Genetic Diversity and Population Structure of Yunnan Rice Landraces and Disease Resistance Assessment

Rice landraces are important resources for resistance breeding and sustainability. To clarify the distribution of the genetic diversity of rice landraces and its relationship with rice blast in Yunnan, China. 141 rice landraces from 7 populations in Yunnan were selected to conduct a genetic diversity analysis based on simple sequence repeat (SSR) markers. The results showed that the 7 rice populations in Yunnan have a rich genetic diversity (expected heterozygosity [He] = 0.63). We also found that the genetic diversity of the 7 rice populations was negatively correlated with geographical longitude in Yunnan (R 2 = 0.70, P < 0.05). Structural analysis and principal coordinate analysis (PCoA) showed that the genetic structure of the rice populations is complex. Analysis of molecular variance (AMOVA) showed that rice landraces in Yunnan have high genetic variation within populations (94%). In addition, evaluations of blast resistance both in the greenhouse and in the eld showed that as the genetic diversity increased, the rice blast occurrence within the rice populations decreased (R 2 = 0.66, P < 0.05). This study not only implicates excellent materials for rice genetic improvement but also provides a theoretical basis for the protection and utilization of Yunnan rice landraces.


Introduction
Genetic diversity is an important part of biological diversity, species diversity and ecosystem diversity [1,2]. Genetic diversity also plays an important role in agroecosystems [3]. Decreasing genetic diversity could easily lead to fragile agroecosystems and increase the risk of disease epidemics [4]. Host plants and pathogens are in a dynamic equilibrium in nature environment [4]. The gradual narrowing of the genetic background of host plants might disrupt this dynamic balance, resulting in disease outbreaks and reducing crop productivity [5]. Therefore, it is vital to increase the richness of crop genetic diversity.
Some studies have reported that in addition to the variation in genetic material, other factors such as human activities, geography, and climatic conditions also affect the genetic diversity of plant populations [6,7]. Previous studies have typically reported that the genetic diversity of most species gradually decreases with increasing latitude and altitude [8][9][10]. However, speci c examples have reported genetic diversity is also signi cantly related to longitude [11]. These results suggest that climate type (due to different longitudes) generated by complex environments might be an important factor affecting genetic diversity.
Yunnan, located in southwestern China, is one of the centers of origin of genetic diversity and Asian cultivated rice worldwide [12]. Yunnan has unique natural environmental conditions and excellent local rice resources [12]. The diverse environments of Yunnan give local rice landraces the capacity to withstand different selection pressures, resulting in genetic variation and rich genetic diversity associated with adaptations to different ecological environments. However, whether these genetic variations would affect the genetic differentiation of rice populations due to climatic factors or geographic distance needs further con rmation [13,14]. Although studies have pointed out that altitude and temperature have impacts on the genetic diversity of plants growing on Yunnan Hani terraces, there is little detailed proof of the impact of natural environmental factors on the genetic diversity of rice landrace resources in different regions of Yunnan, especially with respect to longitude [15,16]. In addition, traditional farming practices (interplanting, mixed planting) and farming activities (seed retention, replacement) in Yunnan, especially the mixed or modi ed planting of japonica and indica rice, has resulted in new variations in local rice resources [17][18][19]. However, it is unclear whether these variations affect the genetic diversity and disease resistance of Yunnan rice landraces. Therefore, the genetic diversity of 141 rice landraces from 7 rice planting areas in Yunnan was analyzed on the basis of simple sequence repeat (SSR) markers to evaluate genetic differentiation and genetic structure. Identi cation of blast resistance for all rice landraces was conducted in both the greenhouse and the eld to explore the relationship between genetic diversity and rice blast resistance. The results are important in terms of the protection of Yunnan's rice landraces and provide a theoretical basis for food security maintenance and sustainable agricultural development worldwide.

Source of rice varieties
Seven rice planting areas in Yunnan (Baoshan, Lincang, Pu'er, Xishuangbanna, Wenshan, Yuxi and Honghe) were selected to collect rice landraces [20]. Two to seven counties were randomly selected in each prefecture. A total of 141 rice varieties belonging to 7 rice populations were collected from the seven planting areas. Five indica and japonica rice varieties were used as references for the genetic structure analysis (Table S1).

DNA extraction and PCR ampli cation
The fresh leaves of rice were collected for genomic DNA extraction according to the cetyltrimethylammonium bromide (CTAB) method [21], and the concentration of genomic DNA was determined by the ethidium bromide uorescent staining method [22]. Forty-three pairs of uorescent primers that can effectively distinguish the genetic differentiation of rice varieties were selected (Table   S2). An Applied Biosystems 2720 thermal cycler (Applied Biosystems, USA) was used for PCR, and the reaction system volume was 10 µL, including 1 µL of DNA template (50 ng), 0.2 µL of each primer (forward and reverse primer), 0.8 µLdNTP mix (TAKARA), 0.05 µL Taq DNA polymerase (Biocolor BioScience, Shanghai), 2 µLPCR buffer (10 ×, including 2 mmol/L MgCl 2 ) and 6.75 µL double distilled water.The PCR conditions were as follows: 1 cycle at 94°C for 4 min; 30 cycles at 94°C for 40 s, 55°C for 30 s andb72°C for 30 s; 1 cycle at 72°C for 7 min. PCR products were separated and analyzed by a uorescent capillary electrophoresis system (ABI 3130xl Genotyper, Applied Biosystems, Genesky, Shanghai). The molecular weights were calculated by the biological software GeneMapper (version 3.7, Applied Biosystems, Inc., USA). SSR data reading was performed according to the size of ampli ed products, and the products with the same size were regarded as an allele.

Genetic diversity analysis
Genetic diversity parameters including the percentage of polymorphic loci (P), observed heterozygosity (Ho), number of alleles (Na), effective number of alleles (Ne), Shannon index (I), expected heterozygosity (He) and polymorphic information content (PIC) were calculated using GenALEx 6.502 software [23]. In addition, we also performed analysis of molecular variance (AMOVA) to divide the results of genetic differentiation into within populations and among populations. Based on AMOVA, we also calculated the amount of gene ow (Nm) and paired genetic differentiation coe cient (Fst) to estimate the level of gene ow and differentiation among rice populations. SPSS 26 software was used to estimate the relationships between genetic diversity and longitude and between genetic diversity and rice blast resistance.

Structure analysis
To infer the genetic structure and genetic relationships between the 7 rice populations, STRUCTURE (version 2.3.4) software [24] was used to analyze the 141 rice landraces. Five typical indica rice varieties and ve japonica rice varieties served as references. The K value was set to 1-7, and 10 replicate calculations were performed. The run parameter burn-in was set to 100000, and the run length was set to 200000. After the operation was complete, the best K value was determined by the Structure Harvester online program (http://taylor0.biology.ucla.edu/structureHarvester/) according to previously reported methods [25]. CLUMPP (version 1.1.2) software was used to perform the best comparison for the 10 replicates of the optimal K value [26]. The calibration results were automatically visualized by distruct 1.1 software. Principal coordinate analysis (PCoA) was used to estimate the genetic composition of the rice landraces.

Resistance evaluation of rice landraces in the greenhouse
Ten rice blast strains donated from Yunnan Academy of Agricultural Sciences (GUY11, 66b, CH1598, CH1139, CH9105a, CH1638, CH1643, CH1633, CH091C, 363) were selected to evaluated their resistance to rice blast. The rice blast strains were cultured at 28°C for 7 days and then sporulated at 28°C under light for 2 days. The spores were washed with distilled water comprising 0.02% Tween 20, and the concentration of the spore suspension was 2×10 5 /mL. Rice seeds were sown in a nursery box after germination. Ten seeds were sown per variety, and 2 replications were included. Rice plantlets at the 3.5leaf stage were used for inoculation. Investigation of the incidence and disease index of rice leaf blast after seven-days inoculation was followed the Standard Evaluation System for Rice [27].

Resistance identi cation of rice landraces in the eld
A eld trial was conducted in Jianshui, Yunnan in 2019 (longitude 102°46′16″, latitude 23°36′2″). The experiment adopted a completely randomized block design. Each plot consisted of 3 rows×20 plants, with a spacing of 5×8 inches, and three replicates were conducted. The highly susceptible variety (Huangkenuo) was used as control variety. Throughout the growth period, water and fertilizer management followed local practices, and pest control instead of disease control measurements was conducted. Investigation of the incidence and disease index of rice diseases (panicle blast at the maturity stage) was followed the Standard Evaluation System for Rice [27].

Data analysis
Microsoft O ce Excel 2016 software was used for statistics and data maintenance. SPSS 26 software was used for performing analysis of variance, correlation analysis, and least signi cant difference (LSD) tests. Graphs were constructed via Origin 2019. ArcGIS 10.2 was used to construct a sampling point map.

Genetic diversity among Yunnan rice landraces is rich
The SSR markers selected in this study were highly polymorphic across the populations, and this level of polymorphism could meet the genetic diversity analysis of rice landrace populations. A total of 387 alleles were detected in these rice varieties, with an average of 9 alleles (Na) at each locus. The average He of the SSR markers was 0.63, which varied from 0.30 (RM463) to 0.81 (RM241). The PIC of the SSR markers varied from 0.34 to 0.85, with an average of 0.65. Among them, there were 34 highly polymorphic sites (PIC>0.5), accounting for 79.07%, and there were nine moderately polymorphic sites (0.25<PIC<0.5), accounting for 20.93% (Table S3). For the genetic diversity of rice populations, seven rice populations collected from seven cities in Yunnan Province also showed high genetic diversity. The He of the rice populations was between 0.58 (Wenshan) and 0.68 (Lincang), with a mean value of 0.63. The Na of the rice populations was 5.01, the Shannon index (I) of the rice populations averaged 1.24, the Ho of the rice populations was 0.09, and the average P of the rice populations reached 99.67% (Table 1). We concluded that rice landraces in Yunnan have a rich basis of genetic diversity.

Genetic diversity in Yunnan rice landraces increases with decreasing longitude
We noticed that the genetic diversity of rice landraces at different locations displays an obvious spatial distribution. The rice landraces in Wenshan with the greatest longitude showed the lowest genetic diversity. As the longitude increases, the genetic diversity gradually decreases. Correlation analysis showed that there was a signi cant negative correlation between genetic diversity and longitude for all rice landraces (R=0.70, P<0.05) (Fig. 1a). We concluded that the genetic diversity of the rice landraces gradually decreased from southwestern to southeastern Yunnan (Fig. 1b).

Relationship between genetic differentiation and gene ow among rice landraces
AMOVA analysis showed that rice landraces in Yunnan have high genetic variation within populations (94%) and 6% genetic variation among populations, indicating that self-pollinated rice has high genetic diversity within populations ( Table 2). In addition, the F statistical calculations of all SSR sites showed that the inbreeding coe cient in the population (Fis) and overall inbreeding coe cient (Fit) were 0.869 and 0.873, respectively, while the Fst was 0.037, indicating that genetic differentiation among populations was limited. Therefore, genetic differentiation among populations could be disregarded.
In addition, paired Fst and gene ow (Nm) were calculated to estimate the level of genetic differentiation among populations and gene ow of the various rice populations. We found that high gene ow mainly occurred between adjacent regions. For example, the Nm of the populations between Baoshan and the adjacent region Lincang was 5.39 but was only 1.92 between Baoshan and Wenshan (the most distant region from Baoshan) (Table 3). Similarly, the paired Fst also increased with a gradual increase in geographical distance among populations. Genetic differentiation among populations became increasingly obvious, which further con rmed that the genetic diversity of the rice populations in Yunnan was signi cantly negatively correlated with longitude.

Genetic structure analysis of rice landrace populations
STRUCTURE software was used to analyze the genetic structure of the populations of the collected rice landraces, and ve japonica and ve indica typical rice varieties were used as references. The results showed that the genetic structure of the rice populations in the 7 localities was complex and that there was no single genetic component. Nonetheless, the genetic structure could be divided into two main genetic components (those associated with indica (yellow) and japonica (blue)) when the K value was determined to be 2 (Fig. 2a). Moreover, both the indica and the japonica rice genetic components were present in all seven rice populations (Fig. S1). There was a subpeak representing a further substructure when K value was 3, indicating that the genetic structure could be divided into three subgenetic components, which are represented by blue, yellow and red colors. The abovementioned three genetic components were found in all 7 rice populations (Fig. 2b). PCoA con rmed that there were three subgenetic components in the genetic structure analysis, which were labeled SP1, SP2, and SP3 (Fig. 3). Notably, typical indica and japonica rice varieties clustered only in SP1 and SP2. In addition, the rice populations of all localities were present in the three components except for the rice population of Yuxi (which was not present in SP3). We could conclude that the rice landraces in the populations were randomly mixed and included each component, indicating that there was large genetic differentiation in the population and thus con rming the high gene ow among the populations.

Genetic diversity is positively correlated with rice blast resistance
To verify the relationship between the genetic diversity and rice blast resistance of the seven rice populations, the 141 rice landraces were used to evaluate rice blast resistance both in the greenhouse and in paddy elds. The results showed that the rice landraces have good blast resistance, the resistance frequency of rice landraces in the greenhouse was 75.62% (Table S4). Moreover, correlation analysis showed that there was a signi cant positive correlation between genetic diversity (Na) and resistance (R 2 =0.66, P<0.05) (Fig. 4a). Further eld experiments showed that seven rice populations had low disease incidence (6.97%) and a low disease index (3.50). However, the disease incidence and disease index of the susceptible variety Huangkenuo (CK) were 53.19% and 14.18%, respectively. Notably, the rice population in Lincang with the highest genetic diversity (He=0.68) had the lowest disease incidence (1.86%) and disease index (0.44), while the local rice population with the lowest genetic diversity (Wenshan) (He=0.58) had the lowest incidence (19.96%) and disease index (12.89). Thus, we concluded that genetic diversity and eld resistance are correlated (Fig. 4b, c).

Discussion
Genetic diversity is an important part of rice germplasm resource utilization. A total of 141 local rice landraces from seven cities in Yunnan were analyzed for their genetic diversity in our experiment, and the results showed that the rice landraces have rich genetic diversity, which are the same as the previous studies [28,29]. Although the genetic diversity of rice populations in the seven localities was high (He = 0.63), the population diversity level was not evenly distributed, showing a clear spatial distribution and a signi cantly negative correlation with geographical longitude. This shows that the longitude gradient has an impact on the genetic diversity of Yunnan rice populations. There was study reported that when populations were divided according to latitude and longitude, the variance of the genetic diversity within populations was small, indicating that latitude and longitude are important factors affecting the genetic diversity of populations [30]. In addition, some people found that the biological genetic diversity throughout the Mediterranean basin is signi cantly correlated with longitude [11]. They also declared that it is di cult to use a single factor to explain the genetic diversity across Mediterranean areas, which have complex environments with speci c climate types, nonetheless, longitude was an important factor [11]. Some studies have shown that factors such as temperature and altitude have varying degrees of in uence on genetic diversity because of the complex terrain of Yunnan, but no speci c analysis has been made from the perspective of longitude [31]. This study found that there is a signi cant correlation between rice genetic diversity and longitude in Yunnan. Thus, we provide a new way to study rice genetic diversity and new ideas for rice germplasm innovation and genetic diversity protection.
Genetic differentiation indicates the degree of variation within a population [32]. This study showed that the genetic variation of the rice population in Yunnan is driven mostly by variation within populations (94%). The Fst was low (0.037), which is consistent with the ndings of previous studies. William [33] used SSR markers to assess the genetic diversity of rice genotypes and found that 70% of the genetic variation originates from variation within populations. Tu [34] also used SSR markers to analyze the genetic diversity of 60 rice germplasms in Yunnan and reported similar results. Wright [32] reported that self-pollinated rice has high genetic differentiation within populations, while genetic differentiation among populations is limited. We found that Yunnan rice populations experienced a high degree of gene ow (Nm) (6.505) and showed obvious distance isolation. In addition, geographically close populations showed a high degree of gene ow, which was limited to geographically separated populations, which are similar to the results of rice populations in Sri Lanka [35]. Some studies have reported that the frequency of landrace seed exchange is more than ten times that of foreign exchanges among farmers in Yunnan, these frequent exchange of seeds between neighboring regions may affect gene ow [36,37]. Thus, we speculated that seed exchange between farmers and regions may be the reason for the formation of high gene ow in geographically close regions.
As one of the basic characteristics of species and populations, genetic structure is the result of multiple factors such as selection, mutation, genetic drift, migration, and gene ow [38]. Structural analysis in this study showed that the genetic structure of rice populations includes components associated with two main groups: japonica and indica rice (Fig. 3a, K=2). Using SSR markers, Chen [39] analyzed the population structure of rice germplasm resources from 16 cities in Yunnan, and the results showed that 908 rice germplasms could be divided into two major groups (japonica and indica), which was consistent with our results. However, Yunnan is divided into three types of rice planting areas, i.e., an indica area, an indica-japonica overlapping area and a japonica area, due to the unique geography and climatic [40]. When the K value in the structure analysis was 3, we found that two genetic components were typical japonica rice and typical indica but that a third component was not typical indica or japonica rice (Fig. 3b,  K=3). PCoA also showed that there were three subgroups, two of which included typical japonica or indica, and the third subgroup was an intermediate type (Fig. 4). In addition, we found that most rice populations (except Yuxi) included all three components (Figure 3b), indicating that there was large genetic differentiation within populations and further con rming that there was high gene exchange within populations.
Rice blast control is an important part of improving the yield and quality of rice [41]. The results of this study showed that rice populations have a rich genetic diversity and that there is a positive correlation between genetic diversity and rice blast resistance (Fig. 5a). Zhu [42] found that genetic diversity was a signi cant factor in the control of rice blast. Tu [34] and He [43] also reported that the higher the genetic diversity of rice varieties was, the higher the resistance of the varieties to rice blast. Therefore, it is highly important to determine the relationship between the genetic diversity of rice landraces and the rice blast resistance of genotypes. Applying these important germplasm resources to rice production is a future endeavor.

Conclusions
In this study, we found that seven rice populations in Yunnan showed an abundance of genetic diversity and that the genetic diversity decreased with increasing longitude. Structural analysis and PCoA showed that the genetic structure of the seven rice populations in Yunnan is complex. When the K value was 2, the rice populations were divided into two genetic components: japonica and indica ones. When the K values was 3, in addition to the two components of indica and japonica, there was a third, unknown genetic component. AMOVA showed that rice landraces in Yunnan have high genetic variation within populations (94%). In addition, evaluation of blast resistance both in the greenhouse and in the eld showed that there was a correlation between genetic diversity and rice blast occurrence: as the genetic diversity increased, the rice blast occurrence throughout the rice populations decreased.

Declarations
Supplementary Materials: The online version contains supplementary material available at https://doi.org/xxx. Fig.S1: The distribution of K-values with ΔK. Table S1: Information of tested rice materials. Table S2: Primer sequences of the 43 simple sequence repeat (SSR) markers used in this study.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.