The discipline of ecology is habitually divided up into a whole range of different specialities, and this thesis forms part of what is commonly known as "community ecology" (Diamond and Case 1986). To be more precise, my work has centred on the degree to which the trophic relationships that exist between the organisms constituting a food web govern the structure and dynamics of the communities they form. I approached the subject from an experimental point of view and all the work was carried out in the field by means of what are referred to as "natural experiments", with all the advantages and disadvantages that this supposes. The chosen ecosystem was the coastal marine benthos or, to be more specific, rocky infralittoral communities, a propitious place in which to work since this setting has seen the production of some highly important contributions to community ecology, above all in the development of experimental techniques (review by Castilla 2000).

Owing to the existence of previous investigations, my work, strictly speaking, does not break any new ground, since the subject matter has already been well studied in other oceans (see Castilla 2000 for a revision). Paradoxically, the Mediterranean, them cradle of the natural sciences (Aristotle, 350 BC) and home to the invention of scuba diving by Gagnan and Cousteau in 1944, has generally not been witness to any of the most important advances in experimental benthic science. Whilst not wanting to question the universality of discoveries regarding benthic communities made in other waters, it can surely not be an excessive precaution to attempt to understand the way regional peculiarities - the Mediterranean is a very singular sea in many ways (Margalef 1985) - affect general principals.

One highly significant way in which benthic experiments in the Mediterranean are playing a part in current debate is the role that marine reserves or protected marine areas (PMAs) can be expected to play as management tools in ecosystems adversely affected by human action. In this sense, my work shares a characteristic with many others: the use of marine reserves as experimental sites that can be compared with nearby unprotected zones in order to try and judge their worth.

As a final reflection on the main object of study in my thesis it is worth remarking that the system under study relates on a trophic level macroscopic algae - the main primary producers in the phyto-benthos - to their herbivores, which in the coastal Mediterranean benthos are essentially a sea urchin (Paracentrotus lividus), a fish, the salema (Sarpa salpa), and a guild of omnivorous fish that include optional grazers of algae and predators on sea urchins. My study investigated the interactions taking place between all the actors in this complex community and not just the interactions occurring between a certain group of species, although inevitably, given its central position, the sea urchin Paracentrotus lividus has been the focal point of our efforts. As a means of simplifying, throughout this paper I often refer to the "triangle" 'algae-sea urchins-fish', which is described in more detail in Annex 1.

I feel that my work has two main applications. Aside from any knowledge gained from the experimental study of the interactions within a complex community under natural conditions, our results have, firstly, a clear role to play in the management of coastal areas and biological conservation in general. It is known that high densities of sea urchins can cause a regression and a loss of diversity in algal communities (that can be compared to the effects of desertification on the biodiversity of forest ecosystems).

Studying experimentally the mechanisms that regulate these interactions is a way of searching for valid management criteria. Secondly, we looked at the relevance of the role played by changes of anthropic origin such as eutrophication or overfishing in the proliferation of sea urchins and the regression of algal communities.