Genetic diversity is not only essential for crop resilience in diverse environments but also fundamental to genetic improvement. Enhancing the genetic diversity of breeding pools that carry optimum combinations of favorable alleles for targeted crop-growing regions is crucial for sustaining genetic gain (Bohra et al., 2022). The evolutionary process of crops typically manifests as dynamic changes in genetic diversity, driven by major forces such as migration, selection, genetic drift, reproductive barriers, and polyploidization (Ladizinsky, 1998; Wu et al., 2020; Cao et al., 2023). Migration is the movement of genes within and between populations. When immigrants from other populations bring unique genes, they can increase the genetic variability of the recipient population. When a plant species is domesticated and becomes a crop to sustain human populations, long-term accumulation of genetic diversity occurs through genetic mutation and migration, leading to diffusion and regional expansion of the crop (Ladizinsky, 1998; Hancock, 2012). Genetic improvement is a co-evolutionary process between human civilization and crop species, where new genetic blocks with beneficial characteristics are continuously selected to meet the evolving demands for improved cultivation conditions and higher yields, in order to support the growing human population. Artificial hybridization and selection allow favorable genetic variants dispersed across different populations to be combined into new genetic combinations, and this introgression process accelerates crop evolution. In Asian cultivated rice (Oryza sativa L.), the recognition that it consists of two subspecies or groups (Kato et al., 1928) has long been used to guide rice breeding practices aimed at exploiting beneficial genetic diversity and heterosis. The success of breeding programs aimed at genetic improvement of temperate japonica rice in Northeast China since the 1950s (Yang et al., 1962) has been demonstrated through inter-subspecific introgression, resulting in significant increases in indica-allele frequencies in the cultivars bred after 1990 (Sun et al., 2012). indica-japonica heterosis, which was expected to show great potential, has been supported by various studies (Yuan, 1987; Fu et al., 2014; Birchler, 2015; Zhang, 2020; Ouyang et al., 2022; Wang et al., 2022a). Advances in genomics have provided deeper insights into the genetic divergence and diversity within the Asian cultivated rice (Garris et al., 2005; McNally et al., 2009; Huang et al., 2012; Xie et al., 2015; Wang et al., 2018; Lin et al., 2020; Qin et al., 2021; Wang et al., 2022b; Ye et al., 2022), highlighting the role of introgression among subgroups and offering a more comprehensive understanding of the genetic makeup and evolutionary history of this crop.

However, the progress of human society and population growth are exerting increasing pressure on rice production and productivity. Therefore, accurately and effectively utilizing the dispersed and rich variations in the Asian cultivated rice to develop new varieties that can address constantly emerging problems in rice production, such as climate change, water scarcity, overuse of fertilizers, insecticides, antibiotics, and herbicides, and thus reduce the cost of rice production and environmental pollution, is a fundamental issue.

Therefore, a scenario for rice genetic improvement that involves integrating elite genetic variations into new combination blocks through inter-subgroup introgression was outlined. Additionally, how the genetic structure within and between subgroups of Asian cultivated rice has evolved and changed over time, highlighting key genetic variations and patterns that have emerged, was reviewed, too. Then, that further emphasis should be placed on mining favorable alleles in previously rarely mentioned subgroups, including aus, basmatic, and rayada (Bin Rahman and Zhang, 2013; Wang et al., 2013; Jing et al., 2023), since these subgroups contain favorable traits that have not been fully identified and utilized worldwide, such as adaptation to water-insufficient upland conditions, robust growth in low-fertilizer environments, and strong competition with weeds, which could be exploited for genetic improvement of other subgroups. Furthermore, it is crucial to conduct investigations into whether hybrid vigor exists among six subgroups of the Asian cultivated rice. Additionally, examining the nature of hybrid sterility between these subgroups will be essential for the future utilization of heterosis among them.

The Asian cultivated rice, O. sativa, was domesticated from the wild species, Oryza rufipogon and O. nivara. During the domestication process, a series of obvious changes occurred in morphological traits, physiological characteristics, and ecological adaptability. These changes include the transition from creeping to erect growth, the loss of grain shattering, the shortening or absence of awns, changes in hull and grain color, reduced grain dormancy, alterations in panicle architecture, increased grain number and weight, and improved regional adaptability. These changes have collectively contributed to the improvement of agronomic traits that were favored by our human ancestors (Xu and Sun, 2021). One typical example is the amylose content of rice grain, which is an important factor in determining the eating quality of rice. The diverse amylose content in rice is mainly attributed to mutations and selections in the Waxy gene, transitioning from the ancestral allele Wx^lv in wild ancestors to the indica-type allele Wx^a , the temperate japonica-type allele Wx^b , the tropical japonica-type allele Wxⁱⁿ , and the intermediate-type allele Wx^op . Furthermore, the Wx^mp and wx alleles were derived from mutations of the Wx^b allele (Zhang et al., 2019). For the last century, modern rice breeding activities have endowed the Asian cultivated rice with more trait improvements. For example, modern rice cultivars with a semi-dwarf plant type that adapts to higher fertilizer inputs and increased planting density without lodging have been bred. These cultivars provide more food to feed the growing population of the world. These improvements are accompanied by a series of favorable mutations and allelic variant accumulations in the crop. Consequently, some favorable alleles, such as various dwarf alleles of Sd1, have become prevalent in modern rice cultivars (Chen et al., 2020; Sha et al., 2022).

The evolution of O. sativa involves the fixation of favorable variants/alleles and the elimination of inferior ones under cultivation conditions, reflecting the dynamic genetic diversity throughout its evolution. Over time, original unique alleles are lost, while new unique alleles are added (Yonemaru et al., 2012). Over decades of breeding, the total number of alleles, the number of unique alleles, and the number of varieties with unique alleles tend to increase in O. sativa (Tang et al., 2021). When compared to O. rufipogon, the diversity of the entire O. sativa population has been reduced by only ~10% of its ancestors (Huang et al., 2012; Wang et al., 2022b).

As one of the most important crops in the world, rice (O. sativa L.) is widely distributed in a range of tropical, subtropical, and temperate climates, from 55°N in China to 36°S in Chile. This far exceeds the distribution range of its ancestors, O. rufipogon and O. nivara, which are primarily limited to swampy habitats in humid tropical Asia. It evolved various ecotypes in different ecosystems, such as irrigated, lowland, upland, and flood-prone areas, and expanded its range to high latitude and altitude temperate climate conditions. Additionally, it accumulated numerous favorable variants, for instance, glutinous rice harboring the wx haplotype, which was absent in its wild progenitors (Chatterjee, 1948; Morishima et al., 1961; Morishima, 1969; Khush, 1997). It was estimated that more than 4,120,000 rice cultivars and germplasm accessions have been recognized worldwide (Song et al., 2021). For a long time, it has been widely accepted and recognized that cultivars of O. sativa are classified into two major types: O. sativa subsp. indica and subsp. japonica, based on morphological, serological characters, as well as inter-varietal hybrid fertility (Kato et al., 1928). Based on morphological characters, some researches proposed that it could be grouped in three types, A, B, and C (Matsuo, 1952), or indica, javanica, and japonica (Morinaga, 1954), or indica, tropical japonica, and temperate japonica (Oka, 1958). One thousand six hundred and eighty-eight traditional rice varieties of the Asian cultivated rice were clustered into six groups based on isozyme polymorphic markers (Glaszmann, 1987), which roughly coincided with different ecological types and geographical distributions. Subsequent researchers proposed that the Asian cultivated rice can be classified into five subgroups: indica, aus, aromatic, temperate japonica, and tropical japonica, using SSR markers (Garris et al., 2005), SNPs, and genomic sequence data (McNally et al., 2009; Wang et al., 2018; Chen et al., 2019). Since the term ‘aromatic’ implies fragrant rice, it is not appropriate to use it to refer to a group of rice that includes both fragrant and non-fragrant varieties, which are specifically found in South Asia and its surrounding regions. Therefore, we refer to this group as “basmatic” instead of “aromatic” (Zhang et al., 2022; Zhou et al., 2022) ( Figure 1 ). A recent investigation suggested that the Asian cultivated rice consists of two separate monophyletic groups, representing the two subspecies. The subspecies indica encompasses two subgroups: indica and aus. Meanwhile, the subspecies japonica includes four subgroups: aromatic (also known as basmatic), rayada, temperate japonica, and tropical japonica. Notably, three minor subgroups—aus, basmatic, and rayada—are genetically distinct, suggesting that they deserve further attention despite being cultivated in limited areas in South and Southeast Asia (Jing et al., 2023). Furthermore, some investigations indicated that both aus and rayada exhibit high levels of genetic diversity and genetic differentiation (Parsons et al., 1999; Cai and Morishima, 2000; Wang et al., 2013; Travis et al., 2015; Islam et al., 2018). While Civán et al. (2015) proposed that the origin of aus was parallel to that of indica and japonica rice.

Generally, subgroup or subpopulation differentiation is a common phenomenon in the Asian cultivated rice, adapting to various ecological conditions at both the species level and the subspecies levels of indica and japonica. Greater genetic divergence and diversity implies further potential for unlocking available genetic variation in future breeding practices.

Artificial hybridization in crop plants, which began about 200 years ago, has allowed breeders to compile traits from various landraces and cultivars into a single plant (Ladizinsky, 1998). For example, the NRT1.1b gene, conferring nitrogen utilization efficiency (NUE), carries a beneficial allele in indica varieties. The DEP1 gene, which influences panicle architecture and yield, has a favorable allele in japonica varieties. When the nrt1.1b allele was incorporated into japonica lines carrying dep1, there was a significant increase in nitrate reductase activity, nitrogen uptake, and grain yield under low nitrogen conditions. Conversely, introducing the dep1 allele into indica lines already carrying the nrt1.1b allele resulted in enhanced glutamine synthetase activity, improved nitrogen transfer, and increased grain yield under both low and high nitrogen conditions, ultimately boosting yield potential in specific planting environments (Zhao et al., 2017).

Artificial hybridization in crop plants began about 200 years ago. It has enabled the assembly of traits, which individually exist in several landraces and cultivars, into one plant (Ladizinsky, 1998). In rice, the frequencies of favorable haplotypes are typically very low in modern indica and japonica varieties for most genes. However, favorable alleles at most cloned genes are present at high frequencies in landraces from specific geographic regions. Unfortunately, accessions from these populations have rarely been used as breeding parents. Specifically, the tropical japonica, subtropical japonica, and aus subgroups have high frequencies of favorable haplotypes at most genes. These haplotypes are present at low frequencies in modern varieties. As a result, most modern varieties currently grown in farmers’ fields do not have the ‘best’ alleles at most gene loci. Therefore, there is a huge potential to improve yield traits and productivity of modern varieties by pyramiding the missing favorable haplotypes at multiple loci from carefully selected breeding parents (Zhang et al., 2021).

indica-japonica introgression is progressing well in rice breeding practices and has achieved remarkable results. However, there has been limited progress in utilizing favorable variations from other subgroups, especially the mini-subgroups such as aus, basmatic, and rayada, in which distinct characters are worthy of more attention (Bin Rahman and Zhang, 2013; Jing et al., 2023). For instance, the favorable allele for higher nitrogen use efficiency (NUE) found in aus and basmatic rice can be utilized to enhance this trait in japonica and indica rice (Liu et al., 2021), thereby alleviating the increasing application of nitrogen fertilizers and corresponding environmental pollution issues. Favorable alleles that affect rice quality, identified in most basmatic rice varieties (Siddiq et al., 2012; Babar et al., 2022), could be exploited to develop new favorable genetic combinations in the backgrounds of other subgroups, thereby improving rice quality. The combination of population growth and restrictions on available arable land has intensified genetic improvement efforts and the intensive cultivation of temperate japonica rice in East Asia, particularly in Northeast China, Japan, and Korea. This has resulted in compact plant types and superior yield traits among temperate japonica varieties in this region (Fei et al., 2020; Cui et al., 2022; Wang et al., 2023a). At the same time, increased fertilizer application and management inputs are required to ensure sufficient biomass and grain production. In comparison, rice varieties from South and Southeast Asia, including indica, aus, basmatic, and tropical japonica, can achieve good biomass and exhibit resistance/tolerance to biotic and abiotic stresses even under low-fertilizer and low-input conditions ( Table 1 ).

Apart from the heterosis observed in hybrids resulting from the genetic differentiation between two indica heterotic groups: short-statured varieties of South China origin and medium-height lines of Southeast Asia origin (Xie et al., 2015), inter-subspecific hybrids between indica and japonica rice varieties exhibited significantly superior performance and higher heterosis compared to intra-subspecific hybrids (Ouyang et al., 2022). With the gradual resolution of hybrid sterility issues (Zhang, 2020; Ouyang et al., 2022; Wang et al., 2023b), the utilization of heterosis from inter-subspecific hybrids between indica and japonica rice has emerged as a feasible strategy for developing hybrid rice with higher yield potential. In South China, a series of indica-japonica inter-subspecific hybrid rice varieties, exampled by Yongyou series of intersubspecific hybrid rice, have been bred through introgression and fixation of advantageous alleles related to hybrid sterility and key yield-determining genes into both male and female parents (Wang et al., 2022a). The effective pyramiding of rare superior alleles with positive dominance effects in hybrids has contributed to higher yields. With the discovery of favorable rare alleles in both indica and japonica rice varieties, it is anticipated that indica-japonica hybrid rice will exhibit even better performance in the future (Wei and Huang, 2019).

Genomic investigation has provided evidence of apparent introgression of heterotic loci from other subgroups into indica hybrid rice parents (Lin et al., 2020), suggesting the presence of potential heterotic loci among subgroups and indicating that hybrid vigor among the six subgroups could be expected.

Genomic advances have provided rich information on genetic diversity and subgroup divergence within the Asian cultivated rice, leading to the proposal of more subgroups beyond the two subspecies, indica and japonica. This provides valuable guidance for exploiting favorable variants for further genetic improvement. Although the six subgroups, including the ‘mini’ groups, aus, basmatic, and rayada, exhibit distinguishable genealogies, genomic information for these subgroups, except for the widely recognized indica and japonica rice, has only been obtained from a limited number of samples from the ‘mini’ subgroups. This may not be sufficient to fully reflect and represent the genetic characteristics and diversity of these subgroups.

Although genomic evidence provides suggestions on genetic divergence in the Asian cultivated rice, it could be further substantiated by additional evidences, such as reproductive barriers, which are important indicators of genetic divergence and speciation in plant evolution. Hybrid sterility is the phenomenon where the F₁ plant produces partially abortive male or female gametes, while its parents possess normal gametes (Kato et al., 1928). It represents a common reproductive barrier among distant varieties and poses a major obstacle to harnessing hybrid vigor between indica and japonica rice. Severe hybrid sterility typically leads to spikelet sterility in F₁ plants derived from crosses between indica and japonica rice. Therefore, studying the phenomenon of hybrid sterility is not only beneficial for understanding the degree of genetic differentiation among subgroups of Asian cultivated rice, but also crucial for overcoming hybrid sterility and achieving the utilization of their heterosis. Twenty-four hybrid sterility loci have been identified between indica and temperate japonica, representing the highest number of such loci among the subgroups compared. Eight loci were found between temperate japonica and tropical japonica, two between tropical japonica and aus, two between indica and aus, five between temperate japonica and aus, and four between tropical japonica and indica. Strikingly, three loci were also discovered within the indica subgroup, suggesting potential genetic divergence within this subgroup (Zhang et al., 2022). Although genomic evidence has suggested that mini-subgroups such as aus, basmatic, and rayada possess unique genetic lineages (Jing et al., 2023), little work has focused on hybrid sterility in these mini-subgroups. Consequently, only a few hybrid sterility loci related to aus and tropical japonica have been detected. To date, there are limited reports (Wang et al., 2023c; Lv et al., 2024) on hybrid sterility associated with basmatic, and even fewer on the rarely studied subgroup, rayada.

Besides hybrid sterility, various other reproductive barriers have been observed in rice, including hybrid weakness (Oka, 1957; Sato and Morishima, 1987), hybrid breakdown (Oka, 1957; Oka and Doida, 1962; Sato and Morishima, 1987; Fukuoka et al., 1998; Kubo and Yoshimura, 2002), transmission ratio distortion (Nakagahra, 1972), certation (Song et al., 2002), and reciprocal gene loss of duplicated genes (Mizuta et al., 2010). Genes responsible for reproductive barrier traits are termed “speciation genes,” which contribute to the cessation of gene flow between populations and can offer clues regarding the ecological settings, evolutionary forces, and molecular mechanisms that drive the divergence of populations and species (Rieseberg and Blackman, 2010).

Past efforts in utilizing rice heterosis have primarily concentrated on intra-subspecific indica-indica hybrid vigor, primarily attributed to rarely hybrid sterility among varieties within it. Advancement on inter-subspecific indica- temperate japonica hybrid vigor in China (Wang et al., 2022a) provided a way for effectively overcoming hybrid sterility, and harnessing heterosis among sub-groups in the Asian cultivated rice. However, little attention has been paid to heterosis among all six subgroups, due to a lack of comprehensive understanding of genetic differentiation and reproductive barriers among these divergent subgroups.

Therefore, future efforts should be focused on the following: 1) identifying more unique advantageous traits/alleles within each subgroup; 2) analyzing whether there exists hybrid vigor/heterotic loci among the six subgroups; 3) studying the genetic nature of reproductive barriers, particularly hybrid sterility, among subgroups in order to deeply understand genetic differentiation, overcome reproductive barriers, achieve effective introgression of target genetic variations, and ultimately utilize the hybridization vitality among subgroups or subpopulations.

Crop diversity underpins the productivity, resilience, and adaptive capacity of agricultural systems. Genetic diversity, on the other hand, represents a dynamic process in crop evolution. Plants can be considered as evolving along two dimensions: anagenesis and cladogenesis. Integrating favorable traits from both evolutionary dimensions to reconstruct new combination blocks for genetic improvement is a major and effective approach in rice breeding, and this method is proving successful in indica-japonica introgression breeding practices, especially exampled in breeding programs in Northeast China. Artificial introgression accelerated evolution. Genomic advances have enhanced our understanding of the greater genetic divergences and diversity among subgroups of the Asian cultivated rice, suggesting that there is further potential for genetic variation awaiting utilization in future breeding practices. This will benefit the targeted and effective exploitation of favorable variants within the germplasm of a specific subgroup. For the mini subgroups—aus, basmatic, and rayada—more samples should be used to further investigate and obtain more genomic and phenotypic information about their genetic differentiation, as well as to exploit favorable variants deposited within them. Research into whether there is hybrid vigor among subgroups; as well as into reproductive barrier traits, primarily manifested as hybrid sterility, will help us better understand subgroup differentiation. It will also aid in overcoming reproductive barriers to achieve effective genetic exchange among subgroups, and even harness hybrid vigor among them. Introgression breeding is also a rational approach to improving the mini subgroups for agricultural purposes.