The kite cover: a new concept for covered codend selectivity studies N. Madsen a,* , K.E. Hansen b , T. Moth-Poulsen a a DIFRES, Danish Institute for Fisheries Research, North Sea Centre, DK-9850 Hirtshals, Denmark b DIFTA, Danish Institute for Fisheries Technology and Aquaculture, The North Sea Centre, DK-9850 Hirtshals, Denmark Received 3 November 1999; received in revised form 13 April 2000; accepted 25 April 2000 Abstract The covered codend method is that most commonly used for estimating the size selectivity of codends for towed ®shing gears. A codend cover held open by kites has been developed to improve the performance and handling properties compared to the hooped covers normally used. The new cover can be wound directly onto the net drum as the catch is hauled. The development and testing of the cover in a ¯ume tank, where it was compared to a hooped cover, are described as well as the practical experience obtained when using the cover at sea. # 2001 Elsevier Science B.V. All rights reserved. Keywords: Codend; Cover; Selectivity methodology; Flume-tank test; Trawl 1. Introduction The covered codend method is most commonly used for estimating the size selectivity of codends for towed ®shing gears. The principal advantage of this method is that the codend selectivity can be estimated directly because the cover will collect ®sh escaping the codend. A serious problem that can be experienced with covers is masking, which occurs when the cover comes in physical contact with the codend, preventing ®sh from escaping through the codend meshes. The potential masking problems have long been recognised by, for example, Pope et al. (1975), who describe experiments where hoops were used to keep the cover clear of the codend. More recently, the use of such hoops when employing a codend cover has been strongly recommended (Wile- man et al., 1996). The hooped cover methodology is, however, often dif®cult to use at sea. The hoops have to be opened and detached from the cover before taking the catches aboard the vessel. It is particularly dif®cult to handle the hoops when the weather is bad, on vessels with a stern ramp, or when working with twin trawls where two covers are operated at the same time. Many commercial Danish ®shing vessels use twin trawl systems which can provide an ideal experimental design for comparing the selectivity of different codends because two codends can be tested at the same time side-by-side under very similar ®shing conditions. Because of the relatively long handling time when using hooped covers, there is the risk that more ®sh escape from the codend at the surface than during commercial conditions. Fisheries Research 49 (2001) 219±226 * Corresponding author. Tel.: 45-33-96-32-00; fax: 45-33-96-32-60. E-mail address: nm@dfu.min.dk (N. Madsen). 0165-7836/01/$ ± see front matter # 2001 Elsevier Science B.V. All rights reserved. PII:S0165-7836(00)00210-1