| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
a USDA-ARS, Dale Bumpers National Rice Research Center, P.O. Box 1090, Stuttgart, AR 72160
b USDA-ARS, National Small Grains Collection, 1691 S 2700 W, Aberdeen, ID 83210
c USDA-ARS, Rice Research Unit, 1509 Aggie Dr., Beaumont, TX 77713
d USDA-ARS, Crops Pathology and Genetics Research Unit, 1308 PES, Dep. of Plant Sciences, Univ. of California, Davis, CA 95616
* Corresponding author (wyan{at}spa.ars.usda.gov).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
0.9, indicating that the RCS was highly representative of the RWC. Correlation coefficients for the other six descriptors were lower (0.65–0.88), but still significant.
Abbreviations: APHIS, animal and plant health inspection service • ARS, Agricultural Research Service • ASV, alkali spreading value • CV, coefficient of variance • DNA, deoxyribonucleic acid • GRIN, germplasm resources information network • IRRI, international rice research institute • L/W, length to width • NCGRP, national center for genetic resources preservation • NGRP, national genetic resources program • NPGS, national plant germplasm system • NSGC, national small grains collection • NSSL, national seed storage laboratory • RCS, rice core subset • RWC, rice whole collection • USDA, United States Department of Agriculture.
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Some 94% of the accessions in the USDA rice germplasm collection were obtained internationally, and the remainder domestically. All public cultivars registered in the USA can be entered in the collection. Foreign germplasm accessions must be grown for one generation in a plant quarantine greenhouse isolated from commercial rice growing areas to prevent accidental introduction of new disease and insect pests (Rutger and Lehman, 1977). Quarantine is managed and conducted by the Animal and Plant Health Inspection Service (APHIS) at Beltsville, Maryland. The resulting seed is increased and distributed to both the base collection in the National Center for Genetic Resources Preservation (NCGRP), formerly known as the National Seed Storage Laboratory (NSSL), at Fort Collins, Colorado for long-term storage, and the working collection in the NSGC at Aberdeen, Idaho for medium-term storage. The accessions in the NSGC are available for public distribution.
Comprehensive evaluation of the collection for numerous descriptors, particularly those involving disease resistance and grain quality, has been hindered by the sheer number of accessions. It also is much harder to analyze such large collections using molecular means. For practical evaluation and effective management of such large collections in crops, Brown (1989a) proposed the core collection concept.
A core collection is a subset of a large germplasm collection that contains chosen accessions that capture most of the genetic variability within the entire gene bank. The first core was developed from the Australian collection of perennial Glycine spp. by Brown et al. (1987). With the strategy of comprehensive evaluation and accurate analysis of the core collection, germplasm curators can accomplish several key tasks. They can assess the genetic diversity of the existing collection, and identify gaps for planning acquisition strategies (Steiner et al., 2001). In particular, calculations of genetic distances can be used to identify special divergent subpopulations that might harbor valuable genetic variation that is under-represented in current holdings. Core collections can monitor changes in heterogeneity and heterozygosity (genetic drift) as accessions are regenerated, and identify and remove duplicate accessions in maintenance. They establish passport data to characterize each accession based on gene, genotype, and genome along with the detailed phenotypic data, which provide accurate and detailed information at both the phenotypic and molecular levels. Crop breeders can then quickly find the traits in which they are interested, e.g., high production-efficiency, premium quality, disease or insect resistance, and stress tolerance. Information on the genetic backgrounds and genetic distances from commercial cultivars is useful for determining strategies for transferring the desirable traits into new commercial cultivars.
A good core collection should not contain redundant entries and should be sufficiently large to achieve reliable conclusions for the whole collection (Brown, 1989b). The origin of its entries should be authentic, unless no information is available. From a genetic aspect, the major species and subspecies and geographic regions should be represented. Also, emphasis should be given to the more broadly adapted rather than narrow-based germplasm. Using this approach, genetic diversity in the core can be maximized. Simple random sampling ensures that every accession in the collection has the same chance of being included in the core. Stratification increases efficiency with right choice of sample sizes for each group (Cochran, 1977). As a result, stratification combining with random sampling from groups of accessions, in logarithmic or absolute proportion to the group size, is the best strategy (Brown, 1989a).
The objectives of this study were to assemble a RCS that adequately represents the germplasm in the USDA RWC, and to effectively and comprehensively phenotype and genotype the RCS to assess the RWC.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Development of the USDA Rice Germplasm Core Subset
The USDA RCS was assembled by sampling the working collection in the NSGC in 1998 and 2002, respectively. A method of stratification by country and then random sampling was adapted by: (i) recording the number of accessions from each country or region of origin; (ii) calculating the logarithm (log) of the number of accessions from each country or region of origin; (iii) randomly choosing the accessions within each country or region based on the relative log numbers, with a minimum of one accession per country or region; and (iv) removing obvious duplications by plant introduction (PI) number and cultivar name. In addition to the stratified sampling, additional emphasis was placed on some newly introduced Chinese germplasm (Yan et al., 2002) and newly released accessions from quarantine programs (Yan et al., 2003).
Evaluation of the Rice Core Subset
The resulting RCS was evaluated at Stuttgart, Arkansas in 2002. Seeds of each accession in the RCS were visually purified by seed shape (long, medium, or short) and hull color (white, straw, golden, or black) as described in GRIN (USDA-ARS, NGRP, 2006). Each accession was grown in a plot consisting of two rows, 0.3 m apart and 1.4 m long, with 3 g of seeds planted in each row by a Hege 500 planter on 16 April 2002. Plots were separated by 0.9 m to avoid biological and mechanical contamination. Seedlings emerged on 28 April 2002, and weeds were controlled with Propanil [N-(3,4-dichlorophenyl)propanamide] at 9.3 L ha–1 mixed with Bolero (S-[(4-chlorophenyl)methyl]diethylcarbamothioate) at 18.5 L ha–1. A permanent flood was established after 67 kg ha–1 of nitrogen as urea was applied at about 5-leaf stage.
Agronomic descriptors were recorded in the field using standard criteria described in GRIN. Grain samples of each entry in the core collection were collected from those entries that produced enough seed in the agronomic evaluation. Rough or paddy rice is the mature rice grain as harvested, and becomes brown rice when the hulls are removed. Only fully filled, mature grains were used for rough rice and brown rice sampling. Rough and brown rice samples were analyzed on an automated grain image analyzer (GrainCheck 2312; Foss Tecator AB, Hoganas, Sweden) to determine rice kernel dimensions (length, width and length/width ratio), hull and seed pericarp (bran) colorations, and 1000 grain weight. A complete color set of rough and brown rice samples were used to calibrate the GrainCheck 2312 analyzer for measuring kernel color and assigning color codes as described in GRIN (USDA-ARS, NGRP, 2006). Samples were milled for determination of apparent amylose content (Pérez and Juliano, 1978; Webb, 1972) and alkali spreading value (ASV) (Little et al., 1958).
Although 28 descriptors for rice are listed in GRIN, only 14 were selected for this study because: (i) some express the same trait in different way, i.e., days to anthesis (coded) vs. days to flower (actual), plant height 1 (coded) vs. plant height 2 (actual); (ii) some contain information about others such as ASV for gelatinization temperature, amylose for endosperm type, kernel length:width ratio for grain type and rough kernel weight for milled kernel weight; (iii) some are greatly affected by environment, e.g., parboil loss, allelopathy, protein, and lodging, and (iv) some are hard to measure, resulting in limited data in GRIN, e.g., aroma, blast, sheath blight, and straighthead reaction.
Statistical analysis was conducted using the univariate and correlation procedures of SAS statistical software, Version 9.1.3 (SAS Institute, 2004). Frequency distributions for each descriptor were determined using Microsoft Office Excel software. Code values for the categorical variables awn, panicle, and plant types, and hull and bran colors, were used with numerical values of other descriptors for correlation analysis. Frequency refers to how often data values occur within a range of values in an Excel bins-array that is an array of data intervals into which the data values are grouped. For example, days to flower had a bins-array of 40, 50, 60, 70, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170, 180, and 190 (Fig. 1A ), e.g., all accessions ranging from 36 to 45 d were grouped in bin 40. Frequencies (%) of the respective bins were 0.02, 0.05, 1.15, 2.91, 7.54, 16.01, 20.33, 21.16, 14.91, 6.65, 4.07, 2.29, 1.83, 0.48, 0.52, and 0.10 among 15097 accessions in RWC, and 0, 0.24, 1.26, 4.56, 10.43, 23.38, 27.40, 13.73, 9.53, 3.54, 2.82, 1.50, 0.96, 0.48, 0.18, and 0 among 1668 entries in RCS. Paired frequencies of the RWC and the RCS on each bin were used for correlation analysis, which measures the correspondence between the two collections.
|
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
USDA Rice Core Subset (RCS)
The USDA RCS was initiated with a 1998 sampling of 970 entries, plus a 2002 sampling of 820 entries for a total of 1790 entries. This RCS of 10%, assembled by the stratified random sampling method, is expected to represent over 70% of the genetic diversity in the RWC (Brown, 1989a).
Countries of Origin
The entries in the core collection originated from 114 countries, one less than the RWC because the seed from one country was dead. Germplasm from China predominated in the core collection (135 entries) due to a large number of germplasm lines that were recently introduced (Yan et al., 2002), followed by germplasm from the Philippines (70 entries) where many accessions were acquired from the International Rice Research Institute (IRRI) collection, followed by Japan having 69 entries. Based on the number of accessions in the collection, from 51 to 60 entries were sampled from two countries, 41 to 50 entries from three countries, 31 to 40 from five countries, 21 to 30 from 25 countries, 11 to 20 from 23 countries, five to 10 from 18 countries, two to four entries from 15 countries and one entry from each of 20 countries.
Rate of Sampling by Country
Sampling rate is defined as the percentage of sampled entries in the core subset divided by the total accessions in the NSGC collection for a particular country. Complete or 100% sampling of germplasm in the NSGC collection was made for 37 countries, in which 20 countries had one accession, four countries had two accessions, and the other four countries had three accessions. Others included: Kazakhstan, 13 accessions; Uruguay, 12; Burkina Faso, seven; Ecuador and Tanzania, six each; Jamaica, Morocco, and Panama, five each, and Unknown origin, four. Fifteen countries had sampling rates from 71 to 99%, 23 countries from 41 to 70%, 18 countries from 21 to 40%, and 21 countries from two to 20%. Two entries, or less than 1%, were sampled from 1421 accessions with uncertain origin. China, which has the most entries in the collection, had a 7% sampling rate.
Species Included in the Core Collection
Nearly 99% of the core collection came from the cultivated Asian species, O. sativa, but representatives of the African cultivated species, O. glaberrima, and ‘O. hybrid’ (defined as the derivatives from crosses between sativa and glaberrima), and ‘O. sp’ (defined as those with no identifiers of species) and species alta, barthii, glumaepatula, latifolia, nivara, officinalis and rufipogon in Oryza were also included.
Time Coverage of Sampled Entries
The sampling was applied from historical to modern germplasm accessions with a slight emphasis on the modern in the collection. Cultivar Ostiglia was included as the earliest existing in the RWC from Germany in 1904, originating in Italy and having the lowest Cereal Investigation (CI) number eight. The new accessions collected from 1991 to 2000 were sampled the most at 24%, whereas previous decades were sampled at rates ranging from 3 to 15%. More than 40% of the accessions (8188) were introduced to the USDA rice germplasm collection in the 1970s, and thus, a majority of the entries (833) to the RCS was sampled from this decade. Conversely, accessions collected during the 1930s were among the least represented because few rice accessions were entered into the collection during this time period.
Comparative Analysis between the USDA Rice Whole Collection (RWC) and Rice Core Subset (RCS)
Days to Flower
This trait was measured as days from seedling emergence to 50% heading. In the RWC, 15097 accessions (82%) have been evaluated for days to flower with a mean of 103 (CV = 20%) and a range of 37 to 183 d. In 2002, 1668 entries of the RCS were recorded for days to flower with a mean of 97 (CV = 20%) and a range of 42 to 180 d. Frequency distributions between RWC and RCS were highly correlated (r = 0.90) although the RCS had more entries for the bin of 90 to 100 d and the RWC had more accessions for the bin of 110 to 120 d (Fig. 1A).
Plant Height
A total of 13052 accessions in the RWC were tested and averaged 118 cm (CV = 22%) with a range of 41 to 208 cm. The 1635 entries in the RCS had a mean of 126 (CV = 20%) with a range of 61 to 206 cm. The RCS had taller cultivars represented at higher frequencies than the RWC (Fig. 1B). The frequency distribution of the RCS was highly correlated with the RWC (r = 0.93), indicating the difference was very minor.
Awn Type
The RCS had approximately 20% higher frequency of awnless types (coded 0) among 1641 entrie, whereas the RWC had more short and partial awn types (coded 1) among 12507 accessions (Fig. 1C). Other types including short and full awn (coded 5), long and partial awn (coded 7), and long and full awn (coded 9) were represented in similar proportions in both the RCS and RWC. The addition of newly introduced accessions in the collection may be partly responsible for the differences since most of them were awnless types (Yan et al., 2002, 2003). Regardless, the awn type distribution of the RCS was highly correlated with that of the RWC (r = 0.93).
Panicle Type
As indicated in Fig. 1D, the distribution of three panicle types, compact (coded 1), intermediate (coded 5), and open (coded 9), was almost identical between the RCS with 1634 entries and RWC with 15374 accessions (r > 0.99).
Plant Type
The RCS had greater frequency for erect types (coded 1) among 1633 entries, whereas the RWC had higher open plant type (coded 5) frequency among 15453 accessions (Fig. 1E). This difference is likely due to the addition of the elite new germplasm accessions since most of them had been selected for erect plant type (Yan et al., 2002, 2003). Intermediate (coded 3) and spreading (coded 7) plant types were represented in similar proportions in both the RWC and RCS. The plant type frequency distribution between the RWC and RCS was significantly correlated (r = 0.69 for df = 4, P < 0.05).
Hull Color
The RWC with 13555 accessions had greater percentage of straw (coded 2) hull color type than the RCS with 1615, entries whereas a greater percentage of spotted or piebald (coded 6) was observed in the RCS as compared to the RWC (Fig. 1F). Frequency differences between the RWC and RCS in other hull color types including white (coded 1), tawny or russet (coded 4), furrowed (coded 5), purple (coded 7), and black (coded 8) were minor. Hull color for the RWC was scored by observation over many years and by different people (Webb et al., 1968) whereas the RCS was rated with an image analysis instrument after the RCS was grown in one field season. Although the method of evaluation may have changed over time, the two distributions were highly correlated (r = 0.88).
Hull Cover
More than 80% of rice accessions had short hairs all over the hull (coded 4), while more than 12% were smooth (coded 1) (Fig. 1G). The frequency distributions of the RCS with 1497 entries and RWC with 14046 accessions were essentially identical (r > 0.99).
Bran Color
The RCS with 1616 entries had about 4% more of light brown accessions (coded 2) and about 4% less of red rice types (coded 5) than the RWC with 18216 accessions (Fig. 1H). No red rice bran types were included in the recently introduced lines, which may explain this proportional difference of bran color between the RWC and RCS (Yan et al., 2002, 2003). Distributions of other color types, white (coded 1), brown (coded 4), and purple (coded 7), were very similar in both the RWC and RCS. Over all, bran color distribution in the RCS was highly correlated with that of the RWC (r > 0.99).
Kernel Length
In the RWC, 7259 accessions, evaluated as brown rice, averaged 6.2 mm (CV = 16%) ranging 3.0 to 9.9 mm. In the RCS, 1616 entries were examined and averaged 6.5 mm (CV = 12%) and ranged 4.2 to 10.0. Over 98% of accessions in both collections had kernel lengths ranging from 5 to 8 mm (Fig. 2I
). The RCS had a higher proportion of long grains than the RWC. However, the two frequency distributions were highly correlated (r = 0.83).
|
2 mm) the frequency distribution of the two collections was highly correlated (r = 0.94).
Kernel Type
Ratio of brown rice kernel length to width (L/W) averaged 2.5 (CV = 29%) with a range of 1.3 to 5.3 among the 7210 accessions examined in the RWC whereas the RCS averaged 2.6 (CV = 20%) and ranged 1.5 to 4.5 for 1616 entries (Fig. 2K). In the RWC, there were 2470 (34.3%) short grain type accessions (L/W 
2.0), 3172 (44.0%) medium grain types (2.0 < L/W 3.0), and 1568 (21.7%) long grain type (L/W > 3.0). In the RCS, there were 282 (17.5%) short grain type entries, 1037 (64.1%) medium grain type, and 297 (18.4%) long grain type. Although the RCS was slightly shifted toward higher length:width ratio as compared to RWC, the two frequency distributions were highly correlated (r = 0.85).
Grain Weight
Weight (g) of 1000 grains of rough rice averaged 24.2 g (CV = 17%) and ranged 8.0 to 45.9 g for 7031 accessions evaluated in the RWC, whereas the RCS averaged 26.2 g (CV = 17%) and ranged 12.5 to 42.7 for 1615 entries. The RWC had more small grain types ( 25 g) and the RCS had more large grain types ( 30 g) (Fig. 2L). These differences were minor, so that frequency distribution in the RCS was highly correlated with that of the RWC (r = 0.91).
Amylose Content
Among 6909 accessions evaluated in the RWC, apparent amylose content averaged 23.0% (CV = 25%) and ranged 9.0 to 40.0% (Fig. 2M). The RCS had a mean apparent amylose content of 20.0% (CV = 25%) and ranged 0.0 to 26.9% among the 1613 entries. In the RCS, 43 accessions had waxy endosperm with amylose contents near zero. Although about a third of the RWC was evaluated for amylose content, no waxy endosperm accessions were evaluated and thus, the lowest amylose content measured was 9.0%. In addition, unlike the RWC in which 2028 accessions (29%) had apparent amylose contents 27% or greater, there were no accessions identified in the RCS with apparent amylose contents greater than 27%. Amylose content data generated for the RWC was determined in the 1960s (Webb et al., 1968), whereas the RCS was evaluated in 2002. Changes in the standards and protocol for determining apparent amylose content might be partly responsible for different distributions between the RWC and RCS. However, the correlation coefficient for the frequency distributions between the two collections was still strong (r = 0.82).
Alkali Spreading Value
The 7234 accessions that were evaluated from the RWC and the 1612 entries from the RCS demonstrated the complete range in ASV values of 2.0 to 7.0. ASV averaged 5.5 ± 1.3 for the RWC and 5.1 ± 1.3 for the RCS. The RCS had 10 and 16% greater frequency than the RWC at ASV 4 and 4.5, respectively (Fig. 2N). The RWC had higher frequencies than the RCS at other ASV values in the range. These differences made the correlation (r = 0.65) of frequency distribution between the RWC and RCS significant at 0.05 probability level.
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Characterization and evaluation of germplasm accessions in collections are essential for both conservation and utilization of these collections (Riley et al., 1995). In spite of the importance, only a small portion of accessions in germplasm collections worldwide has been thoroughly characterized. Most of the world's ex situ plant germplasm resources remain poorly characterized with only passport data (FAO, 1996). Characterizing a core subset of a collection is a means of facilitating the evaluation and utilization of diverse germplasm. This strategy has been adopted for rice landraces in China (Li et al., 2002), wheat (Triticum durum L.) (Spagnoletti Zeuli and Qualset, 1993), soybean (Glycine max L.) (Brown et al., 1987), barley (Hordeum vulgare L.) (Liu et al., 1999), sorghum (Sorghum bicolor L.) (Grenier et al., 2001), alfalfa (Medicago sativa L.) (Basigalup et al., 1995), peanut (Meloidogyne arenaria L.) (Holbrook et al., 2000), sugarcane (Saccharum spontaneum L.) (Tai and Miller, 2001), lentil (Lens culinaris Medik) (Tullu et al., 2001), birdsfoot trefoil (Lotus corniculatus L.) (Steiner et al., 2001), and other crops. To fully realize the advantages of core strategy, the core subset should include most of the genetic diversity in the whole collection and be closely correlated with the whole collection. Correlation for each descriptor, between RWC and the RCS, is an indicator of how well the RCS represents the phenotypic variation of the RWC. Among the 14 descriptors evaluated in this study, eight (hull cover, panicle type, bran color, kernel width, plant height, awn type, grain weight, and days to flower) had correlation coefficients equal to or greater than 0.9 between the RCS and the RWC. Four other descriptors (hull color, kernel type, kernel lengt, and amylose content) correlations ranged from r = 0.82 to 0.88 indicating a strong relationship between the RCS and the RWC. Alkali spreading value and plant type had the smallest RCS vs. RWC correlation coefficients (r = 0.65 and 0.69, respectively), but were significant at the 0.05 level.
The lack of near perfect correlations for some of the descriptors may be due to reasons other than the RCS not truly representing the diversity of the RWC. Although 99% of the RWC has been evaluated for bran color, less than 40% of the RWC has been evaluated for kernel dimensions, amylose content, and ASV traits. Many years ago new entries in the collection were routinely evaluated for as many of the key descriptors as possible. However, as the collection rapidly grew, this became cost-prohibitive and data on all descriptors were no longer collected on a routine basis but rather only from specific accessions of interest. Also, evaluation of one large set of accessions for grain quality descriptors stopped on the retirement of an interested scientist. As a result, data on some of the descriptors are shifted toward accessions that were introduced before 1976 when the RWC included germplasm from only 85 countries. Since 1976 the RWC has increased by 6463 entries and much of this has not been well characterized. Thus the lack of correlation between the RCS and RWC may be due, in part, to the lack and shift of data collected on the RWC. In addition, attempts at systematically characterizing the germplasm collection were started in 1960s and have been performed in sets of 4000 to 5000 accessions at a time (Webb et al., 1968). Thus, over the years that descriptor data have been collected, variation in environmental conditions and cultural management practices, as well as the subjectivity involved with measuring some traits (i.e., hull color, plant type) can influence these correlations. Moreover, methods of measurement for some traits (e.g., amylose content) have improved over time and new instrumentation has allowed more rapid and accurate assessment of the germplasm for traits such as kernel dimensions, which used to be measured by hand on a sample of 20 kernels. Traits such as these are now being measured through image analysis system on hundreds of kernels per accession. As a result, lower correlations for some of the descriptors do not necessarily indicate that the RCS is not representative of the RWC in the USA.
To further elucidate how well the RCS represents the RWC, considering all of the descriptors, the frequency for each bin range of the fourteen descriptors measured in this study were arranged in parallel between the RWC and the RCS, resulting in 152 pairs of data. The correlation for these 152 pairs of data points was determined to be r = 0.94 (P < 0.0001), with a coefficient of determination of r2 = 0.88. The high correlation of the RCS with the RWC demonstrates that a stratified set of 10% of the accessions can be effectively used to assess the variability in the whole rice NPGS collection with 88% certainty. As new accessions are introduced into the rice collection, the RCS will be expanded and updated on a regular basis so that it will continue to be representative of the diversity present in the RWC.
Assessment of the USDA Collection of Rice Germplasm through the Core Subset
Large germplasm collections require extensive resources for characterization and evaluation of the entire collection. In addition, it is virtually impossible to evaluate the entire collection for many agronomic traits under uniform field conditions because of the large number of entries, their diverse growth habits, and wide maturities. Frequently, only portions of the collection are evaluated at any one time. For example, a collection of 20000 accessions might take five to ten years to fully characterize phenotypically. Thus, developing a core subset allows full characterization in a relatively short period of time and requires fewer growing environments that can confound results. In addition, new traits that are important to the research community may be over time identified and should be included as descriptors. Recently, the Rice Germplasm Committee recommended evaluating the collection for resistance to three diseases, blast (caused by Pyricularia grisea), sheath blight (caused by Rhizoctonia solani), and straighthead (a physiological malady) (USDA-ARS, NGRP, 2006). In 2004 in Arkansas, which produces 50% of the USA's rice, fungicides were relied on for controlling blast, sheath blight, and kernel smut diseases on approximately 40% of the state's acreage (Cartwright et al., 2005). Moreover, new races of blast are being identified that overcome major resistance genes that are commonly deployed in U.S. breeding programs (Lee et al., 2005). Out of 18412 accessions in the whole collection of rice, information on resistance to blast and sheath blight diseases is available only for about 300 accessions (<2%) (USDA-ARS, NGRP, 2006). A costly management method, draining and drying the field about 50 d after seedling emergence for preventing straighthead, is applied for 35% of the Arkansas rice area (Yan et al., 2005a). Although genetic resistance to straighthead would be a valuable option, only 25% of accessions in the whole collection have been evaluated for this trait. Premium quality is important for U.S. rice to be competitive in the world market (McClung, 2002), but less than 40% of the accessions in the whole collection have information on grain quality descriptors (amylose, ASV, kernel length, kernel width, ratio of kernel length to width, and kernel weight).
In an effort to better characterize the diversity of the rice collection, the core subset has been evaluated not only for agronomic descriptors (Yan et al., 2005b), but also for kernel dimension traits that impact milling yield and market class (Yan et al., 2005c), resistance to straighthead (Yan et al., 2004), and DNA markers associated with cooking quality and blast resistance (McClung et al., 2004, 2006; Fjellstrom et al., 2006). Evaluation for sheath blight resistance is underway, as is screening for Fe and Zn micronutrient content, which is being conducted in collaboration with the Genetic Institute of Chinese Academy of Sciences. As additional DNA markers are identified that are associated with economically important traits in rice, the core collection will be genotyped for these traits. In addition, research is underway to develop a highly purified core that will be genotyped with some 200 random DNA markers that will serve as the basis for future association mapping studies. This marker assessment can also be used to identify accessions that are genotypically redundant and can be replaced by other more diverse accessions. Having a core subset that is representative of the whole collection is critical for these phenotypic and genotypic assessments to be meaningful.
These results demonstrate that the USDA core subset of rice germplasm, which has been characterized for 14 descriptors, is very representative of the diversity of the whole NPGS rice collection. Phenotypic descriptors that are difficult or costly to measure can be evaluated using a core subset to reveal the diversity in the collection as well as identify accessions and regions of the world that offer unique phenotypes that can be explored in greater detail. In addition, molecular tools that could be cost prohibitive if used on the whole collection can be used on a core subset to provide a better assessment of the allelic diversity and genetic substructure of the collection. Thus, as demonstrated by the USDA rice core subset, a core strategy is an effective tool for characterizing a large germplasm collection at both phenotypic and genotypic levels. It is expected that this will lead to greater utilization of diverse germplasm in the collections for genetic studies, gene discovery, and breeding purposes.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Received for publication July 6, 2007.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
