SEARCH

SEARCH BY CITATION

Keywords:

  • allelic substitution effect;
  • correlation clustering;
  • epinasty;
  • hypertrophied lenticel;
  • multi-environment model;
  • root hypoxia

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIAL AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Quercus robur L. is a mid-European broadleaved tree species that grows readily on temporary waterlogged soils. An experiment aiming to identify potential markers of tolerance to waterlogging in this species and to assess the degree of genetic control over the corresponding traits was conducted. Quantitative trait loci (QTL) were assessed in an F1 progeny for responses to waterlogging, and the relevance of the observed traits as markers of tolerance was investigated using a precise description of the time course of their expression. Five significant QTL involved in the response to waterlogging were identified. In particular, QTL were detected for the development of hypertrophied lenticels and for the degree of leaf epinasty, but not for the formation of adventitious roots. A multi-environment QTL model allowed a detailed description of the time course (7 weeks) of the allelic substitution effect of some of these QTL. Correlation clustering identified significant clusters of QTL, at inter-trait as well as at intra-trait level. These clusters suggest the occurrence of a genetically controlled response cascade to waterlogging.


INTRODUCTION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIAL AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Quercus robur L. is a mid-European broadleaved tree species that favours deep and well-watered soils and is able to grow readily even on temporary waterlogged soils (Lévy, Becker & Duhamel 1992). This contrasts with the closely related Quercus petraea (Matt.) Liebl that requires well-drained soils. Ecological divergence likely plays a role in the maintenance of the separation of these two species despite the absence of a complete reproductive barrier (Petit et al. 2004). Natural selection may enable speciation even in the presence of gene flow (Doebeli et al. 2005; Lexer et al. 2005). In such a model, disruptive selection favours in each species the fixation of alleles with opposite effects on the phenotype (Schluter 2001). In oaks, disruptive selection for the response to waterlogging may partly explain the divergence between Q. robur and Q. petraea. Natural selection may have favoured alleles involved in tolerance to root hypoxia in the Q. robur genome. In consequence, keeping species integrity against a potentially large interspecific gene flow is an important issue to maintain individuals adapted to waterlogged stations.

The identification of traits related to waterlogging tolerance that are under genetic control is a step towards the detection of traits that (1) are relevant for understanding the physiological basis of waterlogging tolerance in young Q. robur trees, and (2) identify genomic regions that differ between the two species. No information is yet available about the degree of genetic determinism of waterlogging tolerance in trees. To detect the degree of genetic control on traits and the genomic regions involved, a quantitative trait loci (QTL) approach is required. The identification of QTL for tolerance-related traits is also a prerequisite for a future search of putative candidate genes for waterlogging tolerance. Furthermore, some alleles of these genes could be specific to the species Q. robur and therefore used for species identification.

Traits related to waterlogging tolerance in trees are expected to differ from those reported in annual plants like rice (Setter et al. 1997; Sripongpangkul et al. 2000; Jackson & Ram 2003) in terms of underlying key mechanisms for survival. The physiological processes involved in tolerance to waterlogging and to the resulting root hypoxia are still poorly known in the case of tree species (Kozlowski & Pallardy 2002; Kreuzwieser, Papadopoulou & Rennenberg 2004). In addition, no consistent model is yet available for hypoxia tolerance (Drew 1997; Vartapetian & Jackson 1997). For instance, in the case of Q. robur, previous studies showed that the response of seedlings to waterloggingincludes short-term responses as well as long-term acclimation processes. Short-term responses include changes in root-cell sugar metabolism, with increased activity of fermentation pathway enzymes (Parelle et al. 2006). Long-term acclimation relies on the formation of hypertrophied lenticels and of adventitious roots (Colin-Belgrand, Dreyer & Biron 1991; Schmull & Thomas 2000; Parelle et al. 2006). The large variability observed in seedling survival during waterlogging could result from an intraspecific diversity in the ability to develop these acclimation traits.

We developed an experiment aiming at detecting QTL for a large number of traits induced by waterlogging. QTL detection is classically used with quantitative traits as recorded at a given time under one or several environmental conditions. However, some traits of interest follow a time course that could provide supplementary information for QTL detection. Detecting QTL for modelled parameters describing the time course of such traits is one possible approach (Zhang et al. 2005). However, such a modelling approach is not always available for complex traits. Direct QTL detection using a multi-environment analysis (Jansen et al. 1995) does not require preliminary modelling. Such analyses are currently used to assess the stability of QTL under different environmental conditions (Jermstad et al. 2003; Reymond et al. 2003) or across years (Lerceteau, Szmidt & Andersson 2001). In the present experiment, successive dates of sampling were used as the different environments in a multi-environment QTL mapping model. This technique increases the power of detection of the QTL and the precision of mapping (Jansen et al. 1995). Additionally, the method provides useful information on the time course of the allelic substitution effect of a given QTL.

The analysis was focused on the morphological response of Q. robur to waterlogging. We monitored the time course of leaf epinasty (i.e. increase of the angle that petioles form with stems), which is known to be highly specific of hypoxia responses (Vartapetian & Jackson 1997). We also measured the time course of the formation of hypertrophied lenticels, which are thought to play a major role in hypoxia tolerance in Q. robur (Parelle et al. 2006). At the end of the treatment, the formation of adventitious roots was recorded as a candidate trait to assess the level of tolerance, as is reported for numerous herbaceous species (Vartapetian & Jackson 1997). As a matter of fact, adventitious roots are known to be formed massively on Q. robur in response to waterlogging (Colin-Belgrand et al. 1991; Parelle et al. 2006). The time courses of leaf chlorophyll content and of diameter at collar, and several traits characterizing plant biomass and leaf structure, were quantified as indicators of general plant performance. These measurements were also aimed at following the time course of the overall plant response to waterlogging and to obtain further information for interpreting the relevance of the detected QTL.

Our main objectives were therefore to document (1) the genetic determinism of responses to waterlogging in Q. robur, (2) the organization of the genomic regions involved in this determinism and (3) the time courses of the allelic substitution effect of some of the detected QTL.

MATERIAL AND METHODS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIAL AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Plant material and waterlogging conditions

Rooted cuttings of Q. robur were grown in a greenhouse from December 2003 to June 2004 in 4 L pots containing peat and sand mixture. The rooted cuttings comprised three clonal replicates of 120 full sibs. These full sibs were a chosen subsample of the mapping population of Q. robur obtained from the controlled cross between two Q. robur individuals located at INRA-Pierroton (South-western France, see details in Scotti-Saintagne et al. 2004). The subsample of individuals was selected among those with the largest number of identified genetic markers. Individuals were randomized in 15 incomplete blocks of 24 plants each. Individual pots were placed in groups of four in larger containers. These containers were filled with water up to the soil level from 21 June to 9 August 2004 (51 d or 7 weeks). The water level was adjusted daily with dis-oxygenized water obtained by bubbling N2 in water tanks (Parelle et al. 2006). Oxygen content was monitored at days 2, 9, 17 and 42 of the waterlogging treatment, in piezometric tubes inserted in each pot. Dissolved oxygen content was already low in tubes close to the roots after 2 d of waterlogging (2.7 ± 0.1 mg L−1) and decreased further to 1.3 ± 0.1 mg L−1 during the treatment. This corresponds to severe hypoxic conditions as compared to water equilibrated with free air (8.5 mg L−1 at similar temperature). The oxygen content in the free water around the pots was higher, but similarly decreased from 5.9 ± 0.1 to 2.3 ± 0.1 mg L−1 during the treatment. The trees were fertilized with Nutricote (N/P/K, 13/13/13; Fertilizer Co Ltd, Chisso-Asan, NY, USA) at 4 g L−1substrate during active growth (before waterlogging, August 2003). During waterlogging, dissolved nitrate was estimated in the water table of 24 randomly selected trees using Quantofix test sticks (Machery-Nagel, Düren, Germany). After 31 d of treatment, a severe N deficiency due to nitrate leaching and denitrification induced by waterlogging was detected. A liquid fertilizer (Plantprod, Brampton, Canada) at 50 mg of nitrogen per litre was applied evenly to all individuals after 31 d of waterlogging.

Recorded traits

The chlorophyll content in leaves was measured twice a week from 6 d before waterlogging to 47 d of treatment. Three to five measurements were made on three fully developed leaves of the last growth flush developed before waterlogging on each tree, using a chlorophyll content metre (CCM; Opti Science, Tyngsboro, MA, USA). The same leaves were monitored during the whole experiment. After 50 d of waterlogging, the three leaves were sampled for water and nitrogen content and leaf mass-to-area ratio (LMA) determination. Three pooled and dried leaves were ground with a ball mill (MM 200; Retch, Haan, Germany), and the nitrogen content was estimated with a Carlo Erba elementary analyser (Carlo Erba, Milan, Italy). The diameter at collar was measured twice a week on all individuals, 2 cm above the water table from 10 d before waterlogging to 46 d of treatment. The occurrence of leaf epinasty was recorded using three ordinal classes: above perpendicular, perpendicular to stem axis and less than perpendicular, and three numerical values were assigned to these classes: 120°, 90° and 60°, respectively, in order to obtain a quantitative variable necessary for QTL analysis.

During the full time course of waterlogging, visible hypertrophied lenticels (i.e. those occurring above and those occurring below the water table and accessible to observation) were monitored on each plant. We used two complementary criteria to describe lenticel development: total count and colour. The latter criterion was used as an estimator of the degree of hypertrophy of the lenticels, and encompassed five levels (see Table 1 for details). Using the colour scale we defined nine different classes to describe the population of lenticels on each individual (Table 1). We used the date at which colour class 9 was reached as an indicator of the rate of development of hypertrophied lenticels. At the end of the experiment (7 weeks of waterlogging), we observed that a large number of hypertrophied lenticels were hidden below the soil level. In consequence, the total amount of hypertrophied lenticels was recorded. In addition, the diameter and the thickness of hypertrophied lenticels above and below the water level were visually assessed.

Table 1.  Development stages for hypertrophied lenticels assessed from their colour
Description of the colour of the hypertrophied lenticelsColour levelLevels observed on whole plantColour classes
  1. Colour classes were defined as a function of the relative fraction (indicated by the symbol >, e.g. 1 > 2 a majority of lenticels at stage 1 and some at stage 2) of the different colour levels observed on the whole plant. *Colour level and classes that were only observed at the end on the treatment, and in consequence not used for the estimation of the development stages for ‘visible’ hypertrophied lenticels.

Black disc visible on the bark, corresponding to the start of hypertrophy and splitting of the bark111
1 > 22
1 > 33
A white spot growing in the middle of the black disc due to appearance of suber cells by the hole formed by the cracking of the bark22 > 14
25
2 > 36
Totally white lenticel33 > 17
3 > 28
3 > 19
3 > 410*
3 > 511*
Large and purple*4*4 > 312*
413*
4 > 514*
Large and black*5*5 > 315*
5 > 416*
517*

After 7 weeks of waterlogging, we recorded the number of severely wilting individuals. Finally, all trees were randomly harvested during 1 week (days 51 to 55 of waterlogging) and the water level was marked on each stem. Small stem sections (corresponding to the zone of formation of hypertrophied lenticels) were sampled and immediately frozen in liquid nitrogen for bark anatomy observations. The roots were cleaned with tap water and adventitious roots were counted, sampled and their length measured. Dry mass was measured on fine, main and adventitious roots separately, and on leaves and stems. Leaf dry mass and LMA of a sample of leaves were used for the calculation of total leaf area of each individual.

Anatomical observations

Twenty-four trees were selected in order to cover independently the range of hypertrophied lenticels and of adventitious roots, together with two replicates of three clones of plants with either no, few or many adventitious roots. We also selected two replicates of three clones grown under normoxic conditions as controls. A levelled cross-section of bark tissue just beneath a hypertrophied lenticel (when there was at least one on the collar of the plant) was observed by variable pressure scanning electron microscopy (VP-SEM, model Leo 1450 VP; Leo Electron Microscopy Inc., NY, USA).

Statistical and QTL analyses

Statistical analyses were performed using Statistica version 7.1 (Statsoft, Tulsa, OK, USA), and QTL detection was done using Multi-QTL version 2.5, available at http://www.multiqtl.com (Haifa, Israel). Charts of linkage maps and QTL were made using the MapChart version 2.1 (Voorrips 2001). Hierarchical linkage clustering (HLC) was performed using TIGR-Mev version 3.1 available at http://www.tm4.org (Saeed et al. 2003). For all analyses (including QTL analyses), the tests were considered significant when P-value was below 0.05.

Statistical analyses for correlations among traits

Correlations among traits were tested with a linear model and comparisons among multiple groups were done with an analysis of variance (anova) followed by Tukey–Kramer mean-comparison test. Adventitious root dry mass was log-transformed in order to respect the postulates of anova. For comparisons between two categories of individuals, weused parametric analyses for quantitative traits, and non-parametric ones for ordinal traits. As the data for diameter and thickness of hypertrophied lenticels did not follow the assumptions of the Student's tests, and as transformation did not improve the distribution, only non-parametric tests were performed for these two traits. Comparisons between above- and below-water hypertrophied lenticels were made with a Student's t-test for dependent samples (number of lenticels) and with sign tests (ordinal traits).

Genetic map

The Q. robur clones used in this study originated from an intraspecific F1 family composed of 278 full sibs. To overcome the unknown allelic phase in offsprings, we used the so-called double pseudo-testcross mapping strategy (Grattapaglia & Sederoff 1994), which only considers alleles segregating in the female background (monomorphic in the father, except for microsatellites) to build the female map and only alleles segregating in the male (monomorphic in the mother, except for microsatellites) to build the male map. Details of the map construction dedicated to QTL location are given in Scotti-Saintagne et al. (2004). A majority of gaps present in an earlier version of the map were filled for the present study. The local order of markers was re-analysed by keeping an LOD threshold of 3. Nineteen co-dominant markers common to both maps were used as bridges to align male and female linkage groups (LGs). For both sexes, 12 LGs were identified, corresponding to the 12 chromosomes of Quercus.

QTL detection

The presence of a block effect was tested using the following statistical model:

  • image

where Yij is the observed value of the trait, µ the overall mean, Ci the random effect of clone i, bj the fixed effect of block j and εij the error term. To estimate the block-corrected clonal means for QTL analysis, the least square means for Ci were used. If no block effect was detected, bj was omitted from the model.

Assumptions of anova were fulfilled for each variable, except for the number and dry mass of adventitious roots, and for oxygen content, which were log-transformed to fulfil the assumptions.

The multi-interval mapping (MIM) method (Jansen & Stam 1994; Zeng 1994; Kao, Zeng & Teasdale 1999) was used to identify QTL, allowing individual QTL to be detected independently of background noise and already detected QTL. Multi-QTL software uses statistical models to estimate the significance of QTL, where models with one and two QTL per LG are proposed. The single QTL model tests the null hypothesis that there is a single QTL on the LG against the absence of QTL, by permutation test (1000 permutations, Churchill & Doerge 1994). The two linked QTL model tests the null hypothesis of the presence of two QTL on the LG: (1) against the hypothesis of the absence of QTL and (2) against the hypothesis of the presence of a single QTL. The two tests were performed by permutations (1000 permutations). The MIM procedure of Multi-QTL allows attributing the appropriate model to each LG. As permutation tests in the Multi-QTL software calculate the significance of the QTL at the chromosome level, the significance for all detected QTL was recalculated at the genome level, as a function of the overall number of markers in the map and the number of markers in each LG (Brendel et al. 2002; Scotti-Saintagne et al. 2004). QTL with a P-value < 0.05 at the genome level were considered to be significant. The confidence intervals of the position were estimated by a bootstrap method (1000 re-samplings, Visscher, Thompson & Haley 1996). The allelic substitution effect of the QTL and the percentage of explained variance (PEV) were calculated with the MIM procedure. When the allelic substitution effect of a QTL was not significantly different from 0, we considered the QTL not significant, even though its P-value at the genome level was < 0.05.

QTL detection was performed in a two-step procedure: (1) a MIM was performed among the LGs, using a two linked QTL model on each LG and (2) a MIM with a two linked QTL model was performed on LGs for which significant two linked QTL were detected, and a single QTL model for the other ones.

For leaf chlorophyll, epinasty, diameter at collar and dissolved oxygen, we used a multi-environment model (Jansen et al. 1995) with each sampling date considered to be a different environment. The model of Jansen et al. (1995) does not require independence among individuals in the different environments, which allows to use this model to assess a time course for the allelic substitution effect of the QTL. The data were normalized in order to take into account the heterogeneity of variances among the different dates. We used the same procedure for the classical single-environment models, except that we considered that the QTL were significant if the effect was significantly different from 0 for at least one date.

As leaf-epinasty was monitored using an ordinal scale, we tested whether the transformation into a quantitative variable did not affect the QTL detection. A marker-by-marker Mann–Whitney U-test was performed, and all QTL detected by MIM were confirmed (P < 0.05 at the chromosome level for at least one date) except for those found on LG 8 on the male map (LG8M).

Correlation clustering of the time patterns of QTL effects

For traits analysed with the multi-environment model, the time courses of allelic substitution effects were compared to detect significant patterns. Absolute values of allelic effects were used and the patterns were normalized to a variance of 1 and a mean of 0. Temporal patterns were sorted by similarity using an HLC algorithm with the absolute Pearson's correlation coefficient as an entry (Plomion et al. 2000). This algorithm builds a hierarchical tree (HLC tree) of the different patterns as a function of their correlation coefficients. Clusters were built by imposing a distance threshold to the HLC tree, corresponding to the minimum absolute Pearson's correlation coefficient necessary to obtain a significant correlation. Finally, the coefficients of determination were reported in a correlation matrix, which was ordered according to the HLC tree and divided in the different significant clusters detected.

This method was used for each trait independently. As the different traits were not measured on the same days, inter-trait comparisons were made after switching to a weekly resolution of the time course of the effects.

RESULTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIAL AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Time course of the responses of Q. robur to waterlogging

The diameter at collar increased during the course of the treatment (Fig. 1a), with two transient slow downs (at the beginning of waterlogging and just before the addition of fertilizer on day 31). The leaf chlorophyll content increased during waterlogging, particularly after fertilization (Fig. 1b). The leaf angles between petioles and stems decreased rapidly during the treatment, revealing an increase in epinasty (Fig. 1c). The number of visible hypertrophied lenticels increased linearly during the treatment (Fig. 2a). Lenticel development, estimated from the changes in colour, took place mainly during the first 20 d of waterlogging (Fig. 2b), after which all lenticels had reached the maximum development stage.

image

Figure 1. Time course during waterlogging of (a) diameter at collar, (b) leaf chlorophyll content [arbitrary chlorophyll content meter metre (CCM) units] and (c) leaf epinasty (defined as the angle that leaves formed with the stems). The arrow indicates the date when liquid fertilizer was added to the waterlogged plants. Means with confidence intervals at 0.95. See text for details.

Download figure to PowerPoint

image

Figure 2. Time course of the number of visible hypertrophied lenticels during waterlogging. (a) Total number (means and confidence intervals). (b) Development stage estimated with the ordinal scale of colours (see Table 1 for details), medians+, quartiles (75% of the individuals) boxes, non-outlier range whiskers (95% of the individuals).

Download figure to PowerPoint

Responses after 7 weeks of waterlogging

Adventitious root formation and hypertrophied lenticels

At the end of the experiment, that is, after 7 weeks of waterlogging, we could not detect any correlation between parameters describing lenticel development and adventitious root formation: the two groups of traits were completely independent. The largest fraction of hypertrophied lenticels (69%) was located below the water table. The diameter and thickness of hypertrophied lenticels were significantly larger below than above water ( Fig. 3). At the end of the treatment, there was a significant correlation between the number of visible and the total number of hypertrophied lenticels (R2 = 0.34, P < 0.001). Nevertheless, only 15% of those under water had been visible during the course of waterlogging.

image

Figure 3. Diameter and thickness of hypertrophied lenticels located on stems and roots above or below the water level in individuals of Quercus robur submitted to 7 weeks waterlogging. Visual estimation, medians+, quartiles (75% of the individuals) boxes, non-outlier range whiskers (95% of the individuals).

Download figure to PowerPoint

Anatomical observations

No lacuna could be detected in the parenchyma beneath lenticels or the equivalent zone on controls of the 30 observed individuals (Fig. 4), regardless of development stage of hypertrophied lenticels or adventitious roots. On the contrary, support tissue was detected in the cortical parenchyma and formed discontinuous concentric circles around the stele in stressed as well as in control individuals (Fig. 4).

image

Figure 4. Anatomy of hypertrophied lenticels and beneath bark tissue observed in Quercus robur roots after 7 weeks of waterlogging (variable pressure-scanning electron microscopy). Support tissue was present in the bark as indicated by an arrow in the figures. (a) Bark tissue beneath a hypertrophied lenticel. (b) Bark of a control plant.

Download figure to PowerPoint

Detected QTL: distribution, number and PEV

QTL for responses to waterlogging were detected on all LGs of the female and male maps, except on LG 6 male (LG6 M) and LG 8 female (LG8F) (Supplementary Information S1 and S2). More QTL were detected with the multi-environment than with the classical single-environment method. Indeed the former allowed to detect QTL with lower PEV (< 9%) than the latter (Supplementary Information 1).

QTL were detected for traits unambiguously induced by waterlogging, many of them being located on the male map (Supplementary Information S2). There were five QTL for epinasty distributed on four different LGs of the male map (3, 8, 9, 11) with maximum PEV (mPEV) ranging from 7.8 to 12.2% depending on the date (Supplementary Information S1). A QTL was found on LG9M for dissolved oxygen content in the water table with an mPEV of 12.2%, and two QTL for hypertrophied lenticels were detected on LG2M (total number) and LG3M (rate of development) with PEV of 14.8 and 10.4%, respectively. Surprisingly, no QTL was detected on either map for adventitious roots, whether mass, number, total length or mean length were taken into account. This occurred despite a significant clone effect for adventitious root mass and number explaining 6.6% of the total variance (data not shown).

Significant QTL were also visible for biomass accumulation on the male map, and for the shoot/root ratio on both male and female maps, with PEV of 15.4 and 12.1, respectively. A QTL for leaf nitrogen content was detected on both maps (PEV of 12 and 16.7%). The traits analysed with the multi-environment procedure produced a large number of QTL. For chlorophyll content, 23 QTL were distributed over all LGs on the male and female maps, except LG 8 (Supplementary Information S2); mPEV (mPEV: maximum of all dates) was between 3.5 and 16.3%. For diameter at collar, 18 QTL were distributed over all LGs except LGs 4 and 8 (Supplementary Information S1 and S2) and displayed mPEV between 3.7 and 20% (see Supplementary Information S1 and S2 for further details on detected QTL).

Important genomic regions involved in control of waterlogging responses

We focused on genomic regions for which at least one QTL was detected for a trait specific to hypoxia response (such as oxygen content, epinasty and lenticel development) with a PEV above 9% (Fig. 5). When several additional QTL presented overlapping confidence intervals and when their localizations were within one adjacent interval of those QTL, they were considered genetically close. Five such genomic regions were identified on the male map, and none on the female map:

image

Figure 5. Genetic map of markers in centimorgan and quantitative trait loci (QTL) position among linkage groups (LGs) on which several QTL were identified at a short distance. Estimation of the position by permutation and bootstrap, and confidence intervals at 0.95. LMA, leaf mass-to-area ratio; DM, dry mass.

Download figure to PowerPoint

Genomic region 1.  On LG2M (at 78.1 cM, Fig. 5), a QTL for the total number of hypertrophied lenticels had a PEV of 14.8%, and QTL for main root dry mass and diameter at collar were close.

Genomic region 2.  On LG3M (at 0 cM, Fig. 5), a QTL for the rate of development of hypertrophied lenticels (PEV = 10.4%) was close to a QTL for epinasty (mPEV = 7.8%), one for leaf chlorophyll content, and four QTL for growth parameters.

Genomic region 3.  On LG9M (at 8.8 cM, Fig. 5), a QTL for dissolved oxygen with an mPEV of 9.3% was close to a QTL for diameter at collar and one for leaf chlorophyll content.

Genomic region 4.  On LG9M (at 54.8 cM, Fig. 5), a QTL for epinasty (mPEV = 9.9%) was close to a QTL for leaf chlorophyll content.

Genomic region 5.  On LG11M (at 57.9 cM, Fig. 5), a QTL for epinasty (mPEV = 11.6%) was close to a QTL for leaf chlorophyll and nitrogen contents, LMA and diameter at collar.

Temporal patterns of QTL effects

The multi-environment QTL analysis provided a variety of temporal patterns for the allelic substitution effect of QTL for epinasty, chlorophyll content and diameter at collar. The number of lenticels monitored during the course of the experiment yielded no QTL. Substitution effects followed very different temporal time courses. For example, the effect patterns of cluster 2 for leaf chlorophyll content were a linear increase during the first 20 d of treatment followed by a constant value while those of cluster 4 linearly increased, or decreased over the whole period (Fig. 6b). Some of these effect patterns were tightly intercorrelated, both on an intra-trait as well as on an inter-trait level (Figs 6 & 7). We found one cluster for epinasty, four for diameter at collar and four for chlorophyll content, each containing two to eight QTL (Fig. 6). Only 10 out of the 45 effect patterns analysed could not be classified (i.e. were not significantly correlated to other patterns). Some of the clustered effect patterns were negatively correlated. For example, in cluster 4 of leaf chlorophyll content (Fig. 6b), the first five QTL followed positively, and the last three negatively correlated temporal patterns. During the inter-trait analysis, six mixed-trait clusters were detected, three (5, 9, and 11; Fig. 7) contained effect patterns of QTL for epinasty, which is highly specific for hypoxia response.

image

Figure 6. Intra-trait clusters of effect patterns. (a) Matrices of correlation among time patterns of quantitative trait loci (QTL) effects, and clusters C1 to C4 evidenced for leaf chlorophyll content, diameter at collar and epinasty (effects normalized to a mean of 0 and an SD of 1 as a function of duration of waterlogging treatment). Colour intensity indicates square Pearson's correlation coefficient between the effect patterns of two QTL. Effect patterns were sorted by similarity using a hierarchical linkage clustering algorithm with the absolute Pearson's correlation coefficient as an entry. Black lines indicate clusters of QTL with correlated effect pattern. (b) Clusters of effect patterns evidenced for leaf chlorophyll content, diameter at collar diameter and epinasty. QTL are abbreviate from their linkage group (LG) while M or F refer to the male and female map, respectively. When two linked QTL were detected on the same LG they were noted −1 for the QTL close to LG start, and −2 for the second. Positively correlated QTL were symbolized with a similar line, and negatively correlated ones with different line (full or dotted lines).

Download figure to PowerPoint

image

Figure 7. Inter-trait clusters of effect patterns. (a) Matrix of correlation among time patterns of effects of quantitative trait loci (QTL) of leaf chlorophyll content, collar diameter and epinasty (effects normalized to a mean of 0 and an SD of 1 as a function of duration of waterlogging treatment). Colour intensity indicates square Pearson's correlation between the effect patterns of two QTL. Effect patterns were sorted by similarity using a hierarchical linkage clustering algorithm with the absolute Pearson's correlation effect pattern. (b) Clusters of effect patterns grouping different traits. QTL are abbreviate from their linkage group (LG) while M or F refer to the male and female map, respectively. When two linked QTL were detected on the same LG they were noted −1 for the QTL close to LG start, and −2 for the second. Positively correlated QTL were symbolized with a similar line, and negatively correlated ones with different line (full or dotted lines).

Download figure to PowerPoint

Some QTL displayed similar locations on the male and female maps: a QTL for diameter at collar on LG7, and a QTL for leaf chlorophyll content on LG10 (Supplementary Information S2). Intra-trait clustering revealed that in the two cases, the QTL clustered with similar effect patterns (positive correlation for the first, and negative for the second). Inter-trait clustering showed that a QTL for leaf chlorophyll content and a QTL for epinasty on LG11M clustered with similar effect patterns and were very close on the genetic map (Figs 5 & 7b).

DISCUSSION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIAL AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Physiological relevance of waterlogging response QTL in Q. robur

Five genomic regions containing each at least one QTL for a trait specific to hypoxia responses were detected. We focused on QTL with a high PEV, keeping in mind that the small number (120) of genotypes from the F1 progeny used here probably induced a threefold overestimation of PEV (Beavis et al. 1994; Brown et al. 2003).

The use of artificial dis-oxygenation of water resulted in severe hypoxia. Nevertheless, we detected an apparent genotypic variability of oxygen content among pots, which led to a QTL (genomic region 3). This QTL can only be explained by the occurrence of a diversity of rates of O2 consumption by roots of the different individuals, which is an indirect index for root activity under hypoxia. As a matter of fact, no such apparent genotypic effect was detected for dissolved oxygen measured in the free water around the pots. Whereas the variability of O2 content in pots may have induced a slight variation of stress among genotypes, no correlation was detected between the presence of adaptive structures (i.e. hypertrophied lenticels and adventitious roots) and oxygen content. It may therefore safely be assumed that the differences in O2 content were too small to induce a significant artefacts for QTL detection.

During hypoxia, adaptive structures such as hypertrophied lenticels and adventitious roots developed at the base of the main stem, the collar and the upper roots, as previously observed on Q. robur and Q. petraea (Colin-Belgrand et al. 1991; Parelle et al. 2006), on Quercus macrocarpa (Tang & Kozlowski 1982b), and on many other trees (e.g. Newsome, Kozlowski & Tang 1982; Tang & Kozlowski 1982a, 1984; Nunez-Elisea et al. 1999). The total lack of correlation between traits characterizing hypertrophied lenticels and adventitious roots was unexpected. In fact, hypertrophied lenticels and adventitious roots are thought to build a highly porous zone conducting oxygen through aerenchyma to the root system of herbaceous (Jackson & Drew 1984) or tropical woody plants (Joly 1991). During anatomical observations, no root primordia were detected beneath lenticels, supporting the hypothesis that adventitious roots do not develop beneath lenticels. These two traits are probably not induced directly by the same signal. In addition, we did not detect any lacuna in the bark beneath the hypertrophied lenticels, confirming observations on different Annona species (Nunez-Elisea et al. 1999), and on Alnus glutinosa, a highly waterlogging tolerant tree of temperate Europe (Armstrong & Armstrong 2005). Lacuna had been observed on other tree species (Ulmus americana, Pinus halepensis) adapted to waterlogging (Angeles, Evert & Kozlowski 1986; Yamamoto, Kozlowski & Wolter 1987). On the contrary, we observed support tissues, which could act as barrier to gas diffusion. All these observations emphasize that oxygen diffusion from lenticels to roots is not very likely to occur in Q. robur under hypoxia. The largest fraction of hypertrophied lenticels was located below the water level, as previously observed on oaks (Tang & Kozlowski 1982b; Parelle et al. 2006) and other tree species (Topa & McLeod 1986b, 1988), and the lenticels formed above the water level were smaller and less developed. In consequence, it may be safely concluded that hypertrophied lenticels are formed mainly in, or close to, a hypoxia zone (below the water level). No clear hypothesis could be formulated about the function of these hypertrophied lenticels. These observations underline that the QTL detected for hypertrophied lenticels (genomic regions 1 and 2) are highly specific to hypoxia stress. However, the detection of QTL for growth parameters in close vicinity to the former on the genome, suggests either that growth was differentially affected by hypoxia among clones, and that the corresponding QTL reflect different levels of tolerance, or conversely, that these QTL reflect clonal and hypoxia-independent differences of vigour among clones. Indeed, the number of lenticels in particular may depend on stem and root dimensions. There is currently no possibility to demonstrate which of these conflicting hypotheses is true. There is no information available in the literature on the variability of hypertrophied lenticels density.

Whereas the function of hypertrophied lenticels is still a matter of debate, the role of adventitious roots in hypoxia tolerance is well understood: they grow close to the top of the substrate where oxygen availability is largest and allow plants to sustain water and nutrient absorption, while deep roots die (Topa & McLeod 1986a; Vartapetian & Jackson 1997; Drew, He & Morgan 2000). Despite the occurrence of clonal effects for the number and biomass of adventitious roots formed on the individuals, no QTL could be detected for adventitious root formation, whatever parameter (number, length or biomass) was taken into account. The clonal variance was low and an environmental background noise may have hindered the detection of a QTL.

Epinasty, the increase of the angle that the petiole forms with the stem, is a highly specific trait for oxygen deprivation and waterlogging (Vartapetian & Jackson 1997), and has never been observed in other stress situations. The QTL detected for this trait (regions 2, 4 and 5) should be highly specific for root hypoxia. Interestingly, the genomic region 2 contained a QTL with a high PEV for the development rate of hypertrophied lenticels as well as for epinasty, suggesting that this region should be of particular interest for future investigations.

The continuous increase of diameter at collar throughout the 7 weeks of waterlogging demonstrated the capacity of Q. robur to sustain growth even under severe root hypoxia. Hypertrophy of collar, resulting also in an increase in diameter, has been reported to be a trait related to hypoxia response (Yamamoto & Kozlowski 1987; Yamamoto et al. 1987). However, most of the detected QTL for diameter growth are probably related to growth control under unstressed conditions such as those detected for oaks (Scotti-Saintagne et al. 2004) and chestnut (Casasoli et al. 2004). The observed increase of chlorophyll content during the first 2 weeks of the stress could be due to the phenological development of leaves. The slight decrease observed after 18 and 28 d, followed by an increase after fertilization, could be linked to nitrogen availability, which is severely lowered in hypoxic soils due to denitrification (Zhang & Wienhold 2002; Venterink, Hummelink & Van den Hoorn 2003). Thus, fertilization probably counteracted this phenomenon, allowing the chlorophyll content to increase. Leaf chlorophyll content is highly linked to leaf nitrogen content (e.g. Niinemets & Kull 1997), and thus depending on nitrogen availability under root hypoxia (Kreuzwieser, Furniss & Rennenberg 2002). In consequence, some of the QTL detected might represent the capacity of plants to maintain their nitrogen uptake despite hypoxic root conditions.

The physiological relevance of most of the QTL for waterlogging response, in particular for collar diameter and leaf chlorophyll content, was difficult to interpret. In order to classify the QTL and obtain further insight in their relevance, we used a clustering method based on multi-environment analysis.

Multi-environment analysis and clustering of effect patterns

Traits analysed using the multi-environment model produced a large number of QTL, especially QTL with small effects, as compared to the single environment analysis. This conformed to the theory described by Jansen et al. (1995) and Korol, Ronin & Nevo (1998), which predicts an increase of detected QTL with low effects, compared to the classical single-environment MIM. Additionally, a large percentage of the QTL detected using the multi-environment approach were precisely located on the map (small confidence intervals).

The estimation of an allelic effect for each environment (in this case the sampling dates) allowed us to estimate the temporal variation for each individual QTL (called ‘effect pattern’). The clustering of the effect patterns produced a few groups with a striking similarity, at the intra- as well as on the inter-trait level.

Some of these QTL with similar effect patterns were QTL for the same trait located in corresponding regions on the male and female maps. This supports the hypothesis that the same QTL were detected on the male and the female map. However, the possibility that different genes display similar effect patterns in the same genomic region cannot be excluded.

In one cluster, two QTL were located at close vicinity on the map. Nevertheless, in the majority of the cases QTL from a given cluster were distributed over the whole map. If it is assumed that each QTL is related to a single gene, then the clustering of effect patterns would suggest a parallel change in the effect of an allele of these genes within the family. Whether this parallelism can be explained by a coordinated response of a gene network to the hypoxic stress signalling, needs to be confirmed by further experiments. In this respect, the correlation coefficients among effect patterns would be an interesting tool for identifying such gene networks.

Furthermore, the inter-trait clustering highlighted two co-localized QTL with similar effect patterns, which supports the hypothesis that the same locus was detected by the QTL of the two traits. Three clusters (C5, C9 and C11) contained QTL for leaf chlorophyll content, diameter at collar (two traits not specific of waterlogging response) and epinasty, which is specific of waterlogging response. This could suggest that the QTL for leaf chlorophyll content and diameter at collar within this cluster are also involved in response of Q. robur to the hypoxia treatment.

CONCLUSION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIAL AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Five main genomic regions containing QTL specific to hypoxia response (epinasty, hypertrophied lenticels development and dissolved oxygen content) were identified, suggesting a genetic determinism for the diversity of waterlogging responses within this pedunculate oak family. Additionally, the multi-environment model and the clustering analyses of time courses of allelic QTL effects suggested further QTL potentially involved in the waterlogging response of Q. robur. In consequence, the strategy developed here is relevant for the identification of loci involved in the waterlogging response of Q. robur, despite the fact that the mechanisms of tolerance to waterlogging are still poorly known in forest tree species. The loci obtained on the genetic map are a powerful resource for the future detection of candidate genes.

ACKNOWLEDGMENTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIAL AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

J. Parelle was supported by a PhD grant from INRA and from the Regional Council of Lorraine. This work was part of the European Union Project ‘Oakflow’ (QLK5-2000-00960). We thank C. Bodénès (INRA Bordeaux) for filling in the missing data of the genetic map, E. Bertocchi (INRA Bordeaux) and G. Roussel (INRA Bordeaux) for the vegetative multiplication of the trees, A. Korol (Haifa University) for the helpful advice on Multi-QTL software and comments on the manuscript, J-M. Gioria (INRA Nancy), Y. Renaud, and all the technicians of the Experimental Unit of INRA Nancy for technical support, J. Marchand (INRA Nancy) for the nitrogen quantifications and D. Le Thiec (INRA Nancy) for SEM. We gratefully acknowledge C. Barlet (Nancy University) for help for the oxygen and diameter at collar measurements, and, above all, for having measured over 53 000 chlorophyll data points.

REFERENCES

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIAL AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information
  • Angeles G., Evert R.F. & Kozlowski T.T. (1986) Development of lenticels and adventitious roots in flooded Ulmus americana seedlings. Canadian Journal of Forest Research 16, 585590.
  • Armstrong W. & Armstrong J. (2005) Stem photosynthesis, not pressurized ventilation is responsible for light-enhanced oxygen supply to submerged roots of alder (Alnus glutinosa). Annals of Botany 96, 591612.
  • Beavis W.D., Smith O.S., Grant D. & Fincher R. (1994) Identification of quantitative trait loci using a small sample of topcrossed and F4 progeny from maize. Crop Science 34, 882896.
  • Brendel O., Pot D., Plomion C., Rozenberg P. & Guehl J.-M. (2002) Genetic parameters and QTL analysis of δ13C and ring width in maritime pine. Plant, Cell & Environment 25, 945953.
  • Brown G.R., Bassoni D.L., Gill G.P., Fontana J.R., Wheeler N.C., Megraw R.A., Davis M.F., Sewell M.M., Tuskan G.A. & Neale D.B. (2003) Identification of quantitative trait loci influencing wood property traits in loblolly pine (Pinus taeda L.). III. QTL verification and candidate gene mapping. Genetics 164, 15371546.
  • Casasoli M., Pot D., Plomion C., Monteverdi M.C., Barreneche T., Lauteri M. & Villani F. (2004) Identification of QTLs affecting adaptive traits in Castanea sativa Mill. Plant, Cell & Environment 27, 10881101.
  • Churchill G.A. & Doerge R.W. (1994) Empirical threshold values for quantitative trait mapping. Genetics 138, 963971.
  • Colin-Belgrand M., Dreyer E. & Biron P. (1991) Sensitivity of seedlings from different oak species to water-logging: effects on root growth and mineral nutrition. Annals of Forest Science 48, 193204.
  • Doebeli M., Dieckmann U., Metz J.A.J. & Tautz D. (2005) What we have also learned: adaptative speciation is theoretically plausible. Evolution 59, 691695.
  • Drew M.C. (1997) Oxygen deficiency and root metabolism: injury and acclimation under hypoxia and anoxia. Annual Review of Plant Physiology and Plant Molecular Biology 48, 223250.
  • Drew M.C., He C.J. & Morgan P.W. (2000) Programmed cell death and aerenchyma formation in roots. Trends in Plant Science 5, 123127.
  • Grattapaglia D. & Sederoff R. (1994) Genetic linkage maps of Eucalyptus grandis and E. urophylla using a pseudo-testcross mapping strategy and RAPD markers. Genetics 137, 11211137.
  • Jackson M.B. & Drew M.C. (1984) Effect of flooding on herbaceous plants. In Flooding and Plant Growth (ed. T.T.Kozlowski), pp. 47128. Academic Press, London, UK.
  • Jackson M.B. & Ram P.C. (2003) Physiological and molecular basis of susceptibility and tolerance of rice plants to complete submergence. Annals of Botany 91, 227241.
  • Jansen R.C. & Stam P. (1994) High resolution of quantitative trait into multiple loci via interval mapping. Genetics 136, 14471445.
  • Jansen R.C., Van Oijen J.M., Stam P., Lister C. & Dean C. (1995) Genotype-by-environment interaction in genetic mapping of multiple quantitative trait loci. Theoretical and Applied Genetics 91, 3337.
  • Jermstad K.D., Bassoni D.L., Jech K.S., Ritchie G.A., Wheeler N.C. & Neale D.B. (2003) Mapping of quantitative trait loci controlling adaptive traits in coastal Douglas fir. III. Quantitative trait loci-by-environment interactions. Genetics 165, 14891506.
  • Joly C.A. (1991) Flooding tolerance of tropical trees. In Plant Life Under Oxygen Deprivation (eds M.B.Jackson, D.D.Davies & H.Lambers), pp. 2334. SPB Academic Publishing, The Hague, The Netherlands.
  • Kao C.-H., Zeng Z.-B. & Teasdale R.D. (1999) Multiple interval mapping for quantitative trait loci. Genetics 152, 12031216.
  • Korol A.B., Ronin Y.I. & Nevo E. (1998) Approximate analysis of QTL-environment interaction with no limits on the number of environments. Genetics 148, 20152028.
  • Kozlowski T.T. & Pallardy S.G. (2002) Acclimation and adaptive responses of woody plants to environmental stresses. Botanical Review 68, 270334.
  • Kreuzwieser J., Furniss S. & Rennenberg H. (2002) Impact of water-logging on the N-metabolism of flood tolerant and non-tolerant tree species. Plant, Cell & Environment 25, 10391049.
  • Kreuzwieser J., Papadopoulou E. & Rennenberg H. (2004) Interaction of flooding with carbon metabolism of forest trees. Plant Biology 6, 299306.
  • Lerceteau E., Szmidt A.E. & Andersson B. (2001) Detection of quantitative trait loci in Pinus sylvestris L. across years. Euphytica 121, 117122.
  • Lévy G., Becker M. & Duhamel D. (1992) A comparison of the ecology of pedunculate and sessile oaks: radial growth in the centre and northwest of France. Forest Ecology and Management 55, 5163.
  • Lexer C., Fay M.F., Joseph J.A., Nica M.S. & Heinze B. (2005) Barrier to gene flow between two ecologically divergent Populus species, P. alba (white poplar) and P. tremula (European aspen): the role of ecology and life history in gene introgression. Molecular Ecology 14, 10451057.
  • Newsome R.D., Kozlowski T.T. & Tang Z.C. (1982) Responses of Ulmus americana seedlings to flooding of soil. Canadian Journal of Botany 60, 16881695.
  • Niinemets U. & Kull O. (1997) Stoichiometry of foliar carbon constituents varies along light gradients in temperate woody canopies: implications for foliage morphological plasticity. Tree Physiology 18, 467479.
  • Nunez-Elisea R., Schaffer B., Fisher J.B., Colls A.M. & Crane J.H. (1999) Influence of flooding on net CO2 assimilation, growth and stem anatomy of Annona species. Annals of Botany 84, 771780.
  • Parelle J., Brendel O., Bodénès C., Berveiller D., Dizengremel P., Jolivet Y. & Dreyer E. (2006) Differences in morphological and physiological responses to root hypoxia between two sympatric oak species (Quercus petraea [Matt.] Liebl., Quercus robur L.). Annals of Forest Science 63, 849859.
  • Petit R.J., Bodenes C., Ducousso A., Roussel G. & Kremer A.(2004) Hybridization as a mechanism of invasion in oaks. New Phytologist 161, 151164.
  • Plomion C., Pionneau C., Brach J., Costa P. & Baillères H. (2000) Compression wood-responsive proteins in developing xylem of maritime pine (Pinus pinaster Ait.). Plant Physiology 123, 959969.
  • Reymond M., Muller B., Leonardi A., Charcosset A. & Tardieu F. (2003) Combining quantitative trait loci analysis and an ecophysiological model to analyze the genetic variability of the responses of maize leaf growth to temperature and water deficit. Plant Physiology 131, 664675.
  • Saeed A., Sharov V., White J., et al . (2003) TM4: a free, open-source system for microarray data management and analysis. Biotechniques 34, 374378.
  • Schluter D. (2001) Ecology and the origin of species. Trends in Ecology & Evolution 16, 372380.
  • Schmull M. & Thomas F.M. (2000) Morphological and physiological reactions of young deciduous trees (Quercus robur L., Q. petraea [Matt.] Liebl., Fagus sylvatica L.) to water-logging. Plant and Soil 225, 227242.
  • Scotti-Saintagne C., Bodenes C., Barreneche T., Bertocchi E., Plomion C. & Kremer A. (2004) Detection of quantitative trait loci controlling bud burst and height growth in Quercus robur L. Theoretical and Applied Genetics 109, 16481659.
  • Setter T.L., Ellis M., Laureles E.V., Ella E.S., Senadhira D., Mishra S.B., Sarkarung S. & Datta S. (1997) Physiology and genetics of submergence tolerance in rice. Annals of Botany 79, 6777.
  • Sripongpangkul K., Posa G.B.T., Senadhira D.W., Brar D., Huang N., Khush G.S. & Li Z.K. (2000) Genes/QTLs affecting flood tolerance in rice. Theoretical and Applied Genetics 101, 10741081.
  • Tang Z.C. & Kozlowski T.T. (1982a) Physiological, morphological, and growth responses of Platanus occidentalis seedlings to flooding. Plant and Soil 66, 243255.
  • Tang Z.C. & Kozlowski T.T. (1982b) Some physiological and morphological responses of Quercus macrocarpa seedlings to flooding. Canadian Journal of Forest Research 12, 196202.
  • Tang Z.C. & Kozlowski T.T. (1984) Water relations, ethylene production, and morphological adaptation of Fraxinus pennsylvanica seedlings to flooding. Plant and Soil 77, 183192.
  • Topa M.A. & McLeod K.W. (1986a) Aerenchyma and lenticel formation in pine seedlings: a possible avoidance mechanism to anaerobic growth conditions. Physiologia Plantarum 68, 540550.
  • Topa M.A. & McLeod K.W. (1986b) Responses of Pinus clausa, Pinus serotina and Pinus taeda seedlings to anaerobic solution culture. I. Changes in growth and root morphology. Physiologia Plantarum 68, 523531.
  • Topa M.A. & McLeod K.W. (1988) Promotion of aerenchyma formation in Pinus serotina seedlings by ethylene. Canadian Journal of Forest Research 18, 276280.
  • Vartapetian B.B. & Jackson M.B. (1997) Plant adaptations to anaerobic stress. Annals of Botany 79, 320.
  • Venterink H.O., Hummelink E. & Van den Hoorn M.W. (2003) Denitrification potential of a river floodplain during flooding with nitrate-rich water: grasslands versus reedbeds. Biogeochemistry 65, 233244.
  • Visscher P.M., Thompson R. & Haley C.S. (1996) Confidence intervals in QTL mapping by bootstrapping. Genetics 143, 10131020.
  • Voorrips R.E. (2001) MapChart, Version 2.0: Windows Software for the Graphical Presentation of Linkage Maps and QTLs. Plant Research International, Wageningen, The Netherlands.
  • Yamamoto F. & Kozlowski T.T. (1987) Effects of flooding of soil on growth, stem anatomy, and ethylene production of Thuja orientalis seedlings. IAWA Bulletin 8, 2129.
  • Yamamoto F., Kozlowski T.T. & Wolter K.E. (1987) Effect of flooding on growth, stem anatomy, and ethylene production of Pinus halepensis seedlings. Canadian Journal of Forest Research 17, 6979.
  • Zeng Z.-B. (1994) Precise mapping of quantitative trait loci. Genetics 136, 14571468.
  • Zhang R. & Wienhold B.J. (2002) The effect of soil moisture on mineral nitrogen, soil electrical conductivity, and pH. Nutrient Cycling in Agroecosystems 63, 251254.
  • Zhang F., Chen G., Huang Q., Orion O., Krugman T., Fahima T., Korol A.B., Nevo E. & Gutterman Y. (2005) Genetic basis of barley caryopsis dormancy and seedling desiccation tolerance at the germination stage. Theoretical and Applied Genetics 110, 445453.

Supporting Information

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIAL AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. REFERENCES
  10. Supporting Information

Supplementary Information S1. List of the QTL detected for the different traits of response to waterlogging in a Quercus robur full-sib family. Supplementary Information S2. Position of QTL on linkage groups (LGs) of the genetic map of the male and female Quercus robur parents.QTL position was estimated by permutation and bootstrap, and confidence intervals at 0.95. Dotted lines between LG of male and female map indicate common markers that allow to compare the relative position of the QTL between the two maps.

FilenameFormatSizeDescription
PCE1629_Supplementary+Information+S1.doc124KSupporting info item
PCE1629_Supplementary+Information+S2.pdf37KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.