Introduction
The most problematic weedy species are those that are most like the crops they infest in terms of morphology, biology, and resource needs. Because weeds are generally more competitive than crops and are often better able to tolerate stress, crop-like weeds are the most difficult species to manage. Such is the case with weedy rice (Oryza spp.). Rice is grown in 120 countries (Figure 1), and weedy rice is a problem wherever rice is grown (Delouche et al. Reference Delouche, Burgos, Gealy, de San Martin, Labrada, Larinde and Rosell2007). Because the rice crop has been domesticated from wild Oryza species, that is, Oryza rufipogon Griff. in Asia and Oryza barthii A. Shiv. in Africa (Molina et al. Reference Molina, Sikora, Garud, Flowers, Rubinstein, Reynolds, Huang, Jackson, Schaal, Bustamante, Boyko and Purugganan2011), it is ecologically logical for the epicenters of rice cultivation to have weedy types that are admixtures of the wild relative, offtypes of the selected and cultivated strains, and various outcrosses, as the crop, weedy, and wild relative coexist in adjoining habitats (Pusadee et al. Reference Pusadee, Schaal, Schaal, Rerkasem, Rerkasem, Jamjod and Jamjod2013; Veltman et al. Reference Veltman, Flowers, Andel and Eric Schranz2019). Rice cultivation originated in the middle region of the Yangtze River basin in China about 8,000 YBP, or possibly as early as 14,000 YBP (Yasuda Reference Yasuda2008). It was deduced that this anthropological development followed the climate warming that favored the growth and expansion of wild rice progenitor populations. Ancient rice cultivation was also evident in the Indo-Gangetic Plain of India around 7,000 to 6,000 YBP (http://ricepedia.org/culture/history-of-rice-cultivation). Integrating genetics and historical linguistics into archaeological data revealed two centers of origin of rice cultivation in China—the middle and lower Yangtze valleys and a separate center of domestication in the Ganges Plain of India (Fuller et al. Reference Fuller, Sato, Castillo, Qin, Weisskopf, Kingwell-Banham, Song, Ahn and van Etten2010; Silva et al. Reference Silva, Stevens, Weisskopf, Castillo, Qin, Bevan and Fuller2015). Weedy rice would then eventually accompany rice culture, via seed-mediated gene flow, around the world into regions with or without wild rice species. Although rice cultivars produced around the world are primarily of Asian lineage and are descendants of O. rufipogon, other wild rice species with the same AA genome as Oryza sativa L. exist in Africa (Oryza longistaminata A. Chev. & Roehr. besides O. barthii A. Chev.), Australia (Oryza meridionalis N. Q. Ng), and Latin America (Oryza glumipatula Steud.) (Vaughan Reference Vaughan1994). As O. sativa cultivars were eventually introduced into Africa, the coexistence of the Asian rice with the locally domesticated rice Oryza glaberrima Steud. and the two wild rice species made the diversification of cultivated rice and weedy rice in Africa even more complex (Veltman et al. Reference Veltman, Flowers, Andel and Eric Schranz2019).
Figure 1. Global map of rice production, 2018 (https://ourworldindata.org/grapher/rice-production?time=2018).
Hybridization between crop, weedy populations, and genetically compatible wild rice species, together with de-domestication (Gealy et al. Reference Gealy, Agrama and Jia2012, Reference Gealy, Burgos, Yeater and Jackson2015; Goulart et al. Reference Goulart, Borba, Menezes and Merotto2014; Grimm et al. Reference Grimm, Sahi, Amann, Vidotto, Fogliatto, Devos, Ferrero and Nick2020; Hoyos et al. Reference Hoyos, Plaza, Li and Caicedo2020; Huang et al. Reference Huang, Young, Reagon, Hyma, Olsen, Jia and Caicedo2017; Kanapeckas et al. Reference Kanapeckas, Vigueira, Ortiz, Gettler, Burgos, Fischer and Lawton-Rauh2016; Subudhi et al. Reference Subudhi, Singh, DeLeon, Parco, Karan, Biradar, Cohn and Sasaki2014; Sun et al. Reference Sun, Qian, Ma, Xu, Liu, Du and Chen2013; Xia et al. Reference Xia, Wang, Xia, Zhao and Lu2011), would spur extensive diversification of weedy populations wherever rice is grown. Morphological and genetic studies of weedy rice around the globe reveal that after having been introduced into areas without wild rice, the introduced weedy types had diversified greatly after centuries of coexistence (Fogliatto et al. Reference Fogliatto, Vidotto and Ferrero2012; Kanapeckas et al. Reference Kanapeckas, Tseng, Vigueira, Ortiz, Bridges, Burgos, Fischer and Lawton-Rauh2018; Shivrain et al. Reference Shivrain, Burgos, Scott, Gbur, Estorninos and McClelland2010).
Weedy rice is a global problem in modern rice culture that is threatening food security (Ziska et al. Reference Ziska, Gealy, Burgos, Caicedo, Gressel, Lawton-Rauh, Avila, Theisen, Norsworthy, Ferrero, Vidotto, Johnson, Ferreira, Marchesan, Menezes and Sparks2015). Its persistence is aided by the high shattering of grains (Nunes et al. Reference Nunes, Delatorre and Merotto2014; Shivrain et al. Reference Shivrain, Burgos, Scott, Gbur, Estorninos and McClelland2010; Thurber et al. Reference Thurber, Reagon, Gross, Olsen, Jia and Caicedo2010) coupled with protracted germination behavior (Shivrain et al. Reference Shivrain, Burgos, Gealy, Smith, Scott, Mauromoustakos and Black2009b) and long life span and deep dormancy of seeds (Baek and Chung Reference Baek and Chung2012; Fogliatto et al. Reference Fogliatto, Vidotto and Fererro2011; Goss and Brown Reference Goss and Brown1939; Tseng et al. Reference Tseng, Burgos, Shivrain, Alcober, Mauromoustakos and Andersson2013). Weedy rice is managed traditionally by transplanting rice seedlings into puddled fields, followed immediately by permanent flooding and hand weeding as needed. Centuries of hand weeding produced weedy rice populations that are morphologically most similar to cultivated rice, making the weed very difficult to distinguish, leaving several escapes to replenish the seedbank. Also, water shortages and human migration to urban areas spurred the development and large-scale adoption of direct seeding and related technologies to make rice production more efficient and economical. Weedy rice infestation rose with the adoption of direct seeding in Japan (Imaizumi Reference Imaizumi2018), Italy (Ferrero and Vidotto Reference Ferrero and Vidotto1998), Malaysia (Abdullah et al. Reference Abdullah, Vaughan, Watanabe and Okuno1996), Thailand (Maneechote et al. Reference Maneechote, Jamjod and Rerkasem2004), the Philippines (Juliano et al. Reference Juliano, Donayre, Martin and Beltran2020), Vietnam (Chauhan et al. Reference Chauhan, Namuco, Ocampo, Son, Thu, Nam, Phuong and Bajwa2015), and other countries in Asia and elsewhere (Chauhan Reference Chauhan2012; Shekhawat et al. Reference Shekhawat, Rathore and Chauhan2020).
Rice yield losses from weedy rice interference could range from 10% to 90% depending on the weedy rice type, rice cultivar, duration of competition, level of infestation, and environment (Table 1). The type of cultivated rice has a large influence on the magnitude of yield loss from weedy rice competition. Hybrid rice is generally more competitive than inbred rice cultivars, and contemporary varieties are more competitive than old ones (Ottis et al. Reference Ottis, Smith, Scott and Talbert2005; Shivrain et al. Reference Shivrain, Burgos, Gealy, Smith, Scott, Mauromoustakos and Black2009b). In addition to loss of production, high weedy rice infestation results in reduction of farm gate price of rice due to seed contamination (Ottis et al. Reference Ottis, Smith, Scott and Talbert2005). There are cases where farmers have had to abandon their fields because of severe weedy rice infestation. The great need for alternative tools to manage weedy rice spurred the development and commercialization of non-transgenic herbicide-resistant (HR) rice technologies, Clearfield® rice in 2002 and ProvisiaTM rice in 2018 (Anonymous 2018; Tan et al. Reference Tan, Evans, Dahmer, Singh and Shaner2005). The former carries resistance to imidazolinone (Group 2) herbicides and the latter to aryloxyphenoxypropanoates (Group 1). The Clearfield® rice technology has met unprecedented level of adoption in the Americas, Europe, and Asia wherever it was introduced (Sudianto et al. Reference Sudianto, Song, Neik, Saldain, Scott and Burgos2013, Reference Sudianto, Song, Neik, Saldain, Scott and Burgos2014). Despite stringent stewardship policies based on scientific data, it was inevitable that HR weedy rice would evolve, for various controllable and uncontrollable reasons. In fact, HR weedy rice populations had evolved in all countries where Clearfield® rice is grown due to pollen- and seed-mediated gene flow, thus reducing the utility of the technology in many places. Rice growers are greatly anticipating ProvisiaTM rice to remedy the problem. The adoption of this new technology is limited only by the low yield potential of current ProvisiaTM varieties (Roma-Burgos et al. Reference Roma-Burgos, Butts, Bottoms and Werle2021). This short review discusses the origin of weedy rice, gene exchange between crop and weed and its impact on the weedy populations, nomenclature associated with weedy rice in the literature, and our recommendation for a unified, scientifically supported nomenclature for weedy rice.
Table 1. Rice yield losses due to weedy rice.
a In competition studies, the duration of competition is generally established from planting and allowed to grow with rice the whole season. In some cases, weedy rice is allowed to compete with rice for only a short period to determine the effect of early-season competition.
The Rise of Weedy Rice
Four non–mutually exclusive hypotheses have been proposed to explain weedy rice origins, as supported by recent genetic studies: (1) de-domestication: feralization of cultivated rice into the weedy form (e.g., Ellstrand et al. Reference Ellstrand, Heredia, Leak-Garcia, Heraty, Burger, Yao, Nohzadeh-Malakshah and Ridley2010; Kanapeckas et al. Reference Kanapeckas, Vigueira, Ortiz, Gettler, Burgos, Fischer and Lawton-Rauh2016; Li et al. Reference Li, Li, Jia, Caicedo and Olsen2017; Qiu et al. Reference Qiu, Zhou, Mao, Ye, Wang, Zhang, Yu, Fu, Wang, Qian, Qi, Wu, Sultana, Cao and Wang2017); (2) intervarietal hybridization: weed emergence via hybridization between different cultivated rice varieties (e.g., Ishikawa et al. Reference Ishikawa, Toki, Imai, Sato, Yamagishi, Shimamoto, Ueno, Morishima and Sato2005; Pusadee et al. Reference Pusadee, Schaal, Schaal, Rerkasem, Rerkasem, Jamjod and Jamjod2013; Qiu et al. Reference Qiu, Zhu, Fu, Ye, Wang, Mao, Lin, Chen, Zhang, Guo, Qiang, Lu and Fan2014); (3) wild-derived origin: incidental selection on the wild ancestor of rice (O. rufipogon, known as common wild rice) or possibly other wild Oryza species for adaptation to and invasion of rice-planting areas (e.g., Huang et al. Reference Huang, Young, Reagon, Hyma, Olsen, Jia and Caicedo2017); and (4) crop–wild hybridization: weed emergence following hybridization between domesticated and wild rice (e.g., Pusadee et al. Reference Pusadee, Schaal, Schaal, Rerkasem, Rerkasem, Jamjod and Jamjod2013; Song et al. Reference Song, Chuah, Tam and Olsen2014). Under scenarios 1 and 2, cultivated rice is the sole progenitor of weedy strains, whereas scenarios 3 and 4 both invoke some role for wild Oryza in the origins and evolution of weedy rice. This difference has potential taxonomic implications, as discussed in the final section.
Many rice-growing regions of the world, including those both within and outside the geographic range of wild Oryza, have two morphologically distinct weedy rice types in terms of seed characteristics: some have seeds resembling cultivated rice (i.e., strawhull awnless, or SH) and others are more wild-like (i.e., blackhull awned, or BHA) (Akasaka et al. Reference Akasaka, Ushiki, Iwata, Ishikawa and Ishii2009; Cao et al. Reference Cao, Lu, Xia, Rong, Sala, Spada and Grassi2006; Chung and Park Reference Chung and Park2010; Grimm et al. Reference Grimm, Fogliatto, Nick, Ferrero and Vidotto2013; Reagon et al. Reference Reagon, Thurber, Gross, Olsen, Jia and Caicedo2010; Shivrain et al. Reference Shivrain, Burgos, Scott, Gbur, Estorninos and McClelland2010; Zhang et al. Reference Zhang, Dai, Wu, Song and Qiang2012; Zhu et al. Reference Zhu, Ellstrand and Lu2012). Because wild Oryza species only occur in a subset of the world’s rice production regions, the following sections first examine weedy rice evolution in regions without wild Oryza, followed by areas where rice crops are planted in proximity to interfertile wild Oryza populations.
Regions without Wild Oryza Species
Many of the world’s major rice production areas occur outside the range of wild Oryza species; these include temperate Asia (northern China, South Korea, and Japan), North America, and Europe. According to a recent global weedy rice population genomics study (Qiu et al. Reference Qiu, Jia, Wu, Weng, Chen, Sun, Chen, Mao, Jiang, Ye, Turra, Guo, Ye, Zhu and Imaizumi2020), 112 of the 163 samples (68.7%) collected from regions without wild rice were morphologically characterized as either SH or strawhull awned (SHA), whereas 51 (31.3%) were either wild-like black/brownhull awnless or black/brownhull awned. Most studies of weedy rice origins in these regions support evolution by de-domestication (scenario 1), with independent origins of the morphologically distinct weed forms (reviewed by Li and Olsen Reference Li and Olsen2020). While blackhull awnless (BH) and BHA weed forms in these regions have seeds with morphological resemblance to wild Oryza species, the presence of the wild-like traits appears to reflect weed descent from unimproved crop landraces (where features such as awns and dark-pigmented hulls can still be found) rather than any direct role for wild rice.
In both China and the United States, where genetics and evolution of weedy rice have been particularly well documented, whole-genome resequencing analyses indicate a clear role for de-domestication of cultivated ancestors in weedy rice origins (Li et al. Reference Li, Li, Jia, Caicedo and Olsen2017; Qiu et al. Reference Qiu, Jia, Wu, Weng, Chen, Sun, Chen, Mao, Jiang, Ye, Turra, Guo, Ye, Zhu and Imaizumi2020). In the southern United States, the genetic makeup of the two predominant weedy rice morphotypes (SH and BHA) are respectively closest to indica and aus rice. As neither of these crop varieties are commercially cultivated in the country, it is likely that the U.S. weedy rice forms originally evolved in Asia and were later introduced through contaminated seed stocks (Reagon et al. Reference Reagon, Thurber, Gross, Olsen, Jia and Caicedo2010). Other non-neutral markers, including adaptive genes influencing fitness (sh4, conferring shattering; Bh4, controlling hull color; Rc, conferring pericarp color and seed dormancy), further substantiate the notion that U.S. weeds are genetically close to rice cultivars or landraces (Gross et al. Reference Gross, Reagon, Hsu, Caicedo, Jia and Olsen2010; Thurber et al. Reference Thurber, Reagon, Gross, Olsen, Jia and Caicedo2010; Vigueira et al. Reference Vigueira, Li and Olsen2013).
Interestingly, in China, many of the most prevalent contemporary weeds appear to have evolved through de-domestication from widely grown twentieth-century cultivars, including ‘Huk Zo’ (temperate japonica) and ‘Nanjing11’ (indica) (Qiu et al. Reference Qiu, Jia, Wu, Weng, Chen, Sun, Chen, Mao, Jiang, Ye, Turra, Guo, Ye, Zhu and Imaizumi2020). For both U.S. and Chinese weeds, genomic analyses have revealed that the evolution of weedy adaptations during de-domestication has generally not involved genetic changes at loci that were targets of selection during rice domestication (Li et al. Reference Li, Li, Jia, Caicedo and Olsen2017; Qiu et al. Reference Qiu, Jia, Wu, Weng, Chen, Sun, Chen, Mao, Jiang, Ye, Turra, Guo, Ye, Zhu and Imaizumi2020). This finding points to fundamentally different genetic mechanisms underlying rice domestication versus de-domestication.
Intervarietal hybridization (scenario 2) has been found to be the main mechanism underlying weedy rice evolution in some parts of world without wild Oryza species. For example, in northeastern China and Bhutan, weedy populations were found to have evolved from indica–japonica hybridizations, creating admixed genomic compositions (Ishikawa et al. Reference Ishikawa, Toki, Imai, Sato, Yamagishi, Shimamoto, Ueno, Morishima and Sato2005; Qiu et al. Reference Qiu, Zhu, Fu, Ye, Wang, Mao, Lin, Chen, Zhang, Guo, Qiang, Lu and Fan2014). In South Korea, similar observations of hybridization between modern cultivars and traditional landraces have been reported (He et al. Reference He, Kim and Park2017).
Regions with Wild Oryza Species
Apart from the two domesticated rice species, O. sativa and the lesser-known African rice, O. glaberrima, there are six wild Oryza species that are phylogenetically close to cultivated rice and share the AA genome that potentially allows some degree of interfertility: O. rufipogon, Oryza nivara S.D. Sharma & Shastry, the annual form of O. rufipogon that is sometimes given its own taxonomic standing), O. barthii (the wild ancestor of African cultivated rice), O. longistaminata, O. meridionalis, and O. glumaepatula (Khush Reference Khush1997). Among these wild species, O. rufipogon is distributed widely throughout the tropics and subtropics in Asia from western India through Southeast Asia and southern China. The remaining wild rice species are found in Africa, tropical Australia, and South and Central America (Oka Reference Oka1988; Vaughan et al. Reference Vaughan, Morishima and Kadowaki2003).
In regions where wild Oryza species are found in proximity to rice-planting areas, most weedy rice evolutionary studies have focused on weed populations where common wild rice (O. rufipogon) is present. To date, there is no definitive genetic evidence that any wild Oryza other than O. rufipogon has contributed to the evolution of weedy rice. Common wild rice has been found to contribute to the evolution of sympatric weedy rice in South and Southeast Asia (Huang et al. Reference Huang, Young, Reagon, Hyma, Olsen, Jia and Caicedo2017; Neik et al. Reference Neik, Chai, Tan, Sudo, Cui, Jayaraj, Teo, Olsen and Song2019; Pusadee et al. Reference Pusadee, Schaal, Schaal, Rerkasem, Rerkasem, Jamjod and Jamjod2013; Song et al. Reference Song, Chuah, Tam and Olsen2014; Sudianto et al. Reference Sudianto, Neik, Tam, Chuah, Idris, Olsen and Song2016; Vigueira et al. Reference Vigueira, Qi, Song, Li, Caicedo, Jia and Olsen2019; Wedger et al. Reference Wedger, Pusadee, Wongtamee and Olsen2019).While there is limited evidence in South Asia for weedy rice evolving directly from O. rufipogon (scenario 3) (Huang et al. Reference Huang, Young, Reagon, Hyma, Olsen, Jia and Caicedo2017), most studies have proposed that weedy rice in these tropical and subtropical regions is evolving through crop–wild hybridization (scenario 4) (Cui et al. Reference Cui, Song, Li, Li, Huang, Caicedo, Jia and Olsen2016; Neik et al. Reference Neik, Chai, Tan, Sudo, Cui, Jayaraj, Teo, Olsen and Song2019; Prathepha Reference Prathepha2009; Pusadee et al. Reference Pusadee, Schaal, Schaal, Rerkasem, Rerkasem, Jamjod and Jamjod2013; Song et al. Reference Song, Chuah, Tam and Olsen2014; Wedger et al. Reference Wedger, Pusadee, Wongtamee and Olsen2019).
In Malaysia and Thailand, where the occurrence and spread of weedy rice have been extensively studied, simple sequence repeats and genome-wide single-nucleotide polymorphisms (SNPs) revealed natural gene flow and hybridization between common wild rice and cultivated rice (Neik et al. Reference Neik, Chai, Tan, Sudo, Cui, Jayaraj, Teo, Olsen and Song2019; Pusadee et al. Reference Pusadee, Schaal, Schaal, Rerkasem, Rerkasem, Jamjod and Jamjod2013; Song et al. Reference Song, Chuah, Tam and Olsen2014; Vigueira et al. Reference Vigueira, Qi, Song, Li, Caicedo, Jia and Olsen2019; Wedger et al. Reference Wedger, Pusadee, Wongtamee and Olsen2019). Such similar crop–wild intercrossing was also found in South Asia (Huang et al. Reference Huang, Young, Reagon, Hyma, Olsen, Jia and Caicedo2017). The observations of a heterozygous state for several adaptive genes, including sh4, Bh4, Rc, and An-1 (controlling awn development), found in Malaysian weed samples further support crop–wild hybridization evolutionary mechanisms in weedy rice populations (Cui et al. Reference Cui, Song, Li, Li, Huang, Caicedo, Jia and Olsen2016; Neik et al. Reference Neik, Chai, Tan, Sudo, Cui, Jayaraj, Teo, Olsen and Song2019; Song et al. Reference Song, Chuah, Tam and Olsen2014). Interestingly, these studies also revealed that the genomes of the Malaysian common wild rice and modern elite cultivars have both been recently introgressed into the local weedy rice populations. Particularly, the brown-striped-like (BR-like) and BHA-like weedy rices have been found to harbor abundant segments of the common wild rice genome. On the other hand, wild introgression detected at domestication genes of the Thai weedy rice samples were likely to be historical gene flow rather than recent introgression (Wedger et al. Reference Wedger, Pusadee, Wongtamee and Olsen2019).
While weedy rice from the tropics and subtropics in Asia has been extensively studied using molecular markers, limited weedy rice population studies have been conducted in regions where wild Oryza other than O. rufipogon is present. For instance, in Central and South America, where rice is planted in proximity to the interfertile wild rice O. glumaepatula, gene flow has been speculated to play a role in local weedy rice evolution. Nevertheless, genome-wide SNP analyses found no evidence for contributions from this native wild Oryza to the co-occurring weedy populations (Hoyos et al. Reference Hoyos, Plaza, Li and Caicedo2020; Qiu et al. Reference Qiu, Jia, Wu, Weng, Chen, Sun, Chen, Mao, Jiang, Ye, Turra, Guo, Ye, Zhu and Imaizumi2020). Instead, the evolutionary trajectory of these Latin American weeds is similar to those of weed populations from the temperate regions without wild Oryza species, wherein intervarietal hybridization between aus and indica is predominant (Hoyos et al. Reference Hoyos, Plaza, Li and Caicedo2020; Qiu et al. Reference Qiu, Jia, Wu, Weng, Chen, Sun, Chen, Mao, Jiang, Ye, Turra, Guo, Ye, Zhu and Imaizumi2020). In Africa, O. barthii and O. longistaminata have been reported to grow sympatrically with cultivar planting areas. As such, the occurrence of weedy rice, some of which have the wild-like feature of red pericarps, may point toward the role of these wild African Oryza species as the weeds’ origin (Delouche et al. Reference Delouche, Burgos, Gealy, de San Martin, Labrada, Larinde and Rosell2007). Comprehensive analysis of genomic data resolved five subpopulations of wild rice in Africa with which O. glaberrima grown in some regions singularly clustered (Wang et al. Reference Wang, Yu, Haberer, Marri, Fan, Goicoechea, Zuccolo, Song, Kudrna, Ammiraju, Cossu, Maldonado, Chen, Lee and Sisneros2014). African weedy rice is, consequentially, a mixture of wild and cultivated species in the region (Orjuela et al. Reference Orjuela, Sabot, Chéron, Vigouroux, Adam, Chrestin, Sanni, Lorieux and Ghesquière2014) and genetically distinct from Asian weedy rice, which descended from O. rufipogon or O. sativa cultivars. Local adaptation, or ecotypic evolution, is yet another means by which rice types could diverge, as demonstrated by the polycentric, geographically influenced divergence of African rice (Meyer et al. Reference Meyer, Choi, Sanches, Plessis, Flowers, Amas, Dorph, Barretto, Gross, Fuller, Bimpong, Ndjiondjop, Hazzouri, Gregorio and Purugganan2016; Veltman et al. Reference Veltman, Flowers, Andel and Eric Schranz2019). Coastal rice, for example, is adapted to a high salt environment and could give rise to a distinct ecotype of weedy rice. Thus, although wild AA genome Oryza species have some degree of interfertility with cultivated rice, they may not contribute to the occurrence of the weedy rice in Latin America and Africa. What has been confirmed, however, is that the closest wild relative of rice, O. rufipogon, does play an important role in the evolution and adaptation of weedy rice in South and Southeast Asian regions.
Crop–Weed–Wild Relative Gene Flow and Weed Population Diversification
The genetic structure of weedy rice as it relates to the crop or wild rice is supported by field-collected gene flow rate data. Numerous pollen flow experiments have been conducted using herbicide-resistance traits as markers to estimate detectable gene flow in “real time.” The gene flow rate from crop to weed is generally low, ranging from 0% to 1.26% (Table 2). The majority of weedy rice used in outcrossing experiments were indica strains. In two studies conducted in China and Korea where the pollen source was an indica cultivar and the recipients were different accessions of japonica weedy rice, the outcrossing rate ranged from 0% to 0.046% (Chen et al. Reference Chen, Lee, Song, Suh and Lu2004). Gene flow data from the United States were derived from indica and aus weedy rice (strawhull and blackhull) with generally tropical japonica rice cultivars as pollen donors. Reverse gene flow from weedy rice to cultivars could range from 0.01% to 0.2% in Arkansas, USA (Shivrain et al. Reference Shivrain, Burgos, Gealy, Sales and Smith2009a). Two exceptions are notable. In Malaysia, the outcrossing rate between the two Clearfield® rice varieties and four unspecified weedy rice accessions ranged from 7% to 28%, averaged across distances of 1 to 5 m from the pollen source (Engku et al. Reference Engku, Mazlan, Juraimi, Rafii, Abdullah and Alam2016). These numbers are the highest on record, globally. Another notable situation is a study conducted in three localities in China (Xia et al. Reference Xia, Wang, Xia, Zhao and Lu2011). Xia and colleagues collected 11 weedy rice accessions from rice fields infested with 2 to 60 weedy rice plants m−2 and determined that the outcrossing rates between these accessions and various rice cultivars ranged from 0.4% to 11.7%. In one study in China where the recipient was wild rice (O. rufipogon) and the pollen donor was an indica cultivar and a glufosinate-resistant transgenic line, the detectable pollen flow was 1.2% to 2.2% (Chen et al. Reference Chen, Lee, Song, Suh and Lu2004). Conversely, O. rufipogon can hybridize with rice cultivars, as demonstrated in controlled crosses with four Thai cultivars as female parents, resulting in 6% to 62% seed set (Niruntrayakul et al. Reference Niruntrayakul, Rerkasem and Jamjod2009). These studies show that pollen flow occurs between crop and weed and between crop and wild relative and vice versa. By extension, one can also expect gene exchange between weedy rice and wild relatives. In rice fields in the United States, populations of tall, pink-awned weedy rice are occasionally spotted. These are outcrosses between blackhull weedy rice and tropical japonica rice cultivars (Vigueira et al. Reference Vigueira, Burgos, Lawton-Rauh and Novak2018) and are potentially bridging gene exchange between the crop and other weedy plants.
Table 2. Some gene flow statistics from cultivated rice to weedy rice.
a CL, Clearfield® rice.
b U.S. weedy rice in these studies was strawhull or blackhull indica.
The lineage of weedy rice, the continuous interaction between crop and weed or crop, weed, and wild relative, and the evolution of some weedy traits from de-domesticated cultivars have led to the remarkable diversification of weedy rice populations regionally and globally. Weedy rice differs from cultivated rice in many morphological (i.e., height; tiller number; color of leaf, hull, and kernel pericarp; awnedness; grain traits) and physiological (i.e., seed dormancy, emergence, flowering time, grain shattering, stress tolerance, photosynthetic capacity) traits. As a consequence of the gene flow continuum, weedy rice populations have also diversified greatly in various regions with or without wild rice. Weedy rice diversity has been documented in various countries in Asia (Chauhan and Johnson Reference Chauhan and Johnson2010; Prathepha Reference Prathepha2009; Pusadee et al. Reference Pusadee, Schaal, Schaal, Rerkasem, Rerkasem, Jamjod and Jamjod2013; Sudianto et al. Reference Sudianto, Neik, Tam, Chuah, Idris, Olsen and Song2016), Colombia (Hoyos et al. Reference Hoyos, Plaza and Caicedo2019), Costa Rica (Arrieta-Espinoza et al. Reference Arrieta-Espinoza, Sánchez, Vargas, Lobo, Quesada and Espinoza2005), Italy (Fogliatto et al. Reference Fogliatto, Vidotto and Ferrero2012), the United States (Kanapeckas et al. Reference Kanapeckas, Tseng, Vigueira, Ortiz, Bridges, Burgos, Fischer and Lawton-Rauh2018; Karn et al. Reference Karn, De Leon, Espino, Al-Khatib and Brim-DeForest2020; Noldin et al. Reference Noldin, Chandler and McCauley1999; Shivrain et al. Reference Shivrain, Burgos, Scott, Gbur, Estorninos and McClelland2010; Tseng et al. Reference Tseng, Burgos, Shivrain, Alcober, Mauromoustakos and Andersson2013), and other parts of the world (Delouche et al. Reference Delouche, Burgos, Gealy, de San Martin, Labrada, Larinde and Rosell2007). Globally, strawhull weedy rice is the predominant type. In Asia, awned and awnless strawhull populations occur in high proportions, but in North America, strawhull weedy rice is predominantly awnless. Brownhulls and grayhulls also exist but occur less frequently. Generally, weedy rice populations are taller than cultivars and have an open canopy, droopy leaves, high tillering potential, a wide range of flowering phenologies, open panicles, high seed shattering, red pericarps, and high seed dormancy. However, closer inspection reveals great variability in these and other traits within the strawhull or blackhull group. Vegetative and reproductive traits vary significantly within each hull-color group, creating subgroups based on various combinations of certain key variant traits, such as plant height, maturation, grain shattering, awn length, pericarp color, or grain size. These generalizations also obscure the fact that some weedy rice types are as short or shorter than the cultivars they infest, flower earlier, and have crop-like grain traits (i.e., long grain with white pericarp, low shattering, not dormant). Atypical weedy rice types may arise from weed–weed gene exchange. For example, genetic analysis of short-awned, intermediate-height weedy rice in Arkansas, USA, indicated that such types could have arisen from hybridization between awned and awnless weedy rice (Gealy et al. Reference Gealy, Agrama and Jia2012). Apart from pollen-mediated gene flow, crop-like weedy rice could have arisen from feralization of cultivars planted, as was the case of most weedy rice in Taizhou, China (Zhang et al. Reference Zhang, Burgos, Ma, Zhou, Geng and Yu2008); rice in abandoned fields of Sri Lanka (Marambe Reference Marambe and Gressel2005); or the first-detected weedy rice in California, USA, after decades of having eradicated the weed (Kanapeckas et al. Reference Kanapeckas, Vigueira, Ortiz, Gettler, Burgos, Fischer and Lawton-Rauh2016).
The hybridization of HR rice with weedy populations, and subsequent selection with the technology-associated herbicides, has further expanded weedy rice diversity. In the southern United States, for example, where the genetic population structure of weedy rice historically consisted of distinct strawhull and blackhull populations, it presently consists of diversified populations of HR strawhull, blackhull, brownhull, and other hull colors containing a significant portion of the crop genome (Burgos et al. Reference Burgos, Singh, Tseng, Black, Young, Huang, Hyma, Gealy and Caicedo2014). The majority of HR weedy rice in Arkansas rice fields with a history of Clearfield® rice are crop-like in terms of plant morphology, flowering phenology, and seed dormancy (Singh et al. Reference Singh, Singh, Black, Boyett, Basu, Gealy, Gbur, Pereira, Scott, Caicedo and Burgos2017). Besides having grains with a variety of hull colors, the HR weedy rice populations also have a range of pericarp colors from red, brown, and light brown to white (Singh et al. Reference Singh, Singh, Black, Boyett, Basu, Gealy, Gbur, Pereira, Scott, Caicedo and Burgos2017), in contrast to historically being all red (Shivrain et al. Reference Shivrain, Burgos, Scott, Gbur, Estorninos and McClelland2010). The hidden consequence of such “disguised” weedy plants is yield loss from competition and loss of grain quality from seed contamination. Overall, this kaleidoscope of traits (crop-like, wild-like, and all sorts in between) is reflected in the confused nomenclature of weedy rice in the literature.
Names Associated with Weedy Rice
As discussed in the previous section, worldwide, weedy rice can be morphologically and genetically classified into two main lineages: crop- and wild-like weedy rice. Based on a literature search, we learned that regardless of the types of weedy rice, the weed has been often referred to as O. sativa. The other commonly used term to describe weedy rice is “O. sativa f. spontanea,” which technically indicates that O. rufipogon, but not other wild species, has contributed wild genetic material to the weedy rice strain. These two names to describe weedy rice could be misleading if the forms and origins of the weeds have not been confirmed. The other less commonly used term is “Oryza spp.,” which refers collectively to multiple undetermined Oryza species.
To determine the usage trend of these scientific names, a Boolean search was conducted in the Web of Science database using the key terms “weedy rice,” “spontanea,” “Oryza spp.,” and “Oryza sativa”; the fields searched were abstract, key words, and title. We categorized the weedy rice studies based on the scientific name used and different time frames (1991 to 2000, 2001 to 2010 and 2011 to 2020). The proportion of weedy rice studies that used O. sativa, Oryza spp., and O. sativa f. spontanea was quantified for the different time frames (Table 3; Figure 2). “Oryza sativa,” as expected, was the most extensively used term in weedy rice studies, while “Oryza spp.” was the least used. Interestingly, the proportion of O. sativa f. spontanea increased from 10.2% (12 papers) in 2001 to 2010 to 17.9% (55 papers) in 2011 to 2020, signifying a growing use of “spontanea” in the past two decades.
Table 3. Scientific names associated with weedy rice in journal articles on the Web of Science and their frequency of use.
Figure 2. Scientific names associated with weedy rice in the literature between 1991 and 2020.
Oryza sativa f. spontanea refers to the “spontanea” type of O. sativa, which is Oryza rufipogon wild-like weedy rice, or a hybrid between wild O. rufipogon and O. sativa, particularly the type found in Asia (Chang 1975, Reference Chang1976a, 1976b; Lu et al. Reference Lu, Naredo, Juliano and Jackson1997; Morishima et al. Reference Morishima, Oka and Chang1961; Oka and Chang Reference Oka and Chang1959). When the term “O. sativa f. spontanea” is used to describe weedy rice in general, it implies that all weedy rice populations originate from the common wild rice (O. rufipogon) or is a product of a wild-cultivated rice hybrid. However, 39 out of the 45 papers (86.7%) we could access that used O. sativa f. spontanea, reported cultivar-like weedy rice or weedy rice from regions where wild rice does not exist. This observation demonstrated the misuse of O. sativa f. spontanea in weedy rice studies, which is problematic, as the scientific name of an organism should be precise and true to the evidence provided, allowing people to communicate about the organism without confusion.
Another notable issue with the term “O. sativa f. spontanea” is that the it does not inherently indicate the involvement of O. rufipogon. Readers might mistakenly assume that O. sativa f. spontanea can be a direct descendant of O. sativa, although in fact some role for O. rufipogon is required for weedy rice to be correctly classified as O. sativa f. spontanea. Moreover, a thorough check of various plant databases, including the International Plant Names Index and the World Checklist of Selected Plant Families, also revealed that O. sativa f. spontanea is not accepted by some databases due to its ambiguous characterizations (Govaerts Reference Govaerts2019; IPNI 2020).
The use of the term “O. sativa” is less problematic but still not universally appropriate, as it gives the impression that O. sativa is the sole contributor to emergence of weedy rice. Given that O. sativa is not the only Oryza species that contributes to the emergence of weedy rice, one should not assume that all weedy rice is O. sativa. We acknowledge that the use of O. sativa to describe weedy rice is not wrong in most contexts, as de-domestication of cultivars is the primary contributor to most of the weeds. However, it becomes problematic when O. sativa is used to describe worldwide weedy rice indiscriminately. The notions and suggestions mentioned in this section are also endorsed by the experts in the field, as listed in Supplementary Table 1.
Conclusion and Recommendation
The common usage of the term “O. sativa” or “O. sativa f. spontanea” in describing weedy rice has confounded the discrimination between crop- and wild-like forms of the weed and a broad spectrum of all other intermediates. Such generalization in naming creates confusion and hinders the weedy rice community from effective scientific communication. In regions where there are no wild rice species present, and where O. sativa is the only Oryza species contributing to the evolution of weedy rice, it is reasonable to use the term “Oryza sativa” for describing crop-like weedy rice. This, however, may not apply to weedy rice in Africa, where rice-growing areas are planted with both O. glaberrima and O. sativa cultivated rice species and where two wild rice species unique to Africa exist. The term should also not be used for wild-like weedy rice introduced into the non–wild rice regions. For regions that have wild Oryza species, the usage of “O. sativa f. spontanea” for weedy rice implies that the rice possesses O. rufipogon characteristics; hence, this name is not appropriate for crop-like weedy rice in those localities, either. Furthermore, not all databases recognize “O. sativa f. spontanea” as a valid taxonomic designation, which may lead to further confusion. Therefore, we advise the use of the more general name, “Oryza spp.,” when describing worldwide weedy rice in general or when the evolutionary history of the weed strain studied is unresolved. As such, authors should be diligent in considering the names to be used when describing weedy rice populations. The precise usage of scientific name not only averts confusion, but also enhances correct understanding by the audience, who may be unfamiliar with weedy rice evolution and biology.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/wsc.2021.22
