350 Assortative Mating and Sexual Size Dimorphism in Black-legged Kittiwakes FABRICE HELFENSTEIN 1,2,3 , ETIENNE DANCHIN 1 AND RICHARD H. WAGNER 2 1 Laboratoire d’Ecologie, Université Pierre et Marie Curie Bat. A 7 ème étage, 7 Quai St. Bernard, case 237, F-75252 Paris cedex 05, France 2 Konrad Lorenz Institute, Austrian Academy of Sciences, Savoyenstrasse 1a, A-1160, Vienna, Austria 3 Corresponding author. Internet: fabrice.helfenstein@free.fr Abstract.—We examined several morphological characters of adult Black-legged Kittiwakes (Rissa tridactyla) breeding in Brittany, France, near the southern limit of the species’ range. Males were significantly larger than fe- males in body mass, head length, wing and tarsus, and the French population differed highly significantly in size from previously studied populations in England and Alaska. There was a strong pattern of assortative mating on tarsus length (r 16 = 0.87) which was also correlated with arrival date in both sexes. After removing the variance pro- duced by arrival date, assortative mating remained significant, suggesting that it may have been produced via sexual selection. Received 28 November2003 accepted 15 April 2004. Key words.—Black-legged Kittiwake, Rissa tridactyla, assortative mating, sexual dimorphism, Brittany. Waterbirds 27(3): 350-354, 2004 Kittiwakes are well-studied colonial sea- birds with sexually monomorphic plumage (Cramp 1983). Morphological measure- ments and sexual size dimorphism have been reported for several populations (Coulson et al. 1983; Cramp 1983; Jodice et al. 2000) but regional variation in body size has not been examined. Additionally, no study has exam- ined whether assortative mating by a mor- phological variable occurs, although Coulson and Thomas (Coulson and Thomas 1983) have reported assortative mating by age. As- sortative mating by size is widespread in birds and has been reported to occur in such fea- tures as wing (Olsen et al. 1998; Wagner and Morton 1997), bill (Coulter 1986; Forero et al. 2001; Wagner 1999), tail (Regosin and Pruett-Jones 2001) and tarsus (Delestrade 2001). Assortative mating may have impor- tant evolutionary consequences (Crow and Felsenstein 1968; Johnstone 1997) and may arise from active mate choice, intra-sexual competition, mating constraints or differen- tial mate availability (Crespi 1989). Assorta- tive mating may provide information on the processes underlying pair formation, a topic of considerable interest, especially for re- searchers of species with long-term pair bonds and high levels of biparental care. Our primary aim is to report a pattern of assortative mating on a morphological char- acter in kittiwakes and to examine the possi- ble reasons for its occurrence. Our second aim is to report morphological data on a population breeding near the southern limit of the species’ range and to compare body size and sexual size dimorphism with that from two other published studies. METHODS Kittiwakes have been intensively banded at Cap-Sizun in Brittany, France (48°50’N, 4°35’W) since 1979 (Cam et al. 1998; Danchin and Monnat 1992). While color- ringing the chicks from mid-June to the end of July, we captured and marked unringed adult birds and also re- captured some adults of known age. Adults were cap- tured using a metal hook fixed to a pole to draw in adults by hooking the birds under their lower mandible. The sex of captured individuals was determined by subse- quent observations of copulation and courtship feeding (Helfenstein et al. 2003b). Measurements were collected on 221 birds of known sex (97 males and 124 females) and dial calipers were used to measure head-plus-bill length (Coulson et al. 1983) and tarsus length (±0.1 mm) and a stop-ended ruler for flattened wing length (±1 mm). We also weighed each individual (±5 g) using a Pe- sola spring balance. A body condition index was estimat- ed using the residuals of the regression of the body mass against tarsus length and wing length. We captured some individuals several times in different years and morpho- metric data were collected at each capture. Repeatability of morphometric data and body condition between years and within individuals can be estimated from an ANOVA with individual identity as a factor (Falconer and Mackay 1996; Lessells and Boag 1987). For analyses other than repeatability, when the same individual had several records of the same morphological trait, we used the first record of this measurement.