J. moll. Stud. (1977) 43, 192-199 PHENOTYPIC VARIATION AND FITNESS OF PERIWINKLES (GASTROPODA:LITTORINIDAE) IN RELATION TO EXPOSURE J. HYLLEBERG & J . TANG CHRISTENSEN Department ofEcology and Genetics, Ny Munkegade, University of A rhus, 8000 A rhus C, Denmark (Received 4 September 1976) INTRODUCTION It has long been recognized, that relationships exist between habitat and parameters like weight, shell thickness, length of shells and radulas. Small limpets, Patella granularis L., are restricted to the lower shore and their size increases gradually higher on the shore to a maximum in the balanoid zone (Branch, 1975). In Patella vulgata L., Brian & Owen (1952), Ebling et ale (1962) and Choquet (1966) observed the opposite trend. In dog-whelks, Nucella lapillus, variation in shell shape and thickness could be related to exposure (Crothers, 1974 a,b). Many examples of variation in bivalves are also published. Galucci & Hylleberg (1976) found variation in bentnose clams, Macoma nasuta, in relation to exposure and Kristensen (1957) called it a curious phenomenon that cockles, Cardium edule, from two localities displayed different characteristics. In other words, it is well documented that for individual species many types of variation can be correlated with environmental factors in specific areas. Unfortunately it is often cryptic, how genetical factors affect visible variation (Berry & Crothers, 1974). The pqrpose of the present study is to show that allometric variation of periwinkles, Littorina littorea, is consonant with the range of variation of the environmental factor exposure. The consonance is studied in terms of phenotypical fitness. The environment of the periwinkles studied is not uniform nor constant thus the optimal phenotype will be the one that is on the avet:age the best fit in the local environment. The variation associated with the optimal phenotype is expressed as standard deviations from the average with respect to the parameter measured. MATERIALS AND METHODS Littorina littorea was sampled at selected stations in June, September and December 1975 in three areas in the vicinity of the Marine Biological Station Ronbjerg, situated in the central part of the Limfjord, Denmark. (Map in Fenchel, 1975.) Area I with the stations 1 to 7 is the Ronbjerg Harbour protected by a northern and southern pier constructed of large boulders embed,ded in concrete. St. 1 is a shallow littoral flat with gravel and small rocks at the end of the southern pier. (Map in Hylleberg & Christensen, 1976.) Sts 2 to 7 are situated at about 10m intervals along this pier. The degree of exposure increases with increasing station number, but \st. 7 differs topographically from the other stations due to a rock slide which has created a small "lagoon" with an outer row of rocks providing a breakwater in the otherwise strongly exposed environment. Area II is a shallow gravel and sand flat on the open coast about 800 m SE of the Laboratory. St. 17 (0·1 m depth) is a rocky habitat, close to a brooklet and the littoral fringe characterized by a , dense growth of Enteromorpha sp. It should be noted that lunar tides are negligible in the Limfjord, so the littoral fringe is rather constant most of the year. Wind pressure may, however, change the water level for unpredictable periods of time. St. 18 is also on gravel and small rocks with a depth of about 0.3 m. In the summer a dense mass of Chaetomorpha /inum accumulates at this station. Area III is a sandy lagoon protected by dikes at Lendrup c. 3 km north of Ronbjerg. Periwinkles were sampled at st. 21 which is the concrete walls and wooden constructions at the dike between the lagoon and the Frederik VII Kanal (Fenchel, 1975). Two methods of sampling were used. In areas I and III all snails were hand-picked from individual rocks. In area II a 1/10 m 2 frame was used and all snails picked from the gravel.