| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Division of Plant Sciences, Univ. of Missouri-Columbia, Columbia, MO 65211
b Crop and Soil Environmental Sciences Dep., Virginia Polytechnic Inst. and State Univ., Blacksburg, VA 24061-0404
* Corresponding author (mckendrya{at}missouri.edu).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Abbreviations: AFLP, amplified fragment length polymorphism • CIM, composite interval mapping • FDK, Fusarium damaged kernels • FHB, Fusarium head blight • DON, deoxynivalenol • MIM, multiple interval mapping • QTL, quantitative trait loci • RILs, recombinant inbred lines • SSR, simple sequence repeat
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Mesterházy (1995) extended the components of resistance originally proposed by Schroeder and Christensen (1963) to include components related to resistance to both DON accumulation and kernel infection. Today, sources of low DON and kernel quality retention are available to wheat breeders developing FHB-resistant varieties in several wheat backgrounds including those of Chinese descent such as ‘Sumai 3’ (Bai and Shaner, 2004), ‘Wangshuibai’ (Ma et al., 2006), and ‘W14’ (Bai et al., 2001; Chen et al., 2006) as well as those in European (Mesterházy et al., 1999) and North American (Bai et al., 2001; McKendry et al., 1995, 2005, 2007) germplasm. However, selection strategies for these traits have not been well investigated.
The possibility of indirectly selecting wheat lines with low DON accumulation through selection for types I and/or II FHB resistances under field conditions has been demonstrated but the interrelationships among these resistance factors and DON have been inconsistent. Arseniuk et al. (1999) found significant deviations from linearity when FHB symptoms were regressed on DON concentration and concluded that the mechanisms for FHB resistance and DON accumulation were independent in wheat. Mesterházy et al. (1999) determined that among highly resistant cultivars, DON levels closely paralleled visual resistance symptoms; however, variation did exist in moderately resistant cultivars that had low visual symptoms but higher DON levels and in moderately susceptible lines, where DON accumulation was lower than expected. Bai et al. (2001) reported similar findings among 116 breeding lines and cultivars. Measures of percentage of scabby spikelets and seed quality were highly correlated and each was also correlated with DON in both greenhouse and field studies, but again exceptions were observed among cultivars with moderate levels of resistance or susceptibility. Conclusions of both studies were that breeders could indirectly select for low DON by selecting for high levels of FHB resistance using visual selection parameters (Mesterházy et al., 1999; Bai et al., 2001). Using both spring and winter populations, Wilde and Miedaner (2006) confirmed that phenotypic selection for high levels of FHB resistance resulted in a significant reduction in DON.
Despite their potential importance for FHB-resistance breeding, there are few studies in the literature to date reporting quantitative trait loci (QTL) associated with low DON, particularly in germplasm adapted to the United States. Additionally, the interrelationships among QTL associated with types I and II resistance and low DON have been inconsistent and their relationship with kernel quality retention remains largely unknown.
Somers et al. (2003) were among the first to directly report QTL associated with DON accumulation using double haploid lines derived from the cross of ‘Wuhan’/‘Maringa’. Although the range of DON was relatively small in this population, they reported significant QTL on chromosomes 2DS, 3BS, and 5AS. The QTL on 2DS and 5AS were associated with low DON and were independent of FHB severity, while that on 3BS appeared to condition both DON and type II FHB resistance. Lines with favorable alleles on 3BS and 5AS reduced DON accumulation by 17%. Their data suggested that low DON was partially independent of type II resistance.
Partial independence of QTL effects for DON and types I and II resistance was also reported for Wangshuibai. Using a set of recombinant inbred lines (RILs) developed from the cross Wangshuibai/‘Annong 8455’, Ma et al. (2006) reported three QTL on chromosomes 5A, 2A, and 3B from Wangshuibai that were associated with low DON. Although QTL effects for both DON and type II resistance were interrelated, the magnitude of the effects differed for each trait. The 3B QTL, which is probably allelic to that in Sumai 3, showed the largest effect on type II resistance (R2 = 17%), but had a relatively small effect on DON (R2 = 6.2%), while that on 5A had a relatively large effect on DON (R2 = 12.4%) but was not consistently significant for type II resistance. The 2A QTL was intermediate and impacted both severity (R2 = 11.5%) and DON (R2 = 8.5%). Semagn et al. (2007) also reported partial independence of QTL for FHB and DON in a mapping population derived from a cross between the European resistant line ‘Arina’ and the Norwegian breeding line ‘NK94603’. Two QTL, on 1AL, and 2AS, explained 27.9 and 26.7% of the phenotypic variation for DON, respectively. While the 1AL QTL impacted both FHB resistance and DON, that on 2AS was only associated with low DON.
In the Sumai 3 genetic background, however, DON and FHB do not appear to be independent. Lemmens et al. (2005) identified a single QTL on 3BS associated with reduced DON in a set of doubled haploid lines developed from the cross of ‘CM82036’ and the susceptible parent ‘Remus’. This locus, which also conditions type II resistance in Sumai 3 (Waldron et al., 1999), accounted for 92.6% of the observed variation in DON.
Kernel infection is an important component of resistance (Mesterházy, 1995) impacting the marketability of grain infected with FHB; however, the only known report of QTL associated with kernel infection is that of Chen et al. (2006). In a greenhouse validation study of the 3BS and 5AS QTL identified in the Chinese landrace W14 they found that these QTL had a significant impact on type II resistance, DON content, and kernel infection, together explaining 33, 35, and 31% of the phenotypic variation in these traits, respectively. This study agrees with earlier findings of a strong interdependence of resistance factors in Chinese germplasm carrying the 3BS QTL.
Ernie, a soft red winter wheat developed and released by the University of Missouri (McKendry et al., 1995) has moderately high type II resistance, coupled with low DON and good kernel quality retention under FHB pressure. It serves as the early resistant check in both the U.S. Northern and Southern Winter Wheat Scab Nurseries coordinated through the U.S. Wheat and Barley Scab Initiative. As a breeding line, Ernie has the added value of being adapted to the U.S. soft red winter wheat region, thus breeding populations may be developed using Ernie that do not suffer from the lack of adaptation associated with the use of Sumai 3 and its derivatives. Recently, QTL were identified that were associated with the type II resistance in Ernie (Liu et al., 2007) and it was determined that the FHB resistance in Ernie differs from that in Sumai 3. However, low DON and kernel quality retention in Ernie have not been mapped. The objectives of this study were to identify QTL associated with low DON and kernel quality retention and to determine whether interrelationships exist among these QTL and those identified for type II resistance in this line.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Phenotypic Evaluation
Experiments were conducted in the greenhouse environment to ensure precision in inoculation technique, a consistently high level of disease pressure and the absence of confounding effects that can often impact field data. Although we acknowledge that this work needs to be verified in the field environment, we elected to conduct this initial work in the greenhouse under more controlled environmental conditions.
Eight plants per RIL per replication were arranged in a randomized complete block design with three replications in 2002 and four replications in 2003. Within replication, RILs were randomized by line. Procedures for inoculation are described in Liu et al. (2005). Infected heads from each plant were harvested at maturity and hand-threshed to ensure all diseased kernels were collected. Kernels from each of the eight plants per replication were separated visually into two groups consisting of sound kernels and a second group of FDK that included from slightly to highly shriveled kernels and tombstones. The number of kernels in each group was counted to precisely determine the proportion of FDK in the infected head. Percentage FDK was determined as the total number of kernels showing FHB symptoms divided by the total number of kernels in the inoculated head, expressed as a percentage. Kernels from the eight heads per RIL within each replication were then bulked to ensure an adequate sample for DON determination. Bulked seed was ground using a Braun coffee grinder and analyzed for DON content at the U.S. Wheat and Barley Scab Initiative–funded laboratory at Michigan State University in East Lansing, MI. Deoxynivalenol content was quantified in micrograms per gram (µg g–1) using the mycotoxin extraction kit Veratoxin for DON 5/5 (Veratox, Lansing, MI).
Molecular Marker Analysis
The map for this population was developed by Liu et al. (2007). A total of 420 SSR markers including Xgwm and Xbarc markers (Röder et al., 1998; Plaschke et al., 1995; Song et al., 2005) and 64 EcoRI/MseI AFLP primer pairs were used to identify polymorphic loci between the two parents, Ernie and MO 94-317. For both DON and FDK, 162 AFLP and 100 SSR polymorphic loci were used to construct the genetic linkage groups.
Statistical Analysis
Before conducting the analyses of variance (ANOVA) data for each trait were tested for normality using PROC UNIVARIATE NORMAL PLOT (SAS Institute, 2006). Deoxynivalenol data were transformed using a log (x + 1) transformation where x represented DON (µg g–1). Error variances were tested for homogeneity using Bartlett's test to determine whether or not data could be combined across experiments. Analyses of variance were conducted for both traits on data from individual years and on data combined over years using PROC MIXED (SAS Institute, 2006). The statistical model used considered RILs fixed while years, replications, and the year x RIL interaction were considered random effects. Variation due to years and RILs was tested for significance against the RILs x year interaction while the RILs x year interaction was tested against the experimental error. Significance levels for ANOVA were declared at the 5% level of probability. Correlation coefficients among FHB-resistance traits were estimated using PROC CORR (SAS Institute, 2006). Broad-sense heritability (H2BS) for log DON and FDK were estimated from the ANOVA for each individual year and for data combined over years. Heritability was determined as: H2BS = 
2g/(2g + 2gxy/y + 2e/ry), where 2g is the genetic variance among RILs, 2gxy is variance due to genotype x year interaction, 2e is variance due to error, and r and y are replications and years, respectively. Exact 95% confidence limits of the estimated heritability were calculated following the procedure of Knapp et al. (1985). The minimum number of genes was estimated using Cockerham's (1983) modification of Wright's (1968) formula.
Linkage Map Construction and QTL Analyses
Linkage maps were constructed using MapMaker 3.0 (Lander et al., 1987). Initial chromosome locations of SSRs were determined based on the reference map of the International Triticeae Mapping Initiative (Röder et al., 1998; Song et al., 2005). The Kosambi mapping function was used to estimate the distance between markers (Kosambi, 1944). Markers were grouped using a logarithms of odds (LOD) value of 3.0 and distance <37 cM. The QTL model used included additive and additive x additive interactions. Because the mapping population was comprised of inbred lines, dominance and dominance-associated genetic interaction were considered negligible. The QTL analyses were conducted using composite interval mapping (CIM) with WinQTLCart 2.5 (Wang et al., 2006). For preliminary analyses of data for individual years, a LOD score of 2.5 was used to declare significant QTL. Across years, threshold LOD scores to declare significant QTL for DON and FDK were determined based on 1000 permutations (Doerge and Churchill, 1994). Cofactors were identified using forward and backward regression. When multiple peaks were found within a single marker interval, the location with the highest LOD score was defined as the QTL peak. A one-LOD drop from the peak position was used as a confidence interval for each QTL location. If another QTL peak appeared 20 cM away from the previous, it was claimed as a separate QTL. Coefficients of determination (R2) of significant QTL were obtained from multiple interval mapping (MIM) from WinQTLCart 2.5 using all significant QTL from CIM (Wang et al., 2006). Interactions between significant QTL were analyzed initially with ANOVA and then using MIM. All possible pairs of marker combinations were tested for significance using PROC GLM in SAS (SAS Institute, 2006) and significant interaction terms with R2 > 5% were included in the model. Phenotypic effects of combinations of the QTL alleles of the two parents were estimated for both DON and FDK using the closest marker for QTL that were significant across the two experimental years. Percentage reductions in DON and FDK due to QTL allele combinations were determined.
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
|
|
Estimated heritabilities varied from 59 to 82% for DON and from 73 to 82% for FDK with mean heritabilities across years of 74% and 78% for DON and FDK, respectively (Table 2). Because these data were determined from inbred lines, and significant transgressive segregation was observed in the population, we concluded that these heritability estimates reflected largely additive genetic variance. Three to four genetic factors conditioned both traits (Table 2). These estimated gene numbers are consistent with those reported earlier for type II resistance in Ernie (Liu et al., 2005).
Correlation coefficients between DON and FDK both within and across years are given in Table 3 . Despite higher disease levels in 2003, DON was significantly correlated (r = 0.41) across years as was FDK (r = 0.63). Deoxynivalenol was highly correlated with FDK in both 2002 (r = 0.75) and 2003 (r = 0.70) and each was highly correlated with the FHB severity data from this population reported earlier (Liu et al., 2007). These results indicate the interdependence of both traits and their relationship with FHB severity in an Ernie genetic background. Although deviations from linearity occurred, our results are consistent with those of Bai et al. (2001) who reported significant correlations under greenhouse conditions between FHB ratings, DON accumulation, and visual seed quality traits in a diverse set of germplasm representing major sources of FHB resistance. Similar correlations have also been reported under field, spray-inoculated conditions (Mesterházy et al., 1999; Arseniuk et al., 1999; Miedaner et al., 2003; 2004). The most thorough report on these interrelationships was a meta-analysis of 163 studies conducted by Paul et al. (2005) in which they found highly significant correlations between DON and FDK and between each of these traits and severity of FHB in the diseased head. Our data suggest that in Ernie there is a somewhat stronger relationship between FDK and FHB severity (r = 0.86 in both 2002 and 2003) than between DON and FDK (r = 0.68 and 0.64 in 2002 and 2003, respectively). The high heritability of both DON, and FDK as well as that for FHB severity (Liu et al., 2007) along with the relatively equal number of genetic factors controlling all three traits suggest similar genetic control of these resistance factors in Ernie. However, further study is needed before this statement can be made conclusively.
|
Based on data combined over years, three regions were identified on chromosomes 3B, 4B, and 5A that were associated with reduced DON (Table 4 , Fig. 2 ). Across years, these three QTL explained 31% of the total phenotypic variation in DON content. Quantitative trait loci effects were predominantly additive and all alleles were from the resistant parent Ernie. A forth QTL on 2B was significant in 2002 but failed to reach significance in 2003 or when data were combined across years.
|
|
The 4BL QTL, which is flanked by Xgwm495 and Xgwm149 (Fig. 2) accounted for 7% of the variation in DON. These flanking markers are 5.5 cM apart and the QTL peak (0.1 cM) is 0.1 cM from Xgwm495 suggesting that this marker once validated, could be immediately useful for marker-assisted selection for low DON. As was the case with the 3BSc QTL, the 4BL QTL in Ernie is coincident with the 4BL QTL, Qfhs.umc-4BL, previously reported for type II resistance in this cultivar (Liu et al., 2007). To our knowledge, a 4BL QTL has not been reported for DON content. Somers et al. (2003) reported a 4BL QTL linked with Xwmc238 that was associated field FHB resistance in the cross Wuhan 1/Maringa, but failed to detect this QTL associated with DON.
Finally, the 5A QTL, associated with Xbarc56 on 5AS accounted for 10% of the variation in DON. The flanking markers, Xbarc56 and Xbarc165 are approximately 21 cM apart and Xbarc56 is 14 cM from the QTL peak (45 cM). Again, this QTL was coincident with a 5AS QTL for type II resistance in Ernie (Liu et al., 2007). Although not conclusive because of the low marker density in this region the data again suggest the interdependence of visual FHB symptoms and DON in this cultivar. Additional markers are being developed to saturate this region and refine this QTL position. The 5AS chromosome region has been found to be involved in FHB resistance in widely diverse germplasm. Quantitative trait loci for type I or field resistance in CM82036 (Buerstmayr et al., 2003), Renan (Gervais et al., 2003), ‘Frontana’ (Steiner et al., 2004), W14 (Chen et al., 2006), and Wangshuibai (Lin et al., 2006) have been detected in this region as well as for type II resistance in ‘Fundulea 201 R’ (Shen et al., 2003), and Ernie (Liu et al., 2007). Our data for Ernie agree with those of Chen et al. (2006), who found interdependence among types of resistance conferred by alleles in the 5AS region. In their study of W14, a 5AS QTL was associated with reduced field incidence, severity, and DON. Flanking markers for this QTL (Xbarc117 and Xbarc186) suggest that this QTL may be the same as that for DON and type II resistance in Ernie, although the QTL effect is somewhat lower. These data, however, conflict with those of Somers et al. (2003) who reported interdependence of DON and type II resistance at the 3BS QTL but not at the 5AS QTL. The latter allele from Maringa had no effect on either type II or field resistance but reduced DON content by 36%. Although the 5AS QTL in Ernie was in the same region as that in Maringa its linkage with Xbarc56 positioned it slightly more distally than that in Maringa, which was linked to Xgwm96 (Somers et al., 2004). Thus, it is unclear whether or not they reflect the same or different genes. Ma et al. (2006) also reported a 5A QTL that was associated with reduced DON in Wangshuibai. However, based on the consensus map (Somers et al., 2004) this QTL, which is linked to Xgwm156, is on the long arm of chromosome 5A, and therefore appears to differ from the 5AS QTL reported here.
A significant epistatic interaction between QTL alleles on 3B and 5A was detected using ANOVA (P < 0.05) that accounted for 8% of the variation in DON. Using the epistasis function in CIM, an interaction between these two alleles was also detected, however, it failed to reach the LOD threshold for significance (LOD = 2.8) and therefore was not reported in Table 4.
For FDK, CIM revealed four chromosome regions on 2B, 3BSc, 4BL, and 5AS, which together over years accounted for 46% of the observed variability in kernel quality. Except for the 2B QTL, all QTL were consistently detected over years. The 2B QTL was significant in both 2002 and in data combined over years but failed to meet the significance threshold in 2003. As with DON, these QTL exhibited predominantly additive effects and QTL x QTL interactions were not significant. All QTL alleles were from the resistant parent Ernie.
The three major QTL on 3BSc, 4BL, and 5AS accounted for 18, 6, and 18% of the observed variability in FDK, respectively (Fig. 2; Table 4). These QTL, linked to Xgwm285, Xgwm495, and Xbarc56 were coincidental with those for DON. They also co-localized with QTL in the same regions for type II resistance (Liu et al., 2007), which confirmed the interdependence of these sources of resistance in Ernie. Reports of QTL for FDK are limited in the literature and to our knowledge this is the first report of QTL for kernel quality on 3BSc and 4BL. Our data for the 5AS QTL are consistent with those of Chen et al. (2006) who reported a QTL in W14 in the same region of 5AS that confers reduced kernel infection. Chen et al. (2006) also found that this QTL conferred type II resistance and reduced DON, although its effect on greenhouse data was smaller than that observed for Ernie (Liu et al., 2007).
The centromeric 2B QTL had a smaller effect on FDK (R2 = 4%) and although it was in the same chromosome region as that for DON, was linked to Xgwm276b with the QTL peak estimated to be 8.2 cM proximal to that for DON. We were unable to resolve whether or not this is the same allele as that for DON, however, the data suggest that these two QTL are coincidental. A 2B QTL with a small effect, Qfhs.umc-2B, linked to Xgwm276b has also been reported for type II resistance in Ernie (Liu et al., 2007). We believe this is the first QTL on 2B reported for FDK.
Both ANOVA and CIM were again used to detect epistatic interactions among QTL alleles. A significant three-way interaction (P < 0.05) among QTL alleles on 2B, 3B, and 5A was detected using ANOVA, which accounted for 6% of the genetic variance, however, this interaction was not detected with CIM. Because of the inconsistency across analytical techniques, this interaction is not presented in Table 4.
Phenotypic Effects of QTL Allele Combinations
To examine allelic effects of QTL for each trait, phenotypic effects of allele combinations were compared using combined data for RILs carrying resistant or susceptible alleles from Ernie (E) or MO 94-317 (M), respectively (Table 5
). For DON, effects of significant QTL on 3BSc, 4BL, and 5AS were considered individually and then in combination with other desirable alleles from Ernie based on associated markers with the largest QTL effect. Variance in the DON data was high as expected, yet significant effects of desirable alleles were observed, which indicate the additive effect of these three QTL alleles and their combinations. Individually, the 5AS, 4BL, and 3BSc QTL alleles reduced DON by 11.8, 16.0, and 20.8%, respectively compared with RILs carrying alleles from MO 94-317 (Table 5). Where lines carried two of the favorable alleles, the combination of QTL on 3BSc and 5AS had the largest effect, reducing DON by 47% while three favorable alleles from Ernie reduced DON by 78% compared to RILs carrying none of the favorable alleles. The magnitude of the effect of 3BSc and 5AS combination may in part reflect an interaction between these two QTL alleles that was detected by CIM but did not reach the LOD threshold required for significance. Although QTL for DON on 3BSc and 4BL have not been previously reported, a similar analysis by Somers et al. (2003) revealed a 6% reduction in DON associated with a 5AS QTL from Maringa that may reflect the same gene as that in Ernie. Whether or not this difference is meaningful is questionable because of the lower range of DON in the Somers study, and the large variance associated with the DON levels themselves.
|
For FDK, four significant QTL were identified in data combined across years. In an analysis similar to that conducted for DON, individual and combined allelic effects were determined (Table 5). Mean FDK of RILs carrying only those alleles from the susceptible parent (M1M1; M2M2; M3M3; M4M4) was 71.8% whereas that for RILs carrying Ernie alleles for all four QTL (E1E1; E2E2; E3E3; E4E4) was 20.8%. With the inclusion of the 2B QTL, which was significant in the combined data, this range more closely paralleled that of the parents than was the case with DON. Individually, the 3BSc QTL allele had the largest effect on kernel quality, reducing FDK by 28%, while the 4BL QTL allele had the smallest effect, reducing FDK by 10.7%. Interestingly, when these two alleles were combined, either with other susceptible alleles or in combination with either or both of the 2B and 3BSc alleles from Ernie their effect appeared to be synergistic. No significant interactions between QTL alleles on 2B and 3BSc, however, were detected by either ANOVA or CIM. When combined with the 5AS allele, FDK was reduced by 56% and where RILs contained all four favorable alleles from Ernie, FDK was reduced by 71% compared to RILs carrying only the susceptible alleles. This data set further suggests the additive effects of these QTL combinations on FDK.
In summary, these results agree with those of Liu et al. (2007) that suggest resistance in Ernie differs from Sumai 3 and other sources of resistance commonly in use in U.S. breeding programs. On-going research will seek to verify these QTL in the field environment and validate them in Ernie-derived populations near-isogenic for these QTL alleles. Additionally, we are working to add markers to the map to saturate regions to refine QTL positions and identify markers linked more closely to the QTL peak. Once this research is completed these markers should be a valuable resource for marker-assisted selection. These data, along with those of Liu et al. (2007), indicate that Ernie will complement other known sources of FHB resistance in breeding programs aimed at pyramiding genetically different FHB alleles into winter wheat varieties. Co-localization of QTL for DON and FDK with those for type II resistance suggests that in this cultivar, resistances to FHB are interdependent. From a breeding perspective, therefore, selection for QTL for type II resistance in Ernie-derived populations should concurrently result in low DON and improved kernel quality retention.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Received for publication January 18, 2008.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Liu, M. D. Hall, C. A. Griffey, and A. L. McKendry Meta-Analysis of QTL Associated with Fusarium Head Blight Resistance in Wheat Crop Sci., October 22, 2009; 49(6): 1955 - 1968. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Bonin and F. L. Kolb Resistance to Fusarium Head Blight and Kernel Damage in a Winter Wheat Recombinant Inbred Line Population Crop Sci., June 26, 2009; 49(4): 1304 - 1312. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
