| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
a Dep. of Agronomy, Univ. of Wisconsin, Madison, WI 53706
b USDA-ARS, Dep. of Agronomy, Iowa State Univ., Ames, IA 50011
c Dep. of Agronomy, Iowa State Univ., Ames, IA 50011. This journal paper of the Iowa Agriculture and Home Economics Experiment Station, Ames, IA, Project no. 3755, was supported by the Hatch Act and State of Iowa funds
* Corresponding author (alorenz{at}wisc.edu).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Abbreviations: H, broad-sense heritability • lpa, low phytic acid • Pi, inorganic phosphorus
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Received for publication March 10, 2007.
a Dep. of Agronomy, Univ. of Wisconsin, Madison, WI 53706
b USDA-ARS, Dep. of Agronomy, Iowa State Univ., Ames, IA 50011
c Dep. of Agronomy, Iowa State Univ., Ames, IA 50011. This journal paper of the Iowa Agriculture and Home Economics Experiment Station, Ames, IA, Project no. 3755, was supported by the Hatch Act and State of Iowa funds
* Corresponding author (alorenz{at}wisc.edu).
Seed P is predominantly bound in the organic compound phytate, which makes the bioavailability of P low for monogastric animals fed maize (Zea mays L.)-based diets. Decreasing phytate and increasing inorganic P (Pi, an available form of P) concentrations in maize grain would be desirable to help ameliorate environmental problems associated with high P in feces. Our objective was to investigate the potential of improving the P profile of maize grain through breeding and selection. Ninety S1 families from the BS31 population were evaluated at two locations for phytate, Pi, and other grain quality and agronomic traits. Phytate concentrations ranged from 1.98 to 2.46 g kg–1, and the broad-sense heritability (H) was relatively low (0.60). Both genetic variance and H (0.84) were much greater for Pi. Few unfavorable genetic correlations were observed between either Pi or phytate and other key economic traits. Also, selection differentials of multiple trait indices indicated that the P profile of maize grain and grain yield and moisture could be improved simultaneously. Many cycles of selection will be needed, however, to reach desirable phytate and Pi concentrations, especially when selecting for multiple traits. Regardless, our results are encouraging given that the families evaluated were related S1 families and the number of families was relatively small.
Abbreviations: H, broad-sense heritability • lpa, low phytic acid • Pi, inorganic phosphorus
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
It would be desirable to increase the concentration of inorganic P (Pi), an available form of P, and reduce the concentration of phytate in maize grain to help alleviate the associated nutritional and environmental problems simultaneously. Low phytic acid (lpa) mutant lines, which have up to a 66% reduction in phytate P and a molar equivalent increase in Pi (five- to 10-fold), have been developed in a variety of crops, including maize (Raboy et al., 2001). Animal feeding trials have shown that P availability is greatly improved in lpa maize compared with wild-type maize, leading to substantial decreases in feces P concentration (reviewed in Knowlton et al., 2004). Reductions in yield potential have been reported, however, for lpa maize (Ertl et al., 1998) as well as for lpa wheat (Triticum aestivum L.) (Guttieri et al., 2006). Pilu et al. (2005) showed that disruption of the gene that causes the lpa1 mutation (the most common lpa mutant) has negative pleiotropic effects on kernel size and germination and thus breeding lpa cultivars equal in yield potential to their wild-type counterparts may not be possible.
Another approach to enhancing the P profile (decreased phytate, increased Pi) in maize grain is through recurrent selection that uses the indigenous quantitative genetic variation for these traits. Significant intraspecific genetic variation for both phytate and total P has been observed in wheat (Raboy et al., 1991), soybean (Israel et al., 2006), dry bean (Phaseolis vulgaris L.) (Lolas and Markakis, 1975) and oat (Avena sativa L.) (reviewed in Maga, 1982). In maize, Raboy et al. (1989) found a 2.5-fold difference in phytate concentration between Illinois high and low protein lines in their 83rd generation of divergent selection. Wardyn and Russell (2004) estimated a broad-sense heritability (H) of 0.82 for total P measured in a maize population of S1 families. Both Raboy et al. (1989) and Wardyn and Russell (2004) concluded that P-related traits in maize could be easily modified through selection. No studies, however, have investigated the genetic variability of both phytate and Pi within a single maize breeding population to determine the potential of enhancing the P profile of maize grain through selection on both components simultaneously. Furthermore, phytate and Pi genetic variability should be studied using methods suitable for maize breeding projects, such as automated sampling during yield trial harvest and simple, rapid protocols for analysis of grain components.
Selection on single traits, such as insect and disease resistance, typically results in unfavorable changes in yield and other important agronomic traits (Hallauer et al., 1988). Yield, moisture, and root and stalk lodging are traits valued by producers and should be included in selection indices to prevent compromising the overall value of the germplasm. Therefore, studying the potential for improving the P profile of maize grain should be done in the context of multiple trait selection.
The objective of our research was to determine the potential of improving the P profile of maize grain alone and along with other economically important traits. This was accomplished by measuring several traits, including phytate and Pi, in a population of S1 families and estimating variance components and H. Genetic correlations were calculated in addition to selection differentials obtained from various multiple-trait selection indices. Grain quality traits were measured on samples obtained through mechanical sampling of yield trial plots. The phytate and Pi laboratory protocol we used is relatively rapid and well suited for breeding (Lorenz et al., 2007).
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
454-g grain sample was obtained from each plot by an automated sampler during machine harvest. A subsample consisting of 30 whole kernels was taken from each field plot sample and milled until 70% of the millings passed through an 80-µm mesh screen. Data collected on plots were machine-harvestable grain yield (megagrams per hectare, adjusted to 155 g kg–1 moisture), grain moisture (grams per kilogram), root lodging (percentage of plants leaning more than 30° from vertical), stalk lodging (percentage of plants broken at or below the primary ear node), test weight (kilograms per cubic meter), and silk emergence (days after planting in which 50% of the plants had emerged silks at Ames, IA). Data collected on field plot samples included kernel weight (grams per 250 whole kernels) and protein, oil, and starch (grams per kilogram; all predicted with near infrared spectroscopy of whole grain).
Phytate and Inorganic Phosphorus Analysis
The colorimetric assays of Vaintraub and Lapteva (1988) and Raboy et al. (2000) were modified as described by Lorenz et al. (2007) and used to measure phytate and Pi on the 30-kernel ground samples. Measurements were performed in triplicate by subsampling ground samples three times and blocking plots from each field replication on 96-well microtitre plates. Six standards were included on each phytate and Pi evaluation plate. Field plots and standards were randomized on each microtitre plate as a resolvable row–column design (John and Williams, 1995). The optical density of colorimetric reactions was quantified with a 96-well spectrophotometer. A single standard curve was developed within each field replication by regressing optical density means on known standard concentrations. The standard curves were used to predict phytate and Pi concentrations in each microtitre plate well. Standard curve R2 values were >0.99.
Statistical Analysis
The number of microtitre plate measurements taken on phytate and Pi was 1080 each. Data from microtitre plate measurements were subjected to an outlier analysis according to Anscombe and Tukey (1963), and outliers were removed from the data sets. The number of outliers removed for phytate and Pi were 14 (1.3%) and 10 (0.9%), respectively. Inorganic P concentration field plot least-squares means were calculated by fitting a mixed model including field plot (fixed) and plate (random). A similar model was used to calculate phytate concentration least-squares means of field plots with the exception of fitting row and column within plate as random effects due to the amount of variation accounted for by these variables. Phytate and Pi field plot least-squares means were subsequently treated as single measurements.
Field plot measurements were fit to a model that included family (random), location (fixed), family x location interaction (random), and field replication nested within location (random). Variance components were estimated using the MIXED (REML) procedure of SAS (SAS Institute, 2003) and tested for significance by the Wald z-test. Broad-sense heritability estimates on an S1 family mean basis were calculated as
 |
F2 is the family variance, FL2 is the family x location interaction variance, e2 is the residual variance, l is the number of locations (l = 2), and r is the number of replications within locations (r = 2). The CV was calculated by dividing the square root of the residual variance by the experiment mean and was expressed as a percentage. Family means across locations were used to calculate phenotypic correlations. Genotypic correlations were calculated by equating mean cross products with their expected values (Bernardo, 2002).
Selection Indices
Selection differentials based on 10% selection intensity were calculated from a series of indices involving phytate, Pi, grain yield, grain moisture, root lodging, and stalk lodging. To avoid problems of scale, family means were transformed to standard deviation units before index values were calculated. Selection differentials are reported in the original units. The heritability index (Smith et al., 1981) and rank summation index (Mulamba and Mock, 1978; reviewed in Hallauer et al., 1988) were used to calculate family index values for various trait sets. The heritability index and rank summation index represent weighted and unweighted indices, respectively, and result in different selection differentials due to differences in H. Trait sets and their selection method are individually described here.
Families were also selected based on phytate and Pi concentrations alone (i.e., the nine families with the lowest phytate concentration were selected and the nine families with the highest Pi concentrations were selected separately to produce two sets of selection differentials). Selection differentials were calculated by subtracting the population mean from the mean of the nine selected families for each trait.
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
Few statistically significant correlations were found between phytate or Pi and all other traits. As in many previous reports (Lorenz et al., 2007; Raboy et al., 1989, 1991), phytate was positively and significantly correlated with protein (0.37 phenotypic, 0.51 genotypic). A negative correlation occurred between phytate and starch (–0.33 phenotypic, –0.45 genotypic). This result reflects the seed deposition pattern of phytate, where approximately 90% of phytate is found in the germ (reviewed in Maga, 1982) and only trace amounts are in the endosperm. We expected a larger correlation given such a large discrepancy in phytate levels between these seed organs. The pattern of phytate deposition in the kernel was used to partially explain a negative correlation between phytate and kernel size in a previous experiment (Lorenz et al., 2007), but no significant correlation was found between phytate and kernel size within this population of S1 lines. This may be due to less genetic variation for these traits among S1 families than inbred lines derived from multiple backgrounds. The phenotypic and genotypic correlations between phytate and Pi were 0.05 and 0.15, respectively, which closely agree with the correlation between these P traits reported in Lorenz et al. (2007). In contrast, a lack of relationship between phytate and Pi differs with a positive correlation found among wild-type soybean lines (Israel et al., 2006) and the repartitioning of total P by the lpa mutations (Raboy et al., 2000). The lpa mutation confers a biochemical lesion in the phytic acid pathway and is inherited as a qualitative trait. Thus, the effects of the lpa mutation cannot be compared with the natural genetic variation of the phytate/Pi ratio. When measurements are taken among wild-type cultivars and breeding families, the correlation between phytate and total P is typically >0.90 (Raboy et al., 2001). This suggests that selection for reduced phytate would decrease total P without repartitioning the P bound in phytate to Pi, an undesirable outcome considering animal nutrition. Our findings reported here and those in Lorenz et al. (2007), however, suggest that phytate and Pi are independent and an improved P profile could be achieved through selection on both traits simultaneously.
Our results on selection indices that included phytate, Pi, grain yield, grain moisture, stalk lodging, and root lodging indicated that progress, albeit slow, could be made for each trait (Table 2 ). Selection differentials could have been increased, with no additional post-harvest labor, by evaluating more families for agronomic traits and truncating the best families before laboratory analysis. Truncating families once with an index including phytate, Pi, yield, and moisture did not achieve greater selection differentials for phytate and Pi compared to the "All" index, but greater pressure was placed on yield. Selection differentials calculated when selection was applied to phytate and Pi only indicates that the P profile could be altered in the desired direction, but yield and moisture are expected to decrease and increase, respectively. The differences observed between the heritability index and rank summation index were expected, with the heritability index placing more pressure on Pi. Realized gains in overall index value would probably be greater for the heritability index due to more emphasis being placed on traits with higher H, but it may not put enough pressure on phytate for significant reduction in the near future. Altogether, selection differentials for phytate were between 2.2 and 7.6% of the population mean when phytate was included in any of the selection methods used. On the other hand, selection differentials for Pi were between 16.2 and 43.2% of the population mean. Furthermore, a greater proportion of the selection differential would be expected to be inherited for Pi. Protein would be expected to decrease while starch is expected to increase for each selection method except when selection is applied to Pi only. An increase in seed phytate through direct selection on protein in maize has been reported (Raboy et al., 1989).
|
On average, swine need 2.35 g Pi kg–1 dietary dry matter to meet their nutritional requirements (Knowlton et al., 2004), and phytate levels should be as low as possible to minimize P pollution. Obviously, many cycles of selection would be needed to attain these goals, but the variation found among related S1 families suggests genetic gain is possible, especially if a larger number of more diverse families is considered for selection and breeding. Nevertheless, any improvement in the P profile of maize grain would have a beneficial impact on the environment and farm economy.
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.
Received for publication March 10, 2007.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
