Crop Science 41:1706-1710 (2001)
© 2001 Crop Science Society of America
CROP BREEDING, GENETICS & CYTOLOGY
Inheritance of Sugarcane Borer Resistance in Sugarcane Derived from Two Measures of Insect Damage
W. H. White*,a,
J. D. Millerb,
S. B. Milliganc,
D. M. Burnerd and
B. L. Legendree
a USDA-ARS-SRRC, Sugarcane Research Unit, 5883 USDA Road, Houma, LA 70360
b USDA-ARS, Sugarcane Field Station, HCR Box 8, Canal Point, FL 33438
c United States Sugar Corp., P.O. Drawer 1207, Clewiston, FL 33440
d USDA-ARS, Dale Bumpers Small Farms Research Center, 6883 State Highway 23, Booneville, AR 72927
e LSU Ag. Center, Plant Science Division P.O. Box 251000, Knapp Hall, Baton Rouge, LA 70894
* Corresponding author (wwhite{at}srrc.ars.usda.gov)
 |
ABSTRACT
|
|---|
The sugarcane borer [Diatraea saccharalis (Fabricius)] is an important insect pest of sugarcane grown in the Americas. Environmental and economic concerns are driving these sugarcane industries to consider alternatives to insecticides for controlling damaging infestations of the borer. Breeding for resistance is a viable option; however, little is known of the inheritance of sugarcane borer resistance. The inheritance of sugarcane borer resistance in sugarcane (Saccharum spp. L.) was investigated in a field study conducted in 1990, 1992, and 1993. We measured resistance by both plant damage response ratings and mean percent internodes damaged. Seedling progeny (F1 plants generated from seed) from 21 to 27 crosses were evaluated each year. These progeny originated from a mating design with females nested within males. Parental genotypes were randomly selected for borer resistance, but were elite cultivars adapted to Louisiana. Data were collected from progeny infested with artificially introduced sugarcane borers. Narrow-sense heritability on a single-plot basis (36 plants measured per plot) for damage ratings 
and for percent damaged internodes 
were high and of comparable magnitude. For both traits, we detected neither dominance nor additive x year interaction; however, dominance x year interaction variance existed. The potential for genetic advance (GA) from direct selection against percent damaged internodes (GA = 33.9% of mean bored internode) was higher than that from direct selection for lower damage rating (13.5% of mean rating). The much greater resources needed to effect selection for percent bored internodes (approx. 24 times that for rating) suggested direct selection for damage rating may be more efficient. Because the traits were highly correlated 
and their heritabilities high, correlated gains in percent damaged internodes by direct selection for damage rating were nearly as high as direct selection for percent damaged internodes (31% indirect vs. 33.9% direct).
|
INTRODUCTION
|
|---|
THE SUGARCANE BORER is historically the most important insect pest of sugarcane in Louisiana (Long, 1969). Larval borers damage cane by boring into cane stalks were they ultimately pupate and emerge as adults. Individual larvae generally stay within a single internode. Since 1969, damaging infestations of the sugarcane borer in Louisiana have been controlled by an integrated pest management program (IPM) comprised of cultural measures, natural enemies, plant resistance, and pesticide application (Hensley, 1971). Hensley (1981) estimated that cultural controls contribute approximately 10% of season-long control, plant resistance and biological control each contribute 25%, and chemical control the remaining 40%. This pest management program has provided effective and stable control of the sugarcane borer for approximately 30 yr. Pest management programs that rely on insecticides as a principal control tactic, however, are under increasing economic and environmental pressures to reduce the IPM's dependency on insecticides.
Plant resistance may provide the additional control needed to supplant insecticides in the IPM program. Studies on plant resistance to the sugarcane borer have been published (Mathes and Ingram, 1944; Long et al., 1978). These studies have included research to determine mechanisms of sugarcane borer resistance and methods to select for resistance (Kyle and Hensley, 1970; White and Hensley, 1987). Little is known of the inheritance of sugarcane borer resistance and breeding methods required to increase resistance in clonal populations. Viator and Henderson (1971) found borer resistance to be quantitative in nature, but provided no measures of genetic variation, heritability, or potential gain from selection.
Sugarcane is a clonally propagated, out-crossing, perennial crop that growers routinely harvest once per year for about 3 yr before replanting. Ratoon crops follow the initial plant cane crop. Commercial sugarcane varieties are generated from material originally derived from interspecific crosses of sweet noble cultivars of Saccharum officinarum L. and vigorous wild relatives, principally S. spontaneum L. The complex polyploid nature of sugarcane commonly generates anueploids although evidence exists that commercial varieties regularly exhibit bivalent pairing (Price, 1963). These observations suggest that sugarcane acts like an allopolyploid and hence the assumption of diploid inheritance might be valid.
Quantitative studies of inheritance on sugarcane have principally focused on agronomic traits and disease resistance (Brown et al., 1968; Tai et al., 1981; Gravois et al., 1991a,b; Hogarth et al., 1983; Hogarth et al., 1993; Milligan et al., 1990; Yin et al., 1996). The objective of our study was to determine the inheritance of sugarcane resistance to the sugarcane borer employing two methods to measure insect damage. Such information will be useful in designing efficient breeding programs to increase borer resistance in sugarcane cultivars.
|
MATERIALS AND METHODS
|
|---|
This study consisted of a series of experiments conducted during the 1990, 1992, and 1993 growing seasons at the USDA-ARS, Ardoyne Research Farm near Chacahoula, LA. We used a parental population of 62 elite clones that had not been selected for borer resistance but which were adapted to Louisiana conditions. The inference population was from the Louisiana cultivar development breeding population for commercial cultivars (Table 1). Each year, we tested (F1) progeny populations from nine genotypes used as males that had been crossed with two to five unrelated genotypes used as females. Fifty-five different female genotypes were used during the study. Some genotypes were used as both males and females. Each year progeny from 21 to 27 biparental crosses were arranged by family and analyzed using a nested mating design model (Design 1) (Comstock and Robinson, 1948; Nyquist, 1991, p. 271). The crosses were either made the year previous to planting or were crosses made during earlier crossing campaigns and the seed stored at -18°C. Parents were not inbred.
F1 individuals from each cross were germinated in a greenhouse in January and February and transplanted to the field during April in a randomized complete block design with four blocks. Specific progeny genotypes were tested for only 1 yr. Approximately 36 seedlings from each family were planted in a single-row plot with an intra-row plant spacing of 45 cm and an inter-row spacing of 1.8 m. Progeny were planted in a two to one skip-row configuration; two rows of cane to one row of maize (Zea mays L.). Infested maize rows were interspersed among the cane rows to act as spreader rows. Maize was planted at the same time as the cane and served as a host for artificial inoculation with sugarcane borer larvae. Maize was drilled and later thinned to a density of 23 000 plants ha-1 (4.2 plants m-1 row). Approximately 20 d after planting, individual maize plants were infested with 10 ± 2 neonate sugarcane borer larvae, a procedure that has proved dependable in screening trials (White, 1993).
Standard Louisiana sugarcane cultural practices were followed for cultivation, fertilization, and weed control. In addition, chlorpyrifos [O,O-diethyl-O-(3,5,6-trichloro-2-pyridinyl) phosphorothioate] was broadcast-applied at a rate of 1.3 kg ai ha-1 to control populations of the red imported fire ant, Solenopsis invecta (Buren). These generalized predators are effective at removing sugarcane borer, and when left uncontrolled may prevent uniform borer populations from developing in small field plots (Reagan et al., 1972).
Single stalk data were collected simultaneously at harvest (in late November or early December) and consisted of determining percent damaged internodes (single-stalk plant-1 sample in 1990; 2-stalk plant-1 sample in 1992 and 1993) and giving damage response ratings (single plant rating). Percent damaged internodes were measured by the ratio of bored internodes to un-bored internodes expressed as a percentage. Sugarcane typically produces 15 to 20 internodes per stalk. Stalks were not split, nor did we determine any degree of internal damage response; however, larval entrance sites are clearly identifiable following removal of leaf-sheaths. Damage response ratings were used to assess a plant's response to borer feeding. The ratings considered the production of lateral buds, and broken or dead tops in addition to the percentage of the leaf sheaths showing feeding sign prior to the larvae entering the stalk. Accumulation of frass at the leaf-sheath and reddening of the sheath are indications of larval feeding activity. Damage response ratings were based on a 1-to-9 scale, where 1 indicated little borer damage and 9 indicated heavy borer damage (Table 2).
Performing the analysis among all 3 yr and all families, we obtained restricted maximum likelihood (REML) variance and covariance components using the random model (Proc Mixed; SAS, 1996):
where Dijklm is the response of Plant m in Year i of Block j of Male k and Female l, µ is the overall mean, Yi is Year i effect, B(Y)ij is Block j in Year i effect, Mk is Male k effect, F(M)kl is Female l within Male k effect, MYik is interaction Male k and Year i effect, F(MY)ikl is Female l within Male k and Year i effect, B(MY)ijk is Block j in Year i and Male k effect, B(FMY)ijkl is Block j in Year i, Male k, Female l, and Year i effect, 
ijlkm error term, i.i.d. N
.
Additive (
2A) and dominance (2D) genetic variance were estimated as 2A = 42M and 2D = 4
. We estimated narrow-sense heritability as:
where b was the number of blocks, y was the number of years, p was the number of plants measured per plot, and 2P was the phenotypic variance. To offer standard bases of comparisons, display relative influences of different sources of variation, and to mimic the selection approach to be used in the breeding program, we calculated three types of heritabilities: a single plant basis 
, a single plot basis 
, and an entry mean basis 
. Standard errors of heritabilities were calculated by Dickerson's approximation (Dickerson, 1969) i.e., std. error of the h2 
. The additive genetic coefficient of variation (ACV) was provided to better compare the relative genetic variation of the two damage traits. It was calculated as: ACV = 100 A/mean. For the three selection scenarios (single plant, single plot and entry mean), expected GA was calculated as a percentage of the mean as GA = 100ih2P/mean. A 10% selection intensity with i = 1.755 was used in calculations (Becker, 1984, p173).
The additive correlation (rA) between percent bored internodes (BI) and the damage ratings (RATE) was calculated as (Becker, 1984, p118):
where A-BI,RATE was the additive genetic male covariance of BI and RATE, 2A-BI was the additive genetic variance for BI, and 2A-RATE was the additive genetic variance for RATE. Expected correlated response to selection was calculated as a percentage of the mean as:
where CRX was the correlated response of Trait X to direct selection for trait Y. The square roots of the heritabilities (hBI and hRATE) and phenotypic variances (PX) were used in the calculations. Two correlated responses were calculated. The expected correlated response to selection was calculated for BI (CRBI) by selecting parents for RATE. Another response was figured for indirect improvement of RATE (CRRATE) in progeny by selecting parents for BI.
Because sugarcane is a clonally propagated crop, a broad-sense genetic analysis is also of interest. Potential gain from clonal mass selection was compared with six selection scenarios that combined family (cross) and individual selection within families. Because of the need to estimate genetic plant-to-plant variation not confounded with nongenetic residual error variance, gain estimates were made for only bored internodes and used data from 1992 and 1993 when two stalks per plant (stool) were measured. We used progeny from families tested in all the years to obtain REML variance components from the random model:
where Dijklm is the response of stalk m in Year i of Block j of Family k and Plant l, µ is the overall mean, Yi is Year i effect, B(Y)ij is Block j in Year i effect, Ck is Family k effect, CYik is interaction Family k and Year i effect, P(CBY)ijkl is Plant l within Family k, Block j and Year i effect, and ijlkm error term, i.i.d. N(0, 2).
We estimated GA using clonal mass selection of individuals as:
This assumed a 10% selection intensity from population of infinite size 
, and Hmass = 
= 
, where p = b = y = 1 and stalks 
= 2.
Expected genetic advance using combined family and individual within family selection was calculated as:
This assumed a 50% selection intensity among a population of 100 families (if = 0.792) and a 20% selection intensity among 125 individuals within each selected family (if =1.388). The family heritability was calculated as
where the number of plants (p), blocks (b), years (y), and stalks (s), varied among the six family selection scenarios. The individual heritability within selected families was estimated as:
|
RESULTS AND DISCUSSION
|
|---|
Seven crosses were common in the 1990 and 1992 studies. Four crosses were common in the 1992 and 1993 studies. We conducted combined analyses of variance for the years 1990–1992 and 1992–1993 and analyses for individual years (analyses not shown). These analyses showed that although variability was high within a year, the 2CY was not important, thereby, justifying the use of a pooled analysis. We assumed that including an estimate of 2CY, poor as it may be, provided an estimate of heritability less biased than assuming no 2CY.
Analyses showed that the male component was more important than the female-within-male component, indicating that additive variance was more important than dominance (Table 3). The standard errors associated with genetic estimates were large and can probably be attributed to the poor commonality of crosses among years and the relatively few crosses used.
View this table:
[in this window]
[in a new window]
|
Table 3. Variance components, means, heritabilities, additive coefficient of variation (ACV), and genetic advance (GA) for sugarcane borer damage measures.
|
|
Single plot heritabilities for ratings and mean percent internodes damaged were 0.73 and 0.76 (Table 3). These heritabilities are high and indicate that success in transferring resistance into progeny populations should be readily possible. The importance of using multiple plants in a plot was exemplified by the dramatic increase in heritability and the doubling of the predicted genetic advance. As expected, replicating the plots further increased predicted response to selection by a smaller degree.
Previous studies indicate that damage response ratings and determining mean percent damaged internodes measure different mechanisms of resistance (White and Hensley, 1987). Ratings measure a plant's response to feeding and are, therefore, indicative of tolerance. Determination of mean percent internodes damaged measure the success of a larva in establishing itself on the plant and thus, to some degree, is a measure of antibiosis. Data reported herein show that selection for both traits is possible and advances in resistance are expected; however, the greatest GA would be expected when selecting for mean percent internodes damaged (Table 3).
The labor requirements needed to select for internodes damaged are considerably greater than collecting damage response ratings because it requires collecting stalks and removing leaf-sheaths. Evaluating tests with approximately 40 selections replicated four times by determining percent damaged internodes on the basis of 10-stalk samples requires approximately 24 laborer hours compared with1 laborer hour when visual ratings are made. Therefore, it is not always practical to obtain damaged internode data in large, segregating populations and in breeding programs initial selections are made with damage ratings (White et al., 1996).
The genetic correlation between these two traits was high (Table 4). Single plot correlated response to selection estimated indirect selection for bored internodes by selecting directly for resistance from ratings (31% of bored internode mean; Table 4). The expected response to selection for ratings was nearly as good as direct selection for bored internodes (34% of bored internode mean; Table 3). Given that recording bored internodes costs approximately 24 times more than damage ratings, indirect selection for bored internodes via selection based on damage ratings is well justified.
View this table:
[in this window]
[in a new window]
|
Table 4. Correlated response to selection for percent bored internodes (BI) and damage rating (RATE) by selecting the other trait.
|
|
The rating system is, however, constrained from usefulness until about September because plants will not usually express damage signs such as lateral bud formation and broken tops until then. As the season progresses and the cane grows in height, the chance of lodging also increases. Thus, waiting too late in the season to rate, improves the chances of losing the opportunity to rate the plants because of lodging. Researchers commonly attempt to increase insect pressure to enhance selection of insect resistant plants. The economic threshold for borer damage in Louisiana is considered to be around 10% damaged internodes. At these levels, plant damage manifestations such as broken tops and lateral bud formation are seldom seen. Thus, the rating system has temporal and damage level constraints not experienced relative to the percent damaged internode measure.
A narrow-sense analysis provides information about potential gain from parental selection. A broad-sense analysis provides information about potential gains from clonal selection once the F1 populations have been developed. Family selection is being used in some sugarcane cultivar development programs for yield (Cox and Hogarth, 1993; DeSousa-Vierira and Milligan, 1999; McRae and Jackson, 1995). Its utility in selecting for borer resistance has not been investigated. We compared simple mass selection (selection of individuals without reference to families) with selection approaches that combined family and individual-within-family selection. Six different family entry mean scenarios were calculated to compare the effects of different types of replication, i.e., blocking vs. number of plants vs. number stalks. The overall selection rate of 10% was maintained to compare mass selection with combined selection.
All combined selection scenarios were better than mass selection (Table 5). Combined selection increased predicted gains from about 36% of the mean for mass selection to 42 to 50% of the mean, depending on the family mean scenario. Increasing the number of sampled stalks from one to two did little to increase gain, whereas increasing the number of sampled plants from 1 to 36 substantially increased gain. After increasing the number of plants to 36, replication did very little to increase predicted gain. The scenarios were chosen to illustrate differences or in some cases to provide typical mean scenarios of what might occur in the breeding program given certain production constraints. They were, however, rather arbitrarily chosen and are not construed as optimal. It seems clear that family selection promises to improve substantially the effectiveness of selection for resistance to percent bored internodes. It is assumed similar gains may be achieved for damage rating.
View this table:
[in this window]
[in a new window]
|
Table 5. Predicted broad-sense genetic advance (GA) for mass selection and family selection scenarios against percent bored internodes.
|
|
|
ACKNOWLEDGMENTS
|
|---|
The authors are especially grateful to Dr. Monica Balzarini, former graduate student, Agronomy Department, LSU Agricultural Center, Baton Rouge, LA (currently, University of Cordoba, Cordoba, Argentina) for helpful suggestions in analyzing the data.
Received for publication April 6, 2000.
|
REFERENCES
|
|---|
- Becker, W.A. 1984. Manual of quantitative genetics, 5th ed. Academic Enterprises, Pullman, WA.
- Brown, A.D.H., J. Daniels, and B.D.H. Latter. 1968. Quantitative genetics of sugarcane. I. Analysis of variation in a commercial hybrid sugarcane population. Theor. Appl. Genet. 38:361–369.
- Comstock, R.E., and H.F. Robinson. 1948. The components of genetic variance in populations of biparental progenies and their use in estimating the average degree of dominance. Biometrics 4:254–266.[Medline]
- Cox, M.C., and D.M. Hogarth. 1993. Progress and changes in the South Queensland variety selection program. Proc. Aust. Soc. Sugar Cane Technol. 15:251–255.
- DeSousa-Vieira, O., and S.B. Milligan. 1999. Intrarow plant spacing and family x environment interaction effects on sugarcane family evaluation. Crop Sci. 39:358–364.[Abstract/Free Full Text]
- Dickerson, G.E. 1969. Techniques for research in quantitative animal genetics. p. 36–79. In A.B. Chapman (ed.) Techniques and procedures in animal science research. Am. Soc. Anim. Sci. c/o Q Corp., New York.
- Gravois, K.A., S.B. Milligan, and F.A. Martin. 1991a. Indirect selection for increased sucrose yield in early sugarcane testing stages. Field Crops Res. 26:67–73.
- Gravois, K.A., S.B. Milligan, and F.A. Martin. 1991b. Additive genetic effects for sugarcane yield components and implications for hybridization. Trop. Agric. 68:376–380.
- Hensley, S.D. 1971. Management of sugarcane borer populations in Louisiana, a decade of change. Entomophaga 16:133–146.
- Hensley, S.D. 1981. Management of sugarcane insect pests. Sugar J. 44:18–18.
- Hogarth, D.M., J.F. Reimers, C.C. Ryan, and P.W.J. Taylor. 1993. Quantitative inheritance of Fiji disease resistance in sugarcane. Field Crops Res. 34:175–186.
- Hogarth, D.M., C.C. Ryan, and J.C. Skinner. 1983. Inheritance of resistance to rust in sugarcane - comments. Field Crops Res. 5:313–316.
- Kyle, M.L., and S.D. Hensley. 1970. Sugarcane borer host resistance studies. p. 55–67. In Proc. Louisiana Acad. Sci. 23–25 April 1970. McNeese State University, Lake Charles, LA.
- Long, W.H. 1969. Insecticidal control of moth borers in sugarcane. p. 149–161. In Pests of sugarcane. Elsevier, Amsterdam.
- Long, W.H., J.M.M. Walder, A.A. Delgado, and M.A.A. Ceasar. 1978. A detailed analysis of the effects of sugarcane borer [Diatraea saccharalis (F.)] damage on the weight and quality of sugarcane in one field experiment. Ecossistema 3:21–28.
- Mathes, R.J., and J.W. Ingram. 1944. Investigations of sugarcane borer (Diatraea saccharalis) control by the use of resistant varieties. Sugar Bull. 22:189–192.
- McRae, T.A., and P.A. Jackson. 1995. Selection of sugarcane families for the Burdekin River irrigation area. Proc. Aust. Soc. Sugar Cane Technol. 17:134–141.
- Milligan, S.B, K.A. Gravios, K.P. Bischoff, and F.A. Martin. 1990. Crop effects on broad-sense heritabilities and genetic variances of sugarcane yield components. Crop Sci. 30:344–349.[Abstract/Free Full Text]
- Nyquist, W.E. 1991. Estimation of heritability and prediction of selection response in plant populations. Crit. Rev. Plant Sci. 10:235–322.
- Price, S. 1963. Cytogenetics of modern sugar canes. Econ. Bot. 1: 97–106.
- Reagan, T.E., G. Coburn, and S.D. Hensley. 1972. Effects of mirex on the arthropod fauna of a Louisiana sugarcane field. Environ. Entomol. 1:588–591.
- Tai, P.Y.P., J.D. Miller, and J.L. Dean. 1981. Inheritance of resistance to rust in sugarcane. Field Crops Res. 4:461–468.
- Viator, D.P., and M.T. Henderson. 1971. Genetic behavior of resistance to sugarcane to the sugarcane borer, Diatraea saccharalis (F.). p. 276–285. In Proc. Int. Soc. Sugar Cane Technol. 22 October–5 November 1971. New Orleans, LA.
- White, W.H. 1993. Cluster analysis for assessing sugarcane borer resistance in sugarcane line trials. Field Crops Res. 33:159–168.
- White, W.H., and S.D. Hensley. 1987. Techniques to quantify the effect of Diatraea saccharalis (Lepidoptera: Pyralidae) on sugarcane quality. Field Crops Res. 15:341–348.
- White, W.H., B.L. Legendre, and J.D. Miller. 1996. Progress in breeding for sugarcane borer resistance. Sugar Cane 5:3–7.
- Yin, Z., J.W. Hoy, and S.B. Milligan. 1996. Evaluation and heritability of resistance to sugarcane red rot. Phytopathology 86:662–667.
This article has been cited by other articles:
|
|
|
|
 
S. B. Milligan, M. Balzarini, and W. H. White
Broad-Sense Heritabilities, Genetic Correlations, and Selection Indices for Sugarcane Borer Resistance and Their Relation to Yield Loss
Crop Sci.,
September 1, 2003;
43(5):
1729 - 1735.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
