Sublethal foot-predation on Donacidae Mollusca: Bivalvia) Carmen Salas * , Cristina Tirado, Maria Eugenia Manjo Ân-Cabeza Departamento de Biologõ Âa Animal, Facultad de Ciencias, Universidad de Ma Âlaga, E-29071 Ma Âlaga, Spain Received 25 September 2000; accepted 6 March 2001 Abstract The incidence of foot nipping was studied on the Donax spp. of the littoral of Ma Âlaga Southern Spain, 2875 specimens collected from February 1990 to January 1991) and of Re  island French Atlantic coast, 262 specimens of Donax vittatus Da Costa, 1778) collected in May 1996). In Ma Âlaga, Donax trunculus L., 1758 was the species most regularly nipped 18% of individuals), with peaks in summer 25% in August and 48% in September) and winter 34% in December). In Re  island, 27% of the specimens showed a nipped foot. Logistic regression shows that in D. trunculus length is the variable that most in¯uences the probability of foot nipping, followed by weight and chlorophyll a. However, the difference in length between damaged and undamaged individuals was not signi®cant U-Mann±Whitney test). The size class frequency and the values of Ivlev's index show that the small size classes were avoided, while for the other size classes predation remained balanced throughout the year. Therefore, the avoidance of the small size classes makes length the most in¯uential variable. The logistic regression indicated a coef®cient B  20.03 for weight. This implies a slightly negative in¯uence on the probability of foot nipping. However, without the data of September, there is a positive correlation r  0:76; p , 0:01 between the monthly percentages of predation and the ¯esh dry weight of a standard individual 25 mm long). The peak in September could be due to the recruitment peak of bivalves, which may have attracted more predators to the area, and/or to the recruitment of predators such as crabs to the swash zone. Logistic regression and test of comparison of percentages indicate that there was not any in¯uence of the sex of an animal on the probability of foot nipping. Only in February was a signi®cantly higher percentage p , 0:05 of females nipped 44.44%) than the total of females in the sample 20.20%). The biomass as ¯esh dry weight) of D. trunculus lost by foot nipping amounts to more than 20% in most of the size classes. There was an increase from the small sizes to the largest ones, in which it reaches 37%, with a positive correlation r  0:84; p , 0:005 between size class and loss of biomass. Possible predators responsible for the foot nipping are crabs. Crab species usually found together with the donacids were Portumnus latipes Pennant, 1777) Liocarcinus vernalis Risso, 1816) and Atelecyclus undecimdentatus Herbst, 1783). In aquarium experiments, they demonstrated an ability to nip the foot of clams. Portumnus latipes was the most active foot nipper, but left alive all the damaged clams. Therefore, we conclude that crabs are the most likely foot-nipping predators in the ®eld. q 2001 Elsevier Science B.V. All rights reserved. Keywords: Sublethal foot predation; Donacidae; Mollusca; Bivalvia; Crabs 1. Introduction Predation is often considered to be the most important cause of mortality in bivalves Dame, 1996; Nakaoka, 1996; Van der Veer et al., 1998). However, the predation is not always lethal. Sublethal predation results in the loss of siphons by `nipping', `cropping' and `browsing' predators, mainly ®sh Edwards and Steel, 1968; Trevallion et al., 1970; Journal of Sea Research 46 2001) 43±56 1385-1101/01/$ - see front matter q 2001 Elsevier Science B.V. All rights reserved. PII: S1385-110101)00064-8 www.elsevier.com/locate/seares * Corresponding author. E-mail address: casanova@uma.es C. Salas).