Genetic variation within a nonweed (coastal) population of poorjoe was measured in an experimental population subjected to soybean and intraspecific competition for space and light. A sib analysis was used to estimate phenotypic and genetic variance/covariance components of aboveground biomass production and five independent measurements of growth defined by a path-coefficient model. A multivariate analysis of response to selection was applied to predict evolutionary change in the growth measurements in response to selection for performance under soybean and intraspecific space/light competition; sufficient genetic variation was present in the experimental population to allow a rapid response to this selection. Selection under competitive conditions was predicted to cause the nonweed population to emerge earlier, grow faster early in development, change in growth form, and grow faster under a soybean canopy. Most of these changes would increase the resemblance of the selected population to a weed population of poorjoe. However, the growth rate of the selected population under a soybean canopy was predicted to become greater than that actually observed in the present-day weed population. Multivariate analysis of response to selection may be generally useful in predicting evolutionary response of plant pest populations to control practices.

Growth of an agricultural weed population and a coastal nonweed population of poorjoe was compared in a competition experiment with soybean and when grown alone. Mean aboveground biomass production of the weed population was roughly twice that of the nonweed population, whether grown alone or with soybean. A path-coefficient model of growth was fitted to the data in order to compare the populations with respect to five independent measurements of weed growth: establishment rate, early and later growth rates, growth form, and final growth rate. These measurements indicated plant growth over successive intervals of the growing season by statistically equating the size of all plants at the beginning of an interval of growth using analysis of covariance. In general, growth measurements were positively associated with aboveground biomass production. The weed population of poorjoe had a greater establishment rate and a greater early aboveground growth rate compared to the nonweed population; however, the final growth rates of the two populations were similar whether soybean was present or absent. Genetic changes leading to earlier establishment and faster early growth, but not increasing tolerance of soybean and intraspecific competition, may have occurred in the adaptation of poorjoe to an agricultural ecosystem.

Feral rye is an agricultural and ruderal weed of the western United States. We investigated how it has phenotypically diverged from its cultivated ancestor, domesticated cereal rye, and across its range since the introduction of its progenitor. Vegetative growth, flowering phenology, and reproductive characters of feral populations from across a 13° range in latitude in the northwestern United States were compared to that of rye cultivars under both vernalized (cold-treated) and nonvernalized conditions. Feral populations as a whole had smaller seeds, thinner culms, and a delay in flowering relative to cultivars, regardless of cold treatment. Vernalized feral populations from northern latitudes (northern California and eastern Washington) produced more, but smaller leaves and more tillers than both vernalized rye cultivars and southern California feral populations. Northern feral populations also flowered significantly later, irrespective of vernalization treatment. We conclude that feral rye is phenotypically distinct from domesticated cereal rye and that feral populations have diverged regionally from one another. Reproductive isolation from domesticated rye, due both to the loss in popularity of the crop and to phenological shifts in feral rye relative to cultivars, may be contributing to the rapid evolution of this weed away from its domesticated ancestor in less than 120 yr since its introduction.