| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Agronomy, University of Wisconsin, 1575 Linden Drive, Madison, WI 53706 USA
b Dep. of Crop and Soil Sciences, Oregon State University, Corvallis, OR 97331 USA
c Novartis Seeds, Inc., Research Triangle Park, NC 27709 USA
d Dep. of Genetics, North Carolina State Univ., Raleigh, NC 27695-7614 USA
smkaeppl{at}facstaff.wisc.edu
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Abbreviations: AM, arbuscular-mycorrhizae • LOD, likelihood of odds ratio • QTL, quantitative trait loci • RIL, recombinant inbred line • RFLP, restriction fragment length polymorphism • SSR, simple sequence repeat
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
In addition to the inherent genetic potential of a plant to extract P from the soil, the ability to interact with microbes can also be a major determinant of the plant's ability to grow in soils with low levels of bioavailable P. Arbuscular mycorrhizal (AM) fungi are plant symbionts that are known to enhance P uptake in numerous plant species. Substantial progress is being made in work to understand the interaction of plants and AM fungi by genetic and molecular analysis (Harrison, 1997; Smith and Hayman, 1997). However, the genetic basis of variation among genotypes and among species for the plant–mycorrhizal interaction is not well understood, with little information available in maize. Studies in wheat (Triticum aestivum L.) suggest that variation for this interaction could be important in agricultural production.
Hetrick and colleagues (Hetrick et al., 1992, 1995, 1996) have published a series of papers describing genetic variation among wheat cultivars for response to mycorrhizal fungi. An interesting finding in these studies was that wheat cultivars developed before 1900 generally showed much more responsiveness to mycorrhizal fungi than cultivars developed more recently (Hetrick et al., 1995). The authors proposed that recent cultivar development in fully fertilized soil may have resulted in selection against genotypes that interact with, or respond to, mycorrhizal fungi. Since mycorrhizal fungi could actually reduce plant growth in situations where nutrients are not limiting (i.e., when the cost of maintaining the mycorrhizae exceeds the benefit to the host), it may be logical that selection under adequate fertilizer levels has selected for nonmycorrhizal genotypes. However, the potential impact of this selection could be that much of the elite germplasm in wheat may not have the alleles necessary to support mycorrhizal associations. If a similar situation exists in maize, it may be difficult to select cultivars for lower-input agriculture from the elite germplasm pool of current cultivars.
The objectives of this research were (i) to characterize variation among a set of Midwest-adapted maize inbred lines for their relative ability to grow in low P soil and to respond to AM symbiosis and (ii) to map QTL controlling differences between Mo17 and B73 for growth in low P soil and mycorrhizal responsiveness.
|
Materials and methods |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
|
The second experiment was the analysis of 197 B73 x Mo17 RILs, the two parents (B73 and Mo17), and the hybrid. These genotypes were grown in factorial combination with nonmycorrhizal–low P (M-P-), mycorrhizal–low P (M+P-), and nonmycorrhizal–high P (M-P+). The 197 lines were grown in two experiments, the first analyzing 98 lines and the second analyzing 99 lines. Each experiment used a randomized block design with three replications. The two parents and the hybrid were included in each experiment. Line means were adjusted based on the experiment means to standardize for differences in growth across the experiments. The parents and the hybrid were excluded from the analysis of genetic parameters and QTL effects.
Greenhouse Analysis Procedure
Individual plants were grown in 500-mL containers in a glasshouse maintained at 27°C day temperature and 24°C night temperature with a 14-h light and 10-h dark lighting regime. The soil for this experiment was a Marshfield silt loam (fine-loamy, mixed, frigid Typic Ochraqualf) with a pH of 7.3 and a Bray P1 soil test of 17 mg kg-1 P. The soil was pasteurized before potting, with no obvious alteration in soil structure. For mycorrhizal treatments, a mycorrhizal inoculum consisting of spores, hyphae, and infected roots generated from plants grown in this same soil was mixed with the pasteurized soil in a 1:3 inoculum/pasteurized soil mixture. This inoculum contained a mixture of five species of mycorrhizal fungi obtained from the International Culture Collection of Arbuscular and VA Mycorrhizal Fungi (INVAM), West Virginia University. The species are Glomus etunicatum UT316, Glomus claroideum IN113, G. clarum and Acaulospora mellea BR147, Glomus intraradices ON103, and Gigaspora rosea FL105. The low P treatments were fertilized with Peter's No-Phos fertilizer solution (Scott's Sierra Co., Allentown, PA) (25-0-25) on a weekly basis. The high P treatments were amended with 150 µg g-1 P at the beginning of the experiment using KH2PO4 on a w/w basis. This treatment resulted in a final Bray P1 soil test of 87 µg g-1.
Plants were harvested 
6 wk after emergence. Dry weight was measured on the total aboveground portion of the plant, and root volume of the belowground portion of the plant was measured by liquid displacement in a graduated cylinder after soil was removed by washing. A subset of roots was cleared and stained using a modification of the technique described by Phillips and Hayman (1970), in which lactoglycerin is substituted for lactophenol. Roots analyzed by staining included all plants from twelve of the inbred lines in the screening experiment and all of the recombinant inbred lines grown at low P with mycorrhizae.
Data Collected
Root volume and shoot weight were measured on each plant grown. The shoot dry weight and root volume data were used to calculate a derivative variable, mycorrhizal responsiveness. Mycorrhizal responsiveness is a percentage measurement that indicates the percentage difference in plant performance in the presence of mycorrhizae relative to the nonmycorrhizal treatment and was calculated for both levels of P as [(shoot weightmycorrhizal - shoot weightnonmycorrhizal)100]/shoot weightnonmycorrhizal (Hetrick et al., 1992). Mycorrhizal responsiveness based on root volume is the same formula substituting root volume for shoot weight. Mycorrhizal responsiveness was calculated on an individual replicate basis. Finally, the variable colonization is used to describe the percentage of roots colonized by mycorrhizae. Colonization was determined on stained roots by assessing the presence of vesicles, arbuscules, and hyphae at 100 intersects per root. Phenotypic correlations and QTL analyses were calculated using colonization data based on detection of any mycorrhizal structure (vesicle, arbuscule, or hypha) at an intersect. Colonization based specifically on the frequency of vesicles or of arbuscules is also reported for twelve inbred lines.
Data Analysis
The data from the inbred screening experiment were first subjected to a combined analysis of variance using a model that included the P and mycorrhizae factorial treatments. Inbreds, P levels, and mycorrhizal treatments were considered fixed effects and replicates were random. Since the P level x mycorrhiza interaction was highly significant, individual ANOVAs were calculated for each of the four mycorrhizae–P treatments, and these individual analyses were used to make mean comparisons among genotypes within treatments. Phenotypic correlations were calculated based on the mean values for each treatment.
Mean values across replicates were calculated for each variable for each RIL. These values were used to determine phenotypic correlations in the recombinant inbred population and to conduct a QTL analysis using PlabQTL (Utz and Melchinger, 1996). Linkages between molecular markers and QTL were determined in a two-step process. First, an analysis was done using simple interval mapping to identify significant chromosome regions with likelihood of odds ratio (LOD) >1.0. Next, the most significant marker in each peak was used as a cofactor for composite interval mapping. All chromosome regions with LOD >2.0 from the composite analysis were considered significant and included in the final model. The additive effect of a marker was calculated as [(mean of the homozygous Mo17 class - mean of the homozygous B73 class)/2]. Tests for digenic epistasis were conducted by analyzing all significant chromosome regions and two-way interactions using Proc GLM in SAS (SAS Institute, 1990). Interactions significant at 
<0.05 were reported. Only traits with multiple QTL detected were subjected to tests for epistasis. Heritability estimates for the recombinant inbred population were calculated using adjusted data across the two experiments as
, where
and
.
|
Results |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
|
Phenotypic correlations were calculated to determine whether growth at low P was simply a reflection of genetic potential for plant size. In the inbred screening experiment, shoot weight of plants grown at low P in the absence of mycorrhizae (M-P-) was not significantly correlated with shoot weight of plants grown at high P in the presence or absence of mycorrhizae (Table 2). Root volume of the M-P- treatment was significantly correlated with the M-P+, but not the M+P+ treatment. These correlations indicate that differences among genotypes grown in low P soil were due primarily to differences in P uptake and efficiency and were not a reflection of genetic potential for plant size in the inbred screening experiment. In the RIL mapping experiment, the correlation between the M-P- and the M-P+ treatments for shoot weight and root volume were low (0.38 and 0.25, respectively) but significant. This correlation indicates that measures of growth at low P may be slightly confounded with genetic potential for plant size in the RIL population.
Mycorrhizal responsiveness at low P for both root volume and shoot weight was significantly correlated with performance under the M-P- treatment, but not the M+P- treatment. Since mycorrhizal responsiveness is a function of both the M-P- and M+P- values, this correlation indicates that differences in mycorrhizal responsiveness measured among inbred lines are determined primarily by variation among genotypes for growth at low P in the absence of mycorrhizae and much less by variation among genotypes for their ability to respond to mycorrhizae. Genetic variation for shoot weight was less in the M+P- treatment than the M-P- treatment (data not shown) due to the extreme differences in growth at the M-P- treatment. Correlation of M-P- and M+P- values with mycorrhizal responsiveness in the RIL population was consistent with a similar contribution of both variation in growth at low P and genetic differences in response to mycorrhizae. In this population, M-P- measurements were significantly negatively correlated with mycorrhizal responsiveness and M+P- values were significantly positively correlated with responsiveness. Therefore, there is a contrasting interpretation of the importance of the M+P- value, a relative measure of the plant–microbe interaction, in explaining mycorrhizal responsiveness in the RIL vs. inbred screening experiment. This contrasting observation perhaps suggests that genetic variation for the ability of plants to respond to mycorrhizae may be important within specific populations, but that genetic variation for growth at low P in the absence of microbes is of most importance across a diverse set of maize genotypes.
Twelve inbreds spanning the range of values for low-P growth and mycorrhizal responsiveness were analyzed in detail for percentage of the root colonized as determined by the frequency of vesicles, arbuscules, and hyphae observed. Significant variation was observed for percentage of root colonized (Table 3) . Correlations between mycorrhizal responsiveness based on shoot weight and percentage of root colonized were not significant for either measure of colonization in plants grown in low or high P soil (Table 2). The correlation between mycorrhizal responsiveness based on root volume and colonization was significant for the low P treatment. Pa36 is a particularly unique line because it was one of the few genotypes that grew poorly at low P yet did not respond much to mycorrhizal colonization (Table 1). Interestingly, Pa36 was the genotype with the highest percentage of root colonized, significantly greater than any other of the twelve genotypes analyzed which were grown in low-P soil.
|
B73 and Mo17 were identified as the inbreds among the 28 analyzed in this study most different for mycorrhizal responsiveness, and among the most extreme for growth at low P, based on shoot weight. A B73 x Mo17 recombinant inbred population of 197 individuals was analyzed to identify QTL controlling these traits. The heritabilities for all traits were relatively low, ranging from 13% for colonization to 42% for root volume in the M-P- treatment (Table 4) . The low heritability for colonization is consistent with the observation that B73 and Mo17 were similar for this trait and transgressive segregation was not large. Low heritabilities for shoot weight and root volume are primarily due to high error variances. Some RILs were as extreme in performance as B73 and Mo17, and these parents were chosen for maximum diversity. The relatively low heritabilities indicate that only a subset of QTL will probably be detected, and the results should therefore be viewed as an incomplete determination of QTL location and effect.
|
The only QTL detected for mycorrhizal responsiveness was based on shoot weight and mapped to chromosome 2. Mo17 contributed the positive allele at this locus, even though B73 was the more responsive parent. This QTL did not map to the same chromosome region as either of the two traits used to calculate this derivative variable, shoot weight (M-P-) and shoot weight (M+P-). Quantitative trait loci detected for the high-P treatments based on shoot weight and root volume were not coincident with any of the low-P QTL detected. Digenic epistasis was not significant among the QTL detected for shoot weight (M-P-), the only trait with multiple significant QTL.
|
Discussion |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Under the conditions of our study, mycorrhizal plants grown at high P accumulated only 88% of the biomass of the same genotypes grown at high P in the absence of mycorrhizae (Table 1). This is consistent with observations of mycorrhizae-induced growth depression at high P in other systems (Graham and Eissenstat, 1992), which indicate that the cost of maintaining the mycorrhizal fungus can exceed the benefit to the host. Additional experiments are needed to determine the responsiveness of maize to mycorrhizae across P levels and under field conditions and in which production systems the AM symbiosis is beneficial, neutral, or detrimental.
The low correlation between percentage of colonization and responsiveness under low-P conditions observed in this study has also been observed by other researchers. Hetrick et al. (1996) found a relationship between colonization and biomass in responsive wheat cultivars but not in nonresponsive cultivars. This indicates that quantification of plant–mycorrhizal interactions can be highly dependent on the method used to make the evaluation. Pa36 is the inbred line most colonized by mycorrhizae but has relatively little responsiveness to mycorrhizae. This genotype is a stark representation of the different interpretations that can be made on the basis of different measures of plant–mycorrhizal interactions.
An important issue in interpreting results from our study is the proportion of genetic variation for mycorrhizal responsiveness that can be attributed to differences in the plant–microbe interaction vs. the proportion that can be attributed to the P requirement of a cultivar at a given soil P test level, that is, the degree to which a genotype can respond. Two numbers are used to calculate mycorrhizal responsiveness: plant growth at low P in the absence of mycorrhizae and plant growth at low P in the presence of mycorrhizae. In our analysis of inbred lines, the correlation between mycorrhizal responsiveness and shoot weight at low P in the absence of mycorrhizae (M-P-) was large (-0.81) and significant. The correlation between mycorrhizal responsiveness and growth at low P in the presence of mycorrhizae (M+P-) was low (0.06) and not significant. Inbreds in the screening experiment showed similar variation at M-P- and M+P-. This indicates that an inbred line's inherent genetic potential to accumulate biomass at low P in the absence of microbes is of much greater consequence in determining responsiveness than its genetic potential to benefit from the symbiosis. The interpretation of mycorrhizal responsiveness is important in plant improvement since approaches to improving P efficiency without respect to the microbe would be different than breeding for optimized plant–microbe interactions.
Hetrick et al. (1992) provided evidence that wheat cultivars developed before 1900 (before the era of extensive chemical fertilization) were more responsive to mycorrhizae than cultivars developed after 1900. These authors interpreted this observation to indicate that selection under conditions of chemical inputs selected against the plant–mycorrhizal symbiosis and that alleles which maximize the symbiosis may not be present in elite cultivars. The alternative hypothesis is that selection has increased the inherent genetic ability of recently developed cultivars to take up nutrients in the absence of microbes, and these cultivars are less responsive because they require less from the symbiosis. The data in the Hetrick et al. (1992) study do support the notion that mycorrhizal responsiveness is largely determined by growth potential at low P in the absence of mycorrhizae and that more recently developed cultivars are less responsive to mycorrhizae because they grow more efficiently at low P in the absence of mycorrhizae relative to wild relatives and cultivars selected before 1900. Therefore, one effect of selection in wheat in the last century appears to be an improvement of P efficiency that does not depend on microbial interactions.
Several arguments would support the logic of such a selection response. First, cultivars of most crops are extensively selected for environmental stability (consistent performance across environments). Microbes, such as mycorrhizal fungi, are themselves affected by environmental factors (Johnson and Pfleger, 1992; Sylvia and Williams, 1992; Ellis, 1998) and vary in density and population composition from environment to environment. Cultivars that take up nutrients well in the absence of microbes may have greater environmental stability. Second, mechanical cultivation of soil apparently disrupts mycorrhizal hyphal networks (Miller et al., 1995). Increased tillage practices may delay the rate at which germinating plant roots are colonized by mycorrhizal fungi and may dictate that plant dependence on mycorrhizae be reduced through selection. Therefore, the artificial evolutionary forces of a plant breeding program are likely to be substantially different than evolutionary forces experienced in nature. Finally, mycorrhizal establishment on plant roots takes several weeks before a beneficial symbiosis takes effect. Genotypes that are inherently better at extracting soil nutrients may show better early seedling vigor and have a selective advantage based on this trait. These hypotheses need further testing but suggest the possibility that improving plants genetically to take up nutrients regardless of which microbes are present may be of benefit to achieve optimum and stable performance in our current production systems.
Our study evaluated maize–mycorrhiza interactions based on differences in the amount of a single element, P. It is not reasonable to assume that performance differences based on a single variable explain the whole story of the interactions of maize and mycorrhizae. The QTL we have detected begin to provide a genetic basis for differences in maize–mycorrhizae interactions, but much more information needs to be collected. Mycorrhizae have been shown to be beneficial to plants in taking up nutrients such as P, Cu, Zn, and Fe; increasing plant drought tolerance; and providing benefits to the plant against pathogens (George et al., 1995; Smith and Hayman, 1997). Further genetic analysis of the maize–mycorrhizal interaction using either segregrating populations or mycorrhizal minus mutants grown across a range of environmental conditions will be necessary to fully define the importance of this interaction in maize production systems and its potential importance as reduced input systems are developed. As reduced input systems are more extensively implemented, knowledgeable management of mycorrhizal populations in soils coupled with use of P-efficient, mycorrhizal-responsive cultivars will become increasingly important. Our data suggest that such cultivars can probably be developed from the elite maize Corn-Belt dent breeding pools used to create current hybrids .DaSilva Gabelman 1993; SAS Institute Inc 1990
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
Received for publication March 1, 1999.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and methodsResultsDiscussionREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. C. Collins, F. Tardieu, and R. Tuberosa Quantitative Trait Loci and Crop Performance under Abiotic Stress: Where Do We Stand? Plant Physiology, June 1, 2008; 147(2): 469 - 486. [Full Text] [PDF]
|
|
|
|
 |
|
 M. Gonzalo, T. J. Vyn, J. B. Holland, and L. M. McIntyre Mapping Density Response in Maize: A Direct Approach for Testing Genotype and Treatment Interactions Genetics, May 1, 2006; 173(1): 331 - 348. [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 | |||
