Cold tolerance adaption is a crucial determinant for the establishment and expansion of invasive alien plants into new cold environments; however, its evolutionary mechanism is poorly understood. Crofton weed (Ageratina adenophora), a highly invasive alien plant, is continuously spreading across subtropical areas in China, north-eastward from the first colonized south-western tropical regions, through cold tolerance evolution. Close relations between the cold tolerance levels of 34 populations, represented by 147 accessions, and the latitude, extreme lowest temperature, coldest month average temperature, and invasion period have provided direct insight into its cold tolerance divergence. A comparative study of the CBF pathway, associated with the cold tolerance enhancement of cold-susceptible CBF1-transgenic plant, among four geographically distinct crofton weed populations revealed that the CBF pathway plays a key role in the observed cold tolerance divergence. Four epialleles of the cold response regulator ICE1 ranged from 66 to 50 methylated cytosines, representing a 4.4% to 3.3% methylation rate and significantly corresponding to the lowest to highest cold tolerance levels among these different populations. The significant negative relation between the transcription levels of the primary CBF pathway members, except for CBF2, and the methylation levels among the four populations firstly demonstrates that the demethylation-upregulated transcription level of CBF pathway is responsible for this evolution. These facts, combined with the cold tolerance variation and methylation found among three native and two other introduced populations, indicate that the ICE1-demethylated upregulation of cold tolerance may be the underlying evolutionary mechanism allowing crofton weed to expand northward in China.

Transgenic herbicide-resistant rice is needed to control weeds that have evolved herbicide resistance, as well as for the weedy (feral, red) rice problem, which has been exacerbated by shifting to direct seeding throughout the world-firstly in Europe and the Americas, and now in Asia, as well as in parts of Africa. Transplanting had been the major method of weedy rice control. Experience with imidazolinone-resistant rice shows that gene flow to weedy rice is rapid, negating the utility of the technology. Transgenic technologies are available that can contain herbicide resistance within the crop (cleistogamy, male sterility, targeting to chloroplast genome, etc.), but such technologies are leaky. Mitigation technologies tandemly couple (genetically link) the gene of choice (herbicide resistance) with mitigation genes that are neutral or good for the crop, but render hybrids with weedy rice and their offspring unfit to compete. Mitigation genes confer traits such as non-shattering, dwarfism, no secondary dormancy and herbicide sensitivity. It is proposed to use glyphosate and glufosinate resistances separately as genes of choice, and glufosinate, glyphosate and bentazone susceptibilities as mitigating genes, with a six-season rotation where each stage kills transgenic crop volunteers and transgenic crop x weed hybrids from the previous season.

Twenty-one grass weeds have evolved resistance to herbicides in Latin America, particularly in rice, soybean, wheat, and orchards. Junglerice, the most widespread and economically important rice weed, evolved resistance to propanil, acetyl-coenzyme A carboxylase (ACCase)-inhibitor herbicides, quinclorac, and imazapyr in Central America, Colombia, and Venezuela. Some junglerice populations are resistant to at least three herbicide modes of action. Other herbicide-resistant (HR) rice weeds are barnyardgrass and gulf cockspur to quinclorac in Brazil, and saramollagrass to ACCase-inhibitor herbicides in Colombia and bispyribac in Venezuela. Populations of weedy rice resistant to imidazolinones are now emerging, most likely originated from gene flow from imidazolinone-resistant rice. Saramollagrass also became resistant to nicosulfuron in corn in Venezuela. Eight species associated with soybean are resistant to ACCase-inhibitor herbicides in Brazil (alexandergrass, goosegrass, and southern crabgrass) and Bolivia (Louisiana cupgrass, itchgrass, sudangrass, and two common wild sorghum species). Four more ACCase-inhibitor–resistant species (hedgehog dogtailgrass, wild oat, rigid ryegrass, and Italian ryegrass) are found in Chile infesting canola and wheat. ACCase-inhibitor–resistant hood canarygrass, littleseed canarygrass, and wild oat are important in wheat in Mexico. Resistance to acetolactate synthase (ALS)-inhibitor herbicides has been reported in itchgrass, goosegrass, and Mexican grass. Italian ryegrass populations resistant to glyphosate have been found in Chile and Brazil. Glyphosate resistance has also evolved in goosegrass in Bolivia and johnsongrass in Argentina. In general, little is done to prevent resistance evolution. An exception is the stewardship programs aiming to prevent gene flow from imidazolinone-resistant rice to weedy rice. Once resistance evolves, HR populations are mostly managed by shifting to herbicides with different modes of action and, in some cases, by slightly modifying agronomic practices. Propanil formulations containing a synergist are used to manage propanil-resistant junglerice. Increased no-till agriculture and planting of glyphosate-resistant crops are likely to select more glyphosate-resistant weeds.

Summary Rice cultivars resistant to broad‐spectrum herbicides have been developed and their commercial release is imminent, especially for imidazolinone and glufosinate resistant varieties in the USA and Latin America. Glyphosate‐resistant rice should follow within a few years. Rice growers throughout the world could benefit from the introduction of herbicide‐resistant rice cultivars that would allow in‐crop, selective control of weedy Oryza species. Other perceived benefits are the possibility to control ‘hard‐to‐kill’ weed species and weed populations that have already evolved resistance to herbicides currently used in rice production, especially those of the Echinochloa species complex. Weed management could also be improved by more efficient post‐emergence control. Introduction of herbicide resistant rice could also bring areas heavily infested with weedy rice that have been abandoned back to rice production, allow longer term crop rotations, reduce consumption of fossil fuels, promote the replacement of traditional chemicals by more environmentally benign products, and provide more rice grain without adding new land to production. There are also concerns, however, about the impact of releasing herbicide‐resistant rice on weed problems. Of most concern is the possibility of rapid transfer of the resistance trait to compatible weedy Oryza species. Development of such herbicide resistant weedy rice populations would substantially limit the chemical weed management options for farmers. Herbicide‐resistant rice volunteers also could become problematic, and added selection pressure to weed populations could aggravate already serious weed resistance problems. Because of the risk of weedy Oryza species becoming resistant to broad‐spectrum herbicides, mitigating measures to prevent gene flow, eventually attainable by both conventional breeding and molecular genetics, have been proposed. With commercialisation of the first herbicide resistant varieties planned for 2001, these mitigating measures will not be available for use with this first generation of herbicide resistant rice products. Release of herbicide resistant rice should depend on a thorough risk assessment especially in areas infested with con‐specific weedy rice or intercrossing weedy Oryza species. Regulators will have to balance risks and benefits based on local needs and conditions before allowing commercialisation of herbicide‐resistant rice varieties. If accepted, these varieties should be considered as components of integrated weed management, and a rational herbicide use and weedy rice control should be promoted to prevent losing this novel tool.