Within-sample variabilities in stomach contents weight of fish — implications for field studies of consumption rate

Synopsis

Stomach contents weights in a study of Arctic charr (Salvelinus alpinus L.) were not normally distributed. This may have implications for the field estimation of consumption rates and also indicates that a more extensive handling of stomach content data is needed. Fish were sampled every 3 h throughout 24 h periods. The stomach contents showed very large variations within each 3 h sample. The distributions of stomach contents weights were skewed to the right, and the variance was dependent on the sample mean, with the variance to mean ratio equaling 1. Although mean values of stomach contents weight seemed to show diurnal variations, there were usually no significant differences between consecutive samples taken throughout each 24 h period, due to the large range in stomach contents weight within each sample. These findings question the accuracy of consumption estimates obtained using the method of Elliott & Persson (1978) and other related methods based on differences between consecutive sample mean values. Due to the large within-sample variations found in the current study, the method of Bajkov (1935) modified by Eggers (1979) is considered to give more reliable consumption estimates than the Elliott & Persson method. The generality of the findings is discussed, and it is pointed out that greater attention should be focused on within-sample variabilities in fish stomach contents studies.