The convergence of the biotas of the five regions of the world with a mediterranean-type climate was actively studied over a decade ago (di Castri & Mooney, 1973; Mooney, 1977; Cody & Mooney, 1978). Already at that time, the large interchange of plant species between the five regions was mentioned (di Castri & Mooney, 1973) but not much documented. The present volume aims to fill that gap. Such an enterprise gains support from the assumption that the success of invasions strongly depends upon the similarity of the source and reception areas in terms of climate, life forms and structure of the biota (Baker, 1986) and that, as a consequence, the five mediterranean-type regions constitute a rather specific network within which invasions are better analysed, understood and amenable to predictions than they would be using larger geographical units composed of several climatic types.

In this chapter, in addition to documenting the invasion patterns of a genus containing ecologically and economically important species, we want to challenge the above assumption. In fact, because of life form, phenology, plasticity or ecotypic differentiation, many species have distributions encompassing regions of several climatic types. For such species, only extremes in climate may prevent invasion. Moreover, climatic similarity is a condition neither sufficient nor necessary for an invasion to occur (as shown by the literature on biological control of insects and weeds, e.g. Harris, 1984).