Extensive niche overlap between closely related species generally leads to aggressive interactions and competition. The short-tailed mouse (Mus spretus Lataste, 1883) and the wood mouse (Apodemus sylvaticus (L., 1758)) show a large habitat overlap without aggressive interactions. The present study investigates the existence of food competition between these species, based on an analysis of carbon and nitrogen stable isotopes. An almost exhaustive sample of plants, which were potential food resources, was taken and analyzed to infer the consumed plants in mouse diets. The main result showed that both species had a similar diet composition, consisting exclusively of seeds and fruits. This suggests that no competition for food between these species is apparent, or if it exists it would be minimized by a differential exploitation of resources. In the absence of food and space competition,the short-tailed mouse may be using the presence of the wood mouse as an indicator of habitat food quality. In the case of wood mice, we hypothesize that the level of competition with short-tailed mice may be low because of the abundance of resources and because the wood mice may perceive the smaller short-tailed mouse as being equivalent to a young of their own species.

RESUME Les activités humaines ont favorisé l'émergence de guildes particulières de micromammifères. La diffusion et la radiation adaptative de la guilde des souris en Europe sont analysées d'un point de vue archéozoologique et écologique. Plus généralement, depuis le Plio-Pléistocène les faunes de mammifères subissent les conséquences des activités humaines. Celles- ci entraînent donc, outre des changements dans la composition des faunes, des modifications radicales dans la structure des communautés. Ces bouleversements fauniques s'apparentent à ceux observés lors des grandes crises écoclimatiques. L'anthropisation de la biosphère apparaît ainsi comme le fait générateur d'une crise écologique majeure.

The black rat Rattus rattus is a successful colonizer that threatens native species on numerous islands. Stable isotopes of carbon and nitrogen were used to quantify the rat diet on islands off the southern French coast. Samples of rat hair and faeces and their potential food were collected on each island. The nitrogen isotopic values of invertebrates and vegetation on Riou (near Marseille) differed from the other two islands (Hyères archipelago). This was attributed to a large colony of gulls on Riou, of which rats take great advantage. A part of the observed variability was due to methodological skews: two kinds of hair gave different results and there was an effect of the season and year of sampling but inter-island differences were noticeable. In certain cases, the food source value from plants and invertebrates could supply the total food needs of rats as there was no evidence of consumption of eggs, chicks or birds.

Les petits rongeurs marques ä l'aide de poudres fluorescentes peuvent etre suivis par les traces qu'ils laissent au sol; leurs deplacements sont ainsi materialises, leurs nids et leurs terriers localises.Les grands deplacements observes grace ä cette technique, chez toutes les especes etudiees, on t amene les auteurs ä modifier leurs protocoles d'echantillonnage des Murides.Cette technique s'est revelee efficace chez de nombreuses especes vivant dans diffe'-rents types de milieux (Mus musculus domesticus dans les dunes sableuses, Mus spre· tus et Apodemus sylvaticus dans la garrigue mediterrane'enne, Proechimys cuvieri en foret tropicale guyanaise, Hylomyscus stella et Hylomyscus fumosus en foret tropicale gabonaise).L'utilisation de ces pigments a egalement permis l'etude de la dispersion des glands par Apodemus sylvaticus dans la chenaic mediterraneenne.

The mating system of murids remains largely unknown, with controversial reports. Reevaluation of previous data reveals monogamous features in the Algerian mouse from the South of France, Mus spretus. To test this hypothesis, we performed dyadic encounters toward three kinds of opponents from the other sex (his/her own mate, a mouse from an unknown pair, and an isolated individual), and compared the results to those of a related acknowledgedly polygynous related species, the house mouse M. musculus domesticus. We expected more affiliative behaviours toward its mate from the M. spretus male, and a higher interest toward unknown females in the house mouse. Females should not exhibit striking differences toward their partner, contrary to their behaviour toward unknown males. Results showed the predicted behaviour for males as well as for females: more amicable, but also sexual, toward its mate for M. spretus male, the unknown protagonists eliciting the greatest social interest in the case of the house mouse.