Agricultural weeds are the most important biotic constraints to global crop production, and chief among these is weedy rice. Despite increasing yield losses from weedy rice in recent years worldwide, the genetic basis of weediness evolution remains unclear. Using whole-genome sequence analyses, we examined the origins and adaptation of Japanese weedy rice. We find evidence for a weed origin from tropical japonica crop ancestry, which has not previously been documented in surveys of weedy rice worldwide. We further show that adaptation occurs largely through different genetic mechanisms between independently-evolved temperate japonica- and tropical japonica-derived strains; most genomic signatures of positive selection are unique within weed types. In addition, some weedy rice strains have evolved through hybridization between weedy and cultivated rice with adaptive introgression from the crop. Surprisingly, introgression from cultivated rice confers not only crop-like adaptive traits (such as shorter plant height, facilitating crop mimicry) but also weedy-like traits (such as seed dormancy). These findings reveal how hybridization with cultivated rice can promote persistence and proliferation of weedy rice.

Field experiments involving eight cultivars were conducted in 1998 and 16 cultivars in 1999 to study the ability of rice ( Oryza sativa L.) to suppress Monochoria vaginalis (Burm. f) Kunth through light competition. Dry weights of M. vaginalis shoots in early season culture exceeded those in normal season culture of any rice cultivars. The relative photosynthetic photon flux density (R-PPFD), which was calculated as the ratio of the photosynthetic photon flux density (PPFD) below the rice canopy to that measured above the rice canopy, varied according to rice cultivar. A strong linear correlation was observed between the mean R-PPFD at 29-35 days after transplanting (DAT) ( r 2 = 0.80; p < 0.01 in 1998; r 2 = 0.63, p < 0.001; and r 2 = 0.93, p < 0.001 in 1999), or 36-42 DAT ( r 2 = 0.66, p < 0.05 in 1998; r 2 = 0.72, p < 0.001; and r 2 = 0.97, p < 0.001 in 1999), and the dry weight of M. vaginalis shoots at approximately 60 DAT. Data from the three experiments could be pooled into one regression line because intercepts and regression coefficients were not significantly different. The r 2 values of the combined regression were highest when R-PPFD was expressed as the mean of measurements taken during 14 days (from 29 to 42 DAT; r 2 = 0.81, p < 0.001). The shortest period for measuring mean R-PPFD in order to obtain a meaningful relationship with M. vaginalis shoot dry weight was 7 days (from 29 to 35 DAT; r 2 = 0.78, p < 0.001). For that same period, relationships between M. vaginalis shoot dry weight at 60 DAT and rice tiller number or leaf area index (LAI) at ground level were weak. However, there were negative relationships between M. vaginalis shoot dry weights at 60 DAT and rice LAI measured 20 cm above the ground, plant heights or rice shoot dry weight, but these coefficients of determination were smaller than those calculated by R-PPFD for the same period. Thus, the ability of rice to suppress M. vaginalis can be evaluated more accurately by measuring mean R-PPFD below the rice canopy for 7 days (from 29 to 35 DAT) than by measuring rice LAI, plant height and shoot dry weight.

No abstract

No abstract

No abstract