Integrating Cropping Practices to Improve Nutrient Management Plans and Ensure Environmental and Economic Sustainability in Dairy Farming Systems

1. Location, Fields, and Cultural Management

This on-farm study was performed at a 425-cow commercial dairy farm located in Crockett, VA. The study was performed in two fields. Field 1 (Dairy) is adjacent to the dairy facilities, whereas field 2 (Remote) is 1.3 km away from the dairy facilities. The distance from fields to the dairy facilities has affected the application rates of manure over the years; therefore, the fertility was quite different between fields leading to two diverse growing environments.

During the springs of 2017 and 2018, a conventional (CONV) corn hybrid (TMF17L86, Mycogen Seeds, Indianapolis, IN) and a brown midrib (BMR) corn hybrid (F2F817, Mycogen Seeds) were planted at a theoretical seeding rate (i.e., treatments) of 60,000, 75,000, and 90,000 seeds/ha (60K, 75K, and 90K, respectively). Four plots (2.3-m wide and 25-m long) were planted for each treatment and hybrid, therefore, leading to 24 plots for each field in each growing season. Within each field, replicates were randomly assigned for each of the three seeding rates. Corn was planted with a 6-row John Deere 1750 no-till planter (Deere & Company, Moline, IL) equipped with a pneumatic dosing machine and with rows separated by 76 cm. Each of the two corn hybrids was placed in three of the six seed hoppers of the planter. In 2017, the preceding crops for fields 1 and 2 were rye for silage and corn for grain, respectively. In 2018, the preceding crop for both fields was rye for silage. Weed and pest control was performed according to on-farm operating procedures.

Fields were also fertilized according to on-farm operating procedures. In both years, 50,000 L/ha of dairy manure slurry (0.4% N) was spread on both fields 10 days before planting, and 55 kg N/ha as urea ammonium nitrate (UAN) was injected in the soil during planting. When the crop showed 6 visible leaves (V6), all plots were split into half plots (hereafter known as subplot), and each subplot was fertilized with either 50 or 100 kg N/ha as UAN (1N and 2N, respectively).

2. Soil Sampling and Analyses

Soil samples were collected at planting and harvesting. At planting, 20 core subsamples (0 to 20 cm) were collected in a random pattern throughout each field and composited providing 1 sample per field each year. At harvesting, due to the extreme hardness of the soil, 5 subsamples were collected from each subplot of the 60K and 90K treatments. For the latter, soil subsamples (0 to 10 cm) were collected from the center rows’ interspace using a drill with a 25.4-mm auger bit (Robert Bosch, LLC; Farmington Hills, MI).  

At arrival to the laboratory, soil samples were spread onto metallic trays and allowed to air-dry at room temperature (25°C). Once dried, crumbs’ size was reduced by crushing the soil with a mortar and pestle and sieving through a 2-mm sieve (Hogentogler & Co., Inc; Columbia, MD). The P concentration in soil was determined after extraction in Mehlich 1 solution (0.05 N HCl + 0.025 N H₂SO₄) as described by Maguire and Heckendorn (2011) with modifications. Briefly, 5 g of soil was placed in 50-mL capped glass tubes and 20 mL of Mehlich solution was added. After shaking for 5 minutes at a rate of 120 oscillations per minute (Eberbach 6000; St. Belleville, MI) the solution was filtered through filter paper (Whatman No. 2). The extract solution was then submitted to the Virginia Tech Soil Testing Laboratory (Blacksburg, VA) for analysis using inductively coupled plasma atomic emission spectrometry. The nitrate concentration in soil was determined after extraction in 2 M KCl solution as described by Knepel (2003).

3. Harvesting and Ensiling

Harvesting time was mainly decided by the owners of the dairy farm and occurred at early-dent stage in 2017 and at ¼ milk-line stage in 2018. For harvesting, 5 consecutive plants from the 2 center rows and at 2 randomly selected spots within each subplot (i.e., 10 plants per subplot) were cut by hand at 15 cm above ground. Whole plants were weighed and chopped with a Stanley CH2 wood chipper (GXi Outdoor Power, LLC, Clayton, NC). After mixing thoroughly, a first subsample of chopped material was collected, immediately frozen in dry ice, and stored at -20°C until analysis. A second 400- to 500-g subsample of chopped material was collected and ensiled into MR-1014 polyethylene embossed pouches (Doug Care, Springfield, CA) and double-sealed anaerobically with a FastVac vacuum sealer (Doug Care) as described by Der Bedrosian et al. (2012). No inoculants were added to enhance fermentation. Mini-silos were stored inside plastic tubs for 90 days.

4. Forage Processing and Analyses

The first subsample of fresh material was thawed and dried at 55°C in a forced-air drying oven (Freas 645, Thermo Electron Corporation, Marietta, OH) until constant weight was reached. The resulting DM concentration was used to calculate DM yield. The dried material was then ground to pass through a 1-mm screen of a Wiley cutter mill (Thomas Scientific, Swedesboro, NJ), and the concentration of P was determined (method 965.17; AOAC International, 2019).

After 90 days of fermentation, the mini-silos were opened, and the pH was determined by blending (Waring, Torrington, CT) 10 g of corn silage with 90 mL of deionized water for 5 min and immediately measuring pH with an Accumet AB 150 pH-meter (Fisher Scientific Company, Suwanee, GA). The remaining corn silage was then dried at 55°C in a forced-air drying oven until constant weight was reached and ground to pass through a 1-mm screen of a Wiley cutter mill. Ash concentration was determined after burning samples in a furnace (Thermolyne 30400, Barnstead International Dubuque, IA) for 2 h at 600°C (method 942.05; AOAC International, 2019). Crude protein concentration was calculated as percent N x 6.25 after combustion analysis (method 990.03; AOAC International, 2019) using a Vario El Cube CN analyzer (Elementar Americas, Inc., Mount Laurel, NJ). Neutral detergent fiber and ADF concentrations were determined using the Ankom²⁰⁰Fiber Analyzer (Ankom Technology, Macedon, NY). Sodium sulfite and α-amylase (Ankom Technology) were used for NDF analysis (Ferreira and Mertens, 2007). Acid detergent fiber and lignin concentrations were determined sequentially. After determining ADF weights, residues were incubated for 3 h in 72% sulfuric acid within a 4-L jar that was placed in a Daisy^IIIncubator (Ankom Technology). Starch concentration was determined using the acetate buffer method of Hall (2009) with α-amylase from Bacillus licheniformis(FAA, Ankom Technology) and amyloglucosidase from Aspergillus niger(E-AMGDF, Megazyme International, Wicklow, Ireland). Sugars concentration was determined as total ethanol/water-soluble carbohydrates as described by Hall et al. (1999).

5. Dry Matter and Neutral Detergent Fiber Digestibility

The Institutional Animal Care and Use Committee of Virginia Tech approved all procedures involving dairy cows for collecting rumen contents (Protocol DASC 15-234). The 30-h in vitro apparent DM digestibility (IVDMD), in vitro true DM digestibility (IVTDMD), and in vitro NDF digestibility (IVNDFD) were determined using a Daisy^IIrotating jar in vitro incubator (Ankom Technology) following the procedures described by Ferreira and Mertens (2005). A composited inoculum was prepared with rumen fluid and rumen solids collected from 3 rumen-cannulated lactating cows (2 Holstein and 1 Jerseys) that were fed a diet containing (DM basis) 37.6% corn silage, 5.4% pearl millet silage, 3.8% alfalfa hay, and 53.2% concentrate mix. The concentrate mix contained corn grain, soybean meal, brewers’ grains with solubles, soy hulls, and a mineral plus vitamin mix.

6. Biomass Yield, Kernel Development, and Stem Width

Standing plants were counted when the crop showed 2 visible leaves (V2). For this, the number of plants within a 13.2-m row was counted in 2 adjacent rows. This procedure was performed in each of the 48 subplots from each of the fields. The single plant biomass (DM basis) multiplied by the number of standing plants were used to determine DM yield.

At harvesting time, kernel lines per ear and kernels per line within an ear were counted for two randomly selected plants within a representative area for each of the 48 subplots from each field. To determine the number of lines per ear, lines were counted at the mid portion of the ear (Rossini et al., 2012). To determine the number of kernels per line, kernels were counted in two opposite kernel lines within the ear. At harvesting time, also, stem width was measured for two randomly selected plants within a representative area for each of the 48 subplots from each field. Stem width was measured placing a digital caliper below the node of the ear-bearing phytomer.

7. Statistical Analysis

The experiment was designed as a split-plot in a completely randomized design. All variables were analyzed using the MIXED procedure of SAS (SAS version 9.3, SAS Institute Inc., Cary, NC). The statistical model included the fixed effect of year (degrees of freedom, df = 1), the fixed effect of field (df = 1), the fixed effect of treatment or planting density (fixed; df = 2), the fixed effect of corn hybrid (fixed; df = 1), the random whole-plot error (df = 72), the fixed effect of fertilization (df = 1), all 2-, 3-, 4- and 5-way interactions (df = 41), and the random split-plot or residual error (df = 72). The 5% least significant difference (LSD) between treatment means was calculated as the product of the standard error of the difference by the t value for a probability of 0.05 with 72 df (Snedecor and Cochran, 1989).