Agronomy Journal 92:644-649 (2000)
© 2000 American Society of Agronomy
SOIL FERTILITY
Effects of Nitrogen and Sulfur on Canola Yield and Nutrient Uptake
Grant D. Jackson
Western Triangle Agric. Res. Ctr., Montana State Univ., P.O. Box 974, Conrad, MT 59425 USA
gjackson{at}montana.edu
 |
ABSTRACT
|
|---|
Spring canola (Brassica napus L. var. napus) is becoming a significant oilseed crop adapted to the western USA. Often N and S limit crop growth. Field experiments were established to study the effects of N and S fertilization on seed yield, oil content, and N, P, K, and S uptake of spring canola. Four N rates in combination with three S rates were evaluated on two irrigated and three rainfed locations in the western triangle area of Montana near Conrad. Seed yields ranged from 0.1 to 3.8 Mg ha-1. Seed yield and oil content N responses were closely related to available N (fertilizer N plus soil NO3–N in 90 cm of soil). Seed oil content varied from 370 to 510 g kg-1 and was depressed by increasing N. Optimum seed and oil yield occurred at about 200 kg N ha-1. Two of the experimental sites responded to S. About 20 kg S ha-1 was adequate for optimum seed and oil yields. At the optimal N and S levels, total plant N, P, K, and S uptake averaged 140, 25, 170, and 60 kg ha-1, respectively. Of the total N, P, K, and S accumulation, about 40% of the N, 30% of the P, and 85% of the K and S remained in the postharvest residue.
Abbreviations: Con95, Conrad, 1995 • Con 96, Conrad, 1996 • Far95, Fairfield, 1995 • ICP, inductively coupled plasma spectrometry • Sun95, Sunburst, 1995 • Sun96, Sunburst, 1996
|
INTRODUCTION
|
|---|
SINCE FERTILIZER PROGRAMS are a major expense and profit margins are very small in canola production, producers and field agronomists need information on nutrient relationships to fine-tune fertilizer recommendations. The literature indicates that N and S relationships are very important in canola production (Grant and Bailey, 1993). Recent fertilizer trials (Jackson et al., 1993; Popove, 1994) and the research summarized by Grant and Bailey (1993) have provided general guidelines for fertilizing canola in Montana. However, in the Montana canola production area, considerable soil and soil parent material variability exists (Veseth and Montagne, 1980), which affects nutrient availability and quantity, particularly that of N and S. Thus, refinement of N and S fertilizer recommendations are needed to ensure optimum productivity, economic vitality, and environmental stewardship. Consequently, field trials were conducted to determine the relationship of canola seed yield and quality to N and S fertilization and soil tests and the effect of N and S fertilization on N, P, K, and S cycling.
|
Materials and methods
|
|---|
Research plots were planted in the western triangle area of Montana near Fairfield (Far95), Sunburst (Sun95), and Conrad (Con95) in 1995 and near Sunburst (Sun96) and Conrad (Con96) in 1996. The Far95 site was flood-irrigated, Con96 was sprinkler-irrigated, and the Con95 and Sunburst locations were rainfed. Sunburst and Conrad sites were planted, no-till, into malting barley (Hordeum vulgare L.) stubble, while the Fairfield location was planted into a tilled seedbed following malting barley. At the no-till sites, windrowed residue was removed prior to planting and standing residue was estimated at 500, 2000, 3000, and 2500 kg ha-1 for Sun95, Con95, Con96, and Sun96, respectively. Planting date for Sun95 and Con95 was 27 April, for Far95 was 19 May, for Sun96 was 29 April, and for Con96 was 2 May. The cultivar Westar was planted at the rate of 7 kg ha-1.
Four N rates (0, 84, 168, and 252 kg N ha-1) and three S rates (0, 22, and 45 kg S ha-1), organized into a randomized complete block design with a 4 x 3 factorial treatment arrangement and four blocks, were applied to plot areas. Plot size was six rows wide (30 cm row spacing) and 6 m long. Each research site received 15 kg P ha-1 as triple superphosphate applied with the seed and 34 kg K ha-1 as KCl broadcast while planting. Nitrogen fertilizer as urea and ammonium sulfate and S fertilizer as potassium sulfate or ammonium sulfate were also applied broadcast while planting.
Irrigated plot sites received enough supplemental water to maintain the soil profile (0–90 cm) at about 50% of the available water holding capacity. Growing season precipitation or precipitation plus irrigation was 24 cm for Sun95, 50 cm for Far95 (flood-irrigated twice at an estimated 15 cm of water per irrigation), 35 cm for Con95, 11 cm for Sun96, and 35 cm for Con96. Con95 required no supplemental irrigation; however, Con96 required six irrigations, totaling 24 cm of supplemental water.
Preplant soil samples were taken for nutrient analysis (Table 1) and soil water content analysis. Organic matter was determined by the colorimetric method published by Sims and Haby (1971). Soil pH was measured from a saturated paste. Phosphorus was determined by the method described by Olsen et al. (1954). Potassium was extracted with ammonium acetate and measured by atomic absorption spectroscopy. Nitrate-N was extracted with KCl (Keeney and Nelson, 1982) and measured by automated Cd reduction (Clesceri et al., 1989). Sulfate-S was determined by inductively coupled plasma spectrometry (ICP) after extraction with ammonium acetate-acetic acid (Bardsley and Lancaster, 1960). Available soil water was estimated by the method described by Cole and Matthews (1954). Available water content in the 0- to 90-cm soil depth at planting time was 9 cm at Sun95, 6 cm at Far95, 3 cm at Con95, 9 cm at Sun96, and 4 cm at Con96.
Soils in the Fairfield area developed from calcareous gravel terraces, while the soils in the Conrad and Sunburst area developed from calcareous glacial till and are dominated chemically by lime and gypsum (Veseth and Montagne, 1980). Soils were classified as follows: Fairfield, Rothiemay loam (fine-loamy, mixed, superactive, frigid Aridic Calciustolls); Sunburst, Vida loam (fine-loamy, mixed, superactive, frigid Typic Argiustolls); and Conrad, Scobey clay loam (fine, smectitic, frigid Aridic Argiustolls).
The Cate-Nelson graphical procedure for determining soil test critical levels as described by Nelson and Anderson (1977) was used to relate S soil test values at the 0- to 30-cm soil depth to canola S response. The method involves one observation per location. The highest yield from a S treatment (with adequate N, P, and K) was divided by the yield of the treatment without S (with adequate N, P, and K), and the result was multiplied by 100. The result, called percentage yield, is plotted against the S soil test for that location, producing a scatter plot that is separated into four quadrants (see Fig. 3)
by a vertical and a horizontal line. These lines are adjusted to maximize the number of data points in the "+" quadrants, segregating soils with low soil tests and low percentage yield (response to added nutrients) from soils with little or no response to added nutrients. The point where the vertical line crosses the x-axis is defined as the critical soil test level. The point where the horizontal line crosses the y-axis will vary, but should range one or two percentage points above or below 90% yield.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 3 Cate-Nelson scatter diagram (as described by Nelson and Anderson, 1977) of canola percentage yield versus SO4–S soil test by ammonium acetate-acetic acid extraction (Bardsley and Lancaster, 1960). Percentage yield is calculated by dividing a location's check treatment (no S, but adequate N, P, and K) by the highest-yielding treatment with S and adequate N, P, and K and multiplying by 100. Horizontal and vertical lines are drawn and adjusted for the maximum number of data points in the (+) quadrants. The point where the horizontal line crosses the x-axis is defined as the critical level
|
|
Whole plant (aboveground) samples (from 0.3 m2 of the two middle rows) were taken at maturity for total dry matter yield and N, P, K, and S analysis. When mature, plots were trimmed to 15 m, and the middle four rows were swathed with a small plot swather (Swift Machine & Welding, Swift Current, SK, Canada) and threshed with a combine (Model 125 C, Hege Equipment, Colwich, KS). Seed samples were dried, weighed, and analyzed for oil, N, P, K, and S content. Oil was determined by nuclear-magnetic resonance as described by Robertson and Morrison (1974), N by the Kjeldahl method, and P, K, and S by ICP after wet ashing. Seed yields were reported at 80 g kg-1 moisture content, and all other measurements were reported on a dry weight basis. Nutrient uptake in the whole plant and seed was calculated by multiplying the respective nutrient concentration by the total dry matter or seed yield . Oil yield was calculated by multiplying seed yield by oil concentration.
All plant and soil samples were dried at 60°C for 72 h in a forced-draft oven. Seed oil analysis was performed at the Eastern Agricultural Research Center, Sidney, MT. All other soil and plant chemical analyses were conducted at the Soil Testing Laboratory, Montana State University, Bozeman.
Nitrogen:S ratios were calculated by dividing the total plant or seed N content by the respective S content. Data were analyzed by analysis of variance (Lund, 1988) and multiple regression (CoHort Software, 1995). The independent N variable was calculated by adding the N fertilizer rate and the amount of soil NO3–N in 0 to 90 cm of soil.
|
Results and discussion
|
|---|
Except for at Sun96, growing conditions were generally excellent for canola production during both growing seasons. Relatively cool maximum daily air temperatures (rarely above 32°C) were conducive to and moisture levels were adequate for flowering and pod set. Seed yields were probably reduced at Far95 by a late planting date (19 May) and at Sun96 by drought conditions and weed competition.
The effects of N and S on seed yield are tabulated in Table 2 . Individual treatment yield varied considerably, ranging from 0.1 Mg ha-1 at Sun96 to 3.8 Mg ha-1 at Sun95. Nitrogen increased yields at all locations. The relationship between seed yield and N was linear and quadratic for all locations.
Con95 and Sun95 had seed yield responses to S. Sulfur response at Sun95 was linear and quadratic, while the S response at Con95 was only linear. The interaction between N and S was significant at Sun95 and Sun96 (yield peaked at 22 kg S ha-1), but the interaction at Con95 was nonsignificant. Figure 1
shows the classical nutrient interaction that occurred at Sun95, where the N versus yield response curves were affected by adequate S or the lack of it. Conversely, Fig. 2
shows the N versus yield response of Con95 without an S interaction but with an S response. In this case, seed yield response to N was similar at all S rates; thus, S affected only the level of response, not the shape of the response curves.
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 1 Nitrogen (fertilizer N plus soil NO3–N in 90 cm of soil) and S effects on dryland canola seed yield near Sunburst, MT, in 1995
|
|
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 2 Nitrogen (fertilizer N plus soil NO3–N in 90 cm of soil) and S effects on dryland canola seed yield near Conrad, MT, in 1995
|
|
Generally, S yield responses were unrelated to S soil test values when soil sampling depths were 60 cm or deeper. This lack of a relationship between canola S response and soil S at the 0- to 60-cm soil depth is consistent with the literature (Grant and Bailey, 1993; Nuttall and Ukrainetz, 1991; and Franzen, 1997). A Cate-Nelson plot of percentage yield versus SO4–S soil test (0–30 cm of soil) is shown in Fig. 3. The use of the Cate-Nelson procedure requires as many data locations as possible; thus four locations, three from the Fairfield area (Popove, 1994) and one from the Sunburst area (unpublished data, 1999), were included with the five locations listed in Table 2. These data from additional experiments were included because each location had variation in S rate with adequate N, P, and K, thus meeting the criteria for the Cate-Nelson procedure. With this procedure, a preliminary critical level indicating a probable S response was found at 70 kg S ha-1. Only nine data points were used to determine this critical level; thus, more canola S response data are needed to refine or support the critical level. Until more data are collected, a critical level of 70 kg S ha-1 should be used.
A S soil test level of 70 kg S ha-1 or less would result in a S fertilizer recommendation (with the Cate-Nelson procedure, experience is used for the actual fertilizer recommendation), and based on the data from Table 2, this would be about 20 kg S ha-1. A thorough evaluation of SO4–S extraction procedures, along with more S response data, are needed. Current guidelines (Lichthardt and Jacobsen, 1991) do not recommend S fertilizer because of the lack of S response data and the unreliability of the S soil test. Similar experiences have been reported in other Great Plains states and provinces (Grant and Bailey, 1993; Franzen, 1997). Much of the frustration with S soil test procedures can be attributed to the variability of S in the soil profile and to the difficulty in estimating SO4–S solubility of the gypsum and pyrite present in most of Montana's cultivated soils (Veseth and Montagne, 1980). Consequently, a starter S fertilizer may be recommended regardless of the S soil test, as is the case with spring wheat P fertilizer recommendations in Montana (Jackson et al., 1997).
As indicated by Grant and Bailey (1993), agronomists suggest a tissue or soil test N:S ratio of 7:1 when using soil test and plant nutrient levels for formulating nutrient management recommendations. Thus, total plant and seed N:S ratios were calculated and reported in Table 3
. Seed N:S ratios varied from 7.9 to 13.0 and were considerably higher than the total plant N:S ratios, which ranged from 1.5 to 7.1. At Sun95, N increased the N:S ratio in both the plant and seed, while S decreased the N:S ratio. The rate of N:S ratio increase was higher when N was increased without S than with either the 22 or 45 kg S ha-1 rate. At the other S response location (Con95), N and S had no effect on either the plant or seed N:S ratio. As at Sun95, adding N increased the N:S ratio of the whole plant at Far95 and Sun96; however, in contrast to Sun95, increasing N lowered the N:S ratio of the seed at Far95, Sun96, and Con96. The seed N:S ratio interaction at Far95 is difficult to interpret because N lowered the seed N:S ratio, and S did not influence the seed N:S ratio. Evidently, the rate of seed N:S ratio decline by increasing N was increased by adding S.
View this table:
[in this window]
[in a new window]
|
Table 3 Effect of N and S on canola whole plant and seed N:S ratio from five locations near Conrad, MT, in 1995 and 1996
|
|
Seed oil levels (Table 4)
ranged from 370 to 508 kg ha-1. In general, N reduced oil content at all locations, consistent with previous reports (Jackson et al., 1993, Popove, 1994, and Grant and Bailey, 1993); this is probably due to N delaying plant maturity. Sulfur increased oil content at Sun95, the site with the lowest soil SO4–S, but had no effect on oil levels at the other locations. Oil content tended to peak at the 22 kg S ha-1 rate. A significant N x S interaction (S increased oil content, while N depressed oil levels) was detected at Sun95 and Sun96.
Nitrogen, P, K, and S content data of the total plant and seed are not shown; however, the average total plant level of N, P, K, and S was 12.4, 2.4, 14.0, and 5.4 g kg-1, respectively. Nitrogen, P, K, and S content of the seed averaged 31, 6.4, 8.3, and 3.3 g kg-1, respectively. These N content levels of both the seed and whole plant are within the range of N concentrations recorded by Hocking et al. (1997). Bolland (1997) reported canola seed P levels of 2 to 6 g P kg-1 and whole plant P levels of 3 to 5 g P kg-1. Nuttall and Ukrainetz (1991) reported similar seed S levels and canola straw S concentrations of 3.5 to 7.1 g S kg-1. Nitrogen increased N and K levels in the plant and N and S in the seed, decreased P in the seed and plant, and had little effect on plant S and seed K. Sulfur effects on N, P, K, and S levels of the seed and plant were inconsistent, but at the S-responsive sites, Con95 and Sun95, S increased plant and seed S.
Total plant yield, seed yield, oil yield, and N, P, K, and S uptake by the total plant and seed from the four N treatments at the 22 kg ha-1 S rate from all locations except Sun96 were regressed against N rate (Table 5)
. Data from the 22 kg ha-1 S rate was selected because this was the rate that resulted in the greatest uptake where S responses were observed to S fertilization (Con95 and Sun95). Data from Sun96 were not included because of severe drought conditions. The relationship between total plant yield at maturity and N rate is shown in Eq. [1] (Table 5). The linear relationship between total plant yield and N reflects the tendency of canola to exhibit an indeterminate growth habit when nutrients and water are essentially unlimited with no heat stress. Seed yield response to available N reflects the more familiar quadratic function (Eq. [2] in Table 5) and is essentially the NO3–N (0–90 cm soil depth) soil test calibration curve. Optimal seed yield of 2.56 to 2.70 Mg ha-1 occurred in the 180 to 220 kg N ha-1 range, consistent with previous data from Jackson et al. (1993), Popove (1994), and Grant and Bailey (1993). The yield difference between Eq. [1] and [2] (Table 5) represents the amount of residue returning to the soil. This yield difference at the 200 kg ha-1 optimal N level was 8.3 Mg ha-1, or about twice the amount of residue of a small grain crop at a similar grain yield level (Kushnak et al., 1992).
View this table:
[in this window]
[in a new window]
|
Table 5 Regression equations relating various canola yield and nutrient uptake dependent variables to soil nitrate-N (0–90 cm soil depth) plus fertilizer N, kg ha-1
|
|
The relationship of oil yield to N rate followed a quadratic expression (Eq. [3] in Table 5). Optimal oil yield occurs in the same N range as seed yield, 180 to 220 kg N ha-1, even though a negative relationship exists between oil content (Table 3) and increasing N. Thus, a producer can expect low oil contents as well as lower oil yields when canola is fertilized at N rates exceeding 220 kg ha-1.
Table 5 also shows the nutrient uptake relationships of the total plant and seed relative to N rate (Eq. [4] to [11]). Nitrogen increased the accumulation of all nutrients in both the whole plant (linear response) and seed (quadratic response). Sulfur also increased nutrient uptake at locations Sun95 and Con95. This effect was not as dramatic as N, and peaked at 22 kg S ha-1 (data not shown). Interesting predictions and observations about canola as a rotation and nutrient scavenging crop can be made by comparing the total plant and seed nutrient uptake equations. The difference between paired equations ([4] and [5], [6] and [7], [8] and [9], and [10] and [11]) at the same N rate represents the amount of N, P, K, and S potentially available for subsequent crops from the postharvest residue. At the 200 kg N ha-1 level, the equations predict the following nutrient amounts returning to the cropping system (difference between total plant and seed uptake): 56 kg N ha-1 (from Eq. [4] and [5]), 7.5 kg P ha-1 (from Eq. [6] and [7]), 147 kg K ha-1 (from Eq. [8] and [9]), and 53 kg S ha-1 (from Eq. [10] and [11]). Substantially more K and S accumulated in the residue than was applied as fertilizer. At a modest N level, 75 kg N ha-1, the whole plant accumulated 15 kg P ha-1, or the initial P fertilization rate. Nitrogen accumulated in the whole plant accounts for about 70% of the available N (NO3–N in the upper 90 cm of soil + fertilizer N). About 60% of the N, 70% of the P, and 15% of the K and S stored in the whole plant was accumulated in the seed.
|
Summary
|
|---|
Canola yield and nutrient uptake are highly dependent on N fertility. Peak seed yields of 2.56 to 2.70 Mg ha-1 occurred with 180 to 220 kg N ha-1, suggesting that canola requires 0.07 to 0.08 kg N kg-1 of yield (3.5 to 4.1 lb N bu-1). Growers should apply about 20 kg S ha-1 to satisfy canola's high S requirement when ammonium acetate-acetic acid–extractable S is 
70 kg S ha-1. Most of the plant N and P is removed from the field with the seed; however, the majority of K and S remains in the residue. At optimal N (200 kg ha-1) and seed yield (2650 kg ha-1), approximately 60 kg N, 8 kg P, 150 kg K, and 55 kg S ha-1 would remain in the canola residue following harvest.
Received for publication December 2, 1998.
|
REFERENCES
|
|---|
- Bardsley C.E., Lancaster J.D. Determination of reserve sulfur and soluble sulfates in soils. Soil Sci. Soc. Am. Proc. 1960;24:265-268.
- Bolland M.D.A. Comparative phosphorus requirement of canola and wheat. J. Plant Nutr. 1997;20:813-829.
- Clesceri L.S., Greenberg A.E., Trussell R.R. Standard methods for the examination of water and wastewater, 17th ed Washington, DC: Am. Public Health Assoc, 1989 APHA/AWWA/WPCF..
- CoHort Software CoStat statistical software. Minneapolis, MN: Version 5.01. CoHort Software, 1995.
- Cole J.S., Matthews O.R. Soil moisture studies of some Great Plains soils: I. Field capacity and "minimum point" as related to the moisture equivalent. Soil Sci. Soc. Am. Proc. 1954;18:247-252.
- Franzen D.W. Fertilizing mustard and canola. Fargo: North Dakota State Univ. Ext. Serv, 1997 SF-1122..
- Grant C.A., Bailey L.D. Fertility management in canola production. Can. J. Plant Sci. 1993;73:651-670.
- Hocking P.J., Randall P.J., DeMarco D. The response of dryland canola to nitrogen fertilizer: Partitioning and mobilization of dry matter and nitrogen, and nitrogen effects on yield components. Field Crops Res. 1997;54:201-220.
- Jackson G.D., Kushnak G.D., Carlson G.R., Wichman D.M. Correlation of the Olsen phosphorus soil test: Spring wheat response. Commun. Soil. Sci. Plant Anal. 1997;28:813-822.
- Jackson G.D., Kushnak G.D., Welty L.E., Westcott M.P., Wichman D.M. Fertilizing canola. Montana AgResearch 1993;10(2):21-24.
- Keeney D.R., Nelson D.W. Nitrogen—Inorganic forms. In: Page A.L., et al. , ed. Methods of soil analysis. Part 2, 2nd ed Madison, WI: ASA and SSSA, 1982:679-682 Agron. Monogr. 9..
- Kushnak G.D., Jackson G.D., Berg R.K., Carlson G.R. Effect of nitrogen and nitrogen placement on no-till small grains: Plant yield relationships. Commun. Soil Sci. Plant Anal. 1992;23:2437-2449.
- Lichthardt J.J., Jacobsen J.S. Fertilizer guidelines for Montana. Bozeman: Montana State Univ. Ext. Serv, 1991 EB 104..
- Lund R.E. MSUSTAT statistical analysis package. Version 5.22. Bozeman: Montana State Univ, 1988.
- Nelson L.A., Anderson R.L. Partitioning of soil test-crop response probability. In: Peck T.R., et al. , ed. Soil testing: Correlating and interpreting the analytical results. Madison, WI: ASA, 1977:19-38 ASA Spec. Publ. 29..
- Nuttall W.F., Ukrainetz H. The effect of time of S application on yield and sulfur uptake of canola. Commun. Soil Sci. Plant Anal. 1991;22:269-281.
- Olsen S.R., Cole C.V., Watanabe F.S., Dean L.A. Estimation of available phosphorus in soils by extraction with sodium bicarbonate. Washington, DC: U.S. Gov. Print. Office, 1954 USDA Circ. 939..
- Popove G.B. Effects of nitrogen, phosphorus and sulfur on the yield, growth and quality of canola (Brassica napus L.).. Bozeman: Montana State Univ, 1994 M.S. thesis..
- Robertson J.A., Morrison W.H., III. Analysis of oil content of sunflower seed by wide line NMR. J. Am. Oil Chem. Soc. 1974;56:911-964.
- Sims J.R., Haby V.A. Simplified colorimetric determination of soil organic matter. Soil Sci. 1971;112:137-141.
- Veseth R., Montagne C. Geologic parent materials of Montana soils. Bozeman: Montana Agric. Exp. Stn, 1980 Bull. 721..
This article has been cited by other articles:
|
|
|
|
 
L. Chen, D. Kost, and W. A. Dick
Flue Gas Desulfurization Products as Sulfur Sources for Corn
Soil Sci. Soc. Am. J.,
August 20, 2008;
72(5):
1464 - 1470.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. T. Berti, B. L. Johnson, R. W. Gesch, and F. Forcella
Cuphea Nitrogen Uptake and Seed Yield Response to Nitrogen Fertilization
Agron. J.,
May 7, 2008;
100(3):
628 - 634.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. S. Malhi, Y. Gan, and J. P. Raney
Yield, Seed Quality, and Sulfur Uptake of Brassica Oilseed Crops in Response to Sulfur Fertilization
Agron. J.,
March 12, 2007;
99(2):
570 - 577.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
L. M. Dosdall, G. W. Clayton, K. N. Harker, J. T. O'Donovan, and F. C. Stevenson
The Effects of Soil Fertility and Other Agronomic Factors on Infestations of Root Maggots in Canola
Agron. J.,
September 1, 2004;
96(5):
1306 - 1313.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
D. Pennock, F. Walley, M. Solohub, B. Si, and G. Hnatowich
Topographically Controlled Yield Response of Canola to Nitrogen Fertilizer
Soil Sci. Soc. Am. J.,
November 1, 2001;
65(6):
1838 - 1845.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
