| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a National Agricultural Research Center for Hokkaido Region, National Agricultural Research Organization, Hitsujigaoka, Sapporo, 062-8555, Japan
b Crop Genetics Research and Development, Pioneer Hi-Bred International, 7300 NW 62nd Avenue, Johnston, IA 50131-1004
c Plant Biotechnology Centre, Primary Industries Research Victoria, Department of Primary Industries, La Trobe University, Bundoora, Victoria 3086, Australia and Cooperative Research Centre for Molecular Plant Breeding, Australia
d Graduate School of Agriculture, Hokkaido University, Kita 9, Nishi 9, Sapporo, 060-8589, Japan
e Pastoral and Veterinary Institute, Primary Industries Research Victoria, Department of Primary Industries, Private Bag 105, Hamilton, Victoria 3300, Australia and Cooperative Research Centre for Molecular Plant Breeding, Australia
f Rhydgoch Genetics, Rhydhir Uchaf, Comins Coch, Aberystwyth, Ceredigion SY23 3BJ, Wales, UK
* Corresponding author (john.forster{at}dpi.vic.gov.au).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The morphogenesis of individual grass plants within a grazed sward plays a key role in determining herbage yield, persistence, and recovery from grazing. In vegetative plants, plant morphogenesis is described by three key variables: leaf appearance rate, leaf elongation rate, and leaf lifespan. The expression of each of these traits is under both genetic and environmental control (Lemaire and Chapman, 1996), and leaf development in Lolium has been demonstrated to be under genetic control in a number of studies (Edwards and Cooper, 1963; Rhodes, 1973; Hazard et al., 1996). Structural characteristics of plants such as tiller number, leaf number, and leaf size are the result of these morphogenetic traits, and their measurement in breeding programs allows a dissection of the complex herbage yield trait as well as predictions of the response to grazing.
Different ecoclimatic regions, or pastures under different grazing regimes, may provide alternative selection goals for these traits. For instance, cool season growth is an important breeding objective in climates with mild winters. Mediterranean genotypes of perennial ryegrass have more rapid rates of leaf appearance and elongation in winter than genotypes from northern Europe and hence better cool season herbage yield (Cooper, 1964). However, this trait may be related to the onset of reproductive development (Kemp et al., 1989) and be negatively correlated with survival during harsh winters. In addition, selection for large leaves has been shown to increase yield under infrequent grazing (Rhodes and Mee, 1980; Hazard and Ghesquière, 1997). However, interactions between grazing management and optimal leaf size have been detected, with short-leaved selections being better adapted to frequent cutting (Hazard and Ghesquière, 1997).
Genetic variation in morphogenetic traits associated with reproductive development is of practical importance in perennial ryegrass breeding not only because of the potential correlation between these traits and vegetative production, but also because of the association of these traits with seed yield (Elgersma, 1990a). Traits such as spike numbers, spikelets per spike, and florets per spikelet have been shown to be heritable (Elgersma, 1990b).
The manipulation of these morphological and developmental traits in a breeding program can be improved by knowledge of the underlying genetic control mechanisms. The detection of QTLs associated with these traits in perennial ryegrass is consequently likely to provide breeders with enhanced possibilities for the development of highly adapted germplasm.
Perennial ryegrass is susceptible to winter stresses caused by both snow molds and low temperatures (Jamalainen, 1974; Jönsson and Nilsson, 1985; Nakayama et al., 2001). In addition to the improvement of growth characteristics, increased winter hardiness of perennial ryegrass is an important breeding objective in northern cold climate regions. Genetic analysis of the components of winter hardiness has been performed in cereals such as wheat (Triticum aestivum L.) and barley (Hordeum vulgare L.) (Cahalan and Law, 1979; Brule-Babel and Fowler, 1988; Doll et al., 1989), allowing the chromosomal location of the genes controlling these characters to be determined. A similar approach may be employed for pasture grasses.
Genetic map-based analysis permits the dissection of complex phenotypes by resolving the locations of interacting and pleiotropic genetic factors. There have been relatively few reports to date of QTL analysis for agronomic traits in perennial ryegrass because of the absence of a sufficiently well developed genetic map. An enhanced molecular marker-based genetic linkage map of perennial ryegrass has recently been constructed through the activities of the International Lolium Genome Initiative (ILGI) (Forster et al., 2001), using the p150/112 one-way pseudo-testcross mapping population. The current map contains 109 restriction fragment length polymorphism (RFLP) loci detected by heterologous probes from wheat, barley, oat (Avena sativa L.), and rice (Oryza sativa L.). Comparative genetic mapping has allowed the alignment of the perennial ryegrass genetic map with those of wheat, rice, and oat, revealing substantial conserved synteny with the genomes of Triticeae species (Jones et al., 2002a). The p150/112 genetic map has been further enhanced by the assignment of nearly 100 polymorphic perennial ryegrass simple sequence repeat (LPSSR) loci (Jones et al., 2001) to locations on each of the seven linkage groups (Jones et al., 2002b). As a consequence, a total of more than 200 codominant genetic markers have been mapped in the p150/112 family, along with about 200 amplified fragment length polymorphism (AFLP) loci, permitting detailed genetic analysis of traits that vary within this population. The SSR loci provide the means to align genetic maps between different mapping families and extrapolate QTL locations between divergent germplasm sources.
Our objective in this study was to use the enhanced genetic map of perennial ryegrass to locate QTLs for a range of agronomic phenotypic traits in the p150/112 population. The emphasis was on morphological and developmental traits associated with pasture productivity, reproductive development traits and winter hardiness characters that may influence survival and subsequent performance in cold climate environments.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Phenotypic Assessment of Morphological and Developmental Characters
Individual plants from the p150/112 mapping family were transplanted at Nagasaka, Japan (35°49' N, 138°22' E) in the field nursery of the Yamanashi Prefectural Dairy Experiment Station (YPDES) in July 1996 with five replicates of each genotype in a randomized complete block design. Morphological characters (plant height, tiller number, tiller size [diameter of the stem of the reproductive tiller], leaf length, leaf width, fresh weight at second harvest, plant type, number of spikelets per spike, and spike length), and developmental characters (heading date and aftermath heading) were measured in the subsequent year. Two cutting regimes were performed in 1997, the first at the heading date for each genotype and the second on 14 July for all genotypes. Plant height (maximum length in cm from the base to the top of the plant) and tiller number were measured in cm on 30 June 1997, while spike length, leaf length and leaf width (in cm) and tiller size (in mm) were measured at the heading date for each genotype. Leaf length and leaf width were measured at the first leaves under the flag leaves of single tillers, and spike length, number of spikelets per spike, and tiller size were measured on the same tillers. Fresh weight at second harvest (14 July 1997) was measured in grams, plant type was measured (on 30 June 1997) on a scale from 1 to 9 with 1 being most erect and 9 being most prostrate, heading date (ear emergence) was measured in days after 1 May 1997 and the degree of aftermath heading (on 14 July 1997) was measured on a scale from 0 to 9 with 0 corresponding to no heads and 9 corresponding to many heads. Measurements were made on one plant per genotype from each of the five replicates for tiller number, fresh weight, plant type, plant height, heading date, and aftermath heading. Measurements were made on five leaves from two replicates (10 leaves in total) for leaf length, leaf width, number of spikelets, spike length, and tiller size at the time of heading.
Phenotypic Assessment of Winter Hardiness Characters
One plant of each p150/112 mapping family genotype was grown outside in pots at the National Agricultural Research Centre for Hokkaido Region (NARCH) located at Sapporo, Japan, (43°00' N, 141°25' E). Electrical conductivity was measured on three leaves from each plant. Leaves were excised in 5-cm-long sections at a 10- to 15-cm height from the base in December 1999, and were shredded, placed into culture dishes, and transferred to a temperature of –2°C. After a 12 h equilibration period, the temperature was lowered manually in 1°C decrements every 1 h to a final value of –6°C and samples were then held for 8 h before being placed in microtubes. Distilled water (1.0 mL) was added to each tube, and samples were held at 5°C for 12 h. The conductivity of the resulting solution was measured using a conductance meter. A comparative value for 100% leakage was obtained by freezing replicate samples from each genotype at –80°C for 4 h.
Individual plants from the p150/112 mapping family were also grown at the field premises of NARCH, with four replications in replicated block design from 1999 onwards. Survival in the field at Sapporo following the winter of 1999–2000 was measured in April 2000 by a visual assessment score (from 1–5) of plant recovery on all four replicates.
Statistical Analysis
Data analyses were performed in SAS (SAS Institute Inc.). The significance of progeny and replicate effects were analyzed by general linear modeling (Proc GLM), broad sense heritabilities were calculated according to Wricke and Weber (1986). The Shapiro-Wilk statistic (W-test) was used to assess the normality of averaged data (normal option in Proc UNIVARIATE). Spearman's rank-order correlation coefficients were determined for pair-wise comparisons of averaged trait data (spearman option in Proc CORR).
QTL Analysis
A framework set of genetic markers from the p150/112-based reference map (Jones et al., 2002a), including the majority of the heterologous RFLP loci, was combined with the perennial ryegrass SSR locus data (Jones et al., 2002b) to produce a composite dataset for QTL analysis of the phenotypic data. Following genetic map construction by MAPMAKER 3.0, a subset of marker loci was selected to provide even coverage of the genome with marker intervals of close to 5 cM, and consensus map distances were subsequently used. Simple linear regression (SMR) was initially employed to identify significant variation with selected genetic markers to provide approximate locations for QTLs. The log-of-odds (LOD) score of association between the genotype and trait data was calculated from simple interval mapping (SIM) in MAPMAKER/QTL (Lander et al., 1987) with the free model of QTL effect. A minimum LOD threshold of 2.0 was selected for significance of location of the QTL for SIM. Composite interval mapping (CIMCIM: Zeng, 1994) was performed by the Windows QTL Cartographer 2.0 application (Basten et al., 1994). Permutation analysis (1000 iterations) was used to establish an experiment-wise significance value at the 0.05 confidence level defined as a minimum LOD threshold for each trait in CIM (Churchill and Doerge, 1994; Doerge and Churchill, 1996). For each form of interval analysis, the maximum LOD value associated with the most closely linked marker, the weight value associated with additive marker allele effects, and the proportion of the phenotypic variance attributable to the QTL were tabulated.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
|
Comparison of Analytical Techniques for Marker-Trait Association
Significant associations between marker and trait data were established for the majority of traits by single marker regression (SMR) on the basis of analysis of variance (ANOVA) at the p < 0.01 significance level. Simple interval mapping (SIM) subsequently identified 17 significant QTLs (maximum LOD 
2) for 11 of the 13 measured traits. Composite interval mapping (CIM) was also performed on the dataset following determination of empirical LOD thresholds, allowing the identification of 20 QTLs for the 11 traits (Table 3). A large proportion of QTL locations were identified by both SIM and CIM. For example, for the heading date trait, genetic markers in the interval from 39.9 to 72.3 cM on linkage group (LG) 4 were significantly associated with the trait data by SMR, while SIM detected a QTL with maximum LOD value of 4.0 close to the xlpssrh01h06 locus at 53.5 cM, and CIM detected a QTL with a maximum LOD value of 3.6 at 56.3 cM with an empirical LOD threshold of 2.9.
|
|
The different results obtained by different forms of analysis are also shown for the trait of electrical conductivity. The terminal marker on LG4 (xr2702Bb) was identified as significantly associated with the trait and SIM detected a QTL with maximum LOD value of 2.0 close to this marker. However, no significant QTLs were detected on this LG by CIM. In contrast, two significant QTLs on LG6 were detected by CIM, with maximum LOD values greater than the threshold value of 2.6. No supporting evidence is available from SMR or SIM analysis for these QTL. The LG6 electrical conductivity QTLs are in close linkage and in repulsion phase, as indicated by weight values of –2.99 and 3.7, respectively, as well as accounting for similar proportions of the phenotypic variance. The similar but opposing effects of these regions may account for the failure of detection by SMR and SIM. For both LGs (4 and 6), but on different criteria, the QTL effects for electrical conductivity are suggestive, but require further validation.
Genetic Control of Morphological and Developmental Traits
Between 1 and 3 QTLs were detected for each of the morphological and developmental characters by either SIM or CIM (Table 3, Fig. 2)
. QTLs for different traits were frequently located in the same chromosomal region, with coincident groups on LGs 1, 3, 4, and 5. The coincident QTLs corresponded to traits that were significantly correlated, and the directions of the QTL effects were in agreement with the sign of the correlations. For instance, coincident QTLs for plant height, tiller size, number of spikelets per spike and spike length were identified by SIM on LG1, and these traits show significant positive phenotypic correlations. The QTL weight values determined by both types of mapping analysis (Table 3) were also all positive.
Coincident QTLs for traits associated with plant size have been identified in three regions, most strikingly on LGs 1 and 3. The positive weight values of the colocating QTLs suggest two possible interpretations. Allelic variation at a single pleiotropic locus could be responsible for the concurrent increase (or decrease) of phenotypic values for the relevant traits. In barley, large QTL effects on plant height are associated with variation at the denso dwarfing gene (Bezant et al., 1996), which maps to 3H, the syntenic counterpart of perennial ryegrass LG3 (Jones et al., 2002a). It is possible that single underlying scale-determining loci may have been detected in this study. Alternatively, a number of cis-linked alleles with similar directions of effect may be present at different loci. In either case, selection for a single set of linked markers would lead to an increase for each trait.
Despite the highly significant correlations between heading date and the morphogenetic traits of plant type, tiller number and plant height, only one QTL (for plant type) colocates with the heading date QTL. This observation supports the previous inference that plant type and tiller number are largely independent of the period of regrowth after cutting, and that the harvesting regime following the heading date did not impair the ability to detect QTLs for these traits. If the QTLs resulting in phenotypic variation for these traits had been a simple reflection of genetic factors segregating for heading date, a large number of coincident QTLs would have been expected.
The only traits in this study that reveal noncoincident QTLs were plant type (LG7) and aftermath heading (LG6). These QTLs offer the potential to select for the corresponding traits without correlated effects on other traits. Plant type is of significance in the breeding of perennial ryegrass for turf quality, which is associated with a more prostrate growth habit. The large plant type QTL on LG7 provides a good target for marker-assisted selection of growth habit. Aftermath heading is usually associated with early flowering perennial ryegrass varieties suitable for hay making, which tend to show reduced perenniality and persistence. For this reason, a high degree of aftermath heading is likely to be disfavored as a breeding objective and the marker allele haplotype associated with reduced aftermath heading could be targeted for selection.
A single QTL for heading date was observed in the current study. However, a number of QTL positions for this trait have been reported from the analysis of single mapping populations in other Poaceae species. Previous mapping studies in perennial ryegrass have produced more complex results (Hayward et al., 1994). The absence of common markers between the population used in this and the present study prevents any inference of common location. The number of QTLs and their relative importance may vary according to the origin of the genotypes used to construct mapping families. Studies on geographical populations of Lolium species covering the climatic range from the Mediterranean region to northern and central Europe revealed a regular cline in flowering responses to temperature and photoperiod (Cooper, 1960). The heterozygous parent of the p150/112 mapping population was derived from a cross between eastern European (Romanian), southern European (north Italian ecotypes) and northern European (‘Melle’ or ‘S23’) genotypes, and might be expected to represent a variety of response genes. It is also likely that small QTLs for heading date have not been detected in this analysis. The single QTL for heading date on LG4 accounts for about 20% of the phenotypic variation based on SIM, but the character shows a high broad sense heritability (0.90), suggesting that a substantial proportion of the heritable variation has not been attributed to specific genomic regions.
Genetic Control of Winter-Hardiness Traits
No significant QTLs were detected for winter survival. It is known that field assessment of winter survival may be confounded by experimental errors (Fowler, 1979). The electrical conductivity method, also known as the ion leakage method, has been used extensively for the evaluation of freezing tolerance through measurement of the release of cellular electrolytes after freezing (Dexter et al., 1930; Dexter et al., 1932). Previous studies of the winter survival of barley, wheat, rye and triticale cultivars in Finland, where frost was considered to be the more important stress factor than snow (Hömmö, 1994), showed good correlations with hardening ability as assessed by the electrical conductivity method. A single QTL for electrical conductivity was detected in the upper part of LG4 by SMR and SIM, but not CIM, accounting for 11.8% of the total phenotypic variation and adjacent to the heading date QTL. Late flowering and reduced freezing tolerance (as indicated by high electrical conductivity) is not significantly correlated in this cross, but the weight values of the linked QTLs were positive, suggesting that cis-linked alleles were causing an increase in phenotypic values of both traits. This is inconsistent with previously identified correlations between higher freezing tolerance and later flowering (Humphreys and Eagles, 1988) and may be due to the complex nature of the cross.
Winter hardiness in wheat (Sutka, 1994; Galiba et al., 1995, 1997) and barley (Pan et al., 1994) is associated with QTLs on the homeologous group 5 chromosomes in the same region as the vernalization response genes that control heading date. In this study, winter hardiness-associated and heading date QTLs were located on LG4 in perennial ryegrass. Perennial ryegrass LG4 has been proposed to correspond predominantly to the homeologous group 4 chromosomes of the Triticeae cereals (Jones et al., 2002a). However, the upper part of LG4 has been shown to contain heterologous RFLP markers that map to the wheat homeologous group 5 chromosomes and its syntenic counterpart, rice chromosome 3. Evolutionary translocations between group 4 and group 5 homeologous chromosomes have been observed for several Triticeae genomes (Devos et al., 1995). A comparative map has been constructed for meadow fescue (Festuca pratensis Huds.) using heterologous anchor RFLPs, many of which are common with the perennial ryegrass study (Alm et al., 2001, 2003). A region syntenic with the portion of the Triticeae 5L chromosomes that corresponds to rice chromosome 3 is present on the upper part of meadow fescue LG4. The Lolium and Festuca genera are closely related, and the high level of recombination observed in triploid F1 hybrids between L. perenne and F. pratensis suggests conservation of gene order (King et al., 1998). An evolutionary translocation between linkage groups corresponding to groups 4 and 5 of the consensus wheat map may have occurred before the divergence of the Poeae grasses. Alm (2001) also performed QTL analysis of frost and drought tolerance in meadow fescue and detected a small QTL for frost tolerance on LG4. This provides further evidence for the presence of genes for winter hardiness on the group 4 chromosomes of the Lolium and Festuca genera as compared with the group 5 homeologous chromosomes of the Triticeae.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The use of several marker-trait association analysis techniques such as SMR, SIM, and CIM provides the means to assess the robustness of QTL detection in a single experimental design. A high degree of congruence between the results of different methods was observed in the present study, with the majority of QTLs detected by the standard IM analysis confirmed by SMR and CIM. However, CIM detected a number of extra QTLs for several traits, and estimates of maximum LOD value with CIM showed both increases and decreases of significance compared with SIM. In addition, the detection of two closely linked QTLs for electrical conductivity with opposing effects on LG6 demonstrates the power of CIM to identify such effects compared to SMR and SIM. These results have provided evidence to support the inference of real genetic effects associated with particular genomic regions, and may be used to priorities the selection of putative marker-QTL combinations for implementation studies.
The identification of QTL locations for selected agronomic traits by molecular-marker based mapping will allow the design of appropriate marker-assisted selection strategies. This will include coselection for desirable characters with coincident QTL locations and counter-selection to break potential unfavorable linkages between negatively correlated traits. The linked marker haplotypes associated with favorable QTL alleles in the current family are available for marker assisted selection. The use of a reference genetic map containing readily transferable SSR and RFLP markers will permit the inference of common QTL locations for the same traits in other experimental pair-crosses, as well as comparative genetic analysis of homologous characters in other Poaceae species.
|
ACKNOWLEDGMENTS |
|---|
Received for publication April 11, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Related articles in Crop Science:
This article has been cited by other articles:
|
|
 |
|
  A. Kosmala, A. Bocian, M. Rapacz, B. Jurczyk, and Z. Zwierzykowski Identification of leaf proteins differentially accumulated during cold acclimation between Festuca pratensis plants with distinct levels of frost tolerance J. Exp. Bot., August 1, 2009; 60(12): 3595 - 3609. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Herrmann, B. Boller, B. Studer, F. Widmer, and R. Kolliker Improving Persistence in Red Clover: Insights from QTL Analysis and Comparative Phenotypic Evaluation Crop Sci., January 16, 2008; 48(1): 269 - 277. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Zhang, M. A. R. Mian, and J. H. Bouton Recent Molecular and Genomic Studies on Stress Tolerance of Forage and Turf Grasses Crop Sci., February 1, 2006; 46(2): 497 - 511. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
