Crop Science Grow Your Career with CSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 31 January 2005
Published in Crop Sci 45:516-523 (2005)
© 2005 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Feil, B.
Right arrow Articles by Stamp, P.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Feil, B.
Right arrow Articles by Stamp, P.
Agricola
Right arrow Articles by Feil, B.
Right arrow Articles by Stamp, P.
Related Collections
Right arrow Irrigation
Right arrow Maize
Right arrow Plant Analysis
Published in Crop Sci. 45:516-523 (2005).
© 2005 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA

CROP ECOLOGY, MANAGEMENT & QUALITY

Mineral Composition of the Grains of Tropical Maize Varieties as Affected by Pre-Anthesis Drought and Rate of Nitrogen Fertilization

B. Feila,*, S. B. Mosera, S. Jampatongb and P. Stampa

a Inst. of Plant Sciences, ETHZ, Universitätstr. 2, CH-8092 Zurich, Switzerland
b National Corn and Sorghum Res. Center (Farm Suwan), Kasetsart Univ., 298 Moo 1, Klandong, Pakchong, Nakhon Ratchasima 30320, Thailand

* Corresponding author (boy.feil{at}ipw.agrl.ethz.ch)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Breeding for higher concentrations of minerals in food crops is one option for improving the health of humans suffering from the consequences of mineral deficiency. The plant breeding approach requires that varietal differences are stable across different environmental conditions. The main objective of our research was, therefore, to determine whether differences in the concentration of grain minerals (P, K, Mg, Ca, Mn, Zn, and Cu) among tropical maize varieties are affected by the level of water and N supply. A 3-yr study with two water regimes (preanthesis drought vs. irrigation throughout the vegetation cycle), three levels of N fertilization (0, 80, 160 kg N ha–1, applied as ammonium sulfate), and four varieties (Suwan 1, La Posta Sequia, KTX2602, DK888) was conducted in the tropical lowlands of Thailand. The water regime did not affect the mineral composition of the grains. Application of N fertilizer reduced the concentrations of Ca and Zn, and increased the concentration of Mn in the grains. The top yielder, DK888, had the lowest concentrations of N, P, Mg, and Cu in the grain. The varietal differences in the concentrations of grain N and minerals were fairly stable across the levels of N and preanthesis water supply. The varieties that differed most in the grain N and P concentrations (DK888 and KTX2602) had almost the same endosperm/germ dry weight ratio. It remains to be determined whether breeding for high grain yield inevitably lowers the concentrations of grain minerals and protein.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE WIDESPREAD INCIDENCE of so-called "Hidden Hunger"—for example, malnutrition due to mineral and vitamin deficiency (Welch and Graham, 2004)—throughout the world has raised the level of interest of nutritionists and agronomists in the minor constituents of food (Peters et al., 2003). High levels of minerals and protein are usually considered to be indicators of a high dietary quality of cereal products for humans and farm animals. However, this may not be true, at least for P. Approximately 80% of the P in maize (Zea mays L.) grains occurs as phytic acid (Lott et al., 2000). With utilization rates of 40% and below, maize phytic acid is a relatively poor source of P for pigs, poultry, and other monogastrics (Lantzsch et al., 1992). Phosphorus from undigested phytic acid may contribute to the eutrophication of surface waters in areas in which large amounts of animal waste are applied (Feil, 2001; Wardyn and Russell, 2004). Furthermore, phytic acid is thought to act as a naturally occurring toxic substance (Feil, 2001). Crops obtain their mineral nutrients from the soil. There is concern that growing cereal varieties, bred for high levels of grain minerals, will lead to a depletion of soil nutrient reserves and will, thus, be unsustainable without the addition of fertilizer (Feil et al., 1992; Graham et al., 1999). One way to avoid this dilemma would be to increase the bioavailability of grain minerals rather than to increase their concentration (Van Campen and Glahn, 1999).

Increasing the grain protein concentration has, thus far, not been a major goal of most maize breeding programs because the nutritional value of ordinary maize protein is low for nonruminants due to its low concentrations of the essential amino acids lysine, tryptophan, and methionine (Pixley and Bjarnasson, 2002; Olsen et al., 2003).

It is well established that the concentrations of N (Kniep and Mason, 1991; Dudley and Lambert, 1992; Bletsos and Goulas, 1999) and minerals (Arnold and Bauman, 1976; Arnold et al., 1977; Raboy et al., 1989; Ahmadi et al., 1993; Wardyn and Russell, 2004) can vary among varieties. Research conducted in the lowlands of Thailand revealed that the grains of tropical maize varieties differ markedly in the concentrations of N, P, and, to a lesser extent, K (Feil et al., 1992, 1993). Little has been published about the genotypic variation in the concentrations of mineral elements other than P and K in the grain of tropical maize. Bänziger and Long (2000) grew more than 1400 improved maize genotypes and 400 landraces in 13 trials in Mexico and Zimbabwe and found genotype differences in the concentrations of grain Fe and Zn. Maziya-Dixon et al. (2000) observed large variation in the concentrations of grain Fe and Zn in a set of 109 inbred lines that were developed for the midaltitude and lowland agroecologies of West and Central Africa.

Since N is the most limiting factor for plant growth on many soils, the application of N fertilizer usually results in marked increases in grain yield. It is suggested that large increments in grain yield, due to N fertilization, tend to dilute the grain minerals. Nevertheless, N fertilization had little or no effect on the concentrations of P and K in maize grains in most studies (Bennett et al., 1953; Thiraporn et al., 1992; Ahmadi et al., 1993; Feil et al., 1993; Alfoldi et al., 1994).

Surprisingly few studies dealt with the effect of water supply on the concentration of minerals in the grain. Harder et al. (1982) imposed various moisture stress treatments on maize after silking. Even though this resulted in grain yield reductions of up to 33%, no changes in the concentrations of grain P and K were observed. Water shortage during the vegetative stages of development limits the grain yield in many maize production areas. Edmeades et al. (1994) assumed that 50% of the losses in maize grain yield in the developing world due to drought result from preanthesis stress. No information seems to have been published about the effects of this preanthesis drought on the mineral composition of maize grains.

Nitrogen and minerals are distributed unevenly throughout the maize kernels. The highest concentrations of minerals are found in the germ, whereas the endosperm is almost void of mineral elements (O'Dell et al., 1972). Varietal differences in the concentration of grain N and minerals may, therefore, be due to variation in the relative size of the major kernel components.

The main objective of our research was to determine whether differences in the concentration of grain minerals (P, K, Mg, Ca, Mn, Zn, and Cu) among tropicial maize varieties are affected by the level of water and N supply. The grains of two varieties that differed in the concentration of N and several mineral elements in the grain were dissected to determine whether the variation in grain N and P concentration can be attributed to variation in the size of germ and endosperm.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Experimental Site
The trials were conducted at the National Corn and Sorghum Research Center, Farm Suwan, Pakchong (14.5° N lat, 101° E long, 360 m asl.), Thailand in the dry seasons of 1994–1995, 1995–1996, and 1996–1997 (hereafter referred to as Experiments 1995, 1996, and 1997, respectively).

The soil is an ustic, isohyperthermic, kaolinitic oxisol. The 5- to 30-cm soil layer can be described as follows: clay, 550 g kg–1; silt, 220 g kg–1; sand, 230 g kg–1; organic matter, 10 g kg–1; bulk density, 1.23 Mg m–3; pH (H2O), 6.6; pH (acetate), 6.9; cation exchange capacity, 163 cmolc kg–1; salinity (KCl), 150 mg kg–1; P, 8 mg kg–1; K, 35 mg kg–1; Mg, 126 mg kg–1; Ca, 1537 mg kg–1; S, 76 mg kg–1; Zn, 0.2 mg kg–1; Mn, 19 mg kg–1; Cu, 1.9 mg kg–1; Fe, 16 mg kg–1; B, 0.17 mg kg–1; and Mo, 0.34 mg kg–1. These values were found for the 30- to 60-cm layer: clay, 740 g kg–1; silt, 130 g kg–1; sand, 130 g kg–1; organic matter, 7 g kg–1; bulk density, 1.23 Mg m–3; pH (H2O), 5.1; pH (acetate), 6.6; cation exchange capacity, 144 cmolc kg–1; salinity (KCl), 153 mg kg–1; P, 8 mg kg–1; K, 23 mg kg–1; Mg, 98 mg kg–1; Ca, 970 mg kg–1; S, 30 mg kg–1; Zn, 0.2 mg kg–1; Mn, 4 mg kg–1; Cu, 1.4 mg kg–1; Fe, 11 mg kg–1; B, 0.09 mg kg–1; and Mo, 0.15 mg kg–1. The data are based on eight randomly collected and bulked samples, which were analyzed in the laboratories of Lonza AG, Basle, Switzerland. The nutrients were extracted with ammonium acetate–EDTA (0.5 M NH4Ac + 0.5 M HAc + 0.02 M EDTA; pH 4.65); the soil/solution ratio was 1:10 and the extraction time 30 min. According to the report of Lonza AG, the cation exchange capacity was low in both soil layers. In the 5- to 30-cm soil layer, the availability of P, K, Zn, Mn, Fe, and B was low, that of Mg and Ca moderate, and that of S, Cu, and Mo sufficient. In the subsoil, the availability of P, K, Ca, Zn, Mn, Fe, B, and Mo was low, that of Mg moderate, and that of S and Cu sufficient. The concentration of Al in the subsoil was relatively high, and the concentration of Co was considered to be toxic to plants.

Cultural Practices and Varieties
The whole field was planted to unfertilized maize in the previous season to remove N from the soil. Shortly before sowing, the field was plowed with a disc harrow, and ridges were formed every 0.75 m. Ten kilograms P ha–1 and 25 kg K ha–1 were applied manually on the tops of the ridges before planting the maize. Nitrogen was applied as ammonium sulfate at three rates: 0, 80 (applied before planting), and 160 kg N ha–1 (80 kg N applied before planting and 80 kg N applied 30 d after plant emergence). The N fertilizer was placed by hand on the side of the ridges and irrigated into the soil. Planting was done on 22 Dec. 1994, on 13 Dec. 1995, and on 18 Dec. 1996. Four cultivars were used: the open-pollinated varieties Suwan 1 and La Posta Sequia as well as the hybrids KTX2602 and DK888. Suwan 1 (C11) is a yellow semi-flint maize variety derived from Caribbean, Mexican, and Philippine populations, which was released in Thailand in 1975. La Posta Sequia (C4) is a late lowland white dent population, which originates from a drought-tolerance selection program of CIMMYT. The KTX2602 variety (also known as Suwan 2602 and KUH2602) is a Thai three-way hybrid, bred by the Kasetsart University and officially released in 1986. The semi-prolific single-cross hybrid DK888 from DeKalb was released in Thailand in 1991 and was the most popular variety in Thailand in the 1990s. Suwan 1 and DK888 were selected for this study because they represent popular local standard varieties from different breeding eras. The grains of KTX2602 were known to be high in N, P, and K (Feil et al., 1993); we wanted to determine whether the levels of minerals other than P and K are also high. La Posta Sequia was used because of its putative tolerance to drought.

Two seeds per hill were sown manually. Approximately 2 wk after emergence, the less vigorous plant was removed; the target plant density was 5.3 plants m–2. Weeds and insects were controlled by applying pesticides.

Immediately after sowing, all the plots were irrigated by overhead sprinklers three (1995 and 1997 experiments) to four (1996 experiment) times every 4 to 5 d to improve the rate of emergence. An estimated 3 to 4 ha cm water were applied at each irrigation. After the stands were established, two water regimes were imposed. Half of the plots (hereafter referred to as well-watered) were furrow-irrigated at weekly intervals at rates that exceeded the demand of the plants (about 4 ha cm water per irrigation event). The remaining plots (hereafter referred to as drought-stressed) were subjected to two drought stress periods. The first drought stress period started 10 d (1995), 16 d (1996), and 9 d (1997) after plant emergence and ended 31 d (1995) and 32 d (1996, 1997) after plant emergence. Thereafter, about 4 ha cm water was applied at once to the whole experimental field by means of sprinklers. Then, a second drought stress phase was initiated during which the drought-stressed plots did not receive irrigation water for 26 d (1995), 31 d (1996), and 24 d (1997). In the 1995 and 1996 experiments, however, rain (3.1 and 4 ha cm, respectively) fell 4 d (1995) and 7 d (1996) before the scheduled end of the second drought stress periods. Irrigation was resumed just before the earliest cultivar (KTX2602) started flowering; from then on, all the plots were furrow-irrigated (about 4 ha cm per irrigation event) at weekly intervals until maturity

Whole Kernel Analyses
At physiological maturity, 3.77 m2 (1995) and 5.66 m2 (1996, 1997) were hand-harvested from the six center rows of each plot. Grains and stover were separated and weighed. Aliquots of the grains (400 g) were dried at 70°C to constant weight to determine grain yield. Subsamples of about 50 g were ground with a Cyclotec Tecator 1093 mill (Tecator, AB, Höganäs, Sweden) and passed through a 1-mm screen. Analyses of the concentration of total N were performed with a LECO CHN-1000 auto analyzer (LECO Corp., St. Joseph, MI). To determine the concentrations of P, K, Mg, Ca, Mn, Zn, and Cu duplicate 1-g samples were ashed in silica crucibles at 550°C for 7 h in a muffle furnace. The residue was taken up in 8 mL of 6.8 M HCl, transferred to a 50-mL volumetric flask, and, after adding 1 mL 0.38 M CsCl solution, diluted to the mark with deionized water. After filtering through a membrane filter (Merck 3558011, E. Merck AG, Darmstadt, Germany), an aliquot of the solution was analyzed with an Inductively Coupled Plasma Atomic Emission Spectrometer, ICP–AES, (Liberty 200, Varian Australia Pty. Ltd., Mulgrave, Victoria, Australia) for mineral element concentration.

Concentrations of Nitrogen and Phosphorus in Germ and Endosperm
In 1995 and 1996, about 40 kernels from each plot were soaked in distilled water for about 24 h. Thereafter, the pericarp was peeled off and discarded. The germ was separated from the endosperm by exerting slight pressure on the kernel. Endosperm material that adhered to the germ was removed with a sharp knife. The germ and endosperm samples were dried, ground, sieved, and assayed for total N as described above. The concentration of P was determined according to a modified version (Feil and Fossati, 1997) of a procedure described by Jones and Case (1990). Five hundred milligrams of ground material were weighed into a digestion tube and 3 mL of concentrated HNO3 were added. The samples were left overnight at room temperature. After adding 1 mL conc. H2SO4, the tubes were placed in the port of a digestion block and first heated to 150°C for 30 min, then to 175°C for 60 min, and finally to 230°C for 90 min. The clear digest was neutralized with NaOH and p-nitrophenol as an indicator (Olsen and Sommers, 1982), and P was determined colorimetrically according to Murphy and Riley (1962).

Experimental Design and Statistical Analysis
The plots consisted of eight rows, 9 m long and 75 cm apart. The experimental design was a split-split-plot design with six replications. Each replication was divided into two main plots that differed in water supply (well-watered and drought-stressed). The three N levels (N0, N80, and N160) were the subplots; the four varieties (Suwan 1, La Posta Sequia, KTX2602, and DK888) were randomly distributed within the subplots. The data were subjected to analyses of variance, which were performed using the PROC MIXED procedure of SAS (SAS Inst., 1997). The replication, water x replication, and water x N rate x replication effects were considered to be random. The plant population was included as covariate in the model. When the F test indicated significant (P < 0.05) differences among the treatments, mean separation was performed using the adjust = Tukey option.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Nitrogen and Minerals in the Whole Grain
The grain yield in 1996 was clearly higher than in 1995 and 1997 under both water regimes. The adverse effect of drought on grain yield was less pronounced in 1996 than in 1995 and 1997. With the exception of Ca and Cu, the concentrations of N and minerals in the grains were similar in all years (Tables 1, 2, and 3).


View this table:
[in this window]
[in a new window]
  		Table 1. Main effects{dagger} of variety, water regime, and rate of N application on the grain yield and the concentrations of grain N and minerals in 1995.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Main effects{dagger} of variety, water regime, and rate of N application on the grain yield and the concentrations of grain N and minerals in 1996.

 

View this table:
[in this window]
[in a new window]
  		Table 3. Main effects{dagger} of variety, water regime, and rate of N application on the grain yield and the concentrations of grain N and minerals in 1997.

 

View this table:
[in this window]
[in a new window]
  		Table 4. Effects of the water regime x N rate interaction on the grain yield (1995, 1997) and the concentration of grain N (1996).

 

View this table:
[in this window]
[in a new window]
  		Table 6. Effects of the N rate x variety interaction on the concentrations of grain N (1996), P (1997) and Mn (1996).

 
Averaged across the rates of N application and varieties, the grain N concentration in 1996 was significantly lower under continuous irrigation than under drought stress (Table 2). An opposite trend was found in 1995 and 1997 (Tables 1 and 3). The effects of N rate and variety on the grain N concentration were significant in all the cropping seasons. However, there were significant interactions (Tables 4, 5, and 6). The effect of the water regime x N rate interaction on the grain N concentration was significant in 1996 (Table 4), that of the water regime x variety interaction in 1996 and 1997 (Table 5), and that of the N rate x variety interaction in 1996 (Table 6). The water regime x variety interaction will be analyzed in more detail, because it was significant in 2 of the 3 yr (Table 5). In 1996, all the varieties had lower grain N concentrations under continuous irrigation than under drought stress; the effect of high soil moisture was least pronounced for DK888. In 1997, however, the grain N concentration of DK888 was affected to a greater extent by continuous irrigation than that of the other cultivars; only the grain N concentration of DK888 increased in response to drought stress. The genotypic variation in the responses of grain N concentration to different water regimes may merely reflect genotypic variation in the responses of grain yield. Indeed, the positive effect of continuous irrigation on the grain yield was strongest for DK888 in 1997, while it was weakest for the same variety in 1996 (Table 5). Although significant water regime x variety interactions occurred in 1996 and 1997, the rank order of the varieties was similar: KTX2602 always had the highest and DK888 the lowest grain N concentration under both water regimes.


View this table:
[in this window]
[in a new window]
  		Table 5. Effects of the water regime x variety interaction on the grain yield (1996, 1997) and the concentrations of grain N (1996, 1997), P (1997), K (1996), Mg (1997), and Cu (1996).

 
The concentration of minerals was not affected significantly by the water regime (Tables 1, 2, and 3). Averaged across the water regimes and varieties, N fertilization resulted in significant decreases in the concentrations of Ca and Zn in all years (Tables 1, 2, and 3). In contrast, N application significantly increased the concentration of Mn in 1996 and 1997 (Tables 2 and 3); a similar but a nonsignificant (P = 0.12) effect was observed in 1995 (Table 1). Averaged over the water and N fertilizer levels, the concentrations of grain minerals were significantly affected by the variety (Tables 1, 2, and 3), with the exception of K in 1995 (Table 1) and Ca in 1997 (Table 3). The main effects of the factor variety on the concentrations of grain N and minerals can be summarized as follows: the grain of KTX2602 had the highest concentrations of N, Mg, Ca, and Mn; that of Suwan 1 had the highest concentrations of Zn and Cu; that of DK888 had the lowest concentrations of N, P, Mg, and Cu; and that of La Posta Sequia had the lowest Zn concentration in all the years (Tables 1, 2, and 3). There were some significant water regime x variety (Table 5) and N rate x variety (Table 6) interactions, while there was only one significant three-way interaction (for P in 1997; data not shown). In general, the relative performance of the varieties was little affected by the water regime and the N level. For example, the breakdown of the data for P from 1997 according to variety, water regime, and the rate of N application revealed that KTX2602 always had the highest and Suwan 1 the second-highest grain P concentration, while the other varieties differed only slightly (data not shown). Since the interactions were inconsistent over the years, they will not be discussed in detail.

The DK888 and La Posta Sequia varieties consistently produced higher grain yields than Suwan 1 and KTX2602 (Tables 1, 2, 3, and 5), which suggests that grain yield and concentrations of grain N and minerals are inversely related. The top yielder, DK888, lodged more N and minerals in the grain than all the other cultivars in 1995 (significant for N, P, K, Mg, Ca, and Zn). In 1996, the grain of DK888 contained the largest amounts of P, K, Mg, Ca, and Zn (significant for P, K, Mg, and Zn), and, in 1997, the largest amounts of N, P, K. Mg, Ca, and Zn (significant for K and Zn) of all the varieties under investigation (data not shown). Thus, the low concentrations of N and minerals in the grains of DK888 were not due to the fact that this variety stored less N and minerals in the grain than the other varieties

Nitrogen and Phosphorus in Germ and Endosperm
Table 7 shows that the ratio of the endosperm to the germ dry matter was slightly higher for DK888 than for KTX2602. The germ had higher concentrations of N and P than the endosperm, but the difference between these grain fractions was smaller for N than for P. Variety KTX2602 had higher concentrations of N and P in both fractions than DK888. Analyses of germ and endosperm samples of KTX2602 and DK888 from 1996 confirmed this result [KTX2602: germ N concentration, 31.6 g kg–1; endosperm N concentration, 16.2 g kg–1; germ P concentration, 25.4 g kg–1; endosperm P concentration, 0.88 g kg–1; DK888: germ N concentration, 31.0 g kg–1; endosperm N concentration, 12.6 g kg–1; germ P concentration, 21.8 g kg–1; endosperm P concentration, 0.80 g kg–1; the variety effects were statistically significant for all the traits except for the endosperm P concentration (P = 0.08)]. The weighted mean N and P concentrations (Table 7) take the differences between the germ and endosperm dry weight into account. The calculated values agree well with the corresponding data in Table 1. There were significant N rate x variety interactions for mean N and endosperm N concentrations (Table 7). The breakdown according to N rate revealed that the N concentration of KTX2602 increased to a greater extent in response to N fertilization than the grain N concentration of DK888. Table 6 shows a similar trend. Compared with KTX2602, the germ of DK888 contributed more N but less P (P = 0.08) to the total amounts of N and P. In summary, KTX2602 did not have higher grain N and P concentrations than DK888 because the grains of the former variety contained larger germs; rather, the differences were due to the higher N and P concentrations in both germ and endosperm.


View this table:
[in this window]
[in a new window]
  		Table 7. Effects of variety on the endosperm/germ dry matter ratio (E/GDM) and the concentrations of N and P in the endosperm and the germ in 1995.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Except for Ca in 1995, the mean concentrations of the minerals in the three experiments (Tables 1, 2, and 3) are similar to those reported by Watson (1987). According to Pietz et al. (1978), Miller (1958) reported Ca concentrations as high as 1100 mg kg–1 in maize grains, indicating that the relatively high mean Ca concentration found in 1995 (136 mg kg–1) is not uncommon. Even though continuous irrigation markedly increased the grain yield in all the cropping seasons (Tables 1, 2, and 3) and thus, the grains' demand for minerals, there was no decrease in the concentrations of grain minerals. It has been reported that drought can alter the root dry matter (Eghball and Maranville, 1993) and the partitioning of roots to the topsoil and the subsoil (Klepper, 1991). In his experiments with maize on Farm Suwan, Camp (1996) found that the zone of maximum water uptake shifted down the soil profile with increasing water stress. This suggests that the drought-stressed plants in our experiment obtained water and minerals from the subsoil, where the concentrations of plant-available nutrients were generally low (see Materials and Methods). Nevertheless, in none of the experimental years were the concentrations of grain minerals affected by the water regime. This finding is in line with the results of Harder et al. (1982), who studied the effects of postanthesis drought on the concentrations of grain P and K.

Water regime significantly affected the grain N concentration in 1996 (Table 2). Comparable responses of grain protein to irrigation have also been reported by Jurgens et al. (1978), Harder et al. (1982), and Kniep and Mason (1991). However, the results obtained in 1995 (Table 1) and 1997 (Table 3) demonstrate that it is impossible to generalize about the impact of drought stress on the level of grain protein. The drought stress was much more severe in 1995 and 1997 (Tables 1 and 3) than in 1996 (Table 2), as indicated by the large increments in grain yield due to continuous irrigation in 1995 and 1997. Nevertheless, grain N was not diluted, that is, continuous irrigation caused similar increases in grain yield and grain N yield.

Application of N fertilizer brought about significant increases in grain yield (Tables 1, 2, and 3), thus increasing the amount of minerals required by the grains to maintain the concentration. A greater demand of the grains for minerals can be met when the plants take up larger amounts of nutrients from the soil. Some studies indicate that heavy N fertilization may reduce root growth (Eghball and Maranville, 1993; Durieux et al., 1994; Oikeh et al., 1999). If this is true, it might be difficult for the plants to acquire adequate amounts of mineral nutrients when N is applied at high rates. Nevertheless, even when the rates of N fertilization were high, the concentrations of most of the mineral nutrients in the grain did not decrease in our experiments (Tables 1, 2, and 3). The well-fertilized plants, probably, exploited the mineral nutrient reserves in the stover more efficiently. Former research in Thailand showed that the P harvest index (equals proportion of grain P to total shoot P) increased as the rate of N application increased (Thiraporn et al., 1992).

In our experiments, N fertilization consistently resulted in a decrease in the concentrations of grain Ca and Zn and in an increase in the concentration of grain Mn (Tables 1, 2, and 3). Pietz et al. (1978) found significant positive correlations of N rate and grain Mn concentration at four of nine sites in Illinois. Significant negative correlations between N rate and Ca were found on three of the fields, while a significant positive correlation was observed on one field. Application of N usually decreased the concentration of grain Zn, but the correlation was significant on three fields only. A dilution effect due to a higher grain yield as a result of N fertilization may have contributed to the inverse relationship between N rate and the concentrations of Ca and Zn in our experiment (Tables 1, 2, and 3). The effect of N fertilization on the mineral composition of plant tissues may depend on the N form. We applied ammonium sulfate, which acidifies the soil. Application of ammonium is considered to be advantageous on neutral to slightly alkaline soils, because it enhances the availability of Zn and Mn to the crop (Schnug and Finck, 1980). Consequently, changes in the soil pH may explain why N fertilization led to higher concentrations of grain Mn. In the case of Zn, the dilution effect caused by increments in grain yield may have been stronger than the ammonium effect.

In small grain cereals, genetic increases in grain yield are often paralleled by decreases in the concentration of grain protein (Feil, 1997) and minerals (Feil and Fossati, 1995). Is this also true for maize? The top yielder, DK888, had lower concentrations of N and many mineral elements in the grain than the other varieties tested (Tables 1, 2, and 3), even though DK888 stored the largest amounts of N and minerals in the grain. Thus, the relatively low concentrations of grain minerals and N for DK888 seem to be due to a dilution effect. In previous studies on Farm Suwan with 12 to 16 tropical maize varieties, the relationship between grain yield and the concentrations of grain P, K, and N was not inverse (Feil et al., 1990, 1993). In agreement with this, grains of old and modern maize hybrids used in Ontario showed no clear differences in the concentrations of these elements, but the concentrations of grain Mg, Cu, Mn, and Se tended to be higher in the old varieties than in newer ones (Vyn and Tollenaar, 1998). On the other hand, Bänziger and Long (2000) found grain yield and grain Zn concentration of tropical genotypes to be negatively correlated. Genetic progress in the grain yield of U.S. maize was associated with a significant drop in the concentration of grain N (Duvick and Cassman, 1999). Selection for high yield under drought resulted in a decrease in the grain N concentration of tropical maize (Bänziger et al., 2002). Consequently, there are indications that in maize, too, breeding for high grain yield tends to lower the concentrations of grain minerals and protein.

The concentrations of grain protein and minerals were affected significantly by the water regime x variety (Table 5) and the N rate x variety (Table 6) interactions in some years. All in all, however, the water regime and the rate of N fertilization had only a small impact on the ranking of the varieties. This indicates that varietal differences in the concentrations of grain N and minerals are fairly stable across wide ranges of N and preanthesis water supply.

The germ contains about 78% of the minerals in maize kernels and 18% of the protein (Watson, 1987). Differences in the composition of minerals in the grains of KTX2602 and DK888 may reflect differences in the proportions of germ and endosperm dry matter to total kernel weight. Watson (1987) reported that endosperm and germ made up on average 82.9% (range 81.8–83.5%) and 11.1% (range 10.2–11.9%), respectively, of the whole kernel dry weight of dent maize kernels of seven U.S. hybrids. Thus, the mean endosperm/germ dry weight ratio was about 7.5. Somewhat higher values were found for KTX2602 (7.7) and DK888 (8.0) (Table 7). The results demonstrate that KTX2602 and DK888 did not differ in the concentrations of grain N and P because KTX2602 had a relatively larger germ and smaller endosperm than DK888. Rather, the N and P concentrations were higher for KTX2602 in both kernel fractions.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
Although the set of varieties used in this study was small and did not represent a broad sample of tropical germplasm, we found significant genotype differences in the concentrations of protein, P, K, Mg, Ca, Mn, Zn, and Cu in the grains. The varieties that differed most in the N and P concentrations had approximately the same endosperm/germ dry weight ratio. Varietal differences in the elemental composition of maize kernels seem to be fairly stable over wide ranges of N and preanthesis water supply. This may facilitate breeding for high or low levels of grain minerals. While our study demonstrated that the concentrations of grain minerals are well buffered against increases in grain yield due to water and N fertilizer applications, it must still be determined whether higher grain yields as a result breeding progress tend to lower the levels of grain minerals and protein.


   ACKNOWLEDGMENTS
 
This project was conducted in collaboration with the Faculty of Agriculture of the Kasetsart University, Bangkok, and was supported by funds from the Swiss Agency for Development and Cooperation (SDC). We thank Dr. Rachain Thiraporn and Mr. Surapol Chowchong for their support and the field staff of the National Corn and Sorghum Research Center, Pakchong, Thailand, for assistance in carrying out the field trials. We also thank Ms. Elisabeth Mayer and Ms. Theres Rösch for conducting the chemical analyses.

Received for publication March 1, 2004.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Feil, B.
Right arrow Articles by Stamp, P.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Feil, B.
Right arrow Articles by Stamp, P.
Agricola
Right arrow Articles by Feil, B.
Right arrow Articles by Stamp, P.
Related Collections
Right arrow Irrigation
Right arrow Maize
Right arrow Plant Analysis

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Vadose Zone Journal
Journal of Plant Registrations Soil Science Society of America Journal
Journal of Natural Resources
and Life Sciences Education		 Journal of
Environmental Quality