Whooping Crane Migrational Habitat Use Draft (2)

Home Browse Search

Aerial Survey Detection Rates Experimental trials were conducted to compare detection rates for the many factors relating to the implementation of the protocol. Detection trials were conducted using a stratified random survey design. Single whooping crane decoys (sandhill crane decoys painted to resemble whooping crane plumage) were randomly placed on accessible conservation lands in the study area in either the channel stratum (river) or upland (return) stratum and their location was either recorded or not recorded during the aerial survey (PRESP 2005). The resulting strata estimates of detection probability are applicable to the types of aerial transects surveyed. Detection trials were conducted in the study area during each of the 11 survey seasons. All trials were conducted using single decoys randomly placed on accessible conservation lands. A total of 176 decoy trials were deployed during the Cooperative Agreement period for detection by the systematic flights: 111 decoy trials were conducted in the river in the wetted channel and 65 decoy trials were conducted within the entire study area in the return strata (Table 1). Flight orientation covariate information was available for 168 of the decoys. A predictive model of detection probability was developed and used to calculate actual sample inctusion probabilities for monitoring flights. Models were parameterized assuming the number of decoys detected during the aerial flights followed a binomial distribution with parameters n (the number of decoy trials) and p(the detection probability). A list of candidate models was generated through listing all possible combinations of the predictor variables: contractor, year, season, flight direction, flight height, flight orientation, and survey strata. Models were fit using logistic regression and the Akaike information criterion (AIC) was calculated to select the most likely model given the data. Movement Pattern Summary The average distance a crane group moved was estimated by averaging each successive known movement, where movement was defined as the straight line distance between two consecutive locations. Crane groups were considered to have moved if the crane group flew to a new location or the crane group walked to a different land cover type. Distances were averaged across all movements observed for a crane group. The average distance moved was estimated across the 13 crane groups with multiple observations, using the crane group designation made by the USFWS. Movements were mapped on color infra-red photographs taken during fall 2003, with the length of time a crane group spent in an area indicated by the size of the mapped syrnbol and arrows to indicate the direction of movement. The average distance moved was used to define the area within which a crane group will make a "local area" selection of habitat within the study area. While there are other ways to define the local area, WEST chose this definition because it could be calculated with the available data and it adequately represents the average distance a crane group flies when intending to continue use of the study area. It is assumed the distances recorded by the ground crews were a random sample of distances moved by crane groups using the study area.