year, the proportion of recoveries increased in the Central Flyway 
from 28 to 45 percent and decreased in the Mississippi Flyway from 
70 to 53 percent (table 37). In 1958, recoveries in the Central 
Flyway declined to 23 percent and in the Mississippi Flyway increased 
to 74 percent. The marked increase in the 1957 recoveries in the 
Central Flyway was associated with a high proportion of band recov¬ 
eries between 44° and 49° of latitude (table 37). January inventories 
of wintering populations revealed more than double the average number 
of waterfowl in the northern part of the Central Flyway, thus suggesting, 
as did the distribution of recoveries, that northward shifting of the 
population accounted for the increased Central Flyway kill. The ratio 
of Central Flyway to Mississippi Flyway recoveries was also higher 
than average for 1957 bandings in southeastern Alberta and southwestern 
Saskatchewan, although in these areas the differences from adjacent 
years are not significant, nor is the relation to latitudinal 
distribution as pronounced. 
In 1955 the ratio of Central Flyway to Mississippi Flyway recov¬ 
eries from bandings in southeastern Alberta and southern Saskatchewan 
tended to be lower than for other years (tables 35 to 37). In 1955, 
November was extremely cold throughout Manitoba, Saskatchewan, and 
Alberta. Aerial surveys by the Canadian Wildlife Service in central 
and western Saskatchewan indicated that most water areas were frozen 
and for practical purposes no ducks remained, whereas in 1954 more 
than 130,000 ducks were counted in early November and 19,000 were 
observed as late as November 23. It seems probable that in 1955 cold 
weather caused mallards to leave the northern portions of the Central 
Flyway, resulting in the increased proportion of recoveries in the 
Mississippi Flyway. In 1957, extremely mild weather throughout the 
hunting season caused mallards to remain much farther north than 
normal, allowing them to be shot in the northern States of the 
Central Flyway. 
We may conclude from the various data that have been discussed 
that (1) the direction of migration from a given reference area is 
probably stable, (2) the timing and ultimate destination of migratory 
movements may vary as a result of weather conditions during the period 
of migration, and (3) where migration routes involve crossing one 
flyway to reach another the timing of migration may cause changes in 
the distribution of kill between the two flyways. Thus, mallards 
moving southwestward from Alberta or western Saskatchewan, a relatively 
short distance across the Central Flyway, may be expected to be con¬ 
sistent in the portion of the population that reaches the Pacific 
Flyway. However, for mallards moving long distances southeastward 
across the Central Flyway, changes in the speed and timing of 
migration may result in varying proportions of birds that reach the 
Mississippi Flyway or that are available to hunters in the Central 
Flyway during their migration. 
18 
