A Test of Morning Covey Call Count Technique


Period: August 10, 2001 – May 12, 2003
Contact: Ronald Drobney
Organization: the University of Missouri and the Missouri Department of Conservation
Funding Source: Prairie Fork Trust

Objectives: The morning covey call count technique is being used to estimate fall northern bobwhite (Colinus virginianus) density in Missouri and other parts of the species range, although it has not yet been validated. I tested several facets of this technique that are potential sources of error. One important aspect of using the technique is the ability to estimate the proportion of coveys that call on a given day (calling rate). I used logistic regression to develop and test models for predicting the calling rate based on year, date, area, and weather conditions (temperature, relative humidity, barometric pressure, barometric status, cloud coverage, and wind speed). Another important component of the morning covey call count technique involves the observer group’s ability to detect and accurately plot calling coveys on an aerial photo. I used a single pen-raised bird at varying locations within a 500 m  500 m grid cell to test the ability of 3 groups of observers to detect and plot a single covey. Because opportunities to collect data for testing observer groups using pen-raised quail are limited, I compared observer accuracy in plotting real birds with that of recorded calls to determine if recorded calls can be used as a substitute for live birds. Using data collected from 229 observations of 60 coveys of quail, I observed an overall calling rate of 78%. Covey calling rates were most influenced by area (Whetstone Creek Conservation Area or Reform Conservation Area) and wind speed. The proportion of coveys that called decreased with increased wind speed, and the calling rate was 12.8% higher on Whetstone Creek Conservation Area than on Reform Conservation Area. I recommend caution in application of my models to new areas, and that all sampling be conducted on mornings with low wind speeds. A one-way analysis of variance showed no significant difference in observer group error when plotting bobwhites calling from different locations within a grid cell. Observer accuracy was generally poor (=75.0, SE=10.9) regardless of where in the grid cell the calling bird was located. Observer accuracy improved with experience, therefore, I recommend observers practice this technique a minimum of 3 times prior to collecting data for use in a quail density estimate. I found that recorded calls can be used as a substitute for live quail when testing observers, but there are potential sources of error that the researcher should be aware of, such as time of day and the possibility of observer accuracy changing as the number of attempts to plot coveys increases in the course of the day. This study did not conclusively validate or invalidate the morning covey call count technique, but it did reveal concerns that should be addressed before important management decisions (e.g. determining hunting regulations) are made based on density estimates obtained through use of this technique.