SEARCH

SEARCH BY CITATION

Keywords:

  • drought;
  • fructan;
  • growth;
  • perennial ryegrass (Lolium perenne);
  • quantitative trait loci (QTL);
  • water-soluble carbohydrate (WSC);
  • water stress

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • • 
    The role of fructan in growth and drought-stress responses of perennial ryegrass (Lolium perenne) was investigated in an F2 mapping family that segregates for carbohydrate metabolism.
  • • 
    A quantitative trait locus approach was used to compare the genetic control of traits.
  • • 
    Growth and drought-stress traits were extremely variable within the family. Most traits had high broad-sense heritability. Quantitative trait loci (QTL) were identified for most traits; the maximum number of QTL per trait was four. Between 11% and 75% of total phenotypic variation was explained. Few growth-trait QTL coincided with previously identified fructan QTL. A cluster of drought-trait QTL was close to two previously identified regions of the genome with tiller base fructan QTL in repulsion.
  • • 
    The high sugar parent contributed few alleles that increased ‘reserve-driven’ growth or performance during drought-stress. Correlation of growth and drought-stress traits with fructan content was low and increasing fructan content per se would not appear to improve drought resistance. Complex patterns of carbohydrate partitioning and metabolism within the cell may explain contradictory relationships between carbohydrate content and growth/stress-resistance traits.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Carbohydrates are often reported to have multiple, functional roles in mediating a wide range of plant growth and environmental responses (Calenge et al., 2006). This is as true of fructans, the water-soluble carbohydrate (WSC) polymers found in several plant groups including the Poaceae (Chalmers et al., 2005), as it is of other carbohydrates. The capacity to accumulate fructans in these temperate plants has been postulated to have evolved to meet a need to adapt to cold winters and dry summers (Hendry, 1993).

Regrowth rates following defoliation have often been linked to vegetative carbohydrate reserve levels (White, 1973; Donaghy & Fulkerson, 1997). In perennial ryegrass (Lolium perenne) the fructans in the leaf sheaths constituting tiller base vegetative reserves may be mobilized to fuel reserve-driven growth (Prud’homme et al., 1992). More generally reserve-driven growth rates have sometimes been shown to be higher when WSC reserves are higher (Turner et al., 2001). This may also apply to early spring growth, which can be higher in ryegrass varieties bred for elevated herbage sugar content. However, these effects are sometimes small and short-lived as photosynthate supply is not often limiting. Indeed, some authors have suggested that nitrogen reserves may be more important (Volenec et al., 1996).

Fructans are widely believed, and frequently quoted, to be involved in resistance to environmental stresses such as cold and drought, although direct correlations have not always been shown (Vijn & Smeekens, 1999). They have been cited as osmoprotectants, in common with proline and glycine-betaine, with postulated roles in membrane stabilisation (van den Ende et al., 2005). It is certainly true that fructan content generally increases during such stresses, but accumulation purely as a byproduct of a reduced growth rate is very different from an active functional role in stress resistance. Species, ecotype and variety comparisons have commonly been used to compare cold tolerance and fructan content in temperate grasses and cereals; in many of these comparisons the more cold-tolerant plants have been shown to be those with higher fructan and/or carbohydrate content (Thorsteinsson et al., 2002; Kerepesi et al., 2004). However, in most cases the genetic backgrounds of the materials being compared have been very different in many other respects, reducing the significance of any conclusions. Moreover, Eagles & Williams (1992) concluded from their work on ryegrass that fructan was not important for cold tolerance. Studies of the same types have been carried out with respect to drought stress. Volaire et al. (1998) reported a positive correlation between the proportion of sugar in the form of polymeric fructan and drought resistance while, by contrast, Thomas & James (1999) concluded that WSC did not appear to be a scarce resource even under severe temperate drought in perennial ryegrass, and that lack of sugar was unlikely to limit survival and regrowth in these circumstances.

As forage fructan content has been demonstrated to have an important role in animal nutrition (Miller et al., 2001), selecting for increased WSC has had high priority in many forage grass breeding programmes. Temperate grasslands support most of the world's milk and meat production so the nutritional value of fodder has a major impact on the efficiency and profitability of livestock production. In excess of 80% of agricultural forage seed usage in the UK is ryegrasses (Burgon et al., 1997) and perennial ryegrass is the most important species. Should increases in the fructan content of perennial ryegrass forage also have positive effects on traits such as regrowth (after grazing and mowing), stress resistance and persistence, then the selection of appropriate alleles for forage quality would have added value.

Further work is necessary to clarify the role of fructan in physiological processes in plants. The objectives of the studies reported here were to characterize regions of the perennial ryegrass genome that have basic control over selected growth and drought-stress traits, and, by comparing quantitative trait loci (QTL) positions, to explore potential functional relationships with fructan content. As the aim was to characterize repeatable effects the strategy of combining replication with year-to-year variation in environmental conditions following the principles advocated by Borevitz & Chory (2004), and employed in a previous WSC QTL study (Turner et al., 2006) and by Price et al. (2002), was also used in this work. Perennial ryegrass populations demonstrate considerable variation for WSC content and fructan QTL have previously been identified in an F2 mapping family (Turner et al., 2006). A range of growth and drought traits have now been measured on the same family. Evidence of co-location for QTL positions for carbohydrate content and physiological traits would be consistent with a close integration of the genetic regulation of the traits and the importance of fructan in these physiological processes, although it should not be forgotten that co-location could simply reflect tight linkage of genes controlling very different traits.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Plant material

The mapping family (WSC F2) has been described previously (Armstead et al., 2002; Turner et al., 2006). A single F1 hybrid plant, generated from crossing individual genotypes of partially inbred-lines of the perennial ryegrass (Lolium perenne L.) varieties Perma and Aurora, was self-pollinated to give a seed yield of something over 200 seeds. Germination was good and initially 200 plants were raised. After the first 3 yr 192 plants remained following the loss of a few weak plants. The F2 population has subsequently stabilized at 188 plants. These plants show good vigour and have been maintained successfully by regular vegetative propagation of small subsets of tillers for over 10 yr. The parents of the F1 plant used were very small and weak, and hence unsuitable for phenotypic analysis. This was a result of inbreeding depression and not surprising with an outbreeding species such as Lolium. Data on some growth traits in the noninbred parent source varieties have previously been published (Turner et al., 2001); these provide an indication of parental responses. Therefore, samples for trait data were taken only from F2 plants.

The mapping family unit in a perennial outbreeding species is the vegetatively-propagated genotype as seed descent is not possible. It is therefore necessary to distinguish between ‘genotype’ and ‘plant’. Replication is achieved by raising separate re-cloned plants of each genotype; replicates are physically separate plants but will always be the same genotype. Replicate measurements for the traits were performed on separate/renewed plants of each genotype and equivalent plants were used for the replicates within each trait. Plants were never more than 1 yr old. The separate sets of trait measurements are therefore not repeated measures and can be treated as replicates.

Trait analysis

Trait analysis was replicated over time within the constraints of experimental resources to avoid interference from confounding environmental and/or physiological variables on any one sampling date, and to identify repeatable effects (Turner et al., 2006). Wherever possible this was replication over years. However, the drought experiment was restricted to 1 yr. As the period when evaporative demand is sufficient to develop a water deficit coincides with flowering the analysis was replicated at approx. 3-wk intervals over the summer to remove interference arising from reproductive responses. All the traits in this study were measured under the conditions (time and environment) most appropriate for the trait concerned and replication details for all traits are summarized in Table 1.

Table 1.  Sampling schedules for traits (one measurement for each of the 188 genotypes was taken at each sampling time)
TraitNumber of sampling timesSampling times
Plant size4From early to late summer in 1 yr
Root amount3Autumn in 1 yr
Leaf extension rate4Early spring and early autumn for 2 yr
Leaf extension time4Early spring and early autumn for 2 yr
Final leaf length4Early spring and early autumn for 2 yr
Regrowth extension rate4Early spring and early autumn for 2 yr
Spring dry matter accumulation3Early spring for 3 yr
Autumn dry matter accumulation3Mid autumn for 3 yr
Wiltiness4End of drought from early to late summer in 1 yr
Water content following rehydration (well-watered)2End of drought from early to late summer in 1 yr
Water content following rehydration (droughted)4End of drought from early to late summer in 1 yr
Osmotic potential following rehydration (well-watered)2End of drought from early to late summer in 1 yr
Osmotic potential following rehydration (droughted)4End of drought from early to late summer in 1 yr
Herbage survival4End of drought from early to late summer in 1 yr
Tiller survival4Few days after end of drought from early to late summer in 1 yr
Regrowth after rewatering4Few days after end of drought from early to late summer in 1 yr

Growth traits were analysed on plants grown in 15-cm diameter pots of Humax John Innes No3 with wetting agent in a frost-free, unlit glasshouse throughout the year. For the drought treatment, single tillers were planted at 8-cm spacing in spring in large, 90-cm deep, brick-built soil bins in an identical glasshouse. The bins contained 5 cm gravel for drainage, 70 cm sterilized silty loam topsoil and 10 cm Humax John Innes No. 3 with wetting agent. The drought-stress was initiated by withholding water from the middle of May onwards, as described by Thomas & James (1999).

Plant size was measured as fresh mass of herbage after 2 months of unrestricted growth from a single tiller. Root amount was scored on a scale of 1 to 5 measured following careful removal of the plant from its pot (1, no roots visible; 5, pot is filled with a solid root ball). Constitutive leaf growth was measured on plants at least 4 wk after they had last been cut back. Leaf length from the leaf tip to the point of emergence (or the ligule) of marked leaves was recorded every other day from emergence until the same measurement was obtained on three consecutive occasions. The last measurement was used as final leaf length. Leaf extension rate was calculated as the slope of a regression line fitted through the central portion of the leaf extension curve, and leaf extension time as leaf length/leaf extension rate. Regrowth rate was calculated as the slope of a regression line fitted through plant height measurements taken daily on day 0 and for 4 d after cutting back to c. 5 cm.

Early and late season growth responses were assessed by dry matter yield. Plants were cut back to a stubble height of 5 cm in late January or late September, respectively, and the herbage removed dried at 80°C. Dry matter was recorded to give a measure of initial plant size. Plants were allowed to grow on in the glasshouse for 8 wk or 5 wk, respectively, and dry matter yield recorded again in the same manner. Data for the regrowth period were analysed both as uncorrected values and as values corrected for initial plant size.

Wiltiness was scored on a scale of 0–6 (0, leaves fully turgid; 6, completely crispy and brown). Leaf water content (hydration) and osmotic potential were measured on leaves rehydrated for 4 h in cold water (Thomas, 1986). Herbage fresh-matter surviving at the end of the drought period was recorded, and considered both uncorrected and corrected for plant size at the beginning of drought. Tiller survival was the proportion of tillers on each plant regrowing following watering at the end of drought. Regrowth after drought was the length of new leaf growth 4–5 d after rewatering.

Data for fructan content were taken from the previously published study (Turner et al., 2006).

Linkage map

The linkage map described previously for this WSC mapping family (Turner et al., 2006) was used in this study.

Statistical analysis

Trait data were characterized and analysed on all genotypes with the number of replicates as shown on Table 2. Correlation analyses were calculated as the product moment correlation coefficient for pairwise combinations. One-way analysis of variance treated year as a random effect and genotype as a fixed effect. Broad sense heritability (H) was calculated from the anova analyses with the formula:

Table 2.  Distributions of traits within the mapping family
TraitNumber of replicatesOriginal dataData used in QTL analysis
Minimum valueMaximum valueMeanSDHeritabilityTreatment of dataMinimum valueMaximum valueMeanSDSkewKurtosis
  1. Variation was characterized before and after normalization and transformation when these have been applied prior to QTL analysis. All data are from mean genotype values for each trait (n= 188).

Plant size (g FW)40.1321.984.063.4620.784normalized and square root0.1792.1940.8890.3990.500–0.272
Root amount31.35.03.70.7330.646none1.35.03.70.733–0.6030.096
Leaf extension rate (mm d-1)43.710.56.301.2730.562normalized0.6061.5811.0000.1940.4840.361
Leaf extension time (d)411.526.917.32.5900.574normalized0.6601.4791.0000.1500.5800.355
Final leaf length (mm)463.8147.5101.915.1760.575normalized0.6201.4471.0000.1460.2040.318
Regrowth extension rate (mm d-1)43.710.76.41.2720.431normalized0.5661.6801.0000.2090.5650.185
Spring dry matter accumulation (g DW)30.103.101.190.5950.710normalized0.0652.8610.9830.5100.5200.529
Autumn dry matter accumulation (g DW)30.162.301.140.3950.620normalized0.1752.0880.9960.3520.2640.127
Spring dry matter accumulation (corrected)30.4724.4851.6720.5190.226normalized0.3442.1760.9920.2670.3211.228
Autumn dry matter accumulation (corrected)30.1060.6690.2030.0580.338normalized and log10–0.6410.394–0.0350.114–0.8955.797
Wiltiness40.56.03.31.2430.775none0.56.03.31.2430.266–0.564
Water content following rehydration (well-watered) (g g-1 DW)23.577.665.250.6980.491none3.577.665.250.6980.3450.002
Water content following rehydration (droughted) (g g-1 DW)42.286.554.440.7120.688normalized0.7071.4211.0000.1260.4260.310
Osmotic potential following rehydration (well-watered) (MPa)20.951.431.180.1030.200none0.951.431.180.1030.030–0.396
Osmotic potential following rehydration (droughted) (MPa)41.121.661.380.1010.180none1.121.661.380.1010.223–0.082
Herbage survival (g FW)40.0253.091.934.6240.895normalized and log10–2.1071.399–0.6610.6830.211–0.365
Herbage survival (corrected)40.0222.4150.3470.3480.508normalized and log10–1.2830.802–0.3530.3890.1510.166
Tiller survival40.001.000.430.3250.747normalized0.0002.4171.0050.7840.211–1.248
Regrowth after rewatering (mm extension growth)40.068.817.214.8340.787normalized and square root0.0001.9720.7390.5270.181–1.067
  • H = Vg/(Vg + Ve/n) (Eqn 1)

(Vg is the genetic variance; Ve is the error variance; and n is the number of replications for each genotype). All statistical analyses were performed according to standard procedures with the menu-driven options within genstat for Windows, version 8.1 (Payne et al., 2005).

QTL analysis

The QTL analysis was carried out with mapqtl version 4.0 (van Ooijen et al., 2002) following the criteria described in Turner et al. (2006). Briefly population structure was set as F2. Kruskal–Wallis nonparametric single locus analysis, interval mapping and composite mapping (MQM mapping in mapqtl 4.0 using cofactors chosen with the help of the automatic cofactor selection option within the software with a threshold of 0.02) were all carried out on the individual replicates and the mean for each trait. A quantitative trait locus was declared when the LOD score was over 3.0 for the mean, the quantitative trait locus was detected in at least two individual replicates with a LOD score of over 2.0 and the locus position was supported by Kruskal–Wallis nonparametric single locus analysis. The comparison of Kruskal–Wallis output with interval mapping included consideration of the pattern of the test statistic (K*) when markers were arranged in map order. Data presented are for MQM mapping of the mean trait values for each genotype. Additive and dominance effects were calculated by the software in the standard manner.

Data were transformed when necessary to reduce any skewed distributions to ‘near’ normal as traits with skewed distributions can give anomalous output with some interval mapping software including mapqtl. Low values for trait distribution skew (below 0.6–0.8) were previously shown not to interfere with mapqtl software (Turner et al., 2006). Few of the traits measured in the current study were suitable for QTL analysis without some preliminary data processing. In most cases the only requirement was to harmonize the means of different replicates (by normalizing to a mean of 1.00 within each replicate) before calculating the phenotype mean for each genotype. This was necessary particularly in respect of growth traits with replicates that were measured at different times of year. All traits containing any element of growth were normalized unless they were recorded as a score rather than as a measurement value. The only other data that were normalized were for the water content (hydration) of rehydrated droughted tissue as there was a twofold difference between the means of the replicates in this data set. The parallel data for well-watered plants were not normalized as the replicates did not show such variation and tests showed that normalization did not change the outcomes (Table 2). Further transformation was necessary for traits that had highly nonnormal distributions. Either Log10 or square-root transformations were used depending on which caused the greatest reduction in the skew (Table 2). All distributions were converted to skew values low enough for single locus analysis and interval mapping to give comparable outcomes.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Trait characterization

The WSC F2 perennial ryegrass mapping family was designed to segregate for WSC content, but has proved to segregate markedly for a number of other traits. This has been illustrated particularly well in the current study, which showed considerable phenotypic variation for growth and drought traits (Table 2). The difference between the minimum and maximum values recorded for the traits varied from twofold for well-watered hydration (water content) to over 2500-fold for uncorrected herbage survival at the end of a drought period. In addition the broad-sense heritability of most traits was moderate to high; the only exceptions being spring and autumn dry matter production and osmotic potential. The original data for a number of traits showed skewed distributions with the mean well below the median of the phenotype range (Table 2). The transformations used reduced the skew to below values of c. 0.600 for all traits except autumn dry matter accumulation.

QTL analysis

The linkage map, with those markers that had a distorted segregation ratio (P < 0.05) indicated, shown on Fig. 1 was used for QTL analysis. Quantitative trait loci were found for most traits (Table 3). The exceptions included osmotic potential, which had a low heritability indicating a strong environmental influence on phenotype. No QTL were found for autumn and spring growth when the data were uncorrected for initial plant size, although these traits appeared to have high heritability. Applying a plant size correction to the data resulted in the identification of one quantitative trait locus for autumn growth, even though the heritability was much lower, but it remained impossible to identify QTL for spring growth that achieved the QTL-declaration criteria used in this study. This may have been because variable weather patterns from year to year made it hard to collect samples in the same physiological time frame. Quantitative trait loci were present in each of the three replicates but were not repeatable from year to year (year 1, QTL with LOD > 2 on chromosomes 4, 6 and 7; year 2, QTL with LOD > 2 on chromosome 4; year 3, QTL with LOD > 2 on chromosome 1; mean of 3 yr, QTL with LOD > 3 on chromosome 5). Final leaf length was another trait for which no QTL were declared, although it had a high heritability. Kruskal–Wallis analysis identified a large number of markers on several chromosomes with moderate to high significance, but no concentrated areas of high significance that might indicate a quantitative trait locus undetected by interval mapping. This suggests that the trait was controlled by a large number of loci spread across the genome, none of which had a large enough effect to be identified as a quantitative trait locus by interval mapping.

image

Figure 1. Locations of quantitative trait loci (QTL) for growth and drought traits in relation to polymeric fructan QTL on the ryegrass (Lolium perenne) genetic linkage map. Chromosome designations conform to the Triticeae numbering. *, Markers with a distorted segregation ratio (P < 0.05). Fructan data are from Turner et al. (2006). Bars represent the 2-LOD interval from composite QTL mapping (MQM). Inheritance is indicated; positive additive effects from the Aurora allele (high-sugar parent; closed bars), and from the Perma allele (open bars). Fructan QTL labels are in normal type (PF, polymeric fructan). Growth trait QTL labels are in italic type (PS, plant size; R, roots; LER, leaf extension rate; LET, leaf extension time; RER, regrowth extension rate; AGC, autumn growth (corrected)). Drought trait QTL labels are underlined (W, wiltiness; HW, hydration (water content: well-watered); HD, hydration (water content: droughted); HS, herbage survival; HSC, herbage survival (corrected); TS, tiller survival; REW, regrowth after rewatering).

Download figure to PowerPoint

Table 3.  Composite quantitative trait loci (QTL) mapping (MQM) of QTL that are detected by interval mapping of mean trait data and that are supported by interval mapping of individual replicates and single locus analysis
TraitChromosome2-LODDistanceFlanking markersLODEffects% Variation explained
Interval (cM)cM at peakAdditiveDominance
  1. The output values for QTL effects are based on transformed data where these have been analysed. Positive values of additive effects indicate that the allele from Aurora, the high-sugar parent, confers the positive effect.

Plant size116–1716.602ga1E38M47096.8−0.227−0.11317.6
46–1816.7E42M3311rv04545.90.208−0.00313.5
50–43.1GSY60.2rv081411.6−0.4740.16628.5
515–2017.0rv0495B61037.40.3120.03215.7
Roots10–31.8rv0913rv13915.6−0.3460.05011.0
261–6865.2E40M5908PSR5407.0−0.3200.43614.4
442–4645.3PSR922E38M47069.50.2870.53120.9
Leaf extension rate228–3130.1BCD855Fp12.25.9−0.070−0.09413.6
367–7372.114ga1rv10465.5−0.093−0.000211.3
Leaf extension time361–6664.2CDO34514ga15.50.0750.01211.2
Regrowth extension rate257–6260.4E39M4906E40M59085.3−0.087−0.0118.9
623–2825.2rv1423E41M57125.6−0.097−0.12715.8
Autumn dry matter accumulation (corrected)322–2725.3rv1133rv11445.7−0.056−0.00411.1
Wiltiness15–137.9rv1391rv06598.40.7940.10419.6
264–7674.8M15-185E41M57054.50.608−0.16310.8
Hydration (well-watered)512–1817.0PSR574rv00826.1−0.318−0.20414.8
Hydration (droughted)68–2312.2CDO542E39M490816.0−0.102−0.07931.1
Herbage survival12–137.9PGIrv065913.5−0.564−0.14833.9
50–43.1GSY60.2rv081411.1−0.7810.09125.0
512–1817.0PSR574rv00826.70.4080.22313.4
Herbage survival (corrected)11–83.4rv0913CDO5809.3−0.241−0.06019.9
Tiller survival12–137.9PGIrv065914.1−0.604−0.17530.2
Regrowth after rewatering12–137.9PGIrv065915.8−0.431−0.15234.8
54–54.4rv0814F29-19.8−0.504−0.02021.1
515–2017.0rv0495B61035.80.2630.19412.1

Quantitative trait loci for growth traits were identified on all chromosomes except for chromosome 7 (Table 3, Fig. 1). Many explained a good part of the phenotypic variation in the trait (up to 29%), especially when more than one QTL was identified for a trait (up to 75% phenotypic variation when the QTL for a trait are summed). There were few examples of trait clusters. Regrowth rate was closely associated with root amount on chromosome 2. Leaf extension rate and leaf appearance time may have been associated on chromosome 3, with high extension rates and short extension times (equals high leaf appearance rate) in the same linkage phase. There were two traits – plant size and root amount – for which alleles giving a positive effect were conferred by both Aurora and Perma at different QTL positions. The QTL where the Perma alleles conferred the positive effect comprised the greater part of the percentage total variation explained by all the QTL for the trait. Indeed, in most cases the allele for increased growth came from Perma, the low sugar parent. Few of the growth QTL coincided with previously identified fructan QTL (Fig. 1).

The QTL for drought traits were concentrated on chromosomes 1 and 5 (Table 3, Fig. 1). In particular there was a large cluster of QTL at the top of chromosome 1, between previously identified QTL in opposite linkage phases for tiller base fructan (Fig. 1). The herbage survival QTL on the top of chromosome 5 were uncorrected for plant size and both aligned with QTL for constitutive plant size. These QTL were not found when the herbage survival data were corrected for plant size. This demonstrates that, in this mapping family, bigger plants were more drought-resistant. It is possible that no further drought responses were involved at these loci. Root amount may also have had some importance as QTL on chromosomes 1 and 2 aligned with drought traits.

On chromosome 6 the large QTL for leaf polymeric fructan content that had been detected previously was associated with a quantitative trait locus for water content after rehydration of leaves from droughted plants. Alleles for high fructan and low water content both came from Aurora. Low water content in hydrated leaf tissue would be a consequence of a greater concentration of solute present per unit dry matter. This is consistent with accumulation of carbohydrate during stress, perhaps as a result of reduced growth. There is no evidence from this QTL mapping study that the presence of this solute affected wilting, and therefore no evidence that it had an osmotic effect, or was used as a means to maintain osmotic potential and therefore turgor.

Trait relationships

Fructan content was not measured in the plants at the same time as the current growth and drought measurements. However, mean fructan data previously obtained over 3 yr analysis are indicative of average fructan content for each genotype and can be used compare relative variation in phenotype values within the mapping family for the traits. Table 4 shows correlation coefficients for some of the growth and drought traits (comprising particularly those traits where relationships with carbohydrate reserves have been postulated) with previously measured polymeric fructan content. None of these traits showed high numerical correlations. Indeed, leaf extension, regrowth after cutting and early spring growth had no relationship with tiller base fructan, which represents the plant's vegetative reserves. Nevertheless, some correlations were significant, particularly for the drought traits. Herbage survival, tiller survival, regrowth on rewatering and wilting were all correlated with spring tiller base fructan reserves although these fructan values had been measured on well-watered plants. However, these correlations all indicated that the high-sugar plants showed the worst performance during drought, being the most wilted and having the lowest herbage and tiller survival, and regrowth on rewatering. Inherent fructan reserves from the tiller base do not appear to have been used to maintain turgor during drought. Although, by contrast, there were positive correlations between leaf fructan content and osmotic potential, the relationship between leaf fructan content and wilting was inconsistent. These trait relationships add weight to the conclusion from the QTL analysis that the high fructan phenotype sourced from Aurora does not confer advantages during drought stress or when growth must be driven primarily from reserves.

Table 4.  Correlations between tissue polymeric fructan content and selected growth and drought traits
TraitTissue polymeric fructan content
LeafLeafTiller baseTiller base
SpringAutumnSpringAutumn
  1. Fructan data are from Turner et al. (2006). n = 188 for all traits. Threshold significance value: 0.144 for P < 0.05; 0.188 for P < 0.01; 0.239 for P < 0.001.

Leaf extension rate (mm d−1)−0.103−0.0790.006−0.091
Regrowth extension rate (mm d−1)0.040−0.230**0.029−0.049
Spring dry matter accumulation (corrected)0.0470.103−0.068−0.055
Autumn dry matter accumulation (corrected)−0.219**0.019−0.015−0.011
Wiltiness0.212**−0.202**0.262***0.007
Osmotic potential following rehydration (droughted) (MPa)0.205**0.256***0.093−0.001
Herbage survival (corrected)−0.0890.159*−0.171*0.021
Tiller survival−0.194**0.120−0.253***−0.017
Regrowth after rewatering (mm extension growth)−0.213**0.171*−0.245***−0.008

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Fructan and growth

Perennial ryegrass is an outbreeding plant and contains enormous genetic variation for a wide range of traits. Variation for leaf appearance, leaf elongation rate, leaf size, tiller number and root amount has been shown previously within the species (Hazard et al., 2001; Crush et al., 2007). Comparable variation was observed here within the WSC F2 mapping family among a range of growth traits. Identification of QTL indicated that genetic control of growth traits was spread widely across the genome. Some QTL were in common with published studies on other ryegrass mapping families, but others differed (Yamada et al., 2004). The only chromosome on which no growth QTL were detected was chromosome 7, where a major heading date QTL was located (Armstead et al., 2004). This suggests that it is unlikely that any of the QTL reported here were influenced by time-to-flowering differences in the mapping family, even when phenotyping measurements were carried out during the course of the summer.

There was little evidence for major genetic regulation of growth from the regions of the genome that were shown previously to control fructan content, even in the case of regrowth after defoliation which in the past had been linked with fructan content. This suggests that regrowth rate can be more an expression of inherent plant growth capacity than of carbohydrate storage, and supports the findings of those studies that concluded fructan was not always a major determinant in regrowth (Morvand-Bertrand et al., 1999). There were only two instances where growth trait QTL co-located with fructan QTL. Positive alleles for fructan content and plant size at these QTL were inherited from opposite parents; large plants were associated with low storage reserves and vice versa. Generally, the positive allele for growth QTL was derived from Perma, the low-sugar parent in the mapping family. This is consistent with carbon being used either for growth or to build up a fructan reserve pool. However, for the QTL at the top of chromosome 1, the alleles for both high fructan content and good rooting were inherited from Perma.

Fructan and stress resistance

The WSC F2 mapping family also showed considerable variation for a range of drought resistance traits. In common with the growth QTL the positive allele was not always derived from the same parent, but within a QTL cluster the positive effect was always provided from only one parent. The QTL clustering suggests that genes for drought resistance were fewer and their distribution more confined. The two major clusters in this study were on chromosomes 1 and 5. In a recent study with fescue Alm et al. (in press) also found clusters of QTL for drought resistance on fescue 1 and 5, in addition to effects on fescue 3 and 4. The alignment on chromosome 5 is difficult to assess because of the lack of common markers, but is very close on chromosome 1 based on the common markers cdo580 and phosphoglucoisomerase (PGI/2). The QTL for drought resistance on chromosome 3 of fescue have been exploited, at least in part, in crop improvement programmes of ryegrass through their independent introgression into homoeologous regions of the ryegrass genome (Humphreys et al., 2006).

There are more uncertainties in interpreting relationships between QTL positions for fructan content and for drought resistance than for fructan and growth traits. The fructan QTL were determined with well-watered plants and reflected the inherent capacity to store fructan rather than the fructan content that might have accumulated as growth slowed following the onset of drought stress. No inherent fructan QTL were identified near the two clusters of drought resistance QTC on chromosome 5. However, inherent plant size QTL did co-locate with these clusters, where herbage survival QTL were found only with data uncorrected for initial plant size. At these loci, survival of drought and subsequent regrowth following irrigation was a reflection primarily of plant size and vigour. Accordingly, it would seem unlikely that if any QTL for fructan accumulation during drought did locate here, that the positive alleles for fructan accumulation and herbage survival would be inherited together. The largest cluster of drought resistance QTL at the top of chromosome 1 was identified for both uncorrected and corrected herbage survival, and this may be a major site for genes controlling drought resistance. This cluster was also in close proximity to tiller base fructan QTL, but the relationships here appear quite complex. Fructan QTL were located just below and just above the cluster of drought trait QTL. Although Perma was the low-sugar parent in the mapping family it conferred the allele increasing fructan content at the upper QTL position and the allele for improved drought resistance. At the lower QTL, the allele for high fructan content came from Aurora (the high-sugar parent) along with the allele for poor drought resistance. In other words, allele(s) for high fructan from Perma at one QTL region appeared to favour improved drought resistance, but allele(s) for high fructan from Aurora at the QTL a few centimorgans away appeared to be associated with increased drought susceptibility. One explanation for this might be that these loci regulate fructan accumulation in different pools and/or cell compartments, and that these pools have different functions in the cell. There are questions concerning the compartmentation of WSC metabolism that remain to be resolved. Sucrose accumulation and metabolism have recently been shown to be more complex than current models would predict (Cairns & Gallagher, 2004). In recent years the most widely accepted model places fructan accumulation and metabolism in the vacuole, but other models have been proposed (Kaeser, 1983). New work on the compartmentation of fructan metabolism is continuing but is already producing some interesting results (Gallagher et al., 2007). However, it would certainly seem that increasing fructan content per se may not improve drought resistance, particularly if the high-fructan trait is derived from Aurora, and perhaps, other related ryegrass varieties with increased herbage sugar content. Complex patterns of sugar compartmentation might also explain the sometimes contradictory results produced by studies seeking to correlate carbohydrate content with stress resistance.

In the current study few QTL were identified for water status related traits, and in only one instance was there any relationship with drought survival traits. The allele for reduced leaf wilting at the major QTL cluster at the top of chromosome came from Perma along with the good drought survival alleles. Alleles from Aurora conferred high fructan (in well-watered plants) and low hydration in droughted plants at QTL on chromosome 6, consistent with a high leaf fructan content also occurring during drought. However there was no evidence that this locus had any further effects on leaf turgor or plant drought survival. In barley, QTL have been found for relative water content (RWC) on five chromosomes at different times and in different studies, namely (Triticeae) 1, 2, 4, 6 and 7 (Teulat et al., 1997, 1998, 2001, 2003). Overlap or close relationships with WSC content measured during the drought stress have been shown only on chromosome 2 (Teulat et al., 2001) and chromosome 7 (Diab et al., 2004). However, QTL for both drought and WSC content have been found individually in other locations, with no good correspondence between them (Teulat et al., 2001; Diab et al., 2004). There do not appear to be any published incidences of co-locating QTL for WSC and osmotic potential in barley or any clear QTL relationships of water-relations traits with drought survival and yield.

This study has not considered cold tolerance, as a number of reports that use QTL analysis to compare fructan and/or WSC accumulation with cold tolerance in ryegrass have already been published (Yamada et al., 2004, 2005). Overall, there would appear to be limited support for a functional role for fructan in cold tolerance and winter survival, perhaps mediated by membrane stabilization (Demel et al., 1998; Hincha et al., 2000, 2002) and possibly restricted to explaining only a small part of the phenotypic variation.

Conclusions

Only a relatively small portion of the phenotypic variation for the traits involved in regrowth responses or drought resistance was explained by co-locating QTL for fructan accumulation. The negative relationship between fructan expression in the high-sugar grass Aurora and drought resistance may in fact imply that deleterious (or at best neutral) consequences on stress resistance follow from the presence and expression of at least some fructan regulating alleles on chromosome 1. Increasing fructan content per se can certainly not be predicted to improve drought resistance. Indeed, complex patterns of carbohydrate partitioning and metabolism within the cell may explain the sometimes contradictory relationships between carbohydrate and growth or stress resistance traits that have been obtained by many studies in the past.

However, other loci on the same chromosome provide an alternative source of genes for drought resistance and an opportunity to combine these with genes for high fructan content, thereby providing a route to combining both desirable agronomic traits into a new sustainable grass variety. Marker selection would be an essential tool for such breeding strategies to break the linkage of the close QTL regions on chromosome 1.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank Markku Farrell and Aneurin James for assistance with the experimental work, and BBSRC and DEFRA for funding.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Alm V, Busso CS, Ergon Å, Rudi H, Larsen A, Humphreys MW, Rognli OA. (in press). A QTL analysis and comparative genetic mapping of frost tolerance, winter survival and drought tolerance in meadow fescue (Festuca pratensis Huds.). Genetics.
  • Armstead IP, Turner LB, Farrell M, Skøt L, Gomez P, Montoya T, Donnison IS, King IP, Humphreys MO. 2004. Synteny between a major heading-date QTL in perennial ryegrass (Lolium perenne L.) and the Hd-3 heading date locus in rice. Theoretical and Applied Genetics 108: 822828.
  • Armstead IP, Turner LB, King IP, Cairns AJ, Humphreys MO. 2002. Comparison and integration of genetic maps generated from F2 and BC1-type mapping populations in perennial ryegrass. Plant Breeding 121: 501507.
  • Borevitz JO, Chory J. 2004. Genomics tools for QTL analysis and gene discovery. Current Opinion in Plant Biology 7: 132136.
  • Burgon A, Bondesen OB, Verburgt WH, Hall AG, Bark NS, Robinson M, Timm G. 1997. The forage seed trade. In: FaireyDT, HamptonJG, eds. Forage seed production. 1. Temperate species. Wallingford, UK: CAB International, 271286.
  • Cairns AJ, Gallagher JA. 2004. Absence of turnover and futile cycling of sucrose in leaves of Lolium temulentum L. Implications for metabolic compartmentation. Planta 219: 836846.
  • Calenge F, Saliba-Columbani V, Mahieu S, Loudet O, Daniel-Vedele F, Krapp A. 2006. Natural variation for carbohydrate content in Arabidopsis. Interaction with complex traits dissected by quantitative genetics. Plant Physiology 141: 16301643.
  • Chalmers J, Lidgett A, Cummings N, Cao Y, Forster J, Spangenberg G. 2005. Molecular genetics of fructan metabolism in perennial ryegrass. Plant Biotechnology Journal 3: 459474.
  • Crush JR, Easton HS, Waller JE, Hume DE, Faville MJ. 2007. Genotypic variation in patterns of root distribution, nitrate interception and response to moisture stress of a perennial ryegrass (Lolium perenne L.) mapping population. Grass and Forage Science 62: 265273.
  • Demel RA, Dorrepaal E, Ebskamp MJM, Smeekens JCM, De Kruijff B. 1998. Fructans interact strongly with model membranes. Biochimica et Biophysica Acta 1375: 3642.
  • Diab AA, Teulat-Merah B, This D, Ozturk NZ, Benscher D, Sorrells ME. 2004. Identification of drought-inducible genes and differentially expressed sequence tags in barley. Theoretical and Applied Genetics 109: 14171425.
  • Donaghy DJ, Fulkerson WJ. 1997. The importance of water-soluble carbohydrate reserves on regrowth and root growth of Lolium perenne (L.). Grass and Forage Science 52: 401407.
  • Eagles CF, Williams J. 1992. Hardening and dehardening of Lolium perenne in response to fluctuating temperatures. Annals of Botany 70: 333338.
  • Van Den Ende W, Yoshida M, Clerens S, Vergauwen R, Kawakami A. 2005. Cloning, characterization and functional analysis of novel 6-kestose exohydrolases (6-KEHs) from wheat (Triticum aestivum). New Phytologist 166: 917932.
  • Gallagher JA, Cairns AJ, Turner LB. 2007. Fructan in temperate forage grasses; agronomy, physiology and molecular biology. In: ShiomiN, BenkebliaN, OnoderaS, eds. Recent advances in fructooligosaccharides research. Kerala, India: Research Signpost, 1546.
  • Hazard L, Barker DJ, Easton HS. 2001. Morphogenetic adaptation to defoliation and soil fertility in perennial ryegrass (Lolium perenne). New Zealand Journal of Agricultural Research 44: 112.
  • Hendry GAF. 1993. Evolutionary origins and natural functions of fructans – a climatological, biogeographic and mechanistic appraisal. New Phytologist 123: 314.
  • Hincha DK, Hellwegge EM, Heyer AG, Crowe JH. 2000. Plant fructans stabilize phosphatidylcholine liposomes during freeze-drying. European Journal of Biochemistry 267: 535540.
  • Hincha DK, Zuther E, Hellwegge EM, Heyer AG. 2002. Specific effects of fructo- and gluco-oligosaccharides in the preservation of liposomes during drying. Glycobiology 12: 103110.
  • Humphreys MW, Yadav RS, Cairns AJ, Turner LB, Humphreys J, Skøt L. 2006. A changing climate for grassland research. New Phytologist 169: 926.
  • Kaeser W. 1983. Ultrastructure of storage cells in Jerusalem artichoke tubers (Helianthus tuberosus L.). Vesicle formation during inulin synthesis. Zeitschrift fur Pflanzenphysiologie 111: 253260.
  • Kerepesi I, Banyai-Stefanovits E, Galiba G. 2004. Cold acclimation and abscisic acid induced alterations in carbohydrate content in calli of wheat genotypes differing in frost tolerance. Journal of Plant Physiology 161: 131133.
  • Miller LA, Moorby JM, Davies DR, Humphreys MO, Scollan ND, Macrae JC, Theodorou MK. 2001. Increased concentration of water-soluble carbohydrate in perennial ryegrass (Lolium perenne L.). Milk production from late-lactation dairy cows. Grass and Forage Science 56: 383394.
  • Morvand-Bertrand A, Boucard J, Prud’homme M-P. 1999. Influence of initial levels of carbohydrates, fructans, nitrogen, and soluble proteins on regrowth of Lolium perenne L. cv. Bravo following defoliation. Journal of Experimental Botany 50: 18171826.
  • Van Ooijen JW, Boer MP, Jansen RC, Maliepaard C. 2002. MapQTL 4.0, Software for the calculation of QTL positions on genetic maps. Wageningen, the Netherlands: Plant Research International.
  • Payne RW, Murray DM, Harding SA, Baird DB, Soutar DM. 2005. GenStat for Windows, 8th edn. Oxford, UK: VSN International.
  • Price AH, Steele KA, Moore BJ, Jones RGW. 2002. Upland rice grown in soil-filled chambers and exposed to contrasting water-deficit regimes. II. Mapping quantitative trait loci for root morphology and distribution. Field Crops Research 76: 2543.
  • Prud’homme M-P, Gonzales B, Billard J-P, Boucaud J. 1992. Carbohydrate content, fructan and sucrose enzyme activities in roots, stubble and leaves of ryegrass (Lolium perenne L.) as affected by source/sink modification after cutting. Journal of Plant Physiology 140: 282291.
  • Teulat B, Borries C, This D. 2001. New QTLs identified for plant water status, water-soluble carbohydrate and osmotic adjustment in a barley population grown in a growth-chamber under two water regimes. Theoretical and Applied Genetics 103: 161170.
  • Teulat B, Monneveux P, Wery J, Borries C, Souyris I, Charrier A, This D. 1997. Relationships between relative water content and growth parameters under water stress in barley: a QTL study. New Phytologist 137: 99107.
  • Teulat B, This D, Khairallah M, Borries C, Ragot C, Sourdille P, Leroy P, Monneveux P, Charrier A. 1998. Several QTLs involved in osmotic-adjustment trait variation in barley (Hordeum vulgare L.). Theoretical and Applied Genetics 96: 688698.
  • Teulat B, Zoumarou-Wallis N, Rotter B, Ben Salem M, Bahri H, This D. 2003. QTL for relative water content in field-grown barley and their stability across Mediterranean environments. Theoretical and Applied Genetics 108: 181188.
  • Thomas H. 1986. Effects of rate of dehydration on leaf water stress and osmotic adjustment in Dactylis glomerata L., Lolium perenne L. and L multiflorum Lam. Annals of Botany 57: 225235.
  • Thomas H, James AR. 1999. Partitioning of sugars in Lolium perenne (perennial ryegrass) during drought and on rewatering. New Phytologist 142: 295305.
  • Thorsteinsson B, Harrison PA, Chatterton NJ. 2002. Fructan and total carbohydrate accumulation in leaves of two cultivars of timothy (Phleum pratense Vega and Climax) as affected by temperature. Journal of Plant Physiology 159: 9991003.
  • Turner LB, Cairns AJ, Armstead IP, Ashton J, Skøt K, Whittaker D, Humphreys MO. 2006. Dissecting the regulation of fructan metabolism in perennial ryegrass (Lolium perenne) with quantitative trait locus mapping. New Phytologist 169: 4558.
  • Turner LB, Humphreys MO, Cairns AJ, Pollock CJ. 2001. Comparison of growth and carbohydrate accumulation in seedlings of two varieties of Lolium perenne. Journal of Plant Physiology 158: 891897.
  • Vijn I, Smeekens S. 1999. Fructan: more than a reserve carbohydrate. Plant Physiology 120: 351359.
  • Volaire F, Thomas H, Lelievre F. 1998. Survival and recovery of perennial forage grasses under prolonged Mediterranean drought. I. Growth, death, water relations and solute content in herbage and stubble. New Phytologist 140: 439449.
  • Volenec JJ, Ourry A, Joern BC. 1996. A role for nitrogen reserves in forage regrowth and stress tolerance. Physiologia Plantarum 97: 185193.
  • White LM. 1973. Carbohydrate reserves of grasses: a review. Journal of Range Management 26: 1318.
  • Yamada T, Forster JW, Humphreys MW, Takamizo T. 2005. Genetics and molecular breeding in Lolium/Festuca grass species complex. Grassland Science 51: 89106.
  • Yamada T, Jones ES, Cogan NOI, Vecchies AC, Nomura T, Hisano H, Shimamoto Y, Smith KF, Hayward MD, Forster JW. 2004. QTL analysis of morphological, developmental, and winter hardiness-associated traits in perennial ryegrass. Crop Science 44: 925935.