1. The feeding rates, diet composition and diel periodicity in feeding activity among larvae of the damselfly Enallagma cyathigerum Charp., exposed to chemical, visual and mechanical stimuli from the predators Ranatra linearis (L.), Notonecta glauca L. and Rutilus rutilus (L.) were investigated. 2. In response to chemical cues produced by all the predator species, larvae reduced their rate of feeding significantly (especially on large, sedentary prey). 3. Small larvae reacted more strongly than large ones to the kairomone produced by N. glauca. 4. The fish kairomone induced a reduction in feeding activity during daylight hours only. This reaction was more intense than the reaction to non‐chemical cues. 5. Observed predator‐induced changes in diet composition, caused directly by reduced activity during feeding, are discussed as an antipredator behaviour of damselfly larvae.

In a four year experiment (1988)(1989)(1990)(1991) carried out in a shallow, eutrophic lake, an increase in youngof-the-year pike (Esox lucius) density was used to control planktivorous fish assemblages consisting mainly of small-sized cyprinids: roach (Rutilus rutilus), Leucaspius delineatus, and white bream (Blicca bjoerkna). Stocking with pike fry of ca 30 mm in three successive springs resulted in large-scale mortality among prey of vulnerable sizes. Rotenone treatment at the end of the fourth year showed that roach and white bream populations contained few 0 + to 2 + fish, but very high numbers of 3 + to 6 + fish. By this time, short-lived, slow-growing L. delineatus, with no size refuge from predators, had been driven almost to extinction.

Only a few studies to-date have dealt with the habitat preferences of the fauna of springs. The main objective of our study was to fill this gap. The research was carried out in springs situated in the Kraków-Częstochowa Upland, Poland. The benthic fauna was collected from fine and coarse substrata using a typical hydrobiological approach. Forty-nine families or subfamilies and four higher taxa (Hydrozoa, Turbellaria, Nematoda, Hydracarina) were found in the springs studied. Invertebrates collected in springs were successfully classified into groups and, using the Analysis of Discriminant Function, were found to diverge in terms of substratum type. This prediction was 96% accurate for samples collected from fine sediments and 91% for samples collected from coarse sediments; the differences were higher in springs located in the southern part of the study area. Substratum type was found to be the main discriminatory factor with regard to the fauna density while faunal composition was related to the geographical position of springs. Our results emphasise that, even in very small habitats such as springs, all major substratum types should be considered for sampling.

Numerous procedures for biological assessment employ different measures of benthos diversity as a main or additional metric. This paper argues that biological diversity of freshwater macrobenthos is not a strong metric for biological assessment. It is extremely difficult to estimate accurately the taxonomic diversity of the whole macrobenthos. In particular, taxonomic diversity estimations are typically based on taxonomic misconceptions. Moreover, even if it is accurately estimated, it has too many drawbacks to be a reliable metric: it is affected by many factors important only for particular taxa and dependent on parameters that are extremely difficult to define and to quantify. Certain requirements are presented and discussed that are necessary to apply macrobenthos diversity in biological assessments.

The