds 5 0.02 50 
Shoreline cover type 
Open 858 0.15 38 
14-23 wooded 205 0.54 38 
“3 + wooded 12] 1.04 44 



the 12 years, nest success was similar in wetlands of dif- 
ferent sizes, but nest densities were greatest in the smaller 
ones, 
Water depths at nests ranged from 0 to 91 em. Average 
water depth at successful nests was 38 em. but was only 
1.8 cm deeper than that at unsuccessful nests. Successful 
nests were an average of 0.9 m farther from shore than 
unsuccessful nests and ranged from 0 to 12.5 m from shore, 
About 61% of the canvasback nests were on wetlands 
in cultivated land; success of these nests was only slightly 
below average. Nests in ponds surrounded by grazed and 
ungrazed grassland were also similar in density and suc- 
cess. Nest densities and nest success were highest in wet- 
lands located in ungrazed woodland. A total of 72% of the 
nests were in wetlands with open shorelines, but nest den- 
sity was greater on wooded ponds, Nests in wooded ponds 
were also more successful than those on more open ponds. 
More nests were found in cattails than in any other cover 
type, but the number of nests per unit cover was higher 
in both willow and hardstem bulrush. During the 12 years, 
nest success was highest in reeds but this cover type con- 
tained less than 1% of the total nests found. Nest success 
in cattail and hardstem bulrush was about the same. 
Brood Production 
An average of 63% of the canvasback broods counted 
during beat-outs occurred on roadside brood counts (Table 
24). High water and dense vegetation reduced observability 
of broods from the road in some years. Some wetlands were 
not observable from the road in late summer because of 
the height of natural vegetation or crops. Beat-out counts 
indicated high numbers of broods in 1963, 1964, and 1966. 
and low numbers in 1967, 1968. and 1972 (Table 24), 
Table 24, Canvashack broods observed by transects and 
beat-out counts in the Minnedosa study area, 1962-72. 





Transects Beat-out areas® 
Year No. No./km2 No. No./km2 
1962» 18 0.6 13 1.7 
1963 32 0.6 15 1:9 
1964 9] 1.8 17 2.2 
1965 43 0.9 12 0.8 
1966 71 1.4 2] 2.7 
1967¢ 24 0.5 5 0.5 
1968¢ 6 0.1 2 0.2 
1969¢ 40 0.8 i) 0.9 
1970¢ 24 0.5 10 1.0 
1971¢ 98 0.5 15 1.5 
1972¢ 24 0.5 2 0.2 
Mean 0.7 1.3 


*“Beat-out counts” see waterfowl census techniques, Stoudt 1971. 
bOnly 28.49 km? surveyed in 1962, 
“10.36 km?*; all previous years 7.77 km2, 
Brood Size 
Mean brood sizes were based on observations of 1.832 
canvasback broods, including those with redheads (Table 
25). These data included only broods in which all ducklings 
were believed to have been seen. Broods were recorded by 
three major classes (I, II, III) to increase sample size. Brood 
size declined from an average of 5.9 Class | ducklings to 
9.3 Class IT ducklings as the breeding season progressed. 
Brood sizes were largest in 1965, 1966. and 1972 and small- 
est in 1963, 1967, and 1968. The number of observations 
in Class IIT broods was small because the fieldwork ended 
on 1 August, and broods had begun to break up. 
