Accurately estimating catch - an illustration of the effects of bias in independent monitoring of fisheries