Michigan Department of Conservation
Research and Development Report No. 22
Institute for Fisheries Research Report No. 1701, 1964
Angling and Trout Populations on the North Branch
of the Au Sable River, Crawford and Otsego Counties, Michigan,
under Special and Normal Regulations, 1958-1963
David S. Shetter and Gaylord Alexander
[...]
Creel census methods
Starting in 1958, the creel census was operated under a stratified random sampling plan which provided estimates of total fishing and total catch for three sections of stream under study. In 1958 and 1959 the census work was done by one clerk. The stream was divided into 18 sub-sections, each slightly longer than 1 mile. The clerk walked sample sub-sections, counting anglers that were fishing. Beginning in 1960, one of two census clerks employed counted anglers by canoeing the entire stream section. These randomly selected instantaneous counts of anglers were the basis for estimating total hours of fishing.
The trout season was stratified into four periods because angling pressure and the response of the fish to various lures differs during the season. Angling pressure varies between weekends (and holidays) and weekdays. It is generally much heavier on weekends. Thus, 50 percent of the samples were taken on weekends and holidays. Each day was divided into three 5-hour sampling periods when anglers were counted and contacted (8 a.m. -1 p.m., 1 p.m. – 6 p.m., 6 p.m. -11 p. m.). The period when a given section was to be sampled was determined from a table of random numbers.
The total angling hours was computed by multiplying the average instantaneous angler counts within a particular stratum by the total hours within that stratum. The sum of the estimated hours for various strata yield total hours of angling for the season. Clerks interviewed as many anglers as possible at the completion of their trips. Each angler was asked how long he had fished. His creeled trout were counted, measured, and sample-scaled. From these data the average catch per hour per angler was computed, and species composition and size frequency of trout were determined. Each angler was also asked how many sublegal trout he had caught and released. The catch was computed by multiplying total hours of fishing by average catch per hour.
Populations study methods
The fish populations of sample sections of stream 700 to 1,300 feet long were estimated by the Petersen mark-and-recapture method. Details of the method used are given by Shetter (1957). Direct-current electrofishing gear was used to capture the fish. Although many minnows and rough fish are present, only trout were included in the estimates. In 1957 and 1958 one sample section was investigated within each of the three stream sections being tested; during 1959-1962 two sample sections were utilized; beginning in the fall of 1962, the number of sample sections within each test section was increased to three. The last six of the nine sections were chosen at random.