Rice
Fungal pathogens cause the most important diseases of rice in the US (Table 2).

Table 2. Important pests of rice in the US and current efforts to provide resistance.

Rice pests in US	Approach to resistance
Blast (Pyricularia orzyzae Cav.)	Germplasm and genomics; Enhancement of native resistance mechanisms via molecular biology
Sheath blight (Rhizoctonia sloani Kuhn)	Germplasm limited
Stem rot (Sclerotium oryzae Catt.)	Related species
Rice water weevil (Lissorhoptrus oryzophillus Kushel)	Germplasm
Bacterial leaf blight *	rDNA
Rhizoctonia*	rDNA
Insect resistance for lepidopteran and coleopteran pests*	rDNA for potato proteinase inhibitor II and various Bt toxins.

* Genetically engineered pest resistant rice field tested in the US as of 1998.

In other parts of the world, bacterial and viral diseases have a greater impact on rice culture. Blast (Pyricularia orzyzae Cav.) is the most important rice disease worldwide and the key fungal disease in the US. There are many races of the blast pathogen, and blast resistant rice germplasm is primarily race specific (McKenzie et al. 1994). Members of the gene family conferring blast resistance, identified by the rice genomics project, are being used to study the mechanism of plant response to pathogens (Ronald 1997). It may not be long before genes for both race-specific and more general disease resistance will be available for evaluation in rice.

In a reference document prepared by the OECD (Cook et al. 1993), the section on wheat cites eyespot resistance. The Pch gene for eyespot resistance was transferred to hexaploid wheat by a wide cross from Aegilops ventricosa. By outcrossing, the Pch gene could be transferred from wheat to jointed goatgrass, A. cylindrica, a related weed that is a host for eyespot fungus. According to this document, if jointed goatgrass acquired the Pch gene from wheat, it would not become a more important weed, but rather become a less important host for eyespot disease. However, members of our working group questioned what is really known of the negative impacts of disease on jointed goatgrass populations. The second statement in the OECD reference document may simply be opinion and is not documented by ecological studies.

• There is no evidence for these crops or weeds that introduced traits affect the establishment, persistence, or ability to spread.
• All these crop/weed complexes are subject to integrated weed management programs; thus, the negative impact is minimized.
• Wheatgrass presents a special case if fields are adjacent to conservation preservation areas and range lands.

The working group asked the question, "Can we apply the successful strategies learned from monocultured crop/weed complexes in managed agro-systems to weeds in the wild?" The consensus was No; these sets of data do not exist and would require difficult and lengthy research to gather. On the other hand, there are inferential data sets that may be derived from floristic survey and weed census reports. For example, in a review for the OECD background paper on plant breeding (Cook et al. 1993), the authors note that easily recognized wheat traits such as red coleoptile and pubescent leaves have never been reported in populations of jointed goat grass.

The biology of johnsongrass has been well described. Although it reproduces and spreads principally by rhizomes, johnsongrass can reproduce sexually by producing selfed, or outcrossed seed with nearby compatible relatives (see Warwick and Black 1983). It is generally tetraploid (2x=40), and may be an allopolyploid result of past hybridization between S. bicolor and S. propinquum (Paterson et al. 1995). North American johnsongrass populations are often believed to contain pools of stable introgressants of wild plants and modern cultivated sorghum (Doggett 1988). In fact, Harlan (1992) suggested that crop-to-wild hybridization has likely been the key to the continued success and aggressiveness of johnsongrass in the United States, though there is no empirical evidence to support this idea.

The gene pools of the wild and cultivated sorghums can be described as unique, but not exclusive due to their common ancestry. Compatibility between these congeners has been well documented (see Hadley 1958). The likelihood of gene flow from crop to weeds is generally considered to be high. Crop sorghum outcrosses at rates as high as 15% (Ellstrand and Foster 1983), and the range of johnsongrass overlaps that of crop sorghum in all areas of cultivation in the US. Arriola and Ellstrand (1996) reported spontaneous weed x crop hybrid formation in the field at rates ranging from 0 - 12% and at distances of 0.5 - 100 meters. Subsequent measurements of hybrid fitness demonstrated no apparent fitness costs to the wild x crop hybrids when compared to the non-hybrid weeds under field conditions (Arriola and Ellstrand 1997). Although hybridization is variable, in this system one can regard the crop sorghum/johnsongrass complex as having a high probability of stable gene transfer in the wild.

The second weed, red rice, is a variant or ecotype of domestic rice, O. sativa. It does not share the perennial nature of O. rufipogon and persists in cultivated rice fields primarily by having highly dormant seed. Seed banks of red rice can be long lived and management of the weed is often based upon depletion of the seed bank. The species can compete with commercial rice and, if not controlled, is considered a weed problem (Craigmiles 1979; Noldine 1998). Red rice mimics crop behavior and often causes reductions of crop yield and quality through the admixture of red grains with the harvest. Red rice has been described as a dominant competitor; in competition studies, as many as three crop rice plants were required to impact yield as much as one red rice plant (Pantone and Baker 1991).

Historically, it is believed red rice originated in the cultivated fields of India where both red and white rice were grown. Its introduction into the US is attributed to a seed mixture imported from the East Indian Seed Company. Red rice was established in the rice fields of the American colonists; in 1850, the USDA published reports that listed four red rice types (Craigmiles 1979). Strict quality standards for seed rice, combined with agricultural practices designed to deplete the red rice seed bank, have eliminated red rice populations from California and sections of the southeastern production area (Hill et al. 1994). Red rice populations continue to hybridize naturally with cultivated rice. Gene flow travels predominately from the cultivated crop into the weedy red rice population. Cultivation of early maturing commercial varieties provides a partial hybridization barrier to the later maturing red rice populations (Langevin et al. 1990). Although the pollination periods for red rice may be later than most of our currently cultivated rice varieties, red rice exhibits an uneven maturity in the panicle and can produce some seed capable of rapid germination. Thus red rice, allowed to produce seed in a commercial rice field, can shatter mature grains in advance of even the early harvested varieties.

Red rice can express a long seed dormancy when submerged and buried in the soil. In field studies of five red rice populations buried at three locations, red rice survived more than 6.5 years, however the length of seed survival varied with location and population (Goss and Brown 1939).

Jointed goatgrass (Aegilops cylindrica) is a major weed of winter wheat in the western United States (Dewey 1996). The species is believed to be indigenous to southern Europe and Russia (Gunn 1958; Donald and Ogg 1991). It was probably first introduced into the United States in contaminated winter wheat seed brought by settlers from the eastern Mediterranean region. Goatgrass is a tetraploid (2N=28) with genomes C and D (Donald and Ogg 1991). Jointed goatgrass and wheat share the D genome in common. The shared genome allows hybridization between the species in the field (Zemetra et al. 1998). Hybrids were once believed to be sterile, but two recent studies found hybrids in the field with viable seed (Mallory-Smith et al. 1996; Seefeldt et al. 1998). Hybrids were not self-fertile. Rather, hybrid plants exhibited a low level of female fertility (approximately 2%) that allowed for natural backcrossing to occur in the field (Zemetra et al. 1998). Greenhouse experiments indicated that percent seed set was similar with wheat or jointed goatgrass as the pollen parent, but that seed set and self-fertility in second generation backcrosses favored jointed goatgrass as the recurrent parent. Based on these results, only two crosses in the field after hybrid formation appear to be sufficient to recover partial self-fertility with jointed goatgrass as the recurrent parent (Zemetra et al. 1998). Thus, if the wheat that produced the hybrid carried a pest resistance gene on the D genome, it would be possible for the pest resistance trait to transfer to jointed goatgrass after only two backcross generations.

In addition to jointed goatgrass, intergeneric hybridization between spring wheat and crested wheatgrasses, Agropyron Gaertn. (sensu stricto), has been reported (Chen et al. 1989, 1990). The crested wheatgrasses constitute a perennial cross-pollinating complex of roughly 10 species with diploid (2n = 14), tetraploid (2n = 28), and hexaploid (2n = 42) forms built on what appears to be one basic genome, P (Dewey 1984; Chen et al. 1990). The species are indigenous to Eurasia but are now widely grown as economically important forage on arid rangelands in United States and Canada (Dewey 1983).

Intergeneric hybridization between wheat and four crested wheatgrasses, Agropyron mongolicum (2n = 14), A. cristatum (2n = 28), A. desertorum (2n = 28), and A. michnoi (2n = 28) has been documented (Chen et al. 1989, 1990). Hybrid seed set varied among wheatgrass species but was always low, ranging from 0.24 to 2.87% seed set. Crossability of diploid species was lower than that of polyploid species. Although most hybrid plants died of hybrid necrosis, a few plants were successfully established. Hybrid necrosis occurred at varying frequencies with different plant combinations suggesting that crossability varies among plants and accessions of a species. Backcross progeny of wheat x A. cristatum and wheat x A. michnoi hybrids were obtained by embryo rescue. Whether hybrids and backcross progeny occur naturally in the field is unknown.