Behavioural ecology is ultimately aimed at understanding the evolutionary significance of behavioural traits, i.e. how behaviour influences fitness. Both theoretical and experimental investigations follow an economic approach: behavioural patterns are analysed in terms of costs and benefits of alternative courses of action, assuming that natural selection would promote optimization of the balance between costs and benefits. It has been progressively recognized that the detailed knowledge of fitness consequences of behavioural acts requires the understanding of mechanisms and physiological processes underlying behavioural decisions. The present study is aimed at testing new techniques for monitoring such variables, which are based on the deployment of small transducers externally mounted on animals. This non-invasive monitoring allows the recording of behavioural patterns while minimizing associated disturbance. Transducers were applied on hard-shelled marine invertebrates: dogwhelks (Nucella lapillus), common mussels Q0:ytilus edulis) and shore crabs (Carcinus maenas), which are also ideal subjects for laboratory studies and have been already extensively studied in behavioural ecology. Papers reported in chapter 2 and 3 deal with the foraging behaviour of dogwhelks. A transducer for recording mechanical vibrations was used to directly record, for the first time, the drilling behaviour of dogwhelks penetrating mussel shells. This in tum allowed the effect of experience on handling time to be quantified. The remaining three studies are based on the use of infrared phototransducers for monitoring cardiac activity. Because of the relationship between circulatory and respiratory systems, measuring heartbeat rate provides information on the consumption of oxygen, and thus, on the energy demand associated with particular performances. In the study reported in chapter 4, this technique was used for monitoring responses of mussels to risk of attack and predation by dogwhelks. In the presence of effluent from dogwhelks, heartbeat rate of mussels significantly increased without any apparent behavioural response. Cardiac activity increased further when under attack. These responses might represent an adaptive trade off between energy budget and risk of predation. In chapter 5, transducers were used to monitor heartbeat of crabs involved in fights against conspecifics of various sizes. The relation between cardiac and repiratory rates was assessed and, consequently, the energy demand of different fighti1tg strategies could be quantified. Costs of aggression were determined by the time spent fighting but not by the fighting strategy. Moreover, post-fighting alertness incurred metabolic costs, independent from the previous behavioural effort, which are probably adaptive against the risk of being injured and/or loosing contested resources. The remaining study focussed on prey-handling behaviour of crabs. By monitoring heartbeat-rate, energy costs of handling mussels of various sizes could be quantified for the first time. Surprisingly, costs represented an almost irrelevant proportion of corresponding gains. Moreover, the tendency of profitability (gross energy gain per unit of handling time) to increase with prey-size was weakened by including energy cost and thus time was judged to be a more appropriate currency for costing. Throughout the study, the mechanistic approach to behavioural analysis was essential for collecting crucial and unique information on how behaviour is related to fitness.