Comparative ecological genetics of Bromus tectorum (cheatgrass, downy brome) from native European, naturalized New Zealand, and invasive North American populations

Within 40 years of introduction in the late 1800s, the Eurasian grass Bromus tectorum L. was abundant across western North America. This annual C3 grass arrived in New Zealand at least 100 years ago, and has yet to become invasive. While differences in invasive ability have often been attributed to environmental differences among ranges or to post-immigration evolution, we hypothesized that inherent genetic differences cause the disparate performances of B. tectorum in North America and New Zealand. We investigated performance differences among populations of B. tectorum from the native range in western and central Europe, naturalized range on New Zealand's South Island, and invasive range in the Intermountain West of North America. In common greenhouse and seed germination experiments, we compared 62 populations under control, low temperature, low nutrient, and low water environments. We assessed 15 seed, growth, and morphometric traits (e.g. germination rate, plant height, tiller number, days to flowering, biomass) and conducted an analysis of phenotypic plasticity for each of these traits. Significant differences among source locations were found for nearly every trait assessed. North American populations of B. tectorum were consistently tallest with the highest biomass values, followed by European and lastly New Zealand populations. Plants from all ranges showed significant levels of phenotypic plasticity for every trait except presence of glume hairs. Where phenotypic plasticity differed among ranges, values for North American populations typically were much higher than those for New Zealand populations. Differences in levels of vigor and phenotypic plasticity between source populations for B. tectorum in North America and New Zealand have likely contributed to the radically different histories of this grass in two of its new ranges. These performance differences cannot be attributed to environmental or maternal effects, and so must be due to heritable differences among plants from each range. In fact, phenotypic trait values for both sets of introduced plants formed discrete subsets of values for the native range. We found no evidence of the evolution of novel phenotypes. These genetic differences are products of the different locales in the donor ranges from which emigrant populations were gathered.