Seed dormancy allows weedy rice (Oryza sp.) to persist in rice production systems. Weedy and wild relatives of rice (Oryza sativa L.) exhibit different levels of dormancy, which allows them to escape weed management tactics, increasing the potential for flowering synchronization, and therefore gene flow, between weedy Oryza sp. and cultivated rice. In this study, we determined the genetic diversity and divergence of representative dormant and nondormant weedy Oryza sp. groups from Arkansas. Twenty-five simple sequence repeat markers closely associated with seed dormancy were used. Four populations were included: dormant blackhull, dormant strawhull, nondormant blackhull, and nondormant strawhull. The overall gene diversity was 0.355, indicating considerable genetic variation among populations in these dormancy-related loci. Gene diversity among blackhull populations (0.398) was higher than among strawhull populations (0.245). Higher genetic diversity was also observed within and among dormant populations than in nondormant populations. Cluster analysis of 16 accessions, based on Nei’s genetic distance, showed four clusters. Clusters I, III, and IV consisted of only blackhull accessions, whereas Cluster II comprised only strawhull accessions. These four clusters did not separate cleanly into dormant and nondormant populations, indicating that not all markers were tightly linked to dormancy. The strawhull groups were most distant from blackhull weedy Oryza sp. groups. These data indicate complex genetic control of the dormancy trait, as dormant individuals exhibited higher genetic diversity than nondormant individuals. Seed-dormancy trait contributes to population structure of weedy Oryza sp., but this influence is less than that of hull color. Markers unique to the dormant populations are good candidates for follow-up studies on the control of seed dormancy in weedy Oryza sp.

A replacement series study was conducted in a greenhouse in 1998 and 1999 to evaluate the interference interactions among two rice cultivars and two red rice ecotypes. Plants were established in proportions of 3:0, 2:1, 1:2, and 0:3 (rice–red rice) plants/pot. Relative yield of Kaybonnet based on the shoot dry weight was lower than that of KatyRR or LA3, whereas PI 312777 was comparable to that of KatyRR and LA3. These results indicate that Kaybonnet was less competitive than PI 312777 when contrasted with KatyRR and LA3 red rice ecotypes. Kaybonnet (commercial rice cultivar) was dominated by both KatyRR (suspected rice × red rice cross) and LA3 (tall red rice ecotype) in tiller production, whereas PI 312777 (weed-suppressive cultivar) was comparable to either KatyRR or LA3. Both KatyRR and LA3 considerably reduced the leaf area of Kaybonnet. In contrast, PI 312777 reduced the growth of KatyRR, and its leaf area was comparable to that of LA3. The data suggest that high tillering capacity, as demonstrated by PI 312777, should be considered when breeding for rice cultivars that are competitive against weeds. This agronomic characteristic of rice may improve the success of reduced herbicide rate application programs.