CANOLA (BRASSICA NAPUS L.) RESPONSE TO SOURCE, RATE AND TIMING OF SULFUR FERTILIZER
J. R Lukach and E. J. Deibert
Langdon Research Extension Center and Soil Science Department
North Dakota State University
ABSTRACTCanola was evaluated for response to sulfur (S) fertilizer on soils with course subsoil in northeast North Dakota in 1998 and 1999. S sources used were ammonium sulfate (AS, 21-0-0-24), K Mag (0-0-20-21-10Mg-2Cl), Sulfur 95 (S95, 0-0-0-95), Tiger 90 (T90, 0-0-0-90), Tiger 90C (T90C, 0-0-0-90), ammonium thiosulphate (ATS, 12-0-0-26) and potassium thiosulphate (KTS, 0-0-25-17). The AS, K Mag, ATS and KTS sources applied pre-plant incorporated (PPI) in the spring had yields from 932 to 1267 lb/a compared to the no S check at 548 and 892 lb/a in 1998 and 1999, respectively. Rates at 40 lb S/a had the highest two year average yields but only 64 lb/a higher than 10 lb S/a. Yields were limited due to low productivity of the soils so 10 lb S/a was adequate to produce optimum yields. Elemental S sources S95, T90 and T90C PPI in the spring had yields from 592 to 1044 lb/a with 40 lb S/a giving small increases in yield over 20 lb S/a. Applications of S as ATS, KTS and AS at 10 or 20 lb S/a during the bud and bolt canola stages gave yields from 767 to 1295 lb/a. ATS and KTS, 40 lb S/a, at bud caused foliage burn and had yields of 702 to 987 lb/a. AS blended with the seed caused stand reductions at rates when more than 10 lb N/a was applied. The thin stands had 801 to 979 lb/a yields indicating the canola plants ability to compensate for low stand.
Sulfur deficiency in canola reduces yields of one of the more economical crops currently grown in North Dakota rotations. This research was initiated to address the concerns among producers about sources, rates and application timing in management of S fertilizer.
MATERIALS AND METHODS
Research in 1998 and 1999 was located near Langdon, ND in Cavalier County, on the northeast corner of Sect.12, T161N, R60W. The site has Vang-Coe soils with 3 to 6 percent slope and is located adjacent to a shale pit. These soils are inclusions in fields with high crop production potential and have exhibited severe S deficiency when cropped to canola without using S fertilizer. The site had been uniformly cropped to small grains with no canola history. The experiment was set up with two replicates each on the shoulder, mid slope and lower slope. Organic matter for the 0-6 inch depth averaged 4.2 shoulder, 3.7 mid slope and 5.0 on the lower slope positions. The lowest position had 8 to 16 inches of black topsoil while the shoulder had 1 to 8 inches of black topsoil. This indicates the variability within slope positions. Spring soil samples were collected on May 8, 1998 and May 12, 1999 with a hydraulic probe and tested at the NDSU Soil Testing Lab using standard procedures. The 1998 soil tests for N, S and Cl in 0-24 inches, was 106, 33 and 37 lbs/a. P, K, Zn, Fe, Mn and Cu in 0-6 inches was 10, 469, 1.1, 53, 25, and 0.5 ppm, respectively. The 0-6 inch depth pH was 6.1. N, P, K and S tests in 1999, on plots which received no S fertilizer in 1998, were 53, 12, 460, 17, respectively. Fertilizer was surface broadcast on the site on May 13, 1998 at 100 lb/a of 11+52+0 and 75 lb/a of 0+0+60. Fertilizer was surface broadcast on the site on May 20, 1999 at 100 lb/a of 46+0+0, 100 lb/a of 11+52+0 and 50 lb/a of 0+0+60. Fertility was adequate for over 2000 lb/a of canola.
S fertilizer treatments were selected that included various sources, application times and rates up to 60 lb S/acre. S sources were ammonium sulfate (AS, 21-0-0-24), K Mag (0-0-20-21-10Mg-2Cl), Sulfur 95 (S95, 0-0-0-95), Tiger 90 (T90, 0-0-0-90), Tiger 90C (T90C, 0-0-0-90), ammonium thiosulphate (ATS, 12-0-0-26) and potassium thiosulphate (KTS, 0-0-25-17).
The pre-plant incorporated (PPI) fertilizer was applied and incorporated on May 20 and June 1 in 1998 and 1999, respectively, and seeded the next morning. The plot size was five by 20 feet. S fertilizer PPI was applied using a spinner cone research plot seeder with seven double disk openers and six inch row spacing. Urea was blended with the S to balance treatments to a 35 lb N/a rate, equal to applying 40 lb S/a as AS. Magnesium in the K Mag treatment was not balanced. ATS and KTS treatments PPI were sprayed on the soil surface with three 20 inch spaced flat fan 8002 nozzles applying 20 gal/a solution on a hand held boom. All plots were roto-tilled once, three inches deep, in the direction of fertilizer application using a five foot wide roto-tiller on a 60 hp tractor at 2.5 mph. The experiment was planted using a plot seeder with seed depth set to one inch. Fertilizer for the drill applied treatments was packeted separately and blended at seeding. Bud and bolt applications of S fertilizer were made on June 27 and July 4 in 1998 and July 1 and 7 in 1999. Plants at the bud stage ranged from four leaf to bud formation. Bolt plant stages ranged from bud to short elongated stems with no flowers open. Variation in emergence date within the experiment occurred in both years. All liquid foliar treatments, ATS and KTS, were diluted with water and applied at 20 gal/a solution using 8002 flat fan nozzles at 35 psi.
The cultivar 'Monarch' was planted in 1998. Weed control in 1998 included liquid trifluralin herbicide at 1 lb/a PPI applied and incorporated twice with a multi-weeder after fertilizer was broadcast but before S was applied. Stinger, 0.3 pt/a plus Poast 1.5 pt/a plus 1.0 qt/a petroleum oil were applied at the 4 leaf stage and resulted in excellent weed control. After harvest, no tillage was done until incorporation of the 1999 fertilizer. Excellent weed control was achieved in 1999 by planting a Round Ready cultivar '3275' and applying one pt/a of Roundup Ultra at the four leaf stage. Crop residue was removed after harvest each year, except lower plant stems, to minimize nutrient accumulation from residue of other treatments. Canola had very low disease incidence in 1998. Canola was planted on canola ground in 1999 so fungicide applications were made for disease prevention. Four oz/a Tilt fungicide was tank mixed with the Roundup and Quadris 9.6 oz/a was applied by plane at 10 percent bloom. The seed was treated with Benlate in both years. No treatment for insects was required in either year.
Plant stand counts were taken after harvest on all plots in both years. Two rows three feet in length were counted. Yield was obtained using a Hege plot combine to straight harvest the experiment. Seed shattering before harvest was very low in 1998 and 3 to 6 percent in 1999.
The experiment was set up as a six by six lattice with six replicates. A check area plot with no S fertilizer was placed between all treated plots in 1998 to evaluate differences expected due to soil variation. Canola was planted on all plots in 1999, consecutive years. S fertilizer for 1999 was applied to the check area plots from 1998. Plots with S treatments in 1998 received no additional S and were evaluated for response to residual S fertilizer in 1999. Data analysis was completed using covariance of the 1998 check plots when appropriate to remove variation due to plot location within the experiment.
RESULTS AND DISCUSSION
Large variations in yield of the 240 no S fertilizer check area plots in 1998 occurred due to plot location within the experiment. Yield of 48 check plots which received no S fertilizer in either year had a correlation of R2 = 0.498 between years. The 1998 mean seed yield of these 48 checks was 615 lb/a with a range of 65 to 1313 lb/a. The same 48 plots with no S applied had a mean seed yield of 917 lb/a with a range of 524 to 1413 lb/a in 1999. S deficiency symptoms were severe in 1998 and minor in 1999. The higher yields in 1999 were likely due to a better distribution of precipitation later in the growing season and/or release of available S by decomposition of canola roots from of the previous growing season. The yield correlation between years indicates that some variability due to soil effects can be removed by using a covariance analysis of 1999 yield by using 1998 data. All data presented for 1998 was adjusted by covariance analysis, of the deviation of the check area on both sides of a fertilized plot, to the mean of all check area plots. 1999 yield was adjusted by covariance of the 1998 yield of the check area plot it was grown on. This procedure reduced the statistical error but did not affect the overall mean data for the trial.
Canola seed yield response to S fertilizer was different between years (Table 1). Only three treatments in 1998, Tiger 90 at 20 and 40 lb S/a and S95 at 40 lb S/a blended with the seed, did not have significantly higher yield than the no S check which had 548 lb/a. The highest yields, 932 to 1109 lb/a, were with sulfate sources applied PPI. The 10 to 20 lb S/a rates of KTS and ATS applied at bud and bolt and AS blended with the seed had 767 to 900 lb/a yield. The second year, 1999, the no S check yielded 892 lb/a with 18 S fertilizer treatments having non-significant yield increases. The highest yielding S treatments, 1101 to 1295 lb/a, included some PPI treatments, the 10 and 20 lb S/a treatments applied at bud and bolt stages and the AS sources blended with the seed. Low yields in both years were associated with the elemental S sources at all rates, ATS and KTS at 40 lb S/a rate applied at bud and when greater than 10 lb N/a was blended with the seed. K Mag at 20 and 40 lb S/a and AS at 40 lb S/a, all PPI, had high yields in both years. The AS rates which optimized canola yields were 20 lb S/a in 1998 and with 10 lb S/a in 1999 though the 40 lb S/a rates had the highest two year average yield. Responses to S fertilizer was somewhat limited by the low productivity of the soils.
The yield response to the S application during the growing season was lower than treatments applied PPI (not significantly) in 1998 but gave some of the highest yields in 1999. The amount of leaf burn associated with ATS and KTS foliar applied S also varied between years. The differences were due to seasonal precipitation distribution. Bud and bolt S applications in 1998 received 0.31 and 0.36 inches the night after application and received 0.13 inch two days after and 0.61 inch the night after application in 1999, respectively. Precipitation after the bolt stage applications until physiological maturity, totaled 1.97 and 5.21 inches in 1998 and 1999, respectively.
A visual percent leaf burn rating due to ATS and KTS foliar applications was made. ATS at 40, 20 and 10 lb S/a at bud stage in 1998 had 26, 18 and 13 percent leaf burn, respectively, while 20 lb S/a at bolt had 10 percent burn. KTS at 40, 20 and 10 lb S/a at bud stage in 1998 had 14, 9 and 4 percent leaf burn, respectively, while 20 lb S/a at bolt was 6 percent burn. ATS at 40, 20 and 10 lb S/a at bud stage in 1999 had 19, 4 and 0 percent leaf burn, respectively, while 20 lb S/a at bolt had no burn. KTS at 40, 20 and 10 lb S/a at bud stage in 1999 had 1, 0 and 0 percent leaf burn, respectively, while 20 lb S/a at bolt had no burn. The differences between years is likely related to timing of development of the heavy cuticle layer the typical of the brassica family or other environmental factors. KTS produced less leaf burn than the ATS in both years.
The experiment was seeded into good soil moisture in both years. The first rain, 1.12 inch, in 1998 was six days after seeding and five days after seeding in 1999, 0.83 inch. Plant stands were lower than for the no S check when S fertilizer was blended with the seed (Table 1). AS at 20 and 15 lb S/a had significantly lower stands in both years. Urea at 15 lb N/a blended with the seed caused a significant stand reduction, only in 1999. K Mag caused a significant stand reduction in 1999 but was not tested in 1998. AS at 10 lb S/a and elemental S at 40 lb S/a caused only small reductions in stand in either year. Canola yield was not significantly reduced by low stand in either year because the low stands were uniformly distributed across the plots with excellent weed control.
An evaluation of the residual S fertilizer effect in the following crop year was difficult to interpret due to yield variability at the site (Table 2). Individual plots, within the six replicates of a treatment, which had very low yield in the 1998 season tended to have high canola yield in 1999. The two year total seed production helps eliminate this variation. S treatments, applied in 1998, which were not significantly different from the highest yielding treatment in either year, all PPI applied, were; K Mag at 20 and 40 lb/a, AS at 20, 40 and 60 lb/a and ATS at 20 lb/a. Others treatments not significantly different from the highest yielding plot for the two year total production were; AS 10 lb/a PPI, AS drill applied at 10, 15 and 20 lb/a and ATS 10 lb/a at bud.
The S treatments,1998, not significantly different from the no S check for the two year total yield were T90 PPI at 20 and 40 lb/a, S95 PPI and through the drill at 20 and 40 lb/a and AS 20 lb/a PPI with 15 lb N/a applied through the drill. The PPI AS with urea through the drill had significantly higher yields than the check in 1998 the year of S application but not in 1999. The drill applied S95 at 40 lb/a was significantly higher in yield in 1999 as residual but not in the year of application.
Thanks to the Northern Regional Canola Grants Program, ND Oil Seed Council, Northern Canola Growers Association, IMC Kalium, Tessenderlo Kerley, Inc who provided financial assistance for sulfur research.
Ammonium Sulfate (AS, 21-0-0-24), K Mag (0-0-20-21-10Mg-2Cl), Sulfur 95 (S95, 0-0-0-95), Tiger 90 (T90,0-0-0-90), Tiger 90C (T90C, 0-0-0-90), Ammonium Thiosulphate (ATS, 12-0-0-26), Potassium Thiosulphate (KTS, 0-0-25-17), KCl =20 lb K/a
Ammonium Sulfate (AS, 21-0-0-24), K Mag (0-0-20-21-10Mg-2Cl), Sulfur 95 (S95, 0-0-0-95), Tiger 90 (T90, 0-0-0-90), Tiger 90C (T90C, 0-0-0-90), Ammonium Thiosulphate (ATS, 12-0-0-26), Potassium Thiosulphate (KTS, 0-0-25-17)