ECOGRAPHY 18: 403-409. Copenhagen 1995 Granivory in the Chilean matorral: extending the information on arid zones of South America Rodrigo A. Vasquez, Ramiro O. Bustamante and Javier A. Simonetti Vdsquez, R A , Bustamante, R O and Simonetti, J A 1995 Granivory m the Chil- ean matorral extending the information on and zones of South Amenca — Ecography 18 403^409 Comparative studies of granivory intensity across different and and semi-and regions of the world have assumed homogeneity of conditions over large geographical areas, despite the existence of evident heterogeneities at local and regional scales For South Amenca, previous evidence from the Monte desert, Argentina, showed that gramvoiy intensity is low compared with other continents and that seed removal rates by taxon (gramvore ranking) are in decreasing order ants > birds = small mammals In this work, we examined the assumed generality of this pattem for another South Amencan region, the Chilean matorral We studied the differences between sparse and dense habitats of the matorral in granivory mtensity throughout the year Although total seed removal did not differ between habitats, differences among granivores did occur Ants were the most important seed removers in both habitats, although they did not differ from birds in the sparse habitat Further, ants were the most important seed consumers dunng summer, but their seed removal rate is almost negligible the rest of the year Birds maintained an lntennediate rate of seed consumption through the year Granivory intensity by small mammals was very low m both habitats dunng the whole year The gramvore ranking obtained for the Chilean matorral was ants > birds > mammals, in dense habitat, and ants = birds > mammals, in sparse matorral Am abundance is correlated with seed removal throughout the year This relationship was not found in the other taxa We discuss possible causes for the differences among hab- itats, taxa, and seasons, as well as disunct granivore rankings from different conti- nents R A Vdsquez, R O Bustamante and J A Simonetti, Dept de Ctencias Ecologicas. Fac de Ciencias, Umv de Chile, Casitla 653. Santiago. Chile (Present address ofR A V Dept of Zoology, Umv ofO^ord. Oxford, UK 0X1 3PS) _ , . curs at larger geographical scales Nevertheless, and re- introdUCtlOn g^j^j j^ present important spatial and temporal hetero- Granivory by ants, birds, and small mammals has been geneiUes (Wiens 1986, Kotler and Brown 1988). Even at used to assess convergent evoluUon across communities a local scale, different habitats can be disUnguished from different and and semi-and regions of the world withm a single ecosystem (Kolasa and Pickett 1991) (Mares and Rosenzweig 1978, Abramsky 1983, Morton This vanability has seldom been addressed in granivory 1985, see also Brown etal 1979) Usually, seed removal studies (but see e.g Brown et al 1975, Morton 1985, rates have been estimated at one or a few sites, and the Diaz 1992), and once taken into account, it could chal- results have been considered to reflect the intensity of lenge current lnter-regional or lnter-continental compari- granivory for a whole region or conUnent The assump- sons If distinct habitats of a given and ecosystem Uon behmd this extrapolation is that the environment is present different pattems of granivory, then the validity homogeneous; i.e what happens at a local scale also oc- of broad scale extrapolaUons would be weakened On Accepted 11 May 1995 Copynght © ECOGRAPHY 1995 ISSN 0906-7590 Pnnted in Denmark - all nghts reserved ECOGRAPHY 18 4 (1995) 403