HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Gravois, K. A. Articles by Bernhardt, J. L. Search for Related Content PubMed Articles by Gravois, K. A. Articles by Bernhardt, J. L. Agricola Articles by Gravois, K. A. Articles by Bernhardt, J. L.

CROP BREEDING, GENETICS & CYTOLOGY

Heritability and Genotype x Environment Interactions for Discolored Rice Kernels

^a Louisiana State Univ. Agric. Ctr., Sugar Res. Stn., P.O. Box 604, St. Gabriel, LA 70776 USA
^b Univ. of Arkansas, Rice Res. and Ext. Ctr., P.O. Box 351, Stuttgart, AR 72160 USA

kgravois{at}agctr.lsu.edu

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results and discussion REFERENCES

 
Discolored kernels are but one component of rice (Oryza sativa L.) quality that requires attention by rice breeders. Discolored kernels are most often caused by damage from rice stink bugs [Oebalus pugnax (Fabricius)] and kernel smut disease (Tilletia barclayana Bref.), but other pathogens and physiological disorders can also contribute to kernel discoloration. Our objective was to further understand the inheritance and genotype x environment (GE) interactions for discolored rice kernels to improve rice kernel quality characteristics. Thirty-seven genotypes, representing southern U.S. rice germplasm, were evaluated for susceptibility to all causes of discolored kernels in field tests conducted during 1993 and 1994 at three Arkansas locations: Stuttgart, Tupelo, and Rohwer. Traits were inherited quantitatively, with single-plot heritability values of 0.07, 0.18, and 0.33 for rice stink bug damage, kernel smut, and other discolorations, respectively. The GE interactions were significant for all traits. Closer inspection of the GE interaction revealed that the causes were primarily related to magnitude changes but also included genotype rank changes. The bias of GE interaction during selection could be reduced by multi-year and multi-location testing. Phenotypic correlations among traits were low, indicating that selection for lower levels of one trait would not adversely raise the levels of another trait. Cultivars, such as Katy, Kaybonnet, and Drew, were identified as having stable, low susceptibility to rice stink bug damage, kernel smut, and other discolorations. Although breeding for improved kernel quality traits will have certain difficulties because of low heritability, the availability of good germplasm and a proper screening program should minimize this problem.

Abbreviations: GE, genotype x environment

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results and discussion REFERENCES

 
RICE QUALITY TRAITS are of vital interest to rice breeders in the southern USA. Unfortunately, kernel discolorations, often called "peck" or "pecky rice", negatively affect rice quality. Collectively, kernel discolorations are caused by insects and pathogens (Fulton, 1908; Douglas and Tullis, 1950), or are now thought to be caused by physiological disorders (i.e., linear damage as described by Douglas and Tullis, 1950). In the mid South, rice kernels are discolored by fungi introduced by the rice stink bug (Lee et al., 1993). Adults and nymphs feed on developing rice kernels shortly after floret fertilization and throughout the soft dough stage. The stage of kernel development determines the amount and type of damage (Swanson and Newsom, 1962). Feeding during the early stages of kernel development prevents kernel development and results in total grain loss. Feeding during the kernel-fill stages often results in a portion of the contents being removed. After the hull is pierced by rice stink bugs, secondary infection by pathogens often results in kernel discoloration, chalkiness, and weakening of the kernel. The amount of damage caused by rice stink bugs affects the acceptability and value of rough rice.

Cultivars vary in the amount of damage caused by the rice stink bug, but one general conclusion is that medium-grain cultivars sustain more damage from rice stink bugs than long-grain rice cultivars (Helm, 1953; Rolston et al., 1966; Bernhardt, 1993). In addition, cultivars within a grain type also vary. For example, among long-grain cultivars, Jefferson, Katy, and LaGrue typically have low amounts of rice stink bug damage, whereas Alan, Newbonnet, and Cypress have high levels of damage.

Another source of rice kernel discoloration is kernel smut. After disease infection, mycelia grow within the developing rice kernel, eventually consume the endosperm, and produce numerous spherical black teliospores (Whitney and Frederiksen, 1975). The teliospores survive on the seed, rice straw residue, and in the soil. These spores overwinter, float to the water surface during the next rice crop, germinate, and produce airborne sporidia that infect the florets. The fungus is capable of infecting florets before anthesis, and infection is enhanced by high moisture during heading (Cartwright et al., 1997). Kernel smut is more severe when high N fertilizer levels are applied (Templeton et al., 1960) and on land with continuous rice or water-seeded rice. Severe kernel smut results in a reduction of rough rice yields, low milling quality, and inferior final milled product. Currently, the only practical means of controlling kernel smut in rice is adoption of resistant cultivars.

Kernel smut is more severe in the mid South in Arkansas, Mississippi, and Missouri than in the gulf coastal rice growing regions of Louisiana and Texas. From 1996 through 1998 in Arkansas, rough rice loads were penalized by the mills for excessive damage from kernel smut. Cultivars such as Cypress and LaGrue are severely affected by kernel smut, while other cultivars such as Bengal, Mars, and Lemont have excellent field resistance.

Discolored kernels negatively affect rice quality during the milling process. When hulls and bran are removed, the abrasion of one kernel upon another causes the breakage of kernels that were damaged by rice stink bugs. Milling seldom completely removes the discolored portion of linear and rice stink bug damaged kernels. Milling will remove damage caused by kernel smut and some other pathogens, but often the kernels are chalky and smaller than normal. Regardless of the cause, the discoloration of damaged kernels is enhanced by parboiling. In addition to the obvious discoloration on individual grain, healthy grain in a heavily smutted load are discolored by smut spores dispersed during milling and by absorption of water-soluble pigments from the spores during the parboiling process. Thus, discolored kernels will either decrease the percentage of whole milled kernels and percentage of total milled rice (broken kernels plus whole kernels), or decrease the quality of white and parboiled rice, or both.

Kernel discolorations are receiving increased attention in the Arkansas rice breeding program. Our objective was to further understand the heritability and GE interactions for discolored kernels so that rice breeders can effectively improve kernel quality characteristics.

	Materials and methods

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results and discussion REFERENCES

 
The Arkansas Rice Performance Trials are conducted annually at five locations in Arkansas and one location in Missouri to assess the adaptability of experimental rice lines and cultivars across the mid-South rice growing region. The traits of interest are rough rice yield, milling yield (levels of head rice and total milled rice), maturity, plant height, lodging, and kernel weight. Head rice refers to the percentage of rice kernels that remain three-fourths their normal length or more after the hulls and bran are removed during milling. Total milled rice includes head rice and the remaining broken milled pieces. The rice harvested from the Arkansas Rice Performance Trials is also inspected for kernel discolorations.

For this study, data were obtained from the Arkansas Rice Performance Trials conducted in 1993 and 1994 at three Arkansas locations: the Rice Research and Extension Center near Stuttgart, Hardin Farms near Tupelo, and the Southeast Branch Experiment Station at Rohwer. Plots were six rows wide on 19-cm row centers and 4.6 m long and seeded at a rate of 430 seed m^-2. The trials were conducted under typical management practices for dry-seeded drilled rice for Arkansas conditions and did not receive insecticide or fungicide treatments.

Each year 72 rice genotypes were grown in a randomized complete block design with three replications per location. Sixteen of the genotypes were cultivars, and 56 genotypes were experimental lines not previously selected against kernel discolorations. A balanced subset of genotypes was used for this study. Thirty-seven genotypes were common among the two years of testing: 16 cultivars and 21 experimental lines. Southern U.S. rice cultivars are derived from a narrow germplasm base as shown by Dilday (1990) by pedigree analysis and Saghai Maroof et al. (1997) by DNA polymorphism. Thus, the population used in this study was deemed representative of southern U.S. rice germplasm.

The plots were hand harvested when the rice attained 180 to 200 g kg^-1 moisture content. The thresher cylinder speed was calibrated to 600 rpm before threshing. The threshed samples were dried overnight at 32°C to 120 g kg^-1 moisture content and stored at 15°C and 45% relative humidity.

After drying, a 250-g sample of rough rice was taken from each sample and hulled. The brown rice was passed three times through an electronic sorting machine that separated discolored kernels from other kernels. The discolored kernels were examined with magnification to determine the cause of discoloration. Discolored kernels were placed into three groups: (i) rice stink bug damaged; (ii) kernels infected by the kernel smut pathogen; and (iii) other kernel discolorations such as minor levels of pink, red, or yellow kernel discolorations, brown discolorations on the bran often caused by Helminthosporium spp. (Lee and Tugwell, 1980), and linear damage. The amount of discolored kernels in a category was expressed as a percentage of the total weight of brown rice adjusted to a 100 g kg^-1 moisture content.

The data were balanced. Restricted maximum likelihood estimates of the variance components were estimated by the following model:

(1)

where T_ijkl was observation of the lth genotype (G) in the kth replicate (R) within the ith year (Y) and jth location (L); µ was the overall mean; and E_ijkl was the residual error. Year and location were considered as fixed effects, and genotype and replication were considered as random effects. All terms in the model were assumed to be uncorrelated to each other and to the residual error. Genotype mean squares were tested for significance by a pseudo-F-test, which constructs appropriate mean squares and degrees of freedom for tests of significance (Hicks, 1982).

Heritabilities were calculated by the formula:

(2)

where ²_G, ²_GY, ²_GL, ²_GYL, and ²_E refer to the G, G x Y, G x L, G x Y x L, and error variances, respectively; y, l, and r refer to the number of years, locations, and replications per location per year, respectively. Heritability on a single-plot basis was estimated with Y = L = r = 1. Heritability on an entry-mean basis reflects the annual selection practice and was estimated with Y = 1, L = 3, and r = 3. Heritability and variance component standard errors were estimated as shown by Dickerson (1969). Genetic advance was expressed as a proportion of the experimental mean to compare among the three traits for potential improvement by selection. Genetic advance was estimated for the entry-mean heritability as . Phenotypic standard deviation (_p) was equated to the square root of the denominator of . A 10% selection intensity with i = 1.76 was used (Allard, 1960).

To test for crossover interactions, all possible quadruple combinations of two genotypes in two environments were constructed (Azzalini and Cox, 1984). The number of possible quadruple combinations for the 37 genotypes and six environments was 9990. The significance of any quadruple depends on the difference of the two genotype means exceeding a critical value LSD (P = 0.05) in one environment, while the negative difference exceeds the critical value in the other environment (Baker, 1990). The LSD values for rice stink bug damage, kernel smut, and all other discolorations was 0.699, 0.149, and 0.437, respectively. The constructed mean square used to test the genotype mean square for significance was used as the error term in determining the LSD. Significant differences between year and location means or at any year–location series were determined by the LSD (P = 0.05), using the mean square for replications within locations and years as the error term. Phenotypic correlations (partial correlation coefficients, r_p, df = 432) were calculated among the three traits after adjusting for replication, year, and location effects.

Stability analyses were conducted for each trait as described by Kang and Magari (1996) after verifying that the genotype x environment (year–location series) interaction was significant. These analyses calculate Shukla's (1972) stability variance statistic and Kang's yield stability statistic. Stability statistics were calculated by a computer program (STABLE) provided by Kang and Magari (1995) that integrates yield and stability performance into a single selection criteria. Their program was run with some modification. For typical yield stability analyses, genotypes are ordered from highest to lowest with low values receiving a Y' value of one. Low values of discolored traits are desirable, unlike yield where high values are desired. For this analysis, genotypes were ordered from lowest to highest for rice stink bug damage, kernel smut, and other discolorations. For ranking of the means (Y'), the highest values received a rank of one.

	Results and discussion

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results and discussion REFERENCES

 
Year and location main effects, as well as their interaction, were significant for the three categories of discolored kernels (Table 1) . Although weather and other environmental conditions were not monitored, conditions necessary for infection by kernel smut or other pathogens and conditions that influence rice stink bug densities probably differed among years and locations within years. These results were consistent with our past experiences with discolored kernels and the mean values of kernel discolorations observed both among years and locations in this study (Table 2) .

For other discolorations, genetic and residual variance components were similar and were larger than the other variance components. The G x Y x L variance was larger than both the G x Y and G x L variance components. Of the three categories for discolored rice kernels, heritability was highest for other discolorations on both the single-plot and entry-mean basis.

We concluded from the analysis of variance that discolorations caused by rice stink bug damage, kernel smut, and other sources are highly subject to GE interactions. The presence of GE interactions will complicate breeding efforts to improve rice kernel quality by decreasing levels of discolored kernels. However, based on genetic advance estimates, selection for improved kernel quality would be easiest for kernel smut followed by other discolorations. Genetic advance for improved resistance to rice stink bug damage through selection would be the most difficult. Multi-year and multi-location screening of new rice experimental lines is essential to adequately account for the GE interactions for the three categories of kernel discolorations.

Further examination of the GE interaction is necessary to determine if the interaction is primarily due to genotype rank changes or magnitude changes. Genotype rank changes would have a more serious effect on selection progress for improving the effects of discolored kernels than would magnitude changes. Rank changes were estimated by the number of crossover interactions. The number of crossover interactions was largest for values of rice stink bug damage and the third category of other discolorations, but <1% of the total possible crossover interactions was found for any category (Table 3). No crossover interactions were detected for kernel smut. The low number of crossover interactions was probably due to the large residual variance and high coefficient of variation (33, 72, and 45% for rice stink bug damage, kernel smut, and other discolorations, respectively). Significant crossover interactions were therefore underestimated due to inflated LSD values for all categories. Both magnitude changes and crossover interactions may be important when screening new experimental lines and cultivars for discolored kernels. To reduce the GE interaction bias for rice stink bug damage, screening in more locations would be helpful. The large G x Y x L variance for kernel smut and other discolorations indicate that additional locations and years would be necessary to decrease bias due to GE interaction.

Although other discolored kernels was significantly correlated with both stink bug damage and kernel smut, the association was weak r_p = 0.11 and r_p = 0.10 (P < 0.05), respectively. Kernel smut and stink bug damage were not significantly correlated. Strong positive correlations among the three categories would have simplified selection because selecting for low levels of one trait would have lowered levels of the others. Each of the three traits appears to be inherited independently, and selection for reduced levels of discolored kernels must be conducted via independent programs. Selection is necessary to improve rice kernel quality because of the rising demand for improved rice kernel quality in new rice cultivars.

Selection progress begins with the identification of resistant parents. Three cultivars (Drew, Katy, and Kaybonnet) and four experimental lines (RU9201124, RU9201179, RU9201191, and RU9301151) were identified as having stable and low levels for all three kernel discolorations (Table 4) . Katy appears to be an excellent source of resistance to rice stink bug damage, kernel smut, and other discolorations and is a common parent for Drew, Kaybonnet, RU9201124, RU9201179, and RU9201191. One obvious void in this sample of germplasm was a medium-grain genotype with resistance to rice stink bug damage. Evaluation of the USDA rice world collection identified Dular as a medium-grain cultivar with resistance to rice stink bug damage (Bernhardt and Dilday, 1992). Dular, a cultivar from India, has other characteristics that may preclude its widespread use as a parent for breeding resistance to rice stink bug damage, such as red bran, shattering, and poor straw strength.

	NOTES

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results and discussion REFERENCES

 
Approved for publication by the Director of the Arkansas Agric. Exp. Stn. Research supported in part by a grant from the Arkansas Rice Res. and Promotion Board.

Received for publication March 22, 1999.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results and discussion REFERENCES

 

• Allard R.W. Principles of plant breeding. New York: John Wiley & Sons, 1960.
• Azzalini A., Cox D.R. Two new tests associated with the analysis of variance. J. Royal Stat. Soc. Ser. B (Methodological) 1984;46:335-343.
• Baker R.J. Crossover genotype–environmental interaction in spring wheat. In: Kang M.S., ed. Genotype-by-environment interaction and plant breeding. Baton Rouge: Louisiana State Univ, 1990:42-51.
• Bernhardt, J.L. 1993. Screening for rice stink bug resistance. p. 96–103. In B.R. Wells (ed.) Arkansas Rice Res. Studies 1992. Arkansas Agric. Exp. Stn. Res. Ser. 431.
• Bernhardt, J.L., and R.H. Dilday. 1992. Management strategies for rice pests: Screening for rice stink bug resistance. p. 46–49. In B.R. Wells (ed.) Arkansas Rice Res. Studies 1991. Arkansas Agric. Exp. Stn. Res. Ser. 422.
• Cartwright, R.D., W.J. Ross, C.E. Parsons, F.N. Lee, R.L. Eason, and G.E. Templeton. 1997. Artificial inoculation of kernel smut of rice in Arkansas. p. 90–93. In R.J. Norman and T.H. Johnston (ed.) Arkansas Rice Res. Studies 1996. Arkansas Agric. Exp. Stn. Res. Ser. 456.
• Dickerson, G.E. 1969. Techniques for research in quantitative animal genetics. In Techniques and procedures in animal science research. Am. Soc. Anim. Sci., Albany, NY.
• Dilday R.H. Contribution of ancestral lines in the development of new cultivars of rice. Crop Sci. 1990;30:905-911.[Abstract/Free Full Text]
• Douglas W.A., Tullis E.C. Insects and fungi as causes of pecky rice. Washington, DC: U.S. Gov. Print. Office, 1950 USDA Tech. Bull. 1015..
• Fulton H.R. Diseases affecting rice. Baton Rouge, LA: Louisiana State Univ. Bull. 105, 1908.
• Helm, R.R. 1953. Research on rice field insects. p. 127–128. In 45th Annual Progress Report. Rice Exp. Stn., Crowley, LA.
• Hicks C.R. Fundamental concepts in the design of experiments. New York: CBS College Publ, 1982.
• Kang M.S., Magari R. Stable: A basic program for calculating stability and yield-stability statistics. Agron. J. 1995;87:276-277.[Abstract/Free Full Text]
• Kang M.S., Magari R. New developments in selecting for phenotypic stability in crop breeding. In: Kang M.S., Gauch H.G., Jr., eds. Genotype-by-environment interaction. Boca Raton, FL: CRC Press, 1996:1-14.
• Lee F.N., Tugwell N.P. Pecky rice and quality reduction in Arkansas rice production. Arkansas Farm Res. 1980;29:2.
• Lee F.N., Tugwell N.P., Fannah S.J., Weidemann G.J. Role of fungi vectored by rice stink bug (Heteroptera: Pentatomidae) in discoloration of rice kernels. J. Econ. Entomol. 1993;86:549-556.
• Rolston L.H., Rouse P., Mayes R. Pecky rice. Arkansas Farm Res. 1966;15:6.
• Saghai Maroof M.A., Yang G.P., Zhang Q., Gravois K.A. Correlation between molecular marker distance and hybrid performance in southern U.S. long grain rice. Crop Sci. 1997;37:145-150.[Abstract/Free Full Text]
• Shukla G.K. Some statistical aspects of partitioning genotype–environmental components of variability. Heredity 1972;29:237-245.[ISI][Medline]
• Swanson M.C., Newsom L.D. Effect of infestation by the rice stink bug, Oebalus pugnax on yield and quality in rice. J. Econ. Entomol. 1962;55:877-879.
• Templeton G.E., Johnston T.H., Henry S.E. Kernel smut of rice. Arkansas Farm Res. 1960;9:10.
• Whitney, N.G., and R.A. Frederiksen. 1975. Kernel smut of rice. Texas Agric. Exp. Stn. Misc. Pub. 1231. College Station, TX.

This article has been cited by other articles:

				X.-M. Fan, M. S. Kang, H. Chen, Y. Zhang, J. Tan, and C. Xu Yield Stability of Maize Hybrids Evaluated in Multi-Environment Trials in Yunnan, China Agron. J., January 1, 2007; 99(1): 220 - 228. [Abstract] [Full Text] [PDF]

				N. A. Slaton, E. E. Gbur Jr., R. D. Cartwright, R. E. DeLong, R. J. Norman, and K. R. Brye Grain Yield and Kernel Smut of Rice as Affected by Preflood and Midseason Nitrogen Fertilization in Arkansas Agron. J., January 1, 2004; 96(1): 91 - 99. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Gravois, K. A. Articles by Bernhardt, J. L. Search for Related Content PubMed Articles by Gravois, K. A. Articles by Bernhardt, J. L. Agricola Articles by Gravois, K. A. Articles by Bernhardt, J. L.