About Us | Help Videos | Contact Us | Subscriptions
 

Agronomy Journal - Article

 

 

This article in AJ

  1. Vol. 105 No. 6, p. 1597-1605
    unlockOPEN ACCESS
     
    Received: Apr 21, 2013
    Published: September 6, 2013


    * Corresponding author(s): qmk2@cornell.edu
 View
 Download
 Alerts
 Permissions
Request Permissions
 Share

doi:10.2134/agronj2013.0203

Can Manure Replace the Need for Starter Nitrogen Fertilizer?

  1. Q. M. Ketterings *,
  2. G. S. Godwin,
  3. S. N. Swink and
  4. K. J. Czymmek
  1. Dep. of Animal Science, Cornell Univ., Ithaca, NY 14853

Abstract

Current New York land grant university fertility guidelines for corn (Zea mays L.) recommend the use of 22 to 34 kg N ha–1 of band-applied starter fertilizer. On-farm research was conducted in 2006 and 2007 on a western New York dairy farm with an alfalfa (Medicago sativa L.)–corn rotation to evaluate the need for starter N for corn in regularly manured fields. In 2009, the study was expanded to include 21 fields varying in soil type, manure history, and soil N supply potential as expressed by the Illinois soil N test (ISNT). At the western New York farm, there was no increase in corn silage yield or quality with starter N use in first-, second-, or fourth-year corn fields. The results of the statewide assessment also showed that for fields with optimal ISNT-N, manure could replace starter N without a decline in corn silage yield or quality. Starter N fertilizer application was needed for optimal yield in fields deficient or marginal in ISNT-N and without a manure history. For such fields, manure could replace the need for starter N as long as sufficient N was applied with the manure, indicated by corn stalk NO3 test (CSNT) results between 750 and 2000 mg NO3–N kg–1. A response of silage yield to starter N was common when CSNT-N was <750 mg NO3–N kg–1. We conclude that manure can replace the need for starter N, but rates should be adjusted to obtain CSNT values between 750 and 2000 mg kg–1.


Abbreviations

    CP, crude protein; CSNT, cornstalk nitrate test; ISNT, Illinois Soil Nitrogen Test; NDF, neutral detergent fiber; PSNT, pre-sidedress nitrate test

Corn is a major agricultural crop in New York, covering a total area of 473,850 ha in 2012, of which 41% was harvested for silage while the remainder was harvested for grain (National Agricultural Statistics Service, 2013). In New York, the greatest N use efficiencies for fertilizer N for corn are typically obtained when the total N application is split into a small starter application at planting followed by sidedress N application when the corn is 15 to 30 cm tall (Lathwell et al., 1966, 1970). Therefore, Cornell University, the land grant university of New York, recommends the application of 22 to 34 kg N ha–1 of band-applied starter fertilizer for all corn, followed by a sidedress N application for second- or higher year corn following alfalfa–grass or for corn in other rotations if crop needs are not met with manure application (Ketterings et al., 2003).

In the past, most common starter fertilizers contained N as well as P and K to supply young seedlings with nutrients in the first 2 to 4 wk in the spring. Recent studies in the northeastern United States have shown that for fields testing high or very high in P, elimination of P in the starter band does not impact corn yield or quality (Jokela, 1992; Roth et al., 2003, 2006; Ketterings et al., 2005). A shift in starter blend from a P-containing fertilizer to a P-free fertilizer can result in substantial fertilizer and labor savings as well as a reduction in the farm’s environmental footprint (Roth et al., 2006; Ketterings et al., 2011; Ketterings and Czymmek, 2012). It is thus not surprising that New York fertilizer sales records have shown a substantial reduction in the use of starter P fertilizer in the past decade (Ketterings et al., 2011; Ketterings and Czymmek, 2012).

Previous studies in unmanured fields showed that a yield response to starter N was common (Ketterings et al., 2005) but that sidedress N could be eliminated without a yield or quality penalty for soils testing optimal in soil N supply potential as estimated by the ISNT (Klapwyk and Ketterings, 2006; Lawrence et al., 2009). Dairy manure applications add inorganic and readily available N as well as organic N. Such manure applications can increase nutrient cycling and build ISNT-N levels with time (Klapwyk et al., 2006). With a growing interest among farmers in reducing their farm’s environmental footprint while also reducing the cost of production, farmers asked whether manure could be used to replace the need for starter N. Eight research station trials and 16 on-farm trials were conducted to evaluate the impact of banded starter N use on corn yield and quality for fields varying in manure history and soil N supply potential.


MATERIALS AND METHODS

Field Trials and Experimental Design

In 2006 and 2007, three on-farm trials were conducted at a western New York dairy farm. The farm implemented three banded starter fertilizer treatments (0, 34, and 68 kg N ha–1) in three replications on three different fields, a second-year corn field after alfalfa in 2006 and first- and fourth-year corn fields in 2007. Fertilizer was applied 5 cm below and 5 cm to the side of the seed at planting. The soils for each of the trials were classified as Langford channery silt loams (fine-loamy, mixed, active, mesic Typic Fragiudepts). Fields were very high in soil test P, K, and Mg, reflecting past manure applications, with pH ranging from 6.5 to 7.0 and soil organic matter ranging from 32 to 35 g kg–1. Each plot was 4.6 m wide (12 rows, 38 cm apart) and at least 285 m long. Consistent with practical limitations with the 38-cm row width used for corn at the farm, no sidedress N was applied; the farm aimed to achieve optimal yield with a 34 kg N ha–1 starter application (urea) and fall- or spring-applied manure. Dairy manure slurry was spring injected at a rate of 84 m3 ha–1 for the second- and fourth-year fields, supplying an estimated 105 kg plant-available N ha–1 while for first-year corn, manure was fall applied at a rate of 39 m3 ha–1 for an estimated 42 kg plant-available N ha–1. Available N estimates assume 35% availability of organic N and 0 vs. 65% availability of inorganic N for fall application and spring injection, respectively (Ketterings et al., 2003, 2013). The manure application rate for second- and fourth-year corn was determined by the farmer-based on soil-derived yield potential, soil N supply potential, and N uptake efficiency. The fall application of manure at a lower rate is a typical practice on dairy farms and takes into account rotation credits for corn following alfalfa–grass sod. Spring application of manure took place 1 to 2 wk before planting, depending on soil moisture.

Soil samples were taken between rows at two depths (0–20 and 0–30 cm, 15 samples per plot) before manure application, when the corn was 15- to 30-cm tall (V4–V6 growth stage), and again at corn silage harvest. Soils were kept cool while sampling in the field. Forage subsamples were taken at harvest (3.78-L subsamples) to determine moisture content and forage quality. End-of-season CSNT samples (15 stalks per plot) were taken at harvest by cutting a 20-cm portion of the stalk between 15 and 35 cm above the ground.

From 2009 through 2011, 21 additional field trials were completed comparing yield and quality of corn as impacted by two banded starter N fertilizer treatments: 0 and 34 kg N ha–1. Fertilizer was applied 5 cm below and 5 cm to the side of the seed at planting. The trials were conducted in 10 agricultural counties of New York, representing different soils, climatic regions, and manure management histories (Table 1). Eight trials were conducted at the Cornell Musgrave Research Farm at Aurora in central New York (2009 and 2010), and 13 trials were conducted on commercial farms in the northern, eastern, western, and southern regions of New York (2009–2011) (Table 1).


View Full Table | Close Full ViewTable 1.

Location, soil series, soil taxonomy, crop history, planting and harvest dates, and manure history for 21 starter N trial sites in New York.

 
Site County Predominant soil
Previous crops Planting date Harvest date Manure history†
Trial year
Soil series Soil taxonomy Application rate and method
Manure N Side-dress N
Trial year 1 yr prior 2 yr prior
m3 ha–1 kg N ha–1
1 Cayuga Lima fine-loamy, mixed, active, mesic Oxyaquic Hapludalfs corn, 2005–2008 12 May 2009 13 Nov. 2009 75 spring aerator 75 spring aerator 89 spring aerator 174 0
2 Cayuga Lima fine-loamy, mixed, active, mesic Oxyaquic Hapludalfs corn, 2005–2008 12 May 2009 13 Nov. 2009 75 spring chisel 75 spring chisel 89 spring chisel 174 0
3 Cayuga Lima fine-loamy, mixed, active, mesic Oxyaquic Hapludalfs corn, 2005–2008 12 May 2009 13 Nov. 2009 none none none 0 134
4 Cayuga Lima fine-loamy, mixed, active, mesic Oxyaquic Hapludalfs corn, 2005–2008 12 May 2009 13 Nov. 2009 75 spring surface 75 spring surface 89 spring surface 88 0
5 Cayuga Lima fine-loamy, mixed, active, mesic Oxyaquic Hapludalfs corn, 2005–2008 11 May 2010 31 Aug. 2010 75 spring aerator 75 spring aerator 75 spring aerator 122 0
6 Cayuga Lima fine-loamy, mixed, active, mesic Oxyaquic Hapludalfs corn, 2005–2009 11 May 2010 31 Aug. 2010 75 spring chisel 75 spring chisel 75 spring chisel 122 0
7 Cayuga Lima fine-loamy, mixed, active, mesic Oxyaquic Hapludalfs corn, 2005–2009 11 May 2010 31 Aug. 2010 none none none 0 90
8 Cayuga Lima fine-loamy, mixed, active, mesic Oxyaquic Hapludalfs corn, 2005–2009 11 May 2010 31 Aug. 2010 75 spring surface 75 spring surface 75 spring surface 35 0
9 Steuben Howard loamy-skeletal, mixed, mesic Glossoboric Hapludalfs alfalfa–grass, 2007; corn, 2008–2009 7 May 2010 26 Aug. 2010 47 spring chisel 47 spring chisel 47 spring chisel NA‡ 0
10 Washington Vergennes very fine, illitic, mesic Glossaquic Hapludalfs corn, 2007–2009 28 May 2010 15 Nov. 2010 112 fall and spring disk >5 d 112 fall and spring disk >5 d 94 fall and spring disk >5 d 82 0
11 Columbia Occum coarse-loamy, mixed, mesic Fluventic Dystrochrepts corn, 2007–2009 11 May 2010 8 Sept. 2010 37 spring chisel 45§ spring chisel 45§ spring chisel 111
12 Albany Angola fine-loamy, mixed, mesic Aeric Ochraqualfs sod, 2007; corn, 2008–2009 27 May 2010 7 Sept. 2010 94 spring incorporated 37 spring incorporated 239 0
13 Washington Hoosic sandy-skeletal, mixed, mesic Typic Dystrudepts corn, 2006–2008 5 May 2009 23 Sept. 2009 56 spring incorporated none 94 spring surface 0
14 Rensselaer Occum–Barbour coarse-loamy/coarse-loamy over sandy or sandy skeletal, mesic Fluventic Dystrochrepts sod, 2007; corn, 2008–2009 10 May 2010 7 Sept. 2010 75 spring surface 84 spring surface 84 spring surface 77 0
15 Lewis Croghan sandy, mixed, frigid Aquic Haplorthods corn, 2008–2010 25 May 2011 3 Oct. 2011 56 spring chisel 1 d 56 spring chisel 1 d 56 spring chisel 1 d 67 0
16 Tompkins Hudson fine, illitic, mesic Glossaquic Hapludalfs corn, 2008–2010 14 May 2011 26 Sept. 2011 63 fall injection 69 spring injection 80 fall surface 99 fall and spring injection 161 79
17 St. Lawrence Hogansburg coarse-loamy, mixed, semiactive, frigid Aquic Eutrudepts corn, 2008–2010 13 May 2011 21 Sept. 2011 103 spring injection 137 spring injection 59 summer surface 82 28
18 Steuben Howard loamy-skeletal, mixed, active, mesic Glossic Hapludalfs corn, 2008–2010 12 May 2011 23 Sept. 2011 56§ winter 2011 surface 56§ winter 2010 surface 56§ winter 2009 surface 0
19 St. Lawrence Swanton coarse-loamy over clayey, mixed, nonacid, frigid Aeric Haplaquepts alfalfa–grass, 2007–2008; corn, 2009 4 May 2010 24 Sept. 2010 107 spring injection none 69 fall 2008 surface 143 0
20 St. Lawrence Malone coarse-loamy, mixed, nonacid, frigid Aeric Haplaquepts sod, 2007–2008; corn, 2009 29 May 2010 23 Sept. 2010 19 spring chisel 19 spring chisel 56 summer surface 61
21 Clinton Malone coarse-loamy, mixed, nonacid, frigid Aeric Epiaquepts corn, 2007–2009 10 May 2010 17 Sept. 2010 38§ winter 2009 surface 150 winter 2008 surface 11§ winter 2007 surface 56 80
Manure applied as a slurry unless otherwise indicated.
NA, not available.
§Solid application.
Manure probably applied, missing farm records for application rate and/or manure analysis.

At the Musgrave Research Farm, trials (Sites 1–8) consisted of two treatments (0 and 34 kg N ha–1 band-applied starter N as urea) with six replications of each treatment in a randomized complete block design. The treatments were imposed on plots varying in manure history and included (i) no manure (fertilizer N only); (ii) surface-applied and unincorporated manure, (iii) manure incorporated using a chisel plow; and (iv) manure incorporated using an aerator, as documented in Lawrence et al. (2008b). Where manure was applied, no sidedress N applications were performed. All plots tested deficient in ISNT-N and those that did not receive manure were sidedressed with 134 kg N ha–1 applied as urea–NH4NO3 using a four-row sidedress unit (CDS-John Blue Co.) that applies the fertilizer solution to every other interrow. All plots were harvested for corn grain.

At the on-farm sites, trials also consisted of the same two band-applied starter fertilizer treatments (0 and 34 kg N ha–1). The starter N sources varied according to individual farm practice. Urea was applied at one farm; the remainder of the farms applied either urea–NH4NO3 or (NH4)2SO4. Fields in second- or higher year corn were selected that had no need for the addition of P based on soil test results (Table 2) and had documented manure applications and manure histories (rates and approximate timings of application; Table 1). The corn trials were conducted using 76-cm-wide rows (except for Site 16 which had twin [56/20-cm wide] rows) and replicated four (Sites 10–12, 14–18, 20, and 21) or five times (Site 9) in a randomized complete block design. Plots were 4 to 16 rows wide depending on planter and chopper width (typically two times the chopper width for machine-harvested sites) and 30 to 600 m long, depending on farm equipment and field size. Soils in each plot were sampled at two depths (0–20 and 0–30 cm) when the corn was 15- to 30-cm tall in June and at harvest, consistent with the sampling and sample processing procedures used for the trials at the western New York farm. Stand density was determined midseason or at harvest time by counting two rows of corn plants in 12 m per plot. Where sidedress N applications occurred, N application rates were documented (Table 1). All on-farm sites were harvested for corn silage.


View Full Table | Close Full ViewTable 2.

Initial soil fertility status (0–20-cm depth) for each of the 21 sites (means of plot values) included in the starter N project. All soils were analyzed for pH (water), organic matter (OM) by loss-on-ignition, Morgan-extractable P, K, Mg, Ca, Al, Mn, Zn, and Illinois Soil Nitrogen Test (ISNT). Trials at Sites 1 to 4 and 13 were conducted in 2009, at Sites 5 to 12, 14, and 19 to 21 in 2010, and all others in 2011.

 
Morgan-extractable elements†
P K Mg ISNT
Site pH OM Conc. Class Conc. Class Conc. Class Ca Al Mn Zn Conc. Ratio‡ Rating§
g kg–1 mg kg–1 mg kg–1 mg kg–1 mg kg–1 mg kg–1
1 7.8 35 14 high 108 very high 371 very high 2971 6 14 1 264 0.92 D
2 7.8 33 14 high 105 very high 348 very high 2833 6 14 1 252 0.89 D
3 7.7 31 8 high 64 high 331 very high 2636 6 13 <1 224 0.81 D
4 7.8 33 16 high 112 very high 353 very high 2806 5 15 1 252 0.89 D
5 7.7 37 16 high 134 very high 343 very high 2806 5 20 1 257 0.87 D
6 7.7 35 15 high 122 very high 334 very high 2876 5 20 1 248 0.86 D
7 7.7 33 9 high 58 high 312 very high 2738 5 18 <1 230 0.82 D
8 7.7 35 17 high 132 very high 343 very high 2946 5 19 1 245 0.85 D
9 6.1 33 5 high 99 high 222 very high 1069 17 10 <1 235 0.83 D
10 7.0 41 13 high 122 very high 292 very high 2674 18 20 1 247 0.81 D
11 6.4 28 40 very high 389 very high 192 very high 1491 10 27 2 216 0.81 D
12 6.6 35 36 very high 167 very high 162 very high 2214 9 11 1 290 1.01 M
13 6.4 49 47 very high 480 very high 213 very high 1782 15 17 3 336 1.05 M
14 6.9 40 18 high 305 very high 212 very high 1868 14 17 2 315 1.05 M
15 6.2 52 3 medium 131 high 157 very high 1108 71 3 2 314 0.97 M
16 7.4 41 40 very high 234 very high 274 very high 2733 6 29 1 324 1.07 M
17 6.7 41 6 high 107 very high 256 very high 1851 8 23 1 356 1.17 O
18 6.8 53 82 very high 783 very high 367 very high 1690 11 16 2 439 1.35 O
19 7.0 42 8 high 75 medium 462 very high 2616 8 13 1 334 1.09 O
20 7.0 41 8 high 32 low 312 very high 2298 10 11 1 344 1.13 O
21 6.9 43 25 very high 288 very high 275 very high 1717 12 13 2 344 1.12 O
Morgan-extractable P, K, and Mg interpretations are defined in Cornell Cooperative Extension (2012). A soil is classified as medium in soil test P when the Morgan-extractable P is between 2 and 4.5 mg kg–1 high in soil test P with a test result between 4.5 and 20 mg kg–1, and very high when the Morgan soil test P exceeds 20 mg kg–1. Soils are very high in Mg when exceeding 100 mg kg–1. Potassium soil test interpretations depend on soil type, as documented in Ketterings et al. (2003).
ISNT-N/critical ISNT-N ratio, where the critical ISNT-N level for a given soil was determined according to Klapwyk and Ketterings (2006) and Lawrence et al. (2009). ISNT-N was measured in 0–20-cm depth samples taken when the corn was 15–30-cm tall (V4–V6 stage). Test results are classified as “deficient in soil N supply potential” when the ratio is <0.93, “marginal in soil N supply potential” when the ratio is between 0.93 and 1.07, and “optimal in soil N supply potential” with a ratio >1.07.
§O, optimal; M, marginal; D, deficient.

For the starter N trial conducted in 2006 at the western New York dairy farm, corn population density data are not available. In addition, in 2010 at the Musgrave Research Farm, corn population density was measured across treatments at each site so only means are reported. For all other trials, corn population density was determined per treatment and replication by counting all plants in two 12-m rows.

Soil Analyses

Following standard procedures for soil preparation (Greweling and Peech, 1965), all soils were oven dried (50°C) for at least 48 h and ground to pass a 2-mm sieve. General fertility was determined from the 0- to 20-cm-depth soil samples taken midseason, prepared following standard soil preparation procedures at Cornell University, and analyses were performed using methods described in Wolf and Beegle (1995). Briefly, soils were analyzed for pH (1:1 w/v water extract), soil organic matter by loss-on-ignition (Storer, 1984), and Morgan (0.72 mol L–1 NaOAc + 0.52 mol L–1 CH3COOH) extractable P, K, Ca, and Mg (Morgan, 1941). For the Morgan extraction, samples were shaken in a 1:5 (v/v) soil/solution ratio for 15 min and filtered through a Whatman no. 2 filter paper. Morgan-extractable PO4–P was measured colorimetrically (Murphy and Riley, 1962) using an Alpkem automated rapid flow analyzer (RFA/2-320) (OI Corp.). Potassium, Ca, and Mg were analyzed by inductively coupled plasma atomic emission spectroscopy (ICP–AES) using a JY70 Type II ICP–AES (Jobin Yvon). Samples were analyzed for ISNT-N according to Khan et al. (2001) with the enclosed-griddle modification (Klapwyk and Ketterings, 2005) and classified for N supply potential based on the ISNT-N/critical ISNT-N ratio (Table 2). Eleven sites with a ratio <0.93 were classified as “deficient in soil N supply potential,” five sites were “marginal in soil N supply potential” (ratio 0.93–1.07), while the remaining five sites were “optimal in soil N supply potential” (ratio >1.07), where the critical ISNT-N level for a given soil was determined according to Klapwyk and Ketterings (2006) and Lawrence et al. (2009).

Soils were classified as high (14 sites) or very high (six sites) in P with the exception of one location where the soil test results classified the site as medium in P (Site 15; Table 2). A response to P was not expected at this location because manure was applied (Table 1). Similarly, all sites were high or very high in soil test K with the exception of Site 19 (medium) and Site 20 (low). At both locations, manure was applied and plants did not display any K deficiencies during the growing season, so both sites were retained for the overall study.

Soil samples taken at the western New York farm, the 0- to 20-cm soil samples (both timings), and the 0- to 30-cm depth samples taken midseason for the statewide project were analyzed for NO3–N using the Morgan extraction (Morgan, 1941). Morgan-extractable NO3–N for 0- to 30-cm-depth samples taken midseason is the standard pre-sidedress NO3 test (PSNT) for New York (Klausner et al., 1993).

Harvest, Silage Quality Analyses, and the End of Season Stalk Nitrate Test

At the Musgrave Research Farm, corn was machine harvested for grain at a targeted grain moisture of 160 to 200 g kg–1. Plots were harvested using a Case IH 2144 combine and grain was transferred to an Unverferth 275 gravity wagon situated on four Intercomp PT300DW-5 wheel load scales. For each plot, a grain subsample was taken and dried for 10 d at 65°C to determine moisture content.

For all other trials, corn was harvested as silage at a targeted whole-plant moisture of 600 to 700 mg kg–1. Trials at the western New York dairy farm and 17 of the statewide sites were machine harvested, while five statewide trials (Sites 9, 11, 12, 18, and 20) were hand harvested. For the machine-harvested trials, choppers harvested the inner six to eight rows of individual plots in one pass (minimum plot length of 83 m), with loads weighed on farm or mobile truck (axle) scales before and after each pass through a plot to determine harvest weight. A calibrated yield monitor was used at two sites (Sites 17 and 19). For the hand-harvested trials, two 10- to 12-m-long rows of corn silage were hand harvested 15 cm above the ground (Sites 9, 11, 12, 18, and 20).

At harvest, a 20-cm portion of stalk (between 15 and 35 cm above the ground according to Binford et al., 1990) was collected from 15 plants per plot. Stalk portions were quartered lengthwise. One of four quarters was retained, dried at 60°C in a forced-air oven for a minimum of 48 h, ground to pass a 2-mm screen, and analyzed for extractable NO3 using a 0.025 mol L–1 Al2(SO4)3 solution, an extraction ratio of 1:100 (w/v), and a shaking time of 15 min. Extractable NO3–N was determined using a VWR SympHony NO3 ion electrode following Miller (1998).

Except for the 2006 trial at the western New York farm, a subsample of approximately 3.78-L volume of the harvested silage per plot was collected at the bunk for all machine-harvested silage trials. Per plot, seven to 10 grab samples were collected at varying depths to get a representative sample for moisture and forage quality. For hand-harvested trials, a five-plant subsample from each plot was chopped in the field using a Model 120312 Mighty Mac, a gas-powered chipper-shredder (Mackissic Inc.). The shredded corn was well mixed, subsampled to fill a 3.78-L plastic bag, sealed, and kept in a cooler during transport to the laboratory, where the samples were dried in a 60°C forced-air oven for a minimum of 48 h.

Forage subsamples were analyzed at Cumberland Valley Analytical Services in Hagerstown, MD. The oven-dried samples were ground to pass a 1-mm screen, subsampled, and analyzed for in vitro neutral detergent fiber (NDF) 48-h digestibility (Goering and Van Soest, 1970). Near-infrared reflectance spectroscopy was used to determine crude protein (CP), soluble protein, acid detergent insoluble CP, neutral detergent insoluble CP, acid detergent fiber, NDF, lignin, sugar, starch, crude fat, ash, Ca, P, Mg, and K contents. Milk2006, a model developed at the University of Wisconsin, was used to estimate yields in milk per megagram of silage and per hectare (Shaver, 2006).

Statistical Analysis

Given the large variability in field characteristics and manure histories across sites, yield, forage quality, and soil and stalk data were analyzed for each site independently using PROC MIXED (Littell et al., 1996, p. 87–134), with N treatment as a fixed effect and block as a random effect using SAS (SAS Institute). Mean separations were done using the LSMEANS procedure with TUKEY adjustment at P ≤ 0.05.


RESULTS AND DISCUSSION

Western New York On-Farm Trials

Starter N application did not increase corn silage yield or impact moisture at harvest for any of the three trials conducted at the western New York farm (Table 3). Eliminating starter N did not impact silage quality parameters in the fourth-year corn site in 2007. For the first-year corn field, adding 67 kg N ha–1 did significantly increase CP; however, the increase in CP did not impact the overall silage quality expressed in estimated milk per megagram of silage or milk per hectare (Table 3).


View Full Table | Close Full ViewTable 3.

Corn silage dry matter (DM) yield expressed at 650 g kg–1 moisture, moisture content at harvest (MC), corn population, and silage quality response to banded starter N applications at a western New York dairy farm, expressed in terms of milk production estimates per megagram DM and per hectare, crude protein (CP), soluble protein (SP), neutral detergent fiber (NDF), digestible neutral detergent fiber at 48 h (dNDF 48h), lignin, and starch contents. Fields received manure injected in the spring at a rate of 85 m3 ha–1.

 
Starter N rate DM yield MC Corn population Milk production† CP SP NDF dNDF 48h Lignin Starch
kg N ha–1 Mg ha–1 g kg–1 plants ha–1 kg Mg–1 kg ha–1 g kg–1 DM g kg–1 NDF g kg–1 DM
Second-year corn after alfalfa (2006)
0 59 a‡ 638 a
34 58 a 652 a
67 58 a 629 a
First-year corn after alfalfa (2007)
0 61 a 595 a 81,400 a 1730 a 37,000 a 62 b 15 a 438 a 701 a 29 a 356 a
34 61 a 597 a 79,600 a 1690 a 36,200 a 67 ab 15 a 460 a 691 a 33 a 329 a
67 62 a 581 a 78,900 a 1680 a 36,500 a 71 a 17 a 463 a 698 a 32 a 325 a
Fourth-year corn after alfalfa (2007)
0 40 a 628 a 79,600 a 1690 a 25,000 a 77 a 21 a 448 a 690 a 32 a 337 a
34 42 a 631 a 78,200 a 1700 a 24,200 a 75 a 20 a 444 a 692 a 31 a 337 a
67 41 a 640 a 80,300 a 1730 a 37,000 a 62 b 15 a 438 a 701 a 29 a 356 a
Milk per megagram and milk per hectare are indicators of overall forage quality estimated using Milk2006 (Shaver, 2006).
Means for starter N rates within a field that are followed by different letters showed a significant change in quality indicators with the use of starter N (P ≤ 0.05).

Soil PSNT levels exceeded 21 mg kg–1, the critical value for corn responsiveness (Klausner et al., 1993), at all three sites including the first-year corn site, suggesting a sufficient N supply through the manure applications, mineralization of organic matter, and sod decomposition (Table 4). Soil ISNT levels were classified as optimal for the second- and fourth-year corn sites, while the first-year corn site was classified as marginal, suggesting that manure can replace the need for starter N for sites at or above the critical ISNT-N level determined by Klapwyk and Ketterings (2006) and validated for corn by Lawrence et al. (2009). Corn stalk NO3 test results and end-of-season soil NO3 data (Table 4) both reflected the higher soil NO3 and ISNT-N levels at the second- and fourth-year corn sites than at the first-year corn site, consistent with a larger number of years of annual manure applications for those two sites. Corn stalk N levels for the first-year corn site were classified as marginal, while no yield response to starter N was measured. This is consistent with previous work that indicated lower threshold levels for PSNT (Morris et al., 1993; Yost et al., 2013a) and for CSNT (Lawrence et al., 2008a; Yost et al., 2012, 2013a, 2013b) for corn following sod in the rotation, and is consistent with the lower manure N application rate and fall application of manure at this site. These results suggest that starter N fertilizer can be eliminated without impacting yield or silage quality for regularly manured fields at or above the critical ISNT-N level for the field.


View Full Table | Close Full ViewTable 4.

Pre-sidedress NO3 test N (PSNT-N), end-of-season soil NO3–N, both measured in 0- to 30-cm depth samples, Illinois soil N test N (ISNT-N), measured in 0- to 20-cm depth samples taken when the corn was 15- to 30-cm tall (V4–V6 stage), and cornstalk NO3 test N (CSNT-N) following three rates of banded starter N applications to silage corn at a western New York dairy farm. Manure was injected in the spring at a rate of 85 m3 ha–1 for an estimated available N application of 105 kg N ha–1 for second- and fourth-year corn, and fall-applied at a rate of 39 m3 ha–1 for an estimated 42 kg plant-available N ha–1. Available N estimates assume 35% availability of organic N and 0 and 65% availability of inorganic N for fall application and spring injection, respectively (Ketterings et al., 2003, 2013).

 
Starter N rate ISNT-N PSNT-N CSNT-N End-of-season soil NO3–N
kg N ha–1 mg kg–1
Second-year corn after alfalfa (2006)
0 322 a† 103 a 9774 a 29 a
34 344 a 98 a 9759 a 27 a
67 340 a 105 a 9857 a 33 a
First-year corn after alfalfa (2007)
0 271 a 24 a 302 a 6 a
34 283 a 22 a 391 a 4 a
67 274 a 24 a 308 a 7 a
Fourth-year corn after alfalfa (2007)
0 302 a 37 a 3073 a 12 a
34 291 a 33 a 4119 a 8 a
67 329 a 43 a 2530 a 14 a
Means for starter N rates within a field that are followed by the same letter are not statistically different (P ≤ 0.05).

Statewide Assessment

Similar to the findings for the western New York sites, at sites with an optimal soil N supply potential as determined by the ISNT-N/critical ISNT-N ratio (Sites 17, 18, 19, 20, and 21), the manure application alone was sufficient to meet the N needs of the crop; none of these five sites showed a yield increase with starter N use, and moisture at harvest was not impacted (Table 5). The CSNT-N, PSNT-N, and end-of-season soil NO3 data


View Full Table | Close Full ViewTable 5.

Corn dry matter (DM) yield expressed at 150 g kg–1 moisture for grain yields (Sites 1-8) or 650 g kg–1 moisture for corn silage harvests (all other sites), moisture content at harvest (MC), corn population, and silage quality expressed in terms of milk production estimates per megagram DM and per hectare, crude protein (CP), soluble protein (SP), neutral detergent fiber (NDF), digestible neutral detergent fiber (dNDF), lignin, and starch contents as influenced by starter N applications of 34 kg N ha–1 at planting in 2009 (Sites 1– 8 and 13), 2010 (Sites 5–12, 14, and 19–21), and 2011 (Sites 15–18). Sites 1 to 8 were harvested for grain. All other sites were harvested for corn silage.

 
Site Treatment Yield MC Corn population Milk production CP SP NDF dNDF Lignin Starch
Mg ha–1 g kg–1 plants ha–1 kg Mg–1 DM kg ha–1 g kg–1 DM g kg–1NDF g kg–1 DM
Sites deficient in ISNT-N
1 starter 7.01 a† 182 a 70,600‡
no starter 6.77 a 174 a 70,600‡
2 starter 7.40 a 183 a 72,900‡
no starter 6.54 b 176 a 72,900‡
3 starter 8.96 a 181 a 72,600‡
no starter 7.89 a 190 a 72,600‡
4 starter 6.43 a 185 a 71,300‡
no starter 5.69 a 182 a 71,300‡
5 starter 9.36 a 166 b 68,900 a
no starter 8.61 b 172 a 60,900 b
6 starter 10.00 a 166 a 71,900 a
no starter 9.42 a 169 a 68,800 a
7 starter 10.77 a 169 b 68,100 a
no starter 9.14 b 173 a 59,500 b
8 starter 8.78 a 167 a 68,400 a
no starter 7.80 b 170 a 64,700 a
9 starter 43.0 a 671 a 63,500 a 1720 a 25,900 a 80 a 16 a 464 a 676 a 35 a 293 a
no starter 44.8 a 670 a 63,300 a 1760 a 27,600 a 79 a 16 a 438 a 665 a 33 a 314 a
10 starter 40.3 a 683 a 72,700 a 1800 a 25,300 a 83 a 20 a 393 a 702 a 28 a 345 a
no starter 42.8 a 677 a 72,100 a 1830 a 27,500 a 78 a 19 b 375 a 702 a 27 a 372 a
11 starter 55.3 a 601 a 93,600 a 1680 a 32,500 a 83 a 22 a 470 a 612 a 36 a 287 a
no starter 55.9 a 612 a 92,800 a 1670 a 32,800 a 83 a 24 a 461 a 606 a 35 a 300 a
Sites marginal in ISNT-N
12 starter 42.8 a 650 a 80,000 a 1780 a 26,600 a 78 a 18 a 405 a 698 a 28 a 346 a
no starter 44.8 a 658 a 76,800 a 1780 a 28,000 a 79 a 20 a 396 a 673 a 27 a 356 a
13 starter 56.9 a 673 a 62,100 a 1730 a 34,500 a 83 a 24 a 422 a 652 a 32 a 336 a
no starter 55.8 a 656 a 61,800 a 1710 a 33,400 a 73 b 21 b 425 a 641 a 30 a 347 a
14 starter 47.5 a 596 a 93,700 a 1790 a 29,700 a 78 a 21 a 400 a 643 a 31 a 404 a
no starter 46.1 a 579 a 93,700 a 1760 a 28,500 a 77 a 22 a 411 a 647 a 31 a 386 a
15 starter 47.5 a 585 a 78,300 a 1650 a 27,300 a 63 a 13 a 426 a 608 b 31 a 380 a
no starter 38.5 b 589 a 77,200 a 1640 a 22,100 b 58 a 11 a 444 a 632 a 29 a 362 a
16 starter 39.2 a 590 a 77,400 a 1670 a 22,900 a 91 a 22 a 414 a 795 b 23 a 349 a
no starter 38.1 a 595 a 77,500 a 1690 a 22,500 a 89 a 22 a 416 a 806 a 23 a 348 a
Sites optimal in ISNT-N
17 starter 56.2 a 494 a 81,600 a 1690 b 33,100 a 77 b 19 b 401 a 567 a 32 a 398 a
no starter 54.9 a 587 a 81,500 a 1730 a 33,200 a 80 a 20 a 388 a 582 a 31 a 409 a
18 starter 51.1 a 653 a 77,300 b 1690 a 30,100 a 88 a 24 a 413 a 638 a 32 a 317 a
no starter 54.0 a 659 a 79,800 a 1660 a 31,400 a 90 a 24 a 413 a 625 b 33 a 314 a
19 starter 50.9 a 498 a 75,300 a 1850 a 43,900 a 81 a 17 a 364 a 748 a 24 a 436 a
no starter 49.7 a 501 a 76,500 a 1860 a 43,100 a 81 a 18 a 342 a 721 a 24 a 461 a
20 starter 44.8 a 679 a 75,500 a 1710 a 26,800 a 79 a 21 a 461 a 646 a 35 a 308 a
no starter 47.3 a 665 a 81,700 a 1700 a 28,000 a 76 a 20 a 460 a 631 a 33 a 317 a
21 starter 53.3 a 679 a 78,800 a 1860 a 34,600 a 92 a 25 a 359 b 775 a 24 a 379 a
no starter 52.9 a 689 a 78,000 a 1780 a 33,000 a 89 b 25 a 396 a 770 a 25 a 341 a
Sites with treatment means followed by different letters showed a significant change in quality indicators with the use of starter N (P ≤ 0.05).
Only one stand density (mean of replications) available for combined starter/no starter at this site.

(Table 6) confirmed that N was not limiting yield at any of these sites. The CSNT-N, PSNT-N, and end-of-season NO3 data also suggest that for the two locations that were sidedressed (Sites 20 and 21), the sidedress application could have been eliminated without impacting yield.


View Full Table | Close Full ViewTable 6.

Illinois soil N test (ISNT) ratio and rating, soil NO3–N (0–20- and 0–30-cm depths), pre-sidedress NO3 test (PSNT), and corn stalk NO3 test (CSNT) as influenced by starter N fertilizer (0 vs. 34 kg N ha–1) in corn trials in 2009 (Sites 1–4 and 13), 2010 (Sites 5–12, 14, and 19–21), and 2011 (Sites 15–18).

 
At sidedress time
At harvest
ISNT
NO3–N
PSNT, 0–30 cm§
NO3–N
CSNT¶
Site Ratio† Rating‡ Treatment 0–20 cm Conc. Class 0–20 cm 0–30 cm Conc. Class
mg kg–1 mg kg–1
Sites deficient in ISNT-N
1 0.91 D starter 6 b# 7 a Class 12 b 5 a 94 a deficient
no starter 9 a 8 a deficient 14 a 7 a 90 a deficient
2 0.90 D starter 9 a 12 a deficient 11 b 6 a 94 a deficient
no starter 11 a 9 a deficient 14 a 7 a 105 a deficient
3 0.88 D starter 2 b 6 a deficient 8 a 4 a 160 a deficient††
no starter 6 a 4 a deficient 9 a 5 a 208 a deficient††
4 0.90 D starter 7 a 9 a deficient 10 a 5 a 104 a deficient
no starter 7 a 7 b deficient 13 a 6 a 94 a deficient
5 0.88 D starter 34 a 28 a sufficient 11 a 15 a 182 a deficient
no starter 37 a 28 a sufficient 10 a 17 a 99 a deficient
6 0.86 D starter 32 a 31 a sufficient 11 a 18 a 80 a deficient
no starter 34 a 26 a sufficient 11 a 15 a 89 a deficient
7 0.82 D starter 18 a 14 a deficient 9 a 16 a 827 a optimal††
no starter 18 a 13 a deficient 9 a 14 a 669 a marginal††
8 0.85 D starter 29 a 24 a marginal 11 a 15 a 129 a deficient
no starter 32 a 25 a sufficient 11 a 15 a 83 a deficient
9 0.84 D starter 42 a 57 a sufficient 7 a 7 a 1661 a optimal
no starter 40 a 54 a sufficient 5 a 5 b 463 b marginal
10 0.81 D starter 33 a 33 a sufficient 20 a 33 a 2552 a excess
no starter 33 a 31 a sufficient 16 a 25 a 1174 a optimal
11 0.81 D starter 65 a 52 a sufficient 79 a 44 a 7838 a excess††
no starter 71 a 45 a sufficient 66 a 53 a 5938 a excess††
Sites marginal in ISNT-N
12 1.01 M starter 48 a 31 a sufficient 10 a 10 a 1225 a optimal
no starter 38 a 33 a sufficient 12 a 9 a 818 a optimal
13 1.07 M starter 30 a 34 a sufficient 32 a 27 a 5154 a excess
no starter 26 a 30 a sufficient 24 a 27 a 5017 a excess
14 1.05 M starter 62 a 55 a sufficient 21 a 18 a 10135 a excess
no starter 59 a 53 a sufficient 13 b 11 a 9164 a excess
15 0.97 M starter 19 a deficient 704 a marginal
no starter 20 a deficient 762 a optimal
16 1.07 M starter 21 a marginal 2129 a excess††
no starter 19 a deficient 1308 a optimal††
Sites optimal in ISNT-N
17 1.17 O starter 20 a deficient 2970 a excess††
no starter 21 a marginal 1353 b optimal††
18 1.35 O starter 48 a sufficient 3449 a excess
no starter 44 a sufficient 5872 a excess
19 1.10 O starter 40 a 29 a sufficient 15 a 14 a 4817 a excess
no starter 41 a 33 a sufficient 16 a 16 a 4164 a excess
20 1.13 O starter 27 a 25 a sufficient 19 a 16 a 4484 a excess††
no starter 29 a 27 a sufficient 19 a 16 a 4599 a excess††
21 1.12 O starter 21 a 24 a marginal 34 a 24 a 9326 a excess††
no starter 25 a 23 a marginal 45 a 33 a 10051 a excess††
ISNT-N/critical ISNT-N ratio, where the critical ISNT-N level for a given soil was determined according to Klapwyk and Ketterings (2006) and Lawrence et al. (2009). ISNT-N was measured in 0–20-cm depth samples taken when the corn was 15–30-cm tall (V4–V6 stage). Test results are classified as “deficient in soil N supply potential” when the ratio is <0.93, “marginal in soil N supply potential” when the ratio is between 0.93 and 1.07, and “optimal in soil N supply potential” with a ratio >1.07.
O, optimal; M, marginal; D, deficient.
§PSNT-N interpretation: <21 mg kg–1 is deficient; 21–24 mg kg–1 is marginal; >24 mg kg–1 is sufficient (Ketterings et al., 2012).
CSNT-N interpretation: <250 mg kg–1 is deficient; 250–750 mg kg–1 is marginal; 750–2000 mg kg–1 is optimal; >2000 mg kg–1 is excessive (Lawrence et al., 2012).
#Sites with treatment means followed by different letters showed a significant change (P ≤ 0.05) in N indicators with the use of starter N.
††Sidedressed in addition to receiving manure (Sites 11, 16, 17, 20, and 21) or sidedressed with no manure history (Sites 3 and 7).

Of the five sites that were classified by ISNT-N levels as marginal in soil N supply potential (Table 2), all received manure and only one (Site 15) showed a yield response to starter N use. The PSNT-N results suggested sufficient N for four of the six sites, while two sites (Sites 15 and 16) indicated a potential deficiency in N. Site 16 was sidedressed to meet N needs (optimal CSNT-N), while at Site 15, the marginal CSNT-N classification was consistent with the yield response to starter N under N-deficient conditions. The additional sites were classified as optimal (Site 12) or excessive (Sites 13 and 14) in N availability based on CSNT-N results (Table 6). We conclude based on these data that manure application can replace starter N for soils with a marginal soil N supply potential as long as sufficient N is added with the manure as confirmed by a CSNT-N >750 mg kg–1.

The sites classified as deficient in soil N supply potential (i.e., soil N alone is not expected to supply sufficient N for the corn crop that year) included two unmanured sites at the Musgrave Research Farm (Sites 3 and 7) and sites with a limited manure history (Sites 1, 2, and 4 in 2009 and 5, 6, and 8 in 2010 at the Musgrave Research Farm) as well as three on-farm sites (Sites 9, 10, and 11). The results at Sites 3 and 7 (significantly higher yield in 2010 with starter N application and a similar trend in 2009 [P = 0.063], a year with below-average precipitation in April and July) suggest that starter N is needed for unmanured fields that are deficient in ISNT-N. The results at Site 7 also suggest that a response to N could have been expected if CSNT values were <750 mg kg–1 (high-producing year on deficient ISNT soil). Sites 3 and 7 represent scenarios typically seen at cash grain operations where corn is grown without manure. In these scenarios, the best management practice remains to use starter N (22–34 kg N ha–1) and sidedress N where needed, consistent with the response to starter N use documented for unmanured sites in Ketterings et al. (2005).

At Sites 1, 2, and 4 in 2009 and 5, 6, and 8 in 2010, liquid manure had been applied at a rate of ∼75 m3 ha–1 yr–1 during the past 5 to 6 yr. Manure application can increase ISNT-N with time (Klapwyk et al., 2006), but after 5 to 6 yr of liquid manure application at this location, the ISNT-N levels of these sites were still classified as deficient, reflecting low solid contents of the manure, consistent with findings documented by Klapwyk et al. (2006). Of these six sites, three showed a significant (P < 0.05) yield increase with starter N fertilizer addition, while a similar trend was seen for two other sites (P = 0.116 and 0.071 for Sites 4 and 6, respectively) (Table 5). The corn grown on these sites exhibited deficient CSNT-N levels as well (Table 6), suggesting that the current-year spring application was insufficient to supply the N needed by the crop, consistent with a yield response to starter N fertilizer. Of the remaining three on-farm sites with low soil N supply potential, two sites had optimal CSNT-N levels (without starter) and one had excessive CSNT-N (Site 11, sidedressed). The latter site also showed the highest PSNT-N and end-of-season soil NO3 levels, consistent with excessive CSNT-N results and suggesting that the sidedress N application that took place at this site was not needed for optimal yield. The lack of a corn yield response to starter N at Sites 9, 10, and 11 illustrated that for these three locations, the current-year manure supplied sufficient N.

The impact of starter N use on corn silage quality was infrequent and inconsistent. Of the 13 silage trials, two locations showed a significant increase in CP with starter N addition (Sites 13 and 21), while at one site, CP declined with starter N addition (Site 17) (Table 5). Soluble protein increased with starter N application at two locations, although the difference was small (an increase of 1 and 3 g kg–1 at Sites 10 and 13, respectively), and decreased at one site (Site 17). Only one site showed a change in NDF (decrease, Site 21). At Site 18, NDF digestibility increased with starter N addition, while at two additional sites, NDF decreased with starter N addition (Sites 15 and 16). Lignin and starch were not impacted by starter N fertilizer use at any of the silage trials. Elimination of starter N did not result in significant differences in milk per megagram of silage estimates except for one site where starter use decreased the estimated milk production (Site 17). Milk per hectare estimates were only impacted at one site (an increase with starter N addition at Site 15), consistent with the yield increase with starter N use at this location.


CONCLUSIONS

Starter N should be used for corn fields with no manure history and no current-year manure applications (sites deficient in ISNT-N). If the ISNT-N is classified as optimal, manure can be used to replace starter N without a yield or quality penalty. Manure can replace starter N for sites deficient or marginal in ISNT-N as well, but only if sufficient N from the manure and other sources is available (CSNT-N between 750 and 2000 mg kg–1). A yield response to starter N is likely if the ISNT-N is deficient and the additional N applied with manure is insufficient. Corn grown in manured fields and with CSNT-N levels between 750 and 2000 mg kg–1, using 20-cm stalks taken between 15 and 36 cm above the ground, did not respond to starter N use. We recommend that producers analyze second- or higher year corn fields for both ISNT-N and CSNT-N to evaluate past-season N management and identify sites where a starter N application can be omitted without impacting yield or silage quality.

Acknowledgments

The project was funded with federal formula funds and Northern New York Agricultural Development Program (NNYADP) funds. We thank B. Boerman (Agricultural Consulting Service), E. Young (Miner Institute), P. Barney (Barney Agronomic Services), and Cornell Cooperative Extension educators C. Albers, P. Barney, S. Canner, P. Cerosaletti, A. Gabriel, M. Hunter, T. Kilcer, J. Lawrence, and A. Wright for their assistance in carrying out the statewide on-farm trials. We also thank the many dairy farmers who collaborated on this project and SUNY Cobleskill interns Joseph Foster and Eun Hong for their help in the field.

 

References

Footnotes


Comments
Be the first to comment.



Please log in to post a comment.
*Society members, certified professionals, and authors are permitted to comment.