Planktivorous predation on copepods. Evaluation of mandible remains in predator guts as a quantitative estimate of predation

Size and shape of mandibles from 8 common species of copepods in Norwegian coastal waters are described, with the aim of using gut contents of copepod mandibles as a precise technique of quantifying predation rates on copepods. The carapace length of 7 of the 8 copepod species could be closely predicted from a general h e a r regression equation of carapace length versus mandible width. The exception, Temora longicornis, had lower carapace length than expected from the equation. Similarly, the 6 copepodite stages of Calanus finmarchicus produced a highly significant species-specific linear regression equation between carapace length and mandible width, although adult males diverged somewhat by showing a lower carapace length than predicted. A correspondence analysis showed that it was impossible to separate all species and developmental stages from each other based on linear measurements of mandible structures alone, and the skill of a well-trained person is therefore needed. The relationship between individual dry weight and mandible width was well described by a regression equation, which hence could be used to calculate the biomass of consumed copepods, based on the contents of copepod mandibles in the predator's stomach. A comprehensive evaluation of the precision and accuracy in the estimates of prey mass is given, including effects of interand intraspecific variability and seasonal variation. By using a power function to regress body mass (mg dry weight) from mandible width (pm), the true prey biomass in the stomach content is estimated with a precision of *20% or better, when based on more than a few mandibles. However, a considerable seasonal variation in the length/weight relationships of the prey copepods may cause considerable bias in the estimates if not compensated for