Journal of Experimental Botany, Vol. 52, No. 355, pp. 203-214,
February 2001
© 2001 Oxford University Press
Original Papers |
Identification of causal relationships among traits related to drought resistance in Stylosanthes scabra using QTL analysis
1 CSIRO Tropical Agriculture, 120 Meiers Rd, Indooroopilly, Qld 4068, Australia
2 School of Agriculture and Horticulture, The University of Queensland, Gatton College, Qld 4345, Australia
3 Indian Grassland and Fodder Research Institute, Jhansi 284003, India
Received 8 May 2000; Accepted 22 August 2000
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
Previous studies have shown that a negative relationship exists between transpiration efficiency (TE) and carbon isotope discrimination (
) and between TE and specific leaf area (SLA) in Stylosanthes scabra. A glasshouse experiment was conducted to confirm these relationships in an F2 population and to study the causal nature of these relationships through quantitative trait loci (QTL) analysis. One hundred and twenty F2 genotypes from a cross between two genotypes within S. scabra were used. Three replications for each genotype were maintained through vegetative propagation. Water stress was imposed by maintaining plants at 40% of field capacity for about 45 d. To facilitate QTL analysis, a genetic linkage map consisting of 151 RAPD markers was developed. Results from this study show that was significantly and negatively correlated with TE and biomass production. Similarly, SLA showed significant negative correlation with TE and biomass production. Most of the QTL for TE and were present on linkage groups 5 and 11. Similarly, QTL for SLA, transpiration and biomass productivity traits were clustered on linkage groups 13 and 24. One unlinked marker was also associated with these traits. There were several markers coincident between different traits. At all the coincident QTL, the direction of QTL effects was consistent with phenotypic data. At the coincident markers between TE and , high alleles of TE were associated with low alleles of . Similarly, low alleles of SLA were associated with high alleles of biomass productivity traits and transpiration. At the coincident markers between trans-4-hydroxy-N-methyl proline (MHP) and relative water content (RWC), low alleles of MHP were associated with high alleles of RWC. This study suggests the causal nature of the relationship between TE and . Phenotypic data and QTL data show that SLA was more closely associated with biomass production than with TE. This study also shows that a cause–effect relationship may exist between SLA and biomass production. Key words: Drought resistance, transpiration efficiency, carbon isotope discrimination, quantitative trait loci (QTL), causal relationships.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
Stylosanthes scabra is a widely sown pasture legume in the semi-arid tropics of northern Australia where water stress is a major problem. Previous studies have shown that transpiration efficiency (TE) is one of the traits contributing to drought resistance in S. scabra (Thumma et al., 1998a
, b). These studies, using several genotypes, demonstrated a significant negative relationship between TE and carbon isotope discrimination (). There was also a negative relationship between TE and specific leaf area (SLA). It was suggested that and/or SLA may be used as indirect measures of TE.
The carbon in atmospheric CO2 is mostly in the form of 12C, but a fraction is also present in the stable isotope form of 13C. During carboxylation, plants discriminate against 13C present in ambient CO2. It has been proposed that plants which show less discrimination against 13C would have high TE (Farquhar et al., 1982). This negative relationship between TE and has been established in many plant species (Farquhar et al., 1989). Carbon isotope discrimination () represents an integrative estimate of TE over the period during which the dry matter was formed (Condon et al., 1987).
In several plant species SLA was negatively correlated with TE and biomass production (Virgona et al., 1990; Wright et al., 1994; Brown and Byrd, 1997). However, the reason for this relationship has not been established. This negative relationship may be due to the fact that plants with low SLA (thicker leaves) will have more mesophyll cells leading to higher rates of CO2 assimilation and, consequently, higher biomass production (Pearce et al., 1969; Nelson, 1988).
Trans-4-hydroxy-N-methyl proline (MHP) is an osmoprotectant (Naidu et al., 1987) accumulating in S. scabra under water stress conditions. Osmoprotectants are the biochemical compounds accumulating in cytoplasm under abiotic stress conditions. These compounds help protect membrane integrity and photosynthetic machinery and may also contribute to cytoplasmic osmotic adjustment (Rhodes and Hanson, 1993). Earlier experiments have shown that MHP is significantly correlated with TE (BR Thumma unpublished results).
DNA-based markers such as RFLP, RAPD, AFLP, and microsatellites provide opportunities to study quantitative traits such as drought resistance through QTL analysis. A QTL is a region of a chromosome linked to a marker, which has a significant effect on the quantitative trait (Tanksley, 1993). So far, most QTL studies have been concentrated on the analysis of morphological and agronomic traits. However, there are a few reports on QTL analysis of physiological traits involved in drought resistance (Martin et al., 1989; Masle et al., 1993; Lebreton et al., 1995; Lilley et al., 1996; Mian et al., 1996; Teulat et al., 1998). QTL analysis not only provides information on the number of loci involved in the expression of a trait, it also indicates the possible location of these loci. The inheritance of water use efficiency, estimated through , has been studied using RFLP markers (Martin et al., 1989). These authors identified four QTL associated with water use efficiency in tomato. Similarly, four QTL associated with water use efficiency have been identified in soybean (Mian et al., 1996).
Recently QTL analysis is being used to test the relationships between physiological traits (Lebreton et al., 1995; Simko et al., 1997). Identification of relationships between two traits using physiological studies may not distinguish whether the traits are causally related or merely varying in association. For two traits to be causally related, the significant QTL effect of one trait should have a measurable effect on the other trait (Lebreton et al., 1995). Coincidence of QTL for two traits, with allelic differences corresponding to the expected causal relationship between the traits, is strong evidence that the two traits are related (Quarrie et al., 1997). However, coincidence of QTL for two traits with QTL effects in the expected direction may not provide conclusive evidence that the two traits are causally related, but it provides circumstantial evidence that the two traits are causally related. Isolation of the genes underlying the traits may provide ultimate evidence that the two traits are causally related.
Using this strategy, it has been shown that xylem ABA concentration is more likely to have a regulatory effect on stomatal conductance than leaf ABA content (Lebreton et al., 1995). It has also been shown that ABA content is unlikely to affect leaf size (Quarrie et al., 1997) even though consistent negative relationships were observed between the two traits using physiological studies (Henson, 1983, 1985). Similarly, it has been shown that an increase in the leaf ABA concentration was associated with low stomatal conductance and low yield (Sanguineti et al., 1999).
Most of the work measuring the relationship between TE and has been based on a limited number of genotypes. A few studies examined the relationship between TE and and/or SLA in a large number of genotypes (Hubick et al., 1988; Ehdaie et al., 1993). However, there are no reports on the identification of a genetic basis of these relationships using QTL analysis. The objectives of the present study were (a) to evaluate the relationships between TE, and SLA in an F2 population to confirm conclusions from earlier studies, (b) to identify the QTL associated with these traits and (c) to study the causal nature of the relationships.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
Evaluation of parents and F2 progeny in glasshouse
Two S. scabra genotypes, CPI 93116 and cv. Fitzroy, were chosen as parents based on level of polymorphism and on differences between them for traits such as TE, SLA and
observed in previous studies (Thumma et al., 1998a, b). An F2 population obtained by crossing these two genotypes was used. One hundred and fifty F2 plants were grown for 3 months in 5.0 l pots. In order to obtain replicated phenotypic data, plants were vegetatively propagated through cuttings. Three cuttings were taken from each F2 line, and roots were initiated by keeping the cuttings in a humid chamber for 2 weeks. Once roots were initiated, cuttings were transferred to 2.0 l pots (one cutting per pot) containing fertile alluvial soil (moisture content 20% of soil mass to field capacity). The soil in the pots was covered by a 2 cm layer of polyethylene beads, which reduced evaporation from the soil to below 10% of total water use. The pots were arranged in three blocks representing three replications in a glasshouse (temperature range 21–30 °C). In each block three empty pots (without plants) covered with a 2 cm layer of beads, were placed to estimate water evaporation from the soil.
Cuttings were grown in soil at field capacity for 20 d before imposing the stress treatment, which involved maintaining the pots at 40% of field capacity. From the original 150 genotypes, 120 (as these genotypes were used for map construction) were subjected to stress treatment. Plants were watered to 40% field capacity every alternate day. On each watering day, pots were re-randomized within each block. The stress treatment lasted for 45 d.
Plants were harvested, separated into shoots and roots (roots washed thoroughly to remove soil), and oven dried at 80 °C for 4 d. Dry weights of shoots and roots were recorded. Transpiration was calculated by deducting pot evaporation from total water use. TE was calculated as total dry matter (TDM) produced per unit of water transpired and expressed as g/kg.
Leaf sampling for SLA and
From each plant, the six youngest fully expanded leaves were sampled (third leaf from the top on six branches). Leaf area was measured using a leaf area meter and the leaves were oven dried at 80 °C for 2 d. SLA was calculated as leaf area per unit leaf dry weight and expressed as cm2 g-1. Leaf samples were ground finely and carbon isotope composition was measured with a ratioing mass spectrometer (Research School of Biological Sciences, Australian National University, Canberra). was calculated according to Hubick et al. (Hubick et al., 1986):
 |
a is carbon isotope composition of air taken as -8
relative to Peedee Belemnite and p is carbon isotope composition of plant material.
Leaf sampling for water status and solute analysis
Relative water content (RWC) and MHP were measured from leaves collected 3 d before final harvest. Leaves were sampled before 07.00 h on three consecutive days (one block of plants per day). Plants were watered to 40% of field capacity 1 d before leaves were sampled.
The three youngest fully expanded trifoliolate leaves from each plant were sampled. One was used for measuring RWC (Wilson et al., 1979), and the remaining two leaves for measuring MHP. The leaf samples used for MHP measurement were sealed in small tubes and frozen in liquid nitrogen before storing at -20 °C. MHP was measured on freeze-dried leaves using the high performance liquid chromatography (HPLC) method described previously (Naidu, 1998).
Statistical and genetic analysis
Analysis of variance (ANOVA) of the different traits was conducted by using GLM. Genetic variance and covariance were estimated by using PROC MIXED functions of SAS statistical package (SAS, 1992) assuming genotype and environments as random effects. Broad-sense heritability was estimated on a genotype mean basis as a ratio of 
2g/(2g+2e/r), where 2g is the genetic variance, 2e is the error variance and r is the number of replicates.
RAPD analysis and map construction
Young leaves were collected and immediately frozen in liquid nitrogen. DNA was extracted from freeze-dried leaves using the method developed previously (Liu and Musial, 1995). PCR amplifications were performed in 15 µl reaction mixture containing 50 mM MgCl2, 0.2 mM of each dNTP, 0.33 µM 10-mer primer (Operon Technologies), 5 ng of genomic DNA, and 1 unit of taq DNA polymerase. PCR products were separated by electrophoresis at 100 V for 3–4 h in 1.6% agarose gels, stained with 0.5 µl of ethidium bromide in 0.5xTBE buffer and photographed under UV light.
The linkage map was constructed using MAPMAKER/EXP (version 3, Lander et al., 1987) and map order was checked by using MapManager program (Manly and Olson, 1999). Chi square analysis was used to test the goodness-of-fit to the expected ratio of 3:1 for dominant markers in F2 generation. All the polymorphic markers were grouped at LOD 3.0 and a recombination frequency (r) of 3.0. The Kosambi mapping function was used to convert the recombination fractions to map distances (Kosambi, 1944).
QTL analysis
ANOVA (P<0.05) was used to detect possible QTL affecting a trait. Interval mapping with a LOD score of 2.0, as the threshold for detecting QTL location, was then carried out using the MapQTL program (Van Ooijen and Malliepaad, 1996). The maximum LOD score along the interval was taken as the position of the QTL, and the region in the LOD score, within one LOD unit of maximum, was taken as the confidence interval. Multiple QTL model (MQM) mapping was used for identification of further QTL using QTL detected with interval mapping as cofactors.
Testing the relationships among traits
Associations among the traits were tested by finding the coincidence of QTL effects initially at P<0.05 and by comparing the genotype class means. When genotype class means between the traits were in the expected direction, to get a clear understanding of the relationships among the traits at other loci, QTL effects significant at a higher probability level (P<0.1) were used to compare the genotype class means. Clearly the QTL obtained using a higher probability level may not be useful in breeding programmes, but they should help in understanding the relationships among traits. Similar significance levels were used by other authors (Lebreton et al., 1995; Prioul et al., 1997; Simko et al., 1997; Davies et al., 1999). QTL effects were calculated as the difference between mean data for identifiable homozygous alleles and the mean data combined for both the other alleles and heterozygotes. Markers that were separated by more than 10 cM, and the QTL effects of which were significant, were treated as separate QTL.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
Genotypic variation
Transgressive segregation resulted in a wide range of values in progeny even though differences between the parental lines were small for some traits such as
, root dry matter (RDM) and shoot dry matter (SDM). Differences between the F2 genotypes were highly significant for all traits (P<0.01; Table 1). Means, standard deviations, ranges and heritabilities of the traits measured in the two parents and F2 genotypes are presented in Table 1. Standard deviations were low for TE, , RWC, and high for MHP and RDM.
|
Relationships among the various traits
Phenotypic and genetic correlations among TE, , SLA, transpiration, and traits related to biomass production (TDM, SDM, and RDM) were highly significant (P<0.01; Table 2). Genetic correlations between the traits were stronger than the phenotypic correlations, but the trends were similar (Table 2). Biomass productivity traits were positively correlated with TE and negatively correlated with and SLA. The correlations between TE and , and between TE and SLA were similar. However, the correlations between biomass productivity traits and SLA were stronger than the correlations between biomass productivity traits and TE or (Table 2). There was a significant positive relationship between transpiration and biomass productivity traits and a negative relationship between transpiration and SLA. However, phenotypic and genetic correlations between TE and transpiration were not significant. Productivity related traits showed high broad-sense heritability (Table 1
). TE and SLA showed intermediate levels (0.69 and 0.68, respectively) of broad-sense heritability while a high level (0.78) of broad-sense heritability was observed for . Broad-sense heritability was intermediate (0.60) for MHP and low (0.36) for RWC.
|
Genetic linkage map and QTL analysis of the traits
One hundred and fifty-one RAPD markers were used to construct the linkage map. Except for three, all other markers segregated according to the expected ratio of 3:1 for dominant markers in the F2 generation. One hundred and twenty markers formed 25 linkage groups (Fig. 1). S. scabra has 20 chromosomes (2n=40). Thirty-one markers remained unlinked. Total genome length covered by markers in the present map is 1406 cM with an average interval of 12 cM.
|
For all the traits, only a few the QTL identified using ANOVA (P<0.01) were identified through interval mapping. However, most of the QTL identified using ANOVA were also identified through interval mapping with the relaxation of the significance level (P<0.05). Most of the QTL associated with TE,
and MHP were present on linkage groups (LG) 5 and 11. All of the QTL associated with SLA, transpiration, TDM, SDM and RDM were present on LG 13 and 24 (Table 3). One unlinked marker (AJ16) was also significantly associated with these traits (Table 4).
|
|
Four QTL were identified for TE using interval mapping (Table 3
). Two of these QTL AA14 on LG5 (P<0.05) and J15 on LG11 (P<0.01) were significant with ANOVA. The percentage of variation explained by these QTL has ranged from 19–42% (Table 3).
Seven QTL were observed for through interval analysis (Table 3). QTL near AA14, A13 and the QTL between K9 and C2 were also observed through ANOVA. However, only the QTL between K9 and C2 was significant at P<0.01. The percentage of variation explained by these QTL has ranged from 21–30% (Table 3). The QTL near AA14 on LG5 was common between TE and .
There were nine QTL for MHP with the percentage of variation explained by these QTL ranging from 8–40% (Table 3). Except for 3 QTL near AF11, B13 and G19-2, all other QTL were significant with ANOVA. However, of the six significant QTL only the QTL present near N13 on LG3 and at AB14 on LG7 were significant at P<0.01.
Four QTL were observed for RWC with the percentage of variation explained by these QTL ranging from 11–33% (Table 3). The QTL on LG19 was significant with ANOVA (P<0.05). The QTL near H12 on LG5 was common for RWC and MHP.
QTL for transpiration, TDM, SDM and RDM identified through interval mapping were clustered near N9, near AF19-2 on LG13, and near AD12 on LG24. Similarly there was a QTL for SLA near AF19-2 on LG13 (Table 3). The QTL for transpiration and TDM near N9 and the QTL for all these traits near AD12 were significant with ANOVA. The QTL present at AD12 on LG24 for TDM and SDM were significant at P<0.01. The QTL for SLA near AF19-2 on LG13 was significant with ANOVA (P<0.05). The percentage of variation explained by the QTL has ranged from 14–36% for TDM, from 14–35% for SDM, from 19–30% for RDM, and from 15–24% for transpiration. The single QTL identified for SLA explained 23% of phenotypic variation (Table 3).
Comparison of QTL effects of different traits
Two markers (AA14 on LG5 and C2 on LG11) were found to be coincident between and TE through ANOVA (P<0.05; Table 4). At these two markers, low alleles of were associated with high alleles of TE. Three more markers (N13 on LG3, N19-2 on LG9 and J15 on LG11) were found to be coincident between and TE when the probability levels were relaxed (P<0.1; Table 4). Again at these three markers, low alleles of were associated with high alleles of TE. There was a common QTL for , TDM, SDM, and RDM at G19 on LG11 (P<0.1). QTL effects for all the traits at this marker were in the expected direction (i.e. low alleles of were associated with high alleles of TDM, SDM and RDM). There were two coincident markers (K9 and C2 on LG11) between TE and MHP (P<0.05). QTL effects at these two markers were opposite to the expected direction (i.e. high TE alleles were associated with low MHP alleles). MHP and RWC have three common markers (P<0.1). At all these three coincident markers, high alleles of MHP were associated with low alleles of RWC (Table 4).
Biomass productivity traits and transpiration have common QTL at AD12 on LG24 and at unlinked marker AJ16 (P<0.05; Table 4). At these two markers there was a positive relationship among the traits, i.e. high alleles of TDM, SDM and RDM were associated with a high level of transpiration. In addition to these, TDM and SDM have 3 common QTL (P<0.05) at B11 on LG5, AX17 on LG23 and at unlinked marker AV5. There was a positive relationship between the two traits at these three markers. SLA appears to be more closely associated with traits related to biomass productivity than with TE. SLA has a common QTL with TDM, SDM, and transpiration at AJ16 (P<0.05; Table 4). At this marker, low alleles of SLA were associated with high alleles of TDM, SDM and transpiration. Two more markers (AF19-2 on LG13 and unlinked marker AJ16) were found to be coincident between SLA, transpiration and biomass productivity traits including RDM using a higher probability level (P<0.1). At these two markers, QTL effects were in the expected direction (i.e. low alleles of SLA were associated with high alleles of transpiration and biomass productivity traits). The coincident markers at AF19-2 were also observed through interval mapping (LOD>2.0; Table 3). In addition to these, SLA has common QTL with SDM at Q17 on LG18, with RDM at AB17 on LG3 and with TE at AA10 on LG1 (Table 4). The direction of the QTL effects at all these loci was as expected (i.e. low alleles of SLA were associated with high alleles of TDM, SDM, RDM, and TE).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
Relationships among the traits
The significant negative relationship of
with TE and TDM and that of SLA with TE and TDM (Table 2) reinforces previous findings obtained using different genotypes of Stylosanthes (Thumma et al., 1998a, b).
There are conflicting reports about the relationship between and TDM. Positive relationships were reported for alfalfa (Ray et al., 1999) and common bean (Zacharisen et al., 1999), while negative relationships were reported for peanut (Hubick et al., 1986) and wheat (Ehdaie et al., 1993). A positive relationship between and TDM suggests that TE is influenced more by stomatal conductance than photosynthetic capacity. Therefore, selection for high TE may result in lower biomass productivity. However, a negative relationship between and TDM indicates that TE is more influenced by photosynthetic capacity than stomatal conductance. Selection for high TE would result in higher biomass productivity (Condon et al., 1987). TE appears to be more influenced by photosynthetic capacity than stomatal conductance in Stylosanthes as indicated by the significant positive relationship between TE and biomass productivity traits and non-significant relationship between TE and transpiration (Table 2).
Transgressive segregation
Transgressive segregation observed for all the traits indicates that alleles affecting these traits were dispersed between the two parents (Table 1). Transgressive segregation will result when high or low alleles of a trait which were dispersed between the parents come together in progeny (Prioul et al., 1997).
Genetic linkage map and QTL analysis
The genetic linkage map (Fig. 1) presented here represents the first detailed map in Stylosanthes even though a partial genetic map consisting of 37 linkage groups was presented at the agricultural biotechnology conference for mapping disease resistance genes (Chakraborty et al., 1998). By taking the average chromosomal length as 100–150 cM (chromosomal length varies between 50–300 cM for a species, Meagher et al., 1988) total length covered by the markers in the present map is estimated to be about 48–70%.
As the total genome is not fully covered by the markers, all the QTL associated with a trait may not have been uncovered in the present study. To get an estimate of the proportion of QTL identified for a trait, total percentage of phenotypic variation explained was obtained by summing the percentage of variation accounted for by each of the QTL. However, with the small population sizes and a low map density, this will inevitably result in overestimation of the amount of variance accounted for by QTL, but at least this may give an approximation of the QTL not detected. Results obtained here show that even though most of the variation for all the traits was explained by the QTL detected here, only a small proportion of phenotypic variation of SLA was accounted for by the QTL identified in the present study (Table 3).
Even though some of the QTL identified with ANOVA were also significant with interval mapping there are a few discrepancies between the two methods. ANOVA is a single point analysis, which looks for QTL on the marker itself whereas interval mapping detects QTL at regular intervals between the two flanking markers. If the QTL is further away from the marker it is less likely to be detected with ANOVA due to recombination between the marker and QTL (Tanksley, 1993). Similarly, in the interval mapping missing data on the marker genotypes and dominance of the markers such as RAPD will lead to lower LOD scores (Van Ooijen, 1992). If the two QTL present in the same linkage group are closer to each other they behave more like one locus and if the two loci are in repulsion there is a possibility that not even one QTL may be detected (Van Ooijen, 1992).
For application of marker-assisted selection (MAS) in breeding programmes, QTL detected with high stringency (P<0.01 and LOD>2.4) may be used to reduce the number of false positives. However, this may result in the loss of potential QTL through false negatives. Recent studies in tomato have shown that some of the QTL detected with low stringency (P<0.1) were consistently detected across different environments while some of the QTL detected with high stringency (P<0.01) were not detected in all the environments (Fulton et al., 2000). Therefore, QTL obtained with low stringency may be treated as potential QTL subject to further testing.
Identification of causal relations among the traits
Significant negative correlation between and TE using phenotypic data, and coincidence of markers between these traits with alleles for high TE being associated with alleles for low , strongly suggest a possible cause–effect relationship between these two traits (Table 4). These markers may represent the mechanistic basis for the negative relationship observed between these two traits in many plant species (Farquhar et al., 1989). Carbon isotope discrimination represents intercellular CO2 concentration (Ci), which is a balance between stomatal conductance and the photosynthetic capacity of a plant. Therefore, Ci can be lower with either high photosynthetic capacity or with low stomatal conductance. The ratio of photosynthetic rate to stomatal conductance (A/g) is negatively related to Ci. Thus, there is a negative relationship between A/g (TE) and (Farquhar et al., 1989). Phenotypic data suggest that of S. scabra is influenced more by photosynthetic capacity than by stomatal conductance (Table 2). Consistent with this, there is one marker coincident between and TDM, SDM and RDM. QTL effects at this marker were in the expected direction (i.e. low alleles of were associated with high alleles of TDM, SDM and RDM; Table 4).
Transpiration efficiency is the biomass produced per unit of water transpired. Therefore, either lower rate of transpiration and/or higher biomass productivity would lead to high TE. QTL effects at the coincident marker between TE and transpiration indicate that a lower rate of transpiration was associated with high TE (Table 4). As there were not many markers coincident between TE, transpiration and biomass productivity traits, it is not possible to determine whether TE is influenced more by transpiration or by biomass production. However, the phenotypic data clearly show that TE is influenced more by biomass production (photosynthetic capacity) than by the amount of water transpired (stomatal conductance. Table 2).
The low correlation coefficient between TE and MHP (Table 2) may be partly due to the contrasting QTL effect of MHP with respect to increase in TE at K9 and C2 on LG11 (Table 4). Coincidence of three QTL between MHP and RWC with allelic differences in the same direction as expected from the phenotypic data (i.e. high MHP alleles were associated with low RWC alleles) shows that low RWC may induce the expression of the MHP genes (Table 4). This is consistent with the accepted hypothesis that these compounds accumulate under water stress conditions.
Relationships among SLA, biomass productivity traits and transpiration
The significant negative correlation between SLA and productivity-related traits using phenotypic data and the coincidence of markers with allelic differences in the expected direction show that SLA may be causally related to biomass productivity. At all the common QTL between SLA, TDM, SDM, and RDM, low alleles of SLA were associated with high alleles of TDM, SDM and RDM (Table 4).
In several plant species SLA was negatively correlated with biomass production (Nelson, 1988; Wright et al., 1994). Although, the reason for this has not been established, this negative relationship may be due to the fact that plants with low SLA (thicker leaves) will have more mesophyll cells per unit area or larger mesophyll cells, leading to higher rates of CO2 assimilation and, consequently, higher biomass production. Indeed, a negative relationship has been observed between SLA and leaf photosynthesis per unit area in many plant species (Pearce et al., 1969; Nelson, 1988). Similarly, nitrogen content per unit leaf area was negatively correlated with SLA in many plant species (Anderson et al., 1996). This negative relationship may be the basis for higher rates of photosynthesis found in plants with low SLA. As most of the leaf nitrogen is accounted for by photosynthetic enzymes, higher rates of photosynthesis are expected in plants with low SLA (high nitrogen content per unit leaf area). If the limiting factor for growth is photosynthetic capacity, low SLA should result in higher biomass production and higher TE (Brown and Byrd, 1996). The phenotypic data and coincidence of markers show that SLA is more closely associated to biomass production than to TE. Thus the significant negative relationship between SLA and TE observed with phenotypic data may be due to the link between biomass productivity and TE rather than SLA and TE having a direct mechanistic relationship. However, this conclusion needs to be treated carefully as only a small proportion of the phenotypic variation in SLA was accounted for by the QTL detected in the present study.
There was a negative relationship between SLA and transpiration (Table 2). Most xeric species of Prunus were found to have the lowest SLA, highest stomatal conductance and highest rate of assimilation (Rieger and Duemmel, 1992). Similarly, a negative relationship has been reported between SLA and stomatal conductance in sunflower (Virgona and Farquhar, 1996). Consistent with the phenotypic data, alleles for low SLA were associated with alleles for high transpiration (Table 4).
Transpiration has always been positively associated with biomass production (Tanner and Sinclair, 1983; Wright et al., 1994). Higher transpiration not only leads to higher photosynthetic rates, but also keeps the leaf surface cool especially under hot conditions (Lu et al., 1994). QTL effects at the coincident markers between transpiration and TDM, SDM and RDM were positive (i.e. alleles for high transpiration were associated with alleles for high TDM, SDM and high RDM; Table 4). Therefore, these markers may form the molecular basis for the positive relationship observed between transpiration and biomass production in several plant species.
These associations among SLA, transpiration and biomass productivity traits may be explained as follows. Plants with low SLA (thicker leaves), with higher photosynthetic capacity may show higher rates of stomatal conductance and photosynthesis, which in turn leads to higher biomass production. Wang et al. have also reported that stomatal conductance correlates with photosynthetic capacity (Wong et al., 1979). They also suggested that this would lead to constancy of Ci. While it remains the case that stomatal conductance and photosynthetic assimilation rate are often correlated, later studies have shown that Ci does vary both within and between species (Farquhar et al., 1989).
The relationship between SLA and transpiration needs to be studied further as it may provide opportunities to improve growth rate during the early stages of plant development under water non-limiting conditions. High stomatal conductance with a consequent high rate of photosynthesis is an important trait which may improve the growth rate of a plant during the early stages of development (Richards, 2000). An associated problem in selecting for low SLA is that it may result in a reduction of leaf area. However, previous studies (Thumma et al., 1998a) have indicated that SLA and leaf area in S. scabra may be independent of each other. Similar results were reported earlier for alfalfa (Barnes et al., 1969). Therefore it may be possible to select for genotypes with low SLA and high leaf area to improve growth rate under favourable conditions.
Breeding implications
For any trait to be used as an indirect selection criterion in breeding programmes, its measurement should be easy and rapid. Such an indirect measure should have high genetic correlation with the trait that is being selected for and it should have high heritability. and SLA may fall into this category. Many studies have shown high heritability and low genotypexenvironment interaction for (Farquhar et al., 1989). A drawback in its use is the cost of measurement, although when compared to using molecular markers it is not very high. Therefore, using itself rather than markers associated with it may be sufficient in breeding programmes to improve TE. However, in some studies was shown to be positively related to yield, indicating selection for low would improve TE, but it could also lead to lower yield (Condon et al., 1987; Ray et al., 1999). Therefore, it was suggested that selection should be applied for both traits simultaneously. In this context molecular markers would be very valuable in pyramiding traits such as low , SLA, high TE, and high biomass production.
|
Conclusions |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
The presence of negative relationships of
with TE and TDM, and of SLA with TE and TDM observed in the present and earlier studies (Thumma et al., 1998a, b) indicates the robust nature of these relationships. All these relationships among different traits are consistent with QTL data. The present study indicates the causal nature of the relationship between and TE. Thus, these results provide a genetic basis for the negative relationship found between and TE in many plant species. There is also evidence that low RWC induces expression of MHP genes. Phenotypic data and QTL data show that low SLA leads to high biomass production and consequently high TE. The present study also indicates that a cause–effect relationship may exist between SLA and biomass production. QTL identified for various traits in this study may be useful in pyramiding the traits together.
|
Acknowledgments |
|---|
This work forms part of BRT's PhD at the University of Queensland. We thank Professor Graham Farquhar for his valuable comments on the manuscript. We also thank Dr Craig Hardner for his help in statistical analysis. We thank the two anonymous referees for their constructive criticism of the manuscript. UQPRS and OPRS scholarships awarded to BRT by the University of Queensland are gratefully acknowledged. This work was supported in part through the grants from the Australian Council of International Agriculture Research (Project No CS 1/95/129).
|
Notes |
|---|
4 To whom correspondence should be addressed. Fax: +61 7 3365 1188. E-mail: N.Bodapati{at}uq.edu.au
5 Present address: School of Land and Food Sciences, The University of Queensland, St Lucia, QLD 4072, Australia.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
Anderson JE, Williams J, Kriedemann PE, Austin MP, Farquhar GD.1996. Correlation between carbon isotope discrimination and climate of native habitats for diverse Eucalypt taxa growing in a common garden. Australian Journal of Plant Physiology 23, 311–320.[Web of Science]
Barnes DK, Pearce RB, Carlson GE, Hart RH, Hanson CH.1969. Specific leaf weight differences in alfalfa associated with variety and plant age. Crop Science 9, 421–423.
Brown RH, Byrd GT.1997. Transpiration efficiency, specific leaf weight, and mineral concentration in peanut and pearl millet. Crop Science 37, 475–480.
Chakraborty S, Liu CJ, Cameron DF, Chandra A, Ramesh CR, Fernandes CD, Charchar MJ, Guodao L, Kelemu S.1998. Application of molecular markers to breed and select anthracnose-resistant Stylosanthes and to characterise endemic and exotic populations of Colletotrichum gloeosporioides. In: Larkin PJ, ed. Proceedings of the 4th Asia-Pacific Conference on Agricultural Biotechnology. Darwin, Australia, 53–56.
Condon AG, Richards RA, Farquhar GD.1987. Carbon isotope discrimination is positively correlated with grain yield and dry matter production in field grown wheat. Crop Science 27, 996–1001.
Davies PJ, Simko I, Mueller SM, Yencho GC, Lewis C, McMurry S, Taylor MA, Ewing EE.1999. Quantitative trait loci for polyamine content in an RFLP-mapped potato population and their relationship to tuberisation. Physiologia Plantarum 106, 210–218.
Ehdaie B, Bernhart D, Waines JG.1993. Genetic analysis of transpiration efficiency, carbon isotope discrimination, and growth characters in bread wheat. In: Ehleringer JR, Hall AE, Farquhar GD, eds. Stable isotopes and plant carbon-water relations. New York: Academic Press, 419–434.
Farquhar GD, O'Leary MH, Berry JA.1982. On the relationship between carbon isotope discrimination and the intercellular carbon dioxide concentrations in leaves. Australian Journal of Plant Physiology 9, 121–137.[Web of Science]
Farquhar GD, Ehleringer JR, Hubick KT.1989. Carbon isotope discrimination and photosynthesis. Annual Review of Plant Physiology and Molecular Biology 40, 503–537.[Web of Science]
Fulton TM, Grandillo S, Beck-Bunn T, Fridman E, Frampton A, Lopez J, Petiard V, Uhlig J, Zamir D, Tanksley SD.2000. Advanced backcross QTL analysis of Lycopersicon esculentumxLycopersicon parviflorum cross. Theoretical and Applied Genetics 100, 1025–1042.[Web of Science]
Henson IE.1983. Abscisic acid accumulation in detached leaves of rice (Oryza sativa L.) in response to water stress: a correlation with leaf size. Annals of Botany 52, 385–398.
Henson IE.1985. Modification of leaf size in rice (Oryza sativa L.) and its effects on water stress-induced abscisic acid accumulation. Annals of Botany 56, 481–487.
Hubick KT, Farquhar GD, Shorter R.1986. Correlation between water use efficiency and carbon isotope discrimination in diverse peanut (Arachis) germplasm. Australian Journal of Plant Physiology 13, 803–816.
Hubick KT, Shorter R, Farquhar GD.1988. Heritability and genotypexenvironment interactions of carbon isotope discrimination and transpiration efficiency in peanuts (Arachis hypogea L.). Australian Journal of Plant Physiology 15, 799–813.[Web of Science]
Kosambi DD.1944. The estimation of map distances from recombination values. Ann Eugen 12, 172–175.
Lander ES, Green P, Abrahamson J, Barlow A, Daly MJ, Lincoln SE, Newberg L.1987. Mapmaker: an interactive computer package for constructing primary genetic linkage maps of experimental and natural populations. Genomics 1, 174–181.[Medline]
Lebreton C, Lazic-Jancic V, Steed A, Pekic S, Quarrie SA.1995. Identification of QTL for drought responses in maize and their use in testing causal relationships between traits. Journal of Experimental Botany 46, 853–865.
Lilley JM, Ludlow MM, McCouch SR, O'Toole JC.1996. Locating QTL for osmotic adjustment and dehydration tolerance in rice. Journal of Experimental Botany 47, 1427–1436.
Liu CJ, Musial JM.1995. Restriction fragment length polymorphism detected by cDNA and genomic DNA clones in Stylosanthes. Theoretical and Applied Genetics 91, 1210–1213.[Web of Science]
Lu Z, Radin JW, Turcotte EL, Percy R, Zeiger E.1994. High yields in advanced lines of pima cotton are associated with higher stomatal conductance, reduced leaf area and lower leaf temperature. Physiologia Plantarum 92, 266–272.
Manly KF, Olson JM.1999. Overview of QTL mapping software and introduction to Map Manager QT. Mammalian Genome 10, 327–334.
Martin B, Nienhuis J, King G, Schaefer A.1989. Restriction fragment length polymorphisms associated with water use efficiency in tomato. Science 243, 1725–1728.
Masle J, Shin JS, Farquhar GD.1993. Analysis of restriction fragment length polymorphisms associated with variation of carbon isotope discrimination among ecotypes of Arabidopsis thaliana. In: Ehleringer JR, Hall AE, Farquhar GD, eds. Stable isotopes and plant carbon/water relations. San Diego: Academic Press, 371–386.
Meagher RB, McLean M, Arnold J.1988. Recombination within a subclass of restriction fragment length polymorphisms may help link classical and molecular genetics. Genetics 120, 809–818.
Mian MAR, Bailey MA, Ashley DA, Wells R, Carter TE, Parrot WA, Boerma HR.1996. Molecular markers associated with water use efficiency and leaf ash in soybean. Crop Science 36, 1252–1257.
Naidu BP, Jones GP, Paleg LG, Poljakoff-Mayber A.1987. Proline analogues in Melaleuca species: response of Melaleuca lanceolata and M. uncinata to salinity and water stress. Australian Journal of Plant Physiology 14, 669–677.[Web of Science]
Naidu BP.1998. Separation of sugars, polyols, proline analogues and betaines in stressed plant extracts by high performance liquid chromatography and quantification by ultra violet detection. Australian Journal of Plant Physiology 25, 793–800.[Web of Science]
Nelson CJ.1988. Genetic association between photosynthetic characteristics and yield: review of evidence. Plant Physiology and Biochemistry 26, 543–554.[Web of Science]
Pearce RB, Carlson GE, Barnes DK, Hart RH, Hanson CH.1969. Specific leaf weight and photosynthesis in alfalfa. Crop Science 9, 423–426.
Prioul JL, Quarrie SA, Causse M, de Vienne D.1997. Dissecting complex physiological functions into elementary components through the use of molecular quantitative genetics. Journal of Experimental Botany 48, 1151–1163.
Quarrie SA, Laurie DA, Zhu J, Lebreton C, Semikhodskii A, Steed A, Witsenboer H, Calestani C.1997. QTL analysis to study the association between leaf size and abscisic acid accumulation in droughted rice leaves and comparisons across cereals. Plant Molecular Biology 35, 155–165.[Web of Science][Medline]
Ray IM, Townsend MS, Muncy CM, Henning JA.1999. Heritabilities of water-use efficiency traits and correlations with agronomic traits in water-stressed alfalfa. Crop Science 39, 494–498.
Rhodes D, Hanson AD.1993. Quaternary ammonium and tertiary sulfonium compounds in higher plants. Annual Review of Plant Physiology and Plant Molecular Biology 44, 357–384.[Web of Science]
Richards RA.2000. Selectable traits to increase crop photosynthesis and yield of grain crops. Journal of Experimental Botany 51, 447–458.
Rieger M, Duemmel MJ.1992. Comparison of drought resistance among Prunus species from divergent habitats. Tree Physiology 11, 369–380.[Abstract]
Sanguineti MC, Tuberosa R, Landi P, Salvi S, Maccaferri M, Casarini, E, Conti S.1999. QTL analysis of drought-related traits and grain yield in relation to genetic variation for leaf abscisic acid concentration in field-grown maize. Journal of Experimental Botany 50, 1289–1297.
SAS.1992. SAS technical report. SAS/STAT software: changes and enhancements, Release 6.07. (SAS institute inc: Cary, NC).
Simko I, McMurry S, Yang HM, Manschot A, Davies PJ, Ewing EE.1997. Evidence from polygene mapping for a causal relationship between potato tuber dormancy and abscisic acid content. Plant Physiology 115, 1453–1459.[Abstract]
Tanksley SD.1993. Mapping polygenes. Annual Review of Genetics 27, 205–233.[Web of Science][Medline]
Tanner CB, Sinclair TR.1983. Efficient water use in crop production: research or re-search? In: Taylor HM, Jordan WR, Sinclair TR, eds. Limitations to efficient water use in crop production. Madison, Wisconsin, USA: ASA-CSSA-SSSA, 1–27.
Teulat B, This D, Khairallah M, Borries C, Ragot C, Sourdille P, Leroy P, Monneveux P, Charrier A.1998. Several QTLs invoved in osmotic-adjustment trait variation in barley (Hordeum vulgare L.). Theoretical and Applied Genetics 96, 688–698.[Web of Science]
Thumma BR, Naidu BP, Cameron DF, Bahnisch LM.1998a. Carbon isotope discrimination and specific leaf weight estimate transpiration efficiency indirectly in Stylosanthes under well-watered conditions. In: Proceedings of the 9th Australian Agronomy conference 1998, 167–169.
Thumma BR, Naidu BP, Cameron DF, Bahnisch LM.1998b. Transpiration efficiency and its relationship with carbon isotope discrimination under well-watered and water-stressed conditions in Stylosanthes scabra. Australian Journal of Agricultural Research 49, 1039–1045.[Web of Science]
Van Ooijen JW.1992. Accuracy of mapping quantitative trait loci in autogamous species. Theoretical and Applied Genetics 84, 803–811.[Web of Science]
Van Ooijen JW, Malliepaad C.1996. MapQTL VERSION 3.0: software for calculation of QTL position on genetic maps. Wageningen: CPRO-DLO.
Virgona JM, Farquhar GD.1996. Genotypic variation in relative growth rate and carbon isotope discrimination in sunflower is related to photosynthetic capacity. Australian Journal of Plant Physiology 23, 227–236.[Web of Science]
Virgona JM, Hubick KT, Rawson HM, Farquhar GD, Downes RW.1990. Genotypic variation in transpiration efficiency, carbon isotope discrimination and carbon allocation during early growth in sunflower. Australian Journal of Plant Physiology 17, 207–214.[Web of Science]
Wilson JR, Fisher MJ, Schulze ED, Dolby GR, Ludlow MM.1979. Comparison between pressure-volume and dew point hygrometry techniques for determining water-relation characteristics of grass and legume leaves. Oecologia 41, 77–88.[Web of Science]
Wong SC, Cowan IR, Farquhar GD.1979. Stomatal conductance correlates with photosynthetic capacity. Nature 282, 424–426.[Web of Science]
Wright GC, Rao RCN, Farquhar GD.1994. Water-use efficiency and carbon isotope discrimination in peanuts under water deficit conditions. Crop Science 34, 92–97.
Zacharisen MH, Brick MA, Fisher AG, Ogg JB, Ehleringer JR.1999. Relationship between productivity and carbon isotope discrimination among dry bean lines and F2 progeny. Euphytica 105, 239–250.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J.I.L Morison, N.R Baker, P.M Mullineaux, and W.J Davies Improving water use in crop production Phil Trans R Soc B, February 12, 2008; 363(1491): 639 - 658. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. G. Condon, R. A. Richards, G. J. Rebetzke, and G. D. Farquhar Breeding for high water-use efficiency J. Exp. Bot., November 1, 2004; 55(407): 2447 - 2460. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. W. Cowley Jr. Genomics and homeostasis Am J Physiol Regulatory Integrative Comp Physiol, March 1, 2003; 284(3): R611 - R627. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||