Visibility bias in waterfowl brood surveys and population ecology of dabbling ducks in central Washington / John H Giudice
Material type:
TextPublication details: Moscow, ID : University of Idaho, 2001.Description: xiv, 134 p. : 28 cm. illOnline resources: Abstract: I investigated the feasibility of using sightability models to estimate and correct visibility bias in waterfowl brood surveys. Visibility bias is an important source of error in brood surveys, but most surveys do not account for imperfect detection or sightability covariates. I used 93 radio-marked mallard (Anas platyrhynchos) broods and a point-count method to develop predictive sightability models and assess the relative importance of sightability covariates. Mean brood visibility was 24%,but detection rates ranged from 13 to 51% depending on year and scale of application. Brood visibility was negatively correlated with percent visual obstruction and brood age. Observer experience, ambient temperature, and wind speed also influenced detection probabilty, but effects were less important than percent visual obstruction or brood age. Brood visibility was not strongly correlated with wetland-basin characteristics (e.g., size, shape, and percent surveyed). Brood and survey-area charact
| Item type | Current library | Collection | Call number | Status | Date due | Barcode |
|---|---|---|---|---|---|---|
Electronic Report
|
Electronic Library | Non-fiction | GIU (Browse shelf(Opens below)) | Available | 4698 |
Thesis(Ph.D.)--University of Idaho, 2001.
Includes bibliographical references (leaves 102-114).
I investigated the feasibility of using sightability models to estimate and correct visibility bias in waterfowl brood surveys. Visibility bias is an important source of error in brood surveys, but most surveys do not account for imperfect detection or sightability covariates. I used 93 radio-marked mallard (Anas platyrhynchos) broods and a point-count method to develop predictive sightability models and assess the relative importance of sightability covariates. Mean brood visibility was 24%,but detection rates ranged from 13 to 51% depending on year and scale of application. Brood visibility was negatively correlated with percent visual obstruction and brood age. Observer experience, ambient temperature, and wind speed also influenced detection probabilty, but effects were less important than percent visual obstruction or brood age. Brood visibility was not strongly correlated with wetland-basin characteristics (e.g., size, shape, and percent surveyed). Brood and survey-area charact