| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
International Crops Research Institute for the Semi Arid Tropics (ICRISAT), Patancheru 502324, Andhra Pradesh, India
* Corresponding author (V.MAHALAKSHMI{at}CGIAR.ORG)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Abbreviations: % GLA, percent green leaf area • DAF, days after flowering • DAS, days after sowing
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Recent research has provided a better understanding of the genetics and the physiology of stay green. Greater green-leaf-area duration during grain filling appears to be a product of different combinations of three distinct factors: green leaf area at flowering, time of onset of senescence, and subsequent rate of senescence (van Oosterom et al., 1996; Borrell et al., 2000a). Further, all three factors appear to be inherited independently (Van Oosterom et al., 1996), and thus sources expressing different components can be combined easily in breeding programs (Borrell et al., 2000a). Stay-green hybrids have been shown to produce significantly greater total biomass after anthesis, to retain greater stem carbohydrate reserves, to maintain greater grain growth rates, and to have significantly greater grain yields under terminal drought stress than related but senescent hybrids (Borrell et al., 1999, 2000b). Stay-green genotypes also appear to have higher leaf-nitrogen concentrations (specific leaf nitrogen) at flowering and maintain these during grain filling (Borrell and Hammer, 2000), which is possibly associated with a higher transpiration efficiency in the best stay-green hybrid (Borrell et al., 2000c).
Conventional breeding for stay green has been based primarily on two sources for this trait, B 35 and KS 19. KS 19 is a selection from a cross of short Kaura, an improved landrace cultivar from northern Nigeria, with Combine Kafir 60 (Henzell et al., 1984). B 35 (PI 534133) was selected from a converted (dwarf height, early flowering) version of IS 12555, an Ethiopian landrace (Rosenow et al., 1983, 1996). KS 19 has been used commercially primarily in the breeding program of Queensland Department of Primary Industries, whereas B 35 is widely used in both public and private sector breeding programs in the USA. Recent research suggests that the two sources differ in the mechanisms by which they prolong leaf-area duration during grain filling. Both sources delay the onset of senescence. They differ in that B 35-derived lines have a greater leaf area at flowering and a normal rate of leaf senescence, whereas KS 19-derived lines have a smaller leaf area at flowering and a slower rate of leaf senescence (Borrell et al., 2000a).
Although the ability of leaves to delay senescence has a genetic basis in sorghum (van Oosterom et al., 1996), the expression of the character is strongly influenced by environmental factors. The trait expresses best in environments in which the crop is dependant upon stored soil moisture, but where this is sufficient to meet only a part of the transpiration demand. Sufficient expression of the trait for selection is thus dependant upon the occurrence of a prolonged period of drought stress during the grain-filling period of sufficient severity to accelerate normal leaf senescence, but not of sufficient magnitude to cause premature death of the plants. Because of this precise requirement for the trait expression, field environments do not offer ideal conditions for selection and molecular markers associated with this trait may offer the better alternative (Crasta et al., 1999; Xu et al., 2000).
The limited number of sources of stay green currently in use in sorghum breeding programs contrasts with its importance in improving adaptation to postflowering drought stress, and with the effort being invested in identifying molecular markers to transfer it to new lines more effectively. A search for additional sources of stay green, which may be different genetically or physiologically, would thus be worthwhile. The objective of the research reported in this paper was to evaluate critically a number of stay-green sources available at ICRISAT for their patterns of leaf senescence under terminal stress conditions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
|
In the 1998-1999 experiment, the 81 genotypes were arranged in a 9 x 9 lattice design with four replications. In the 1999-2000 experiment, 72 of the original 81 genotypes were arranged in a 9 (genotypes per block) x 8 (blocks per rep) alpha design, with three replications. The 1998-1999 experiment was planted on 10 November in 3-row plots, each row 4 m long and spaced 0.60 m apart. The 1999-2000 the experiment was planted on 21 October in 2-rows plots, each row 4 m long and 0.60 m apart. A basal application of 20 kg ha-1 N and 20 kg ha-1 P2O5 as di-ammonium phosphate was banded before sowing. The seeds were machine planted and the field irrigated with overhead sprinklers to ensure germination. The crop was thinned 10 d after emergence to about 60 000 plants ha-1. Twenty days after emergence, an additional 45 kg ha-1 N, as urea, was side-dressed and the field given a light (15 mm) sprinkler irrigation. Flowering was recorded as the time of stigma emergence in 50% of the main-shoot panicles. At maturity (about 60 d after flowering), panicles from one meter of the central row (1998-1999), or from two meters of both rows (1999-2000),were harvested, dried, and threshed to estimate grain yield. The crop was protected from both leaf feeding insect pests and stem borers with appropriate insecticides. Regular prophylactic measures were taken to prevent leaf rust, which is common during this season.
Estimation of Senescence
At the time of emergence of the flag leaf, three uniform plants in each plot were tagged, the length and breadth of upper six leaves measured, and the area of each estimated as: leaf length x leaf width x 0.70. (This factor was determined by measuring the leaf length, breadth, and actual area of 50 randomly chosen leaves.) Beginning at flag leaf emergence, the percentage remaining green of each of the upper six leaves of each tagged plant was visually estimated at weekly intervals, on a linear 0-to-9 scale, where 0 = 0 to 10% green-leaf area, and 9 = 90 to 100% green-leaf area. Weekly green-leaf area of each tagged plant was calculated by multiplying the percent green-leaf area by the measured area of each leaf, and summing across the six measured leaves. Percentage green-leaf area (% GLA) for each plant, for each week, was calculated by dividing the estimated GLA for that week by its measured leaf area at flowering. Plot values for% GLA were calculated by averaging the individual plant values for each plot.
The weekly % GLA data were used to fit an appropriate equation to describe the pattern of leaf senescence during the period of observations; -6 to +55 d after flowering (DAF) in 1998, and -10 to +55 DAF in 1999-2000. For the 1998-1999 data, a logistic fit was satisfactory (R2 > 90%) for majority of the plots. For those where logistic equation did not provide a satisfactory fit, a second order polynomial fit was used. For the 1999-2000 data linear fits were satisfactory for most plots, for those where a linear fit was not adequate, a second order or logistic fit was used. In both years, the coefficient of determination for all plots were not less than 90%. In both years, the fitted equation for each individual plot was used to estimate % GLA at 15, 30, and 45 DAF. These estimated values were used in the analysis of genotype differences in stay green and in the clustering of genotypes.
Data Analysis
Nine entries in the experiment in 1998-1999 that failed to flower in a reasonable time were excluded from the analysis, and the data for the 72 remaining genotypes were analyzed as an unbalanced lattice by means of REML option of GENSTAT (ver. 5.5). The 1999-2000 data were analyzed by the same program, with all effects, including genotype, considered as random. The data for the two years were combined and reanalyzed to partition variances into genotype, year, and genotype x year effects and to calculate heritabilities for the expression of stay-green trait at various times after flowering, by the REML option of GENSTAT. Genotype data reported in all tables and figures are the Best Linear Unbiased Predictors (BLUP) of genotype means.
To separate the 72 genotypes into groups on the basis of their stay-green expression at various times after flowering, cluster analysis (Wards method) on standardized data (SAS Inst., 1996) was done using six variables: % GLA at 15, 30, and 45 DAF from both the years. Clustering based on individual year's data takes into account the differences in leaf-senescence patterns in the two years.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
|
Leaf Senescence Patterns
Leaf senescence patterns could be described either by a logistic or a linear curve (Fig. 2)
. Most stay-green lines, such as B 35, exhibited a delayed onset of senescence, and, in some cases, a slower rate of senescence, once leaf senescence began, than did the known senescent line SPV 786 (Fig. 2). The combined effect of differences in onset and rate of senescence often resulted in large differences among lines in % GLA at maturity (Fig. 2). Differences in patterns of senescence are common in many species that exhibit stay green, even among genotypes that may retain same level of leaf area at maturity (Thomas and Howarth, 2000). For example, Borrell et al. (2000a), in a study of nine sorghum hybrids based on two different sources of stay green (B 35 and KS 19), reported that KS 19 hybrids had a delayed onset and reduced rate of senescence, whereas B 35 hybrids had only delayed onset of senescence. KS 19 hybrids had a smaller leaf area at flowering than the B 35 hybrids, however; so the green-leaf areas at maturity for both were similar. The nitrogen concentration in the green leaves of both the B 35 and KS 19 hybrids was also reported to be higher than in senescent hybrids at midgrain fill and maturity, yet this was associated with thicker leaves only in hybrids with the B 35 source (Borrell and Hammer, 2000).
|
Differences among genotypes in % GLA were significantly affected by time to flowering in 1998-1999. Later-flowering genotypes had a significantly lower % GLA at 15 DAF (r = -0.42, P < 0.003), a similar % GLA at 30 DAF (r = -0.11, P > 0.33), but a significantly higher % GLA at 45 DAF (r = 0.43, P < 0.001). Apparently, grain yields were more affected (directly) by the stress in later-flowering genotypes, and as a result, there was less translocation from the leaves. This reduced senescence rate ultimately resulted in a higher % GLA at maturity in the late-flowering genotypes than in early flowering ones, despite the fact that senescence began at an earlier developmental stage in the case of the later flowering genotypes. This explanation is supported by a negative relationship between grain yield and %GLA at 45 DAF (r = -0.45, P < 0.001), i.e., the less the yield, the greater the % GLA, though there was a meaningful relationship at % GLA at 30 DAF (r = -0.22, P < 0.08). Thus, differences among genotypes in escape from stress, which result in differences in grain number or grain growth under grain filling stress, and, consequently, in the amount of nitrogen translocated from the leaves, affect leaf senescence in unanticipated ways. Clearly, comparisons of leaf senescence among genotypes with different flowering times, especially under severe stress, must be made with care.
In contrast, in 1999-2000, under a somewhat milder stress, there was no effect of time to flowering on % GLA at 15 DAF (r = 0.04), 30 DAF (r = 0.03), or at 45 DAF (r = 0.03), or on grain yield (r = -0. 11, P > 0.35), despite the time-to-flowering range of 25 d among the 72 genotypes. Consequently, there were no significant correlations between % GLA at any of the times of measurement and grain yield.
Grouping of Genotypes on the Basis of Stay-Green Trait
To classify the 72 genotypes by type and level of expression of stay green across the two years, clustering was done on the basis of six variables: % GLA at 15, 30, and 45 d after flowering in each of the years. This method allows for differences in response to the two years (i.e., genotype x year interactions) in both the time of onset and rate of senescence. The clustering procedure was truncated at five groups that retained 74% of the original variation in these traits (Fig. 3)
. Cluster 1 and Cluster 2 represented the stay-green genotypes, whereas Cluster 4 and Cluster 5 contained the most senescent genotypes (Table 1).
On the basis of combined data from the two years, Cluster 1 genotypes had a higher % GLA at 15 DAF (delayed onset of senescence) than did Cluster 2 genotypes, but senesced at a greater rate thereafter, particularly between 30 and 45 DAF, to reach approximately the same % GLA at 45 DAF as Cluster 2 genotypes (Table 1). There was, however, a major difference between the two stay-green clusters in their response to the two years. In 1998-1999, final mean % GLA (45 DAF) was lower than that in 1999-2000. Consistent with the more severe stress in 1998-1999, Cluster 1 genotypes clearly delayed onset of senescence much more effectively than Cluster 2 genotypes. In contrast, in 1999-2000, Cluster 1 and Cluster 2 genotypes had similar patterns of senescence (Fig. 4) . Cluster 1 and Cluster 2 genotypes differed significantly in mean time to flowering in both years (74 vs. 95 d in 1998-1999 and 71 vs. 81 d in 1999-2000). This means that Cluster 1 genotypes had two advantages over Cluster 2 genotypes: (i) a lower seasonal total transpiration because of a shorter growing season, and (ii) a less severe stress after flowering in years when the evaporative demand increased with time. Both of these advantages were likely factors in this experiment. In 1998-1999, both the preflowering (301 mm) and postflowering (288 mm) cumulative evaporative demand, to which Cluster 1 genotypes were exposed, were considerably less than that to which Cluster 2 genotypes (417 mm preflowering and 342 mm postflowering) were exposed (Table 3). The advantages of early flowering in 1998-1999 were especially marked in the first 15 d of grain filling when leaf senescence begins: Cluster 1 genotypes were exposed to a cumulative evaporative demand of 79 mm, compared with 103 mm for Cluster 2 genotypes (Table 3). In the less stressed year, the differences in cumulative evaporative demand between Cluster 1 and Cluster 2 genotypes were less marked in both the preflowering period (353 mm for Cluster 1 vs. 399 mm for Cluster 2) and the grain-filling period (255 mm for Cluster 1 vs. 297 mm for Cluster 2; Table 3). Thus, the differences between Cluster 1 and Cluster 2 in 1998-1999 are likely to be largely a consequence of drought escape. In 1999-2000, where early flowering conferred less of an advantage, the senescence patterns of the two clusters were similar (Fig. 4). Therefore, Cluster 2 genotypes are likely to have the same level of per se nonsenescence, as do Cluster 1 genotypes, despite being grouped into different clusters on the basis of observed % GLA.
|
|
The pattern of senescence in individual genotypes is influenced by both the time of onset and the rate of development of stress in an individual environment. It is, however, possible at least to understand qualitatively the effects of different times of onset of senescence and rates of senescence (because of environmental differences or drought escape) and to identify nonsenescent genotypes across different patterns of stress development. It is also evident that genotypes with intermediate levels of nonsenescence (such as those in Cluster 3) will be classified as more or less senescent, depending upon the pattern of stress development. This is also consistent with the hypothesis that there are more than one mechanism by which leaves stay green (Thomas and Howarth, 2000) which are likely controlled by different genes that, in turn, are triggered by the specific pattern of stress development (Dunwell, 2000).
Stay-Green Sources
Cluster 1 genotypes included the two most widely used stay-green sources—B 35 and KS 19 (in the form of QL 10 and QL 27), which combined both stress escape and stay green (primarily by delayed onset of stress) in the post-rainy season in India (Table 1). Their drought escape is likely to be an artifact of the short daylength of November-December at ICRISAT (17 N), and may not be a major factor in longer day lengths in main growing seasons elsewhere. The other Cluster 1 genotype, IS 22380, is a landrace from Sudan that was classified as stay green (Van Oosterom et al., 1996). It also benefitted from early flowering caused by the short daylength, but had a marked delay in onset of senescence in both years (Table 2). The line may be worth evaluating further in other environments in which stay green is of value, as it is of a different biological race (caudatum) than either B 35 or KS 19.
Three of the four nonsenescent lines in Cluster 2 are ICRISAT breeding lines derived from the ICRISAT B/R line multifactor resistance population (ICSP-2 B/R MFR); two of these trace back to the same S2 line from the population (S2 407-1), and two of the lines also have E 36-1 in their pedigree, suggesting these lines may have conferred the stay-green trait. The multifactor resistance population includes both sources of stay green and resistance to charcoal rot (Dr. B.V.S. Reddy, ICRISAT, 2000, personal communication). E 36-1 is a recognized source of stay green (Van Oosterom et al., 1996), and has been widely used in the drought-breeding program at ICRISAT. E 36-1 itself, however, clustered with genotypes in Cluster 3, where its % GLA at 45 DAF was approximately half of that of its derivatives under the more severe stress in 1998-1999, although it was nonsenescent when the stress development was more gradual in 1999-2000 (Table 1). The Cluster 2 line E36 x R 16 8/1 is a deliberate stay-green selection from a cross between stay green and senescent (R 16) parents, so E36-1 would appear to transmit the trait to its progeny with adequate selection. The two senescent checks R16, and SPV 783 (Van Oosterom et al., 1996), appear in Cluster 5, as expected.
A small quantity of seed of these genotypes is available on request from the ICRISAT Genebank Curator. The seed request forms and other details are available on line at http://www.icrisat.org/text/research/grep/homepage/grephomepage/mta.htm (verified November 12, 2001).
|
ACKNOWLEDGMENTS |
|---|
Received for publication March 23, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  V. Pegadaraju, C. Knepper, J. Reese, and J. Shah Premature Leaf Senescence Modulated by the Arabidopsis PHYTOALEXIN DEFICIENT4 Gene Is Associated with Defense against the Phloem-Feeding Green Peach Aphid Plant Physiology, December 1, 2005; 139(4): 1927 - 1934. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. D. Subedi and B. L. Ma Nitrogen Uptake and Partitioning in Stay-Green and Leafy Maize Hybrids Crop Sci., February 23, 2005; 45(2): 740 - 747. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
