Efficacy of artificial shelters for roach, Rutilus rutilus, against predators in the presence of reeds

Cormorants, Phalacocorax spp., have great potential to affect recreational fisheries and have increased substantially in abundance over the last decades in UK inland waters. Fabricated refuges provide a potential means of reducing fish losses, yet the benefits of such structures may be marginal if natural shelters are abundant and favoured, or if strong density-dependence limits refuge use. This study examined the efficacy of artificial shelters in mesocosm enclosures that allowed standardised and replicated observations of roach, Rutilus rutilus (L.), distribution. When given a choice between occupying open water, simulated reedbeds and artificial brushwood shelters, roach used brushwood shelters extensively across a range of fish densities. When fish had a choice of occupying open water or reedbeds offering no overhead cover, they actively avoided reedbeds and used open water almost exclusively. Occupation of reedbeds was positively related to the amount of overhead cover they provided. When artificial brushwood shelters and reedbeds offered complete overhead cover, brushwood shelters were occupied twice as much as reedbeds. Artificial shelters may therefore have wide application in stillwaters with abundant reedbeds unless the reedbeds are coupled with extensive overhead cover.