ORIGINAL PAPER Fish length-weight relationships in the Weddell Sea and Bransfield Strait Received: 10 January 2002 / Accepted: 2 April 2003 / Published online: 14 May 2003 Ó Springer-Verlag 2003 Abstract Length-weight relationships are provided for 49 Antarctic fishes from 9 families. The fish were from three research cruises of 1996, 1998 and 2000 to the Weddell Sea and Antarctic Peninsula. Samples were collected by benthopelagic net, Agassiz trawl and bot- tom trawl at depths of 200–2,300 m. The exponent b in the length-weight relationship W=aL b ranged from 2.53 to 4.05. With a mean of 3.33, 50% of the values ranged between 3.157 and 3.480. In 31 species, the parameters were determined for each sex. For 64.5% of species, b was higher in females than in males. Exceptions were the Channichthyidae where b for six species was higher for males. Introduction Knowledge on length to weight relationships (LWR) in Antarctic fish is scanty, even though it is useful in studies of fish growth and for comparing populations from different localities (Pauly 1993). In fish, the relationship between the weight (W) and the length (L) can generally be expressed by the equation: W=aL b . When b=3, the length to weight relationship is isometric and when b is not 3, the relationship is allometric. This work provides estimates of the LWR parameters for 49 Antarctic fish species from the Bransfield Strait and Weddell Sea. The available LWRs are mainly for fish from the Antarctic Peninsula area (Kock et al. 1985; Ekau 1988, 1991). The present paper provides a comprehensive revision of LWRs for the fish of the high Antarctic and of the Antarctic Peninsula. Materials and methods Samples were collected during the R/V Polarstern EASIZ I, II and III cruises. Sampling was undertaken in three areas of the Weddell Sea during each cruise, and in the Antarctic Peninsula during the third cruise. The cruise dates, study areas, the three types of gear used and the depths sampled are shown in Table 1. Gear and fishing characteristics are summarised in Table 2. Fish total length (L) was measured to the nearest millimetre, weight (W) to the nearest gramme. In macrourids, preanal length (PL) was determined. For abundant species, the data were analysed separately for each sex. The data of fish whose sex could not be determined were added to the male and female data to calculate the LWR. Least squares regressions of W (g) on L (cm) were derived after log transformation of the two variables (log e W=log e LÆb+log e Æa) to correct for non-linearity and heter- ogeneous variances. A t-test was used to compare regression lines and to test whether b values were significantly different to 3 at the 0.05% significance level. Results The number and weight of fish caught per cruise are summarised in Table 1. Of the 21,214 fish from 64 spe- cies and 13 families, there were sufficient (N‡10) with a wide range in length to determine the LWR for 49 spe- cies from 9 families (7,719 fish). Table 3 shows the results grouped by families and species. Some 84% of species were Notothenioidei. The sample sizes varied from 10 fish for Notolepsis coatsi (Paralepididae) to 930 fish, for Pleuragramma antarcticum (Nototheniidae). The r 2 values ranged from 0.855 for Gerlachea australis to 0.996 for Cygnodraco mawsoni and Lepidonotothen kempi. All regressions were significant (P<0.05). The mean b parameter of the LWR for the nototheioids was significantly greater than 3 [Artedidraconidae b=3.26 (SE=0.30), Bathydraconidae b=3.37 (SE=0.56), Channichthyidae b=3.47 (SE=0.30), Nototheniidae b=3.3 (SE=0.13)]. Polar Biol (2003) 26: 463–467 DOI 10.1007/s00300-003-0505-0 Bernat Artigues Æ Beatriz Morales-Nin Eduardo Balguerı´as B. Artigues Æ B. Morales-Nin (&) CSIC/UIB-Institut Mediterrani d’ Estudis Avanc¸ats, Miguel Marque´s 21, 07190 Esporles, Spain E-mail: ieabmn@uib.es Tel.: +34-971-611721 Fax: +34-971-611761 E. Balguerı´as IEO-Centro Oceanogra´fico de Tenerife, ap. 1373, 38080 Santa Cruz de Tenerife, Spain