Table 29. Number and area of wetlands with water in 
mid-July in the Minnedosa study area, 1962-72. 
19 
Table 30. Use of different habitat types by canvas- 
back broods in the Minnedosa study area, 1962-72. 



Roadside transects Beat-out areas 

Year® No. Km? No. Km? 
1962 261 10.0 
1963 275 3 
1964 735 14.3 197 25.5 
1965 1,212 23.6 301 38.6 
1966 1,130 22.0 214 27.4 
1967 785 15.1 141 18.1 
1968 760 14.7 102 13.1 
1969 1,540 29.7 308 29.7 
1970 1,430 27.8 286 27.8 
1971 1,450 28.2 290 28.2 
1972 1,130 22.0 296 91.6 
Mean — 22.0 — 25.1 

4aNo data for roadside transects in 1962 or 1963. 
Brood use per hectare of wetland was greatest on Type 5 
wetlands (Table 30). Of all canvasback broods observed, 
82% were on Type 5 wetlands; only 2% were seen on 
Type 3 and none on Type 1. The heavy use of Type 5 wet- 
lands by canvasback hens with broods indicates the need 
for this permanent type of wetland in canvasback breed- 
ing habitat. 
Wetlands larger than 0.4 ha were used most extensively 
by hens with broods (Table 30) — as they were by breeding 
pairs (Table 10), Wetlands 0.2 ha or smaller were used least 
for brood rearing but heavily for nesting (Fig. 11). Because 
brood use was greater on the larger wetlands it was also 
greater on the deeper wetlands (Table 30). 
Hens with broods appeared to select wetlands located in 
wooded pastures but the pattern was not consistent 
(Table 30). In some years brood use was greatest in open 
pasture. The rather inconsistent rate of brood use of various 
land use types between years suggests that factors other than 
land use may be important in determining wetland pref- 
erence for canvasback hens with broods. However, when 
the data were summarized by major land use types (Table 
30), use of wetlands in pasture was heaviest. Although 85 % 
of all broods were found on wetlands with less than one- 
third of their shoreline wooded, brood densities were higher 
on half-wooded wetlands (Table 30). 
Canvasback broods most often used wetlands with less 
than one-third of the water area covered with emergent 
aquatic vegetation (Fig. 12, Table 30). If Class Ia broods 
were omitted from the analysis, use of the more open wet- 
lands would be the greater. Brood use was greatest in wet- 
lands that contained a mixture of cattail and hardstem bul- 
rush (Table 30), 
Productivity 
Average annual production of canvasback ducklings from 
1961 to 1972 was 9.7/km? (Table 31), including adjusted 


Broods — Habitat types 
Habitat No./ha Mean size 
classification of water ha/year@ 
Wetland type 
1 0.00 2.0 
3 0.01 16.6 
4 0.10 23.1 
5 0.12 72.9 
Pond size (ha) 
0.1-0.5 0.03 13.4 
0.6-1.0 0.09 12.2 
1.1-2.0 0.13 21.1 
2.1-4.0 0.12 51.4 
>4 0.15 16.6 
Land use type 
Grassland pasture 0.21 11.3 
Wooded pasture 0.21 4,] 
Ungrazed woodland 0.12 3.2 
Ungrazed grassland 0.11 7.3 
Grain crops 0.13 53.9 
Summer fallow 0.11 30.4 
Hay crops 0.06 4.1] 
Pasture 0.17 15.4 
Idle land 0.10 11.0 
Cultivated 0.11 84,2 
Hayland 0.06 4.1 
Shore cover type 
Open 0.11 101.3 
Half-wooded 0.12 10.1 
Wooded 0.07 3.2 
Percent of emergent vegetation 
0-10 0.15 50.6 
11-33 0.15 37.3 
34-66 0.04 13.8 
>66 0.02 13.4 
Emergent vegetation species 
Bulrush 0.12 29.6 
Mixed 0.10 46.6 
Carex 0.03 2.4 
Cattail 0.14 10.9 
Open 0.16 7.3 
Whitetop 0.08 17.8 
Pond depth (cm) 
1-30 0.01 10.5 
30-60 0.10 27.1 
>60 0.14 76.1 


4For wetland type, “mean (ha/year)” refers to mean hectares of 
water; for land use and shore cover type, the value refers to mean 
hectares per year of the land use or above cover type specified. 
data on nest success and clutch size. Production was below 
average in 7 of the 12 years, Production was poor in 1961 
and 1962 because the breeding population was low and 
nesting success was poor, and in 1967, 1970, and 1972 be- 
