Realm: Terrestrial
Climate: Temperate
Biome: Temperate grasslands, savannas and shrublands
Central latitude: 50.845447
Central longitude: -107.446257
Duration: 26 years, from 1952 to 1977

392 records

13 distinct species

Across the time series Anas americana is the most frequently occurring species

Methods

Census dataBecause of their historical economic importance.ducks afford the opportunity to examine very largedata sets collected over large geographic distances.long time periods. and through changing. but measurable.levels of breeding-habitat availability. In 1952 inmixed prairie near Redvers. Saskatchewan. and in 1953in aspen-parkland near Lousana. Alberta (Fig. 1). theUnited States Fish and Wildlife Service undertookyearly collections of waterfowl census and habitat dataon 0.2 km wide. permanently established transects. TheRedvers Waterfowl Study Area was a reversed-L transect.each arm roughly 32.2 km long; the total areawas 12.9 km2. The terrain was gently rolling and reliefgenerally <3 m. The pothole basins were generallysaucer shaped rather than bowl shaped (Stoudt 1971).The Lousana Waterfowl Study Area was a rectangulartransect roughly 16.1 x 6.4 km. or 9.3 km2. The terrainwas knob-and-kettle-like. A larger percentage ofponds at Lousana. compared with Redvers (45 vs.31%). were classified as permanent (Smith 1971).Each pothole on the transects was numbered andvisited several times yearly when breeding pairs of allspecies were counted. Dzubin (1969a) provided a reviewof the potential biases in transect-collected censusdata for ducks. Violations of these assumptionspose problems for deterministic calculations; the valuein these data is that. while they may be crude by thosestandards. the amount of data and the time span theycover outweigh disadvantages imposed by the mannerof data collection. Two assumptions. for purposes ofthis analysis. are important: first. the relative populationsizes of ducks were as the observers recordedthem. and second. proportional use of different pondtypes by different species was represented by the censusdata. The second assumption is made more plausibleby validation of the first; the first is reasonablebecause considerable effort was made to census allponds thoroughly (Smith 1971:14. Stoudt 1971:14). Onebiologist collected most of the data. although in recentyears. other United States Fish and Wildlife Servicepersonnel and I collected data.Each pothole was scored for each of the followingcharacteristics: permanence classification. size.depth. degree of openness of surrounding vegetation.percent coverage of pond surface by emergentvegetation. and type of vegetation. All values of habitatvariables were of ordinal level; type of vegetationwas not a nominal-level variable because it tookon values that corresponded to structural complexity(height. density) of each plant type or association.Twenty-six years of data (1952-1977) from Redverswere available for analysis. Four censuses were conductedyearly from 1952 to 1965; commencement datesof censuses varied between 1-11 May. 1-4 June. 29June-13 July. and 22-26 July. Eight and six censuseswere conducted in the drought years 1961 and 1962.respectively. After 1965. one July census was begunbetween 11 and 23 July; dates of May and June censusesvaried as in previous years. Nine years (1968-1976) of data from the Lousana area were availablefor analysis. Commencement dates for censuses. whichwere conducted three times yearly. varied between Iand 7 May. between 31 May and 8 June. and between14 and 17 July. Unit of abundance = Count, Unit of biomass = NA

Citation(s)

Vickery, W. L. & Nudds, T. D. (1984) Detection of Density-Dependent Effects in Annual Duck Censuses. Ecology, 65, 96.