4.2 Multi-way gene flow among Oryza types
Gene flow analysis revealed significant gene flow among all sixOryza types (Table 1). Among the rice groups, O. nivarashowed higher sensitivity to introgression, and asymmetric gene flows were observed between wild and weedy rice groups, particularly betweenO. rufipogon and weedy rice (Table 1). The open panicle structure of O. rufipogon makes it more receptive to foreign pollen, resulting in higher gene flow from other Oryza groups. TheO. nivara populations studied were in natural habitats near commercial rice fields. Long-term gene flow from DWWC components toO. nivara populations may promote evolution and local adaptation, transitioning their reproductive system from high selfing to slight outcrossing, increasing its sensitivity to introgression.
In Sri Lanka, rice cultivation is carried out in two seasons,Maha and Yala , coinciding with the inter-monsoon and monsoon rains. Approximately 450,000 hectares of paddy are cultivated during the Yala season (March-September), while 850,000 hectares are cultivated in the Maha season (October-February) (Punyawardana, 2008). Most inbred rice varieties and landraces have a maturity duration of three to four and a half months, leading to overlap in flowering periods with wild rice, particularly during the Mahaseason. Concurrently, O. rufipogon being perennial, can be observed throughout the year, while the annual O. nivara seeds germinate at the beginning of the Maha season. Both species overlap in flowering time with inbred rice varieties. This increases the possibilities of pollen-mediated gene flow facilitated by monsoon wind currents and insects, contributing to admixtures between wild and weedy rice groups. Further, volunteer weeds may evolve into weedy rice over generations and interact with the DWWC components over time. Long-term introduction and persistence within the DWWC has resulted in complex multi-directional gene flow among its components, shaping the complex rice ecosystem in Sri Lanka (Vaughan et al ., 2005).
Our findings support higher gene flow from weedy rice to O. rufipogon (M=7.299) than vice versa (M=4.334), and significant gene flow was observed from cultivated rice to feral rice (M=8.1597) (Table 1), implying an early stage of de-domestication (Ellstrand et al., 2010). Pollen-mediated gene exchange occurs between crops and weeds or wild relatives and between weedy and wild relatives. Reverse gene flow from weed to the crop has been well documented (Langevinet al., 1990; Majumder et al., 1997; Song et al.,2002). The presence of wild Oryza and greater diversity of crop varieties and landraces in the region (Song et al., 2014) contribute to the evolutionary dynamics of South Asian weeds. Domesticated rice can mate with wild relatives, leading to gene flow that can influence the genetic structure of wild rice (Wang et al., 2017). Similarly, gene flow between cultivated carrots (Daucus carota ssp. sativus) and wild relatives (D. carotassp. carota) has caused adjacent wild populations to become genetically similar to the cultivars, contributing to the development of more aggressive weeds in carrot fields (Magnussen and Hauser, 2007).
Weedy rice in the U.S.A. is genetically distinct from inbred rice varieties due to de-domestication processes during its evolution and accidental introductions (Reagon et al., 2010; Li et al.,2017). It predominantly descends from wild rice populations in tropical Asia (Londo and Schaal, 2007; Reagon et al ., 2010), partly incurred by an extended flowering period and weak post-zygotic reproductive barrier, which promotes gene flow between cultivated and wild rice (Craig et al ., 2014). The gene flow patterns elucidated in the present study concerning weedy, wild, and cultivated rice corroborate the scenario (Table 1). Within the context of weedy rice origins and evolution in Sri Lanka, multiple factors exert influence, such as the feralization of cultivated rice, hybridization between landraces and/or inbred rice varieties, the adaptation of wild O. nivara , and the hybridization between domesticated and wild rice (Rieseberg et al ., 1993; Arnold, 1997; Rosenthal et al.,2008; Ellstrand, 2009; Ellstrand et al., 2010). Wedger et al . (2019) discovered that 23% of the studied weedy rice plants in Thailand exhibited introgressed alleles at one or more loci, derived from three domestication gene sequences of DWWC. Similarly, a preceding study done in Thailand (Wedger et al., 2019; Pusadee et al., 2013; Wongtamee et al., 2017) reported that 17.6% of 29 weedy rice plants showed evidence of introgression from wild or cultivated rice groups, indicating gene flow as a significant driving force for the evolutionary dynamics of the Oryza complex. Based on pairwise migration data, He et al . (2014) reported a significantly high level of gene flow within distinct weedy rice populations in Sri Lanka. Similarly, gene flow from Malaysian weedy rice to O. rufipogon substantiates the hypothesis positing potential introgression with other indigenousOryzas within the context of wild rice populations (San Sudoet al ., 2021). These studies indicate that weedy rice in the Asian region has undergone a complex evolutionary process involving gene flow with wild and cultivated rice, contributing to its genetic distinctiveness from cultivated rice varieties.