Abstract
Freeze coring is a commonly used method for the investigation of the bed sediment fauna of rivers. It is considered to produce quantitative numbers of invertebrate abundance in different depth layers. Calculations of abundance use total volume of the freeze core sample as spatial reference. This definition of sample volume is incorrect. In the present, study freeze core samples are shown to consist of two parts: (1) a cylindrical inner core in which the pore water has turned into ice and (2) all parts of sediment, which protrude from this inner core. Invertebrates are fixed only within the inner core since the protruding parts contain no pores and therefore no invertebrate habitat. In samples from gravel rivers, the protruding parts form a considerable bias, which depends on the size of the core and the coarseness of the sediment. In the present study, total volume of individual core segments varied between 97 and 248% of actual sample volume. The inner core can be measured directly to avoid the bias. The procedure is proposed for future studies to produce unbiased, comparable values of invertebrate abundance and consequently reliable data on vertical distribution.