Bruijnzeel / Agriculture, Ecosystems and Environment 104 (2004) 185–228

Invasive species have received considerable attention in recent years, but research has primarily focused on invasive species of natural habitats. Furthermore, cropland weeds have often been viewed as possessing a “general-purpose genotype ” and therefore exhibiting relatively static genetics. However, a more current view is that weeds are capable of rapid genetic change, thereby making analysis of their evolutionary ecology a potentially valuable component for the development of sustainable weed management systems. In particular, further analysis of ongoing evolutionary change in cropland weeds is important be-cause (1) most cropland weed species exhibit considerable adaptability, (2) cropland agriculture is continuously changing, and (3) further research on weed adaptability is needed to design cropping systems to address evolutionary change. In this review, we examine the potential of cropland weeds to evolve so as to affect their invasiveness. There is abundant evidence of genetic variation within and among weed populations in traits relevant to invasiveness, including seed germination patterns, life history traits, physiological adaptability and adaptation to disturbance and resource fluctuations. Approximately half of cropland weed species are primarily selfing, and species with a high degree of selfing tend to exhibit homogeneity within populations but divergence among populations. We identified four critical areas for future research: conflicting selection pressures on weed