Journal of Fish Biology (2010) 77, 2454–2459 doi:10.1111/j.1095-8649.2010.02800.x, available online at wileyonlinelibrary.com BRIEF COMMUNICATIONS Scale circuli patterns differentiate between hatchery-reared and wild Rutilus rutilus during evaluation of fish stocking J. R. Britton Centre for Conservation Ecology and Environmental Science, School of Applied Sciences, Bournemouth University, Poole, Dorset BH12 5BB, U.K. (Received 11 February 2010, Accepted 5 September 2010) Hatchery-reared and wild roach Rutilus rutilus were correctly classified to their rearing environment in a discriminant function using data on their scale circuli patterns. The discriminant function was then used to classify recaptured R. rutilus by their origin (hatchery-reared and wild) during a riverine stocking evaluation. At 8 and 18 months post-stocking, 10 and 7% of fish captured in the vicinity of the stocking site were classified as hatchery-reared.  2010 The Author Journal of Fish Biology  2010 The Fisheries Society of the British Isles Key words: discriminant functions; hatchery; roach; stocking. Fish hatcheries breed, rear and stock-out substantial numbers of fishes annually in order to supplement wild fish populations, conserve threatened species, compen- sate for disturbance and support recreational fisheries (Brown & Day, 2002; Eby et al., 2006). Stocking evaluations are commonly completed through monitoring the contribution of marked fishes to the fishery or spawning stock (Barnett-Johnson et al., 2007). Marking methods, including tag implantation and chemical marking, are often associated with high application costs, tag loss and high mortality (Barlow & Gregg, 1991). An alternative method is the use of scale circuli patterns and otolith microstructure as these have successfully discriminated between wild and hatchery conspecifics in many fishes (Unwin & Lucas, 1993; Barnett-Johnson et al., 2007; Taylor & Piola, 2008). In salmonids, the rearing environment effectively produces a natural tag via growth checks on the body structures formed through high correlations between environmental conditions, fish growth rates and structural growth (Barnett-Johnson et al., 2007). This provides an effective and efficient method to separate hatchery-reared fishes from wild fishes using data on these ‘natural tags’ through discriminant functions (Ross & Pickard, 1990; Barnett-Johnson et al., 2007). The application of this technique to cyprinids has not been demonstrated, yet in many regions, hatchery-reared cyprinids are regularly stocked into rivers to supple- ment wild stocks (Aprahamian et al., 2004). In the U.K., roach Rutilus rutilus (L.) is Tel.: +44(0)1202 965384; email: Rbritton@bournemouth.ac.uk 2454  2010 The Author Journal of Fish Biology  2010 The Fisheries Society of the British Isles