Is invasiveness a legacy of evolution? Phylogenetic patterns in the alien flora of Mediterranean islands


*Correspondence and present address: Global Programmes Department, Royal Society for the Protection of Birds, The Lodge, Sandy, Bedfordshire SG19 2DL, UK. E-mail:


  • 1The Mediterranean region has been invaded by a wide range of introduced plant species which differ greatly in their ecology, morphology and human utilization. In order to identify a suite of traits which characterize invasiveness, recent studies have advocated the use of evolutionary relationships to unravel highly confounded influences.
  • 2This study attempts to identify an evolutionary component to invasiveness and other complex invasion-related traits in the Mediterranean alien flora using an autocorrelation technique, the ‘phylogenetic association test’. I compared a traditional hierarchical taxonomy with the recent phylogeny of the Angiosperm Phylogeny Group.
  • 3Invasiveness did not have a significant phylogenetic component. Any weak clustering was generally at the genus level.
  • 4Several associated ‘meta-traits’ (high introduction frequency, adaptation to several habitat types and favourability for different modes of introduction), exhibited stronger phylogenetic components. Although each of these conveys some of the attributes of invasiveness, their clustering patterns differed considerably, suggesting that they arise from independent evolutionary pressures. Furthermore, within each meta-trait, different clusters may have been selected for different reasons.
  • 5Other reasons for the lack of a detectable evolutionary component to invasiveness are discussed. Firstly, the results of our test simulations suggested that incorrect phylogeny could result in a moderate degree of error. Secondly, over evolutionary time, complex or stochastic events such as ecosystem change could radically alter the adaptive advantages of particular traits.
  • 6Synthesis. Since invasiveness has little phylogenetic component, I argue that it is less likely to be predictable from as yet unidentified traits in any simple way. Although trait syndromes could develop without leaving a phylogenetic pattern, its absence probably indicates that the dominant selective forces are responses to short-term ecological shifts, and a greater mechanistic understanding of these is needed.