Crop Science Grow Your Career with CSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (9)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lambrides, C. J.
Right arrow Articles by Shorter, R.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Lambrides, C. J.
Right arrow Articles by Shorter, R.
Agricola
Right arrow Articles by Lambrides, C. J.
Right arrow Articles by Shorter, R.
Related Collections
Right arrow Water Stress
Right arrow Crop Physiology & Metabolism
Right arrow Sunflower
Published in Crop Sci. 44:1642-1653 (2004).
© 2004 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA

CROP PHYSIOLOGY & METABOLISM

Genetic Variation for Carbon Isotope Discrimination in Sunflower

Association with Transpiration Efficiency and Evidence for Cytoplasmic Inheritance

C. J. Lambridesa,c,*, S. C. Chapmanb and R. Shorterb

a School of Land and Food Sciences, Univ. of Queensland, St. Lucia 4072, Brisbane, Australia
b Commonwealth Scientific and Industrial Research Organisation Plant Industry, 306 Carmody Rd., St. Lucia 4067, Brisbane, Australia
c Commonwealth Scientific and Industrial Research Organisation Plant Industry, 306 Carmody Rd., St. Lucia 4067, Brisbane, Australia

* Corresponding author (chris.lambrides{at}uq.edu.au).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Plants accumulate isotopes of carbon at different rates because of discrimination against 13C relative to 12C. In plants that fix carbon by the C3 pathway, the amount of discrimination correlates negatively with transpiration efficiency (TE) where TE is the amount of dry matter accumulated per unit water transpired. Therefore, carbon isotope discrimination ({Delta}) has become a useful tool for selecting genotypes with improved TE and performance in dry environments. Surveys of 161 sunflower (Helianthus spp.) genotypes of diverse origin revealed a large and unprecedented range of genetic variation for {Delta} (19.5–23.8{per thousand}). A strong negative genetic correlation (rg) between TE and {Delta} (rg = –0.87, P < 0.001) was observed in glasshouse studies. Gas exchange measurements of field grown plants indicated that {Delta} was strongly correlated with stomatal conductance to water vapor (g), (rg = 0.64, P < 0.01), and the ratio of net assimilation rate (A) to g, (rg = –0.86, P < 0.001), an instantaneous measure of TE. Genotype CMSHA89MAX1 had the lowest TE (and highest {Delta}) of all genotypes tested in these studies and low yields in hybrid combination. Backcrossing studies showed that the TE of this genotype was due to an adverse effect of the MAX1 cytoplasm, which was inherited from the diploid perennial H. maximiliani Schrader. Overall, these studies suggested that there is an excellent opportunity for breeders to develop sunflower germplasm with improved TE. This can be achieved, in part, by avoiding cytoplasms such as the MAX1 cytoplasm.

Abbreviations: {Delta}, carbon isotope discrimination • A, net assimilation rate • CHLadj, chlorophyll content adjusted for leaf thickness • CMS, cytoplasmic male sterility • EMS, error mean square • g, stomatal conductance to water vapor • HGT, plant height • HI, harvest index • LA, leaf area • LDW, leaf dry weight • RDW, root dry weight • SDW, stem dry weight • SLW, specific leaf weight • TDW, total dry weight • TE, transpiration efficiency • VPD, vapor pressure deficit • WUE, water use efficiency • WUSE, total water use


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
IN AUSTRALIA, the majority of sunflower is grown in the summer in dry subtropical environments (<500 mm rainfall) on soils that vary in plant available water holding capacity from 90 to 250 mm. Seasonal rainfall varies greatly across years, and large evaporative demands occur because of daily solar radiation of more than 25 MJ m–2 d–1 and maximum temperatures frequently exceeding 35°C. Water is the major abiotic limitation to growth and yield, such that mean yields in Australia are about 990 kg ha–1 (1993–2002; FAOSTAT, 2003). There are few agronomic options, other than conservation tillage, to improve water supply in these environments and so genetic improvements in the ability of sunflower to access water or to use it more efficiently are appropriate. Water use efficiency (WUE) is a term used to express the ratio of total dry matter production to evapotranspiration:

[1]
where TEa is the transpiration efficiency (above ground dry weight transpiration–1), Es is the water lost by evaporation from the soil surface, and T is water transpired by the crop (Richards, 1991). If rainfall is infrequent and the crop is reliant on water stored in the subsoil, as are many sunflower crops grown in Australia, then Es is low and increasing TEa provides the greatest opportunity to increase WUE (Richards et al., 2002).

Plants incorporate isotopes of carbon into their tissue at different rates because of discrimination against 13C relative to 12C. This difference in discrimination has been negatively correlated with TE in many C3 species and so carbon isotope (13C/12C) discrimination ({Delta}) of leaf tissues has been proposed as a potential tool for selecting genotypes with improved performance under water limited conditions (Farquhar and Richards, 1984; Hubick et al., 1988; Ehleringer et al., 1991; Condon and Richards, 1992; Condon et al., 1993; Rebetzke et al., 2002). Rebetzke et al. (2002) demonstrated that early generation divergent selection for {Delta} affected grain yield among high and low {Delta}, BC2F4:6 progeny of wheat (Triticum aestivum L.). The low {Delta} lines had a yield advantage compared to the high {Delta} lines of up to 11% in low rainfall Australian environments.

Generally, a negative correlation exists between leaf {Delta} and TE (see review by Hall et al., 1995). Virgona and Farquhar (1996) observed a negative correlation (r = –0.98, P < 0.001) between leaf {Delta} and TE for a range of sunflower genotypes grown in a glasshouse. Two factors are potentially associated with improved TE in crop plants, improved photosynthetic capacity (the amount of photosynthetic machinery per unit leaf area) and differences in stomatal conductance (the regulation of water and CO2 through stomata) (Farquhar et al., 1982). For glasshouse grown sunflower, differences in photosynthetic capacity were thought to underlie the variation observed for {Delta} and, therefore, TE (Virgona and Farquhar, 1996) because no association was observed between {Delta} and conductance.

A project was initiated to develop sunflower germplasm with enhanced TE with leaf tissue {Delta} as an indicator. On the basis of glasshouse studies, Virgona et al. (1990) and Virgona and Farquhar (1996) identified genetic variation for {Delta} and TE in a small, though broad, selection of sunflower accessions. In the present study, the analysis of variation for {Delta} in sunflower was extended to field conditions in a greater range of germplasm. The results of several surveys for leaf {Delta} in sunflower sampled from Australian and USA field sites were evaluated. On the basis of these field surveys, a subset of genotypes was selected to investigate the relationship between {Delta} and TE in a glasshouse experiment. Apart from reporting on {Delta} and TE, experimental and genetic parameters (repeatability, variances and correlations) are computed to help evaluate the potential of {Delta} for use in breeding programs. In addition, the effect of the cytoplasm on TE was examined by investigating the {Delta} values of a set of alloplasmic lines that share a common nuclear genome but have different ancestral cytoplasms.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Genetic Materials for Field Surveys of {Delta}
A total of 161 germplasm lines and hybrids (Table 1) were sampled for leaf {Delta} in at least one of three field locations: Fargo, ND, USA (sown May 1997) at the USDA summer nursery site; Gatton, Queensland, Australia (sown February 1999) at the Pacific Seeds Research Station (referred to as GattPac hereafter); and Gatton, Queensland, Australia (sown August 1999) at the CSIRO Research Station (referred to as Gatton hereafter). Material was grown in two replicates at each Australian location and one replicate in Fargo. Our aim was to sample a wide range of genetic diversity in cultivated (H. annuus) background without sampling wild species directly. The genetic material surveyed represented a diverse set of (i) nuclear backgrounds, including accessions derived from wild species (H. anomalus Blake, H. argophyllus Torrey and Gray, H. bolanderi Gray, H. debilis Nuttall subsp. cucumerifolius Heiser, H. debilis Nuttall subsp. silvestris Heiser, H. deserticola Heiser, H. exilis Gray, H. hirsutus Rafinesque, H. paradoxus Heiser, H. petiolaris Nuttall, H. praecox Engleman and Gray, H. resinosus Small, and H. tuberosus L.) and (ii) cytoplasmic backgrounds including a set of eight HA89 alloplasmic lines (see top of Table 1). The alloplasmic lines were also evaluated across several seasons at Gatton and were sown in February 2000 (two replicates), August 2000 (two replicates), February 2001 (one replicate) and February 2002 (two replicates). All field sites were given adequate nutrition for normal plant growth and were kept insect and disease free before sampling. The Fargo site was sown on a full profile of moisture and exposed to ambient rainfall. The Australian trials were sown on fallow ground and given 25 to 50 mm of irrigation if required, during establishment or before sampling for {Delta} and then exposed to ambient rainfall thereafter. Genotypes were planted in single row plots (3.5 m long at 1-m row spacing) and hand thinned to a density of 16 plants per row. Genotypes were sown in a randomized complete block design that also incorporated an {alpha} lattice.


View this table:
[in this window]
[in a new window]
  		Table 1. {Delta}{dagger} Values (blup, best linear unbiased predictor) for 161 sunflower genotypes (Helianthus spp.) grown in at least one field location (Fargo 1997, GattPac 1999, and Gatton 1999). Genotypes that were subsequently used in a glasshouse experiment to study the association between {Delta} and transpiration efficiency are also indicated.

 
Traits Measured in the Field
Carbon Isotope Discrimination ({Delta})
At the Fargo location, two upper canopy leaves per plant from five plants in the late vegetative phase were sampled from the single replicate of each genotype. For all Australian locations, one fully expanded sunlit leaf (leaf 8–12) from each of 5 to 10 plants (5–10 leaves in all) was sampled during the early vegetative phase from each of two replicates of each genotype. Leaves were dried overnight at 80°C and ground through a 0.5-mm screen to a fine powder for carbon isotope analysis. All analyses were done by mass spectrometry at the Research School of Biological Sciences, Australian National University, Canberra, Australia. The {Delta} value is a measure of the 13C/12C ratio in plant material relative to the value of the same ratio in ambient air and was calculated according to Farquhar and Richards (1984). The {Delta} values are expressed in per mil ({per thousand}).

Chlorophyll Content
At the 1999 Gatton trial, chlorophyll content of each genotype was estimated with a SPAD meter (Minolta, Osaka, Japan). Twenty readings were averaged per genotype (four readings per fully expanded sunlit leaf per plant times five plants per replicate) for each of two replicates. Measurements were made at the time of sampling for {Delta}.

Instantaneous Gas Exchange Measurements
Measurements of rates of leaf transpiration and CO2 assimilation with a LI-COR 6400 (LI-COR, NE) were made on 12 Nov. 1999 during the vegetative phase (16–20 leaf stage and approximately 2 wk before anthesis) for each of 48 genotypes (Table 2) grown at the Gatton site. These genotypes had been sampled for {Delta} as previously described. Gas exchange measurements were made near the middle of the day on a single fully expanded sunlit leaf per replicate x two replicates in natural sunlight when photosynthetic active radiation was between 1500 to 2000 µmol m–2 s–1. The flow rate was set between 500 to 750 µmol s–1 to achieve rapid equilibration (40–60 s). Chamber temperature was set to ambient temperature (24–28°C). Net assimilation rate (A) and stomatal conductance to water vapor (g) were logged at each measurement.


View this table:
[in this window]
[in a new window]
  		Table 2. Instantaneous gas exchange measurements for 48 genotypes measured in the field at Gatton in spring 1999 with a LI-COR 6400 photosynthesis gas analyzer. Data presented are mean values for two replicates.

 
Glasshouse Evaluation of Transpiration Efficiency
Fourteen lines spanning the range of {Delta} observed in the field surveys were selected to investigate the relationship between {Delta} and TE in a glasshouse experiment conducted at CSIRO, St. Lucia, Brisbane. In addition, five commercial hybrids (Suncross 53, Monosun 140, Advantage, Hysun 36, and Hysun 46) and the parents of Hysun 36 and 46 were included. The commercial hybrids were selected for their low (Hysun 36, Hysun 46, Suncross 53), moderate (Advantage), or high (Monosun 140) {Delta} values on the basis of surveys of Australian commercial variety trials (unpublished data).

The glasshouse experiment was conducted during September and October 1999. Day temperatures averaged 26.5°C (maximum of 34°C) and night temperatures averaged 23°C (minimum of 18°C). Four seeds of each genotype were planted in pots (0.15-m diameter, 0.6 m high) containing 11 kg of air dry soil with a water holding capacity (after free drainage) of 18% (w/w) and a lower limit of 11%. The soil mix included a complete nutrient fertilizer. Four replicate pots were planted per genotype, and pots were spaced five pots wide in the glasshouse in a randomized complete block design that also incorporated an {alpha} lattice.

The pots were watered to keep them at field capacity until 21 d after planting. The plants were then thinned to one plant per pot. The pots were weighed and watered to field capacity, and the soil surface covered to 0.02-m depth with white alcathene (polyethylene) beads that minimized evaporation from the soil surface. By monitoring several pots with different plant size, water was added every 2 to 3 d in 100- to 300-mL quantities. All pots were weighed each week and at the end of the experiment. Total water use (WUSE) was estimated by subtracting final pot weight (less above ground plant) from initial pot weight (less plant weight at 21 d after sowing) and adding the amount of water that had been applied to the pots.

Plants were harvested 50 d after sowing (5 Nov. 1999) when plants were at the 16- to 21-leaf stage. On the final day, plant height (HGT) was measured from the top of the apical meristem to the base of the stem and leaf number per plant (LNo) determined. Harvested plants were separated into stems and leaves and oven dried to estimate stem dry weight (SDW) and leaf dry weight (LDW). Leaf area (LA) was determined on fresh leaves before drying. The soil was washed from the pots to recover roots, which were dried and weighed to provide an estimate of root dry weight (RDW). Total dry weight (TDW) was estimated as SDW + LDW + RDW.

TE was calculated by dividing total dry matter accumulation during growth (TDW minus initial plant weight at 21 d after sowing) by WUSE and multiplying by the average of the vapor pressure deficit (VPD) for the duration of the experiment. Values of daylight averaged VPD were estimated from temperature and relative humidity (Tanner and Sinclair, 1983) logged every 10 min and coefficients reported by Abbot and Tabony (1985). The average value of VPD (1.5 kPa) was not weighted for diurnal variation in stomatal conductance (Farquhar et al., 1982). Estimates of TE were adjusted for VPD to enable comparisons with future experiments under different VPD conditions. Initial weight of plants was estimated from seedlings removed at the time of thinning. Specific leaf weight (SLW) was calculated as LDW LA–1. Values of {Delta} for each plant were determined from leaf tissue bulked at the final harvest, processed and analyzed as described above.

Harvest Index (HI), SLW, HGT of Inbred Lines Evaluated in the Field at Gatton 2000
The relationship between {Delta}, HI, SLW, and HGT was investigated in a summer sown field trial at Gatton in 2000. Fourteen of the 20 inbreds used in the glasshouse trial were used in this experiment. Urea at 150 kg ha–1 was applied before sowing, and the soil contained a full profile of moisture (about 270 mm). Fifty millimeters of irrigation was applied at sowing with 57 mm of rainfall received after irrigation. Genotypes were planted in single row plots (4 m long with 1-m row spacing) and hand thinned to a density of about 16 plants row–1. Four replicates were sown in a randomized complete block design that also incorporated an {alpha} lattice. Plants were kept disease and pest free with the use of appropriate chemicals. At physiological maturity (when the back of the capitulae had turned yellow), three representative plants per plot were harvested and oven dried. Values of HI were calculated as the ratio of grain yield to total above ground dry matter production averaged for the three plants. SLW was estimated from 5 to 10 fully sunlit leaves sampled during the early vegetative phase (leaves 8 to 12). These leaves were also used for {Delta} analyses. HGT was measured as the maximum length of the main stem.

Yield Evaluation of Alloplasmic Material
Seven alloplasmic, cytoplasmically male sterile (CMS), HA89 female lines based on the MAX1, PET1, PET2, ANN2, ANN3, GIG1, and PEF1 cytoplasms were crossed with four male parents, HAR4 seln 1, SA52, PAR-1673-2 seln 1, and DES-1474-2. At Gatton in 2000, hybrid seed was planted in single row plots (4 m long with 1-m row spacing) and hand thinned to a density of about 16 plants row–1. Plots were arranged in a randomized complete block design with four replicates that also incorporated an {alpha} lattice. At physiological maturity, the capitulae from the center 3.5 m of each plot were hand harvested, oven dried and threshed in a single plot thresher.

Statistical Analysis
For the {alpha} lattice designs, restricted maximum likelihood methods were applied using the mixed models approach and the PROC MIXED option (SAS Institute, 1990) with genotypes treated as fixed effects, while replicates and blocks nested within replicates were treated as random effects. Least square means (blues—best linear unbiased estimator) were derived for each genotype by using the LSMEANS option of PROC MIXED. For the glasshouse experiment, it was necessary to use spatial analysis (Gilmour et al., 1997) to adjust for spatial variation among pots. Fitting the AR1 (autoregressive 1) x AR1 model was sufficient to account for the extraneous variation.

Across Site Analysis to Express All Genotypes on a Common Scale
For genotypes grown at Fargo in 1997, GattPac in the summer of 1999 and Gatton in the spring of 1999, variance components for genotype , genotype x environment interaction and phenotype were determined for {Delta} using an across-site analysis with the ASREML software (Gilmour et al., 1997) treating all sources of variation as random. The data input for this analysis included blues of each genotype obtained at the GattPac and Gatton sites and raw data from Fargo. Values of {Delta} (blups—best linear unbiased predictor) were determined for each genotype by using the ‘Predict’ and ‘wt’ options of ASREML. The ‘wt’ is calculated from the trial error mean square (EMSi) and is the same for all entries in a trial. For example, trial i, wti = no. reps x AvgEMS EMSi–1, where AvgEMS is the averaged EMSi across all trials. EMSi was 0.09 for GattPac, and 0.10 for Gatton. For the unreplicated data from Fargo, the EMS was set to the highest value observed in other {Delta} trials, (i.e., 0.12). Calculating {Delta} values in this way ensured that more weight was given to experiments that were grown and measured with greater precision, (i.e. lower EMSi or higher heritability).

Repeatability Estimates and Correlations
Repeatability (Fehr, 1987, p. 97) was used as a measure of the chance of detecting the same differences among genotypes in future experiments and can be used to decide how the trait may be utilized by breeders. For individual experiments, variance components were estimated with the COVTEST option of PROC MIXED. Broad sense repeatability on an individual plot (or pot) basis was calculated for traits within an experiment from variance components as the ratio of genetic variance to total phenotypic variance :

[2]
where

[3]
and {sigma}2e is the environmental variance.

Estimates of broad sense repeatability on a genotype mean basis were calculated from variance components as follows

[4]
where r is the number of replicates. These estimates may be biased upward because {sigma}2ge was confounded with the estimate of {sigma}2g. Genetic (rg) and phenotypic (rp) correlations were calculated using variance components. For example,

[5]
and

[6]
where Covg is the genetic covariance between traits x and y and Covp is the phenotypic covariance between traits x and y.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
{Delta} Surveys
Estimates of repeatability of {Delta} on an individual plot basis were high in this study, h2 ≥ 0.77, for all replicated field experiments (Table 3), indicating that the assay developed for field grown sunflower was robust. The strategy was to sample early in the vegetative phase (but late enough to avoid seed derived effects) under well-watered conditions. Previous studies with wheat (Condon and Richards, 1992) and cowpea [Vigna unguiculata (L.) Walp.] (Hall et al., 1993) showed that this phase of growth was optimal for identifying genotypic differences for constitutive {Delta}, (i.e., before the effects of differences in soil water availability).


View this table:
[in this window]
[in a new window]
  		Table 3. Phenotypic (rp) and genetic correlations (rg) among traits measured on a range of inbred and hybrid genotypes in two field experiments (Gatton spring 1999, summer 2000) and one glasshouse experiment (1999). Broad sense repeatability values on an individual plot/pot basis and on a genotype mean basis are given for each trait.

 
A large and unprecedented range of almost 4.4{per thousand} (19.5–23.8) was observed for sunflower {Delta}s in these surveys (Table 1), exceeding the range of about 2.5{per thousand} observed in glasshouse studies of sunflower (Virgona et al., 1990; Virgona and Farquhar, 1996). The large range of {Delta} identified in the present study suggested that there is potential to select sunflower germplasm with improved TE. The observed range was in part due to the large amount of interspecific material sampled, and also to the extreme value of one genotype, CMSHA89MAX1. This genotype consistently exhibited the highest {Delta} value in all experiments of this study, exceeding the next highest genotype by about 1.8{per thousand}. Consequently, with CMSHA89MAX1 excluded, the range identified in this study was similar to that obtained by Virgona et al. (1990) and Virgona and Farquhar (1996). This genotype will be discussed in greater detail below.

The genotype with the lowest {Delta} and potentially the highest TE was ANO-1509-2 seln 1, a line derived from H. anomalus. This rare species originates in south central and west central Utah and in northeastern Arizona, USA. The habitats in Utah where this species is found are arid and typically consist of deep, fine sand supporting xeric flora (Rogers et al., 1982). Several genotypes derived from H. paradoxus (e.g., PAR-1673-2 seln 1) and H. hirsutus (e.g., HIR-1734-2 seln 1) also exhibited low {Delta} values (Table 1). Helianthus paradoxus is an endangered species found in marshy saline soils of Ft. Stockton, TX, USA (Rogers et al., 1982). Helianthus hirsutus is found in dry and open habitats extending from Texas to Minnesota in central USA and from Pennsylvania to Florida in the eastern USA (Rogers et al., 1982).

Genotype x Environment Interaction for {Delta}
The across-site analysis of {Delta} showed that {sigma}2g (0.276) was five times larger than {sigma}2ge (0.056) or 83% of {sigma}2g + {sigma}2ge. The low level of genotype x environment interaction for {Delta} in sunflower using the protocols of this study was consistent with studies of wheat (Condon and Richards, 1992) and peanut (Arachis hypogaea L.) (Hubick et al., 1988), and suggests that selection for {Delta} can occur in a range of environments if sampling strategies are followed that minimize nongenetic components of variation.

Relationship between TE, {Delta}, and Other Traits in Different Germplasm Types
In the 1999 glasshouse study, heritability on a genotype mean basis was high for all traits evaluated (Table 3). High negative genetic correlations between TE and {Delta} were observed (rg = –0.87, P < 0.01 for hybrids + inbreds, and rg = –0.92, P < 0.01 for inbreds). The phenotypic correlation between TE and {Delta} for the hybrid group was rp = –0.49, but this correlation was derived from only five data points. The correlations between TE and {Delta} in this study are consistent with the studies by Virgona and Farquhar (1996), who observed a phenotypic correlation of rp = –0.98 (P < 0.001). These results suggest that {Delta} may be a very useful surrogate for indirect selection of TE in sunflower improvement programs, although the association between TE and {Delta} will need to be confirmed in segregating populations. Studies involving random inbred lines, such as the one described here, provide useful but not definitive proof of the genetic basis for an association between TE and {Delta}.

When inbreds and hybrids were analyzed together, strong positive genetic correlations between TE and biomass (TDW, SDW, LDW, RDW) were also observed, ranging from rg = 0.47 to 0.56. However, the genetic correlations dropped substantially when the inbred group was analyzed separately. For example, the genetic correlations between TE and TDW decreased by almost half from rg = 0.56 (P < 0.01) when hybrids and inbreds were analyzed together to rg = 0.29 (P < 0.05) when inbreds were analyzed alone (Table 3). Hybrid variety performance is characterized by heterosis for a range of traits including plant growth. In this study, an analysis of Hysun 36, Hysun 46, and their respective parent lines indicated heterotic effects (superior performance compared with both parents) for traits including TE, {Delta}, LA, TDW, LDW, SLW, and HGT (Table 4). Combined analyses of hybrids and inbreds should be undertaken with caution in physiological studies of growth. Virgona and Farquhar (1996) observed a phenotypic correlation between TE and biomass of rp = 0.99, which may be an overestimate because their study of TE consisted of a combined analysis of four hybrids, four open pollinated varieties (which also express heterosis for many traits), and seven inbreds.


View this table:
[in this window]
[in a new window]
  		Table 4. Genotypic means (four replicates) of 20 inbred lines and five hybrids for a range of traits measured in a glasshouse experiment conducted at Brisbane, Australia in November 1999.

 
A small but significant genetic correlation (rg = –0.26, P < 0.05) was observed between TE and LNo (Table 3), suggesting that improved TE in sunflower may be related to slightly slower development rate as measured by rate of leaf appearance. No association was observed between TE and LA or HGT (taken at the completion of the glasshouse experiment).

The strong genetic correlation between SLW (a measure of leaf thickness) and TE (rg = 0.63, P < 0.01) for the inbred lines (Table 3) in the glasshouse experiment was unexpected in sunflower, although such a correlation was observed previously in peanut (Wright et al., 1993). Virgona and Farquhar (1996) found no association between specific leaf area (SLA, the inverse of SLW) and TE in sunflower, possibly because of the confounding effects of analyzing hybrids and inbreds together as mentioned above.

The lowest {Delta} genotype, ANO-1509-2 seln 1, was expected to have a higher value of TE than that observed in this glasshouse experiment (Table 4). While useful for developing correlations between {Delta} and TE, the glasshouse screening of germplasm for {Delta} per se is not recommended, given that glasshouse values (Table 4) correlated only moderately with those obtained under field (Table 1) conditions (r = 0.44). The basis of the lack of correlation is unknown but may be associated with a number of factors. A reranking of genotypes in the glasshouse may be caused by genotypic differences in growth rate that may result in some entries depleting a greater proportion of the water supply in the pot between waterings compared with the field situation where water is freely available. For example, some slow growing genotypes such as ANO-1509-2 seln 1 may have been overwatered during some phases of growth. Reranking may also be associated with microenvironmental effects such as temperature and CO2 gradients that can exist in glasshouse environments depending on the relative positioning of pots and ventilation systems.

Previous studies with peanut showed that for some germplasm undesirable associations were observed between {Delta} and HI (Wright et al., 1993). When selecting for low {Delta}, it is important that there be no adverse pleiotropic effects on other key traits. Consequently, a sample of the inbred lines used in the glasshouse experiment was evaluated for {Delta}, HI, SLW, and HGT in a replicated field trial under dryland conditions at Gatton in 2000. In this experiment, no associations were found between {Delta} and HI and HGT (Table 3). Branching habit is a trait used for the development of male restorer sunflower lines and is widely selected by sunflower breeders. There was no effect on {Delta} for a pair of isogenic lines differing in branching habit grown at GattPac in 1999. TUB-1789 seln 1 is branched, and TUB-1789 seln 2 is unbranched and both gave identical {Delta} values (Table 1). More research is needed to verify these relationships and the relationship between {Delta} and yield and yield components in segregating populations.

Factors underlying Variation for {Delta} and TE in Sunflower
Differences in the diffusion of CO2 to the photosynthetic apparatus (conductance) and/or photosynthetic capacity may be responsible for genetic variation in {Delta} (Farquhar et al., 1982). Virgona et al. (1990) concluded that photosynthetic capacity rather than stomatal conductance was driving variation in TE of the sunflower genotypes they studied. At one location (GattPac in 1999), the leaves sampled for {Delta} from 87 genotypes were also assayed for SPAD units (a measure of chlorophyll content). In some species, e.g., peanut, SPAD measures correlated strongly with specific LA and specific leaf nitrogen and, therefore, photosynthetic capacity (Nageswara Rao et al., 2001). Despite the high repeatability (h2 = 0.9) and significant genetic variation for the SPAD characteristic, no association was found between chlorophyll content and {Delta} in this experiment. While the effect of chlorophyll content adjusted for leaf thickness (CHLadj) was not determined in this study, later studies showed that CHLadj was also not associated with {Delta} (unpublished data). Variation for other traits such as maturity, HGT, a visual estimate of LA and branching habit of the 87 genotypes was also unrelated to {Delta} (data not shown).

Measurements of gas exchange allow the determination of instantaneous parameters directly related to TE such as the ratio A g–1. Gas exchange was measured on the 48 genotypes grown in the Gatton 1999 trial where moisture was not limiting and light levels were uniform and high (1500–2000 µmol m–2 s–1). The precision of these measurements was moderate to high, with repeatability h2 ≥ 0.68 (Table 3) for A, g, and A g–1. A strong genetic correlation was observed between {Delta} (sampled from these plots at the early vegetative phase) and A g–1 (rg = –0.86, P < 0.01) (Table 3), providing additional evidence that {Delta} is a useful surrogate for TE in sunflower. In this experiment, {Delta} showed a strong positive genetic correlation with g (rg = 0.64, P < 0.01) but no correlation with A. In addition, a significant correlation between A and g was also observed (rg = 0.50, P < 0.01). These results suggest that variation in g contributes significantly to variation in TE in sunflower. Virgona and Farquhar (1996) found no relationship between g and TE in their study where most of the variation in TE was attributed to variation in photosynthetic capacity. Interestingly, the lowest {Delta} genotype (ANO-1509-2 seln 1), discussed above, had very low stomatal conductance, high A g–1 (Table 2) and thick leaves (Table 4).

Selecting Low {Delta} Lines in Sunflower Improvement Programs
When considering germplasm for a sunflower breeding program, lines such as ANO-1509-2 seln 1 should be avoided, despite its low {Delta} and potentially high TE. This line was slow growing and lacked vigor in field plantings. Simultaneous selection for lines with low {Delta} and high plant vigor would be preferred (Condon et al., 2002). Accordingly, genotype HAR4 seln 1 with low {Delta} (Table 1) and high plant vigor, i.e., high TDW (Table 2), was selected. This genotype was used as a parent to make segregating populations for future experiments designed to understand the inheritance of {Delta}, identify molecular markers linked to {Delta}, determine the linkage relationship between {Delta} and other key breeding traits of sunflower, and develop germplasm that can be used in commercial breeding programs. HAR4 has the additional benefit of being highly resistant to two major pathogens: sunflower rust (caused by Puccinia helianthi Schw.) and downy mildew (caused by Plasmopara halstedii Farl.).

Evidence for Cytoplasmic Inheritance of {Delta}
The high {Delta} value of CMSHA89MAX1 (Table 1) was unexpected. This entry was one of a set of alloplasmic lines included in the survey. Members of an alloplasmic set share a common nuclear genome but have different ancestral cytoplasms. The {Delta} of CMSHA89MAX1 was almost 3{per thousand} higher than that of the other members of the set, with no significant difference among the other non-MAX1 cytoplasms (Table 5). To test the effect of the MAX1 cytoplasm on other nuclear genomes, lines PAR-1673-2 seln 1 and DES-1472-2 were backcrossed into the MAX1 cytoplasm. When evaluated in several field trials, the effect of the MAX1 cytoplasm on {Delta} was again apparent (Table 5). Backcross lines containing the MAX1 cytoplasm had {Delta} values 2 to 3{per thousand} higher than their recurrent parents. These results clearly suggest a cytoplasmic role in the control of {Delta} at least in some sunflower germplasm lines.


View this table:
[in this window]
[in a new window]
  		Table 5. {Delta} values of alloplasmic lines based on data collected in the field at Gatton, Australia, in 1999, 2000, 2001, and 2002 for the HA89 materials and in 2001 and 2002 for the PAR-1673-2 seln 1 and DES-1472-2 materials.

 
To confirm these observations and to understand the physiological basis of the large differences in {Delta}, lines CMSHA89MAX1, CMSHA89PET1 and HA89ANN1 were included in the glasshouse experiment in which TE was measured. These data (Table 4) indicated that the TE of CMSHA89MAX1 was only 83% that of CMSHA89PET1 and HA89ANN1. The lower TE of CMSHA89MAX1 was associated with significantly lower specific leaf weight (SLW) (thinner leaves). However, no significant differences were observed in HGT or LA (Table 4). Field measurements of gas exchange indicated that A of CMSHA89MAX1 was markedly lower than the other entries (Table 2). The g of CMSHA89MAX1 was near average, but A g–1 was substantially lower than all other entries. These observations are consistent with the low value of TE observed in the glasshouse experiment for this genotype.

Seven alloplasmic, cytoplasmically male sterile, HA89 lines based on the MAX1, PET1, PET2, ANN2, ANN3, GIG1, and PEF1 cytoplasms were crossed with each of four male parents HAR4 seln 1, SA52, PAR-1673-2 seln 1, and DES-1474-2 to test the effect of cytoplasm on grain yield. Hybrids based on CMSHA89MAX1 were 23% lower yielding (0.84 ± 0.16 Mg ha–1) than hybrids made on non-MAX1 females (1.09 ± 0.08 Mg ha–1, P < 0.05, Table 6). These data provide further evidence that the inheritance of {Delta} in this material has a cytoplasmic basis. The effect of the MAX1 cytoplasm on {Delta} (Tables 1, 5), TE (Table 4) and grain yield (Table 6) may be related to the lower levels of photosynthesis that this cytoplasm exhibits (Table 2). Jan (1992) described a "reduced vigor" effect that can occur when cultivated genomes are backcrossed into "perennial" cytoplasms such as MAX1.


View this table:
[in this window]
[in a new window]
  		Table 6. Mean grain yield (Mg ha–1) (two replicates) and mean {Delta}{dagger} values (four replicates) for hybrids made by crossing a HA89 alloplasmic set of females with four male lines (HAR4 seln 1, SA52, PAR-1673-2 seln 1 and DES-1474-2). These hybrids were evaluated in a dryland field trial at Gatton in summer 2000.

 
Since the discovery of cytoplasmic male sterility (CMS) in sunflower (Leclercq, 1969), breeders around the world have exclusively used the PET1 and ANN1 cytoplasms to develop commercial sunflower hybrids. The vulnerability of using a narrow range of cytoplasms has been recognized and has led to the search for several other CMS systems. To date, more than 40 CMS systems have been derived and characterized (Jan, 1997). Our study suggests that breeders developing drought-tolerant hybrids may need to consider selection of cytoplasmic genomes, in addition to nuclear genomes. The adverse effect of one cytoplasm (MAX1) on TE in cultivated sunflower has been demonstrated, but a more comprehensive evaluation of other cytoplasms within the genus Helianthus may reveal cytoplasms with a beneficial effect on TE. Commercial sunflower hybrids retain the cytoplasm of the female parent. If female parent lines of sunflower can be developed with a cytoplasm that has a beneficial effect on TE, then this benefit will be inherited by any hybrids made with this female. A comprehensive study of TE among the cytoplasmic variants within the genus Helianthus is clearly needed.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Considerable genetic variation for {Delta} was identified in field surveys and trials. Values of {Delta} measured in the glasshouse were highly correlated with TE but were only moderately correlated with field values. Large differences in plant growth were observed with a mixture of germplasm types (inbreds and hybrids) used in glasshouse studies. These results suggest that genotypic ranking for {Delta} based on glasshouse studies should be interpreted with caution. Low levels of genotype x environment interaction and high levels of genetic variation for {Delta} suggest scope for breeders to select sunflower with high levels of TE.

Instantaneous gas exchange measurements revealed a close relationship between {Delta} and g, an association not previously reported in sunflower. Analysis of an alloplasmic set of sunflower lines provided clear evidence of cytoplasmic effects on {Delta} and TE suggesting that breeders may need to consider these effects when developing drought tolerant hybrids. Genotype HAR4 seln 1, identified as having low {Delta} and high TE, was selected as a parent to make segregating populations for future experiments.


   ACKNOWLEDGMENTS
 
This project was funded in part by the Grains Research Development Corporation of Australia. We acknowledge the contribution made by Dr. Naidu Bodapati on the design and execution of this study, and the critical reading and valuable suggestions made by Dr. Tony Condon and Prof. Graham Farquhar. We also thank Dr. C.C. Jan and Lisa Brown for sampling field trials in the USA, Drs. C.C. Jan, Gerald Seiler and Jerry Miller for supplying germplasm and Pacific Seeds Australia for providing a field site and parent lines for use in the glasshouse experiment. We also appreciate the valuable suggestions made by the journal reviewers.

Received for publication October 16, 2003.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 Am. J. Bot.Home page
C. A. Wu and D. R. Campbell
Leaf physiology reflects environmental differences and cytoplasmic background in Ipomopsis (Polemoniaceae) hybrids
Am. J. Botany, November 1, 2007; 94(11): 1804 - 1812.
[Abstract] [Full Text] [PDF]

Home page
 J Exp BotHome page
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]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (9)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lambrides, C. J.
Right arrow Articles by Shorter, R.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Lambrides, C. J.
Right arrow Articles by Shorter, R.
Agricola
Right arrow Articles by Lambrides, C. J.
Right arrow Articles by Shorter, R.
Related Collections
Right arrow Water Stress
Right arrow Crop Physiology & Metabolism
Right arrow Sunflower

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Vadose Zone Journal
Journal of Plant Registrations Soil Science Society of America Journal
Journal of Natural Resources
and Life Sciences Education		 Journal of
Environmental Quality