Skip to main content
SLU publication database (SLUpub)

Report2011

Distribution Estimation for Fishing Time

Yu, Jun; Leonardsson, Kjell

Abstract

Information on visitation frequencies in recreational fishing is important when dealing with fishing tourism in order to make prognoses for the future upon changes in the management of the aquatic environment (flow regime or habitat restoration), fishing regulations, or to estimate the total harvest of fish. Therefore interviews have been performed in a number of streams and sections of streams, throughout the last twenty years in the Jämtland-Härjedalen region in Sweden. In this work the probability distribution of total fishing hours a day (possibly on different periods) is considered. We found that both Gamma and Weibull distributions can be considered as approximate distributions that generate the data. Gamma distribution fits very well for summer season while Weibull distribution is more appropriate for the other periods. In general, the gamma model is easier to interpret and better fits the mode of the distribution, and therefore, is preferred. Having parameters estimated, we are able to calculate probabilities of different fishing times. It is also suggested to use two periods: mid-summer – the end of August and other dates. The modelling at section level seems to be successful. Both the Gamma and Weibull distributions fits well the data for all periods, providing that the number of observations are not less than 20. The mean fishing hours, however, varies from section to section, even within the same watercourse

Keywords

Fishing time; density estimation; goodness-of-fit test; randomisation

Published in

Research report (Centre of Biostochastics)
2011, number: 2011:02
Publisher: Centre of Biostochastics, Swedish University of Agricultural Sciences