Microhabitat use and preferences of juvenile and adult bullhead Cottus gobio, from the River Voer, Flanders, were studied and compared across different seasons. Water depth, water velocity near the substratum, surface water velocity and substratum type used by C. gobio differed between seasons. These differences, however, were not attributable to differences in microhabitat availability. Adults appeared to prefer higher water velocities and coarser substrata than the average ones available in the basin. Although water depth appeared to have little influence on seasonal variation of microhabitat use in adult C. gobio, juveniles preferred deeper water and coarser substrata in winter, whereas in summer they appeared to use shallower water. There was a difference in microhabitat use between juvenile and adult bullhead only in summer. # 2005 The Fisheries Society of the British Isles

The existence of two morphotypes, broadheaded and narrowheaded, in European eels Anguilla anguilla is common knowledge among fishermen and eel biologists in Europe. To test whether European eels really are dimorphic in head shape, a total of 277 specimens from two locations in Belgium (Scheldt-Lippenbroek and Lake Weerde), in combination with a larger data set of 725 eels from river systems across Flanders (the northern part of Belgium) were examined. Our biometric data support the hypothesis that a head shape variation in 'Belgian' European eel is best described as having a bimodal distribution. Literature data suggest that this may be the result of phenotypic plasticity related to trophic segregation between morphs.

The spatial variation in the fish communities of four small Belgian rivers with variable habitat diversity was investigated by electric fishing to define the minimum sampling distance required for optimal fish stock assessment and determination of the Index of Biotic Integrity. This study shows that the standardised sampling distance of 100 m was not always sufficient to collect most species present. The required minimum sampling distance seems to be correlated with habitat diversity. In homogeneous streams, a mean sample distance of 282, 452 and 572 m is necessary to capture 80, 90 and 95% of all species present, respectively. In heterogeneous streams, these sample distances decrease to 217, 380 and 503 m. Hence, at least 300 m should be sampled to catch most species present with a single-pass sampling method. However, our results show that a 100 m sampling distance as presently used in the Flemish monitoring programs is sufficient to accurately describe the ecological quality since differences in IBI evaluation between adjacent stretches could at least for some rivers be explained by differences in habitat heterogeneity

No abstract