Early theoretical analyses and observations of community structure have led
to the hypothesis that a high diversity of communities increases their res
istance to invasions. However, to this date, observation and experimental d
ata have shown positive or negative relationships between plant community d
iversity and invasibility. Studies of biological invasions proposed several
others mechanisms explaining invasions: disturbances, presence of empty ni
ches and interactions with others community trophic levels. The role of pla
nt diversity can be decomposed in two factors: is specific richness per se
and/or the diversity of functional groups that lead to invasion resistance
of communities? A study carried out in old fields of Montpellier country sh
owed a general picture of invasion and its complexity in this country. Our
observations of these invasion patterns confirmed the difficulties to obser
ve clear relationships with their diversity. We then resorted to semi-contr
olled field experiments. We sowed annual old field communities with differe
nt levels of diversity: variation in species richness, variation in functio
nal richness and variation in functional identities for a fixed Functional
richness. Demographic and vegetative parameters of two exotic probes transp
lanted into these communities, Conyza bonariensis and C.canadensis, were me
asured along their life cycle. Primary community functioning parameters of
the model ecosystems were measured simultaneously. Species richness had lit
tle effect on performance of the two Conyza species. Functional composition
appeared more relevant than functional richness per se to explain communit
y invasibility. Mechanisms of functional composition effects are proposed.