Arlington Agricultural Research Station Research Articles

Row unit down force and coulter effects vary by environmental conditions in organic no‐till soybeans

AbstractConcerns over tillage intensity and soil erosion in organic cropping systems have prompted the development of reduced tillage organic systems that use cover crops rather than soil disturbance as the primary weed control tool. However, planting into high residue cover crop mulches may inhibit crop establishment due to poor seed placement. Adaptation of agricultural equipment to high residue planting conditions could reduce variability in crop stands, thereby improving grain yield. Field trials at the University of Wisconsin Arlington Agricultural Research Station, Marshfield Agricultural Research Station, and three on‐farm locations representing seven site‐years of data from 2019 to 2020 comprised a 2 × 2 × 2 × 2 factorial experiment comparing (1) a low (457,135 seeds ha−1) versus high (555,986 seeds ha−1) soybean [Glycine max (L.) Merr.] seeding rate, (2) a 13 fluted‐wavy coulter versus no coulter attachment, (3) low (667 N) versus high (1334 N) row unit down force, and (4) spiked versus rubber closing wheels on soybean stand establishment and grain yield in an organic cover crop–based reduced tillage system. Combined across site‐years, high row unit down force and coulters increased soybean stand establishment by 2.3% and 2.4%, respectively; however, the magnitude and direction of effects varied within each site‐year. Winter rye (Secale cereale L.) biomass accumulation negatively correlated to soybean stand establishment and demonstrated trends toward increasing soybean yield. Significant correlations between soybean stand establishment and yield were noted under high rye biomass conditions, indicating potential improvements to soybean yields by increasing plant stands.

PSVIII-30 Reduced-lignin alfalfa digestibility and effects on performance of growing beef steers

Abstract Novel alfalfa varieties have been developed to have less lignin by genetic modification, (HarvXtra, Forage Genetics International) and by conventional breeding, (HiGest 360, Alforex Seeds). Second crop (d 29) of these alfalfas and a control, LegenDairy XHD (Winfield Solutions LLC), were harvested as high moisture wrapped bales (45.9–51.6 % DM) at the Arlington Agricultural Research Station (AARS) located near Arlington, WI, on July 5, 2017. The objectives of these experiments were to characterize novel reduced-lignin alfalfa varieties, assess their effects on growth of beef steers, and determine digestibilities using a total fecal collection trial. Treatments were fed ad libitum as alfalfa baleage to 300 kg (initial weight) black-hided beef steers, (n = 72, 4 pens/treatment, 83 d). Steers were fed solely alfalfa baleage and offered a trace mineral salt block by way of fence line feed bunks. Alfalfa lignin values were 7.56, 7.18, and 6.3% DM (P = 0.34) for LegenDairy, HiGest and HarvXtra, respectively. Steers gained 0.94, 1.00, and 1.07 kg hd-1 d-1 (P = 0.25) for LegenDairy, HiGest and HarvXtra, respectively. The same harvested alfalfa was used in a total collection fecal trial with fecal collection bags (Table 1). Numerical differences between alfalfa digestibilities were observed but no significant treatment effects were detected. Total tract NDF digestibilities (TTNDFD) for harvested alfalfas were 33.9, 35.0 and 40.3 % NDF (P = 0.079) for LegenDairy, HiGest and HarvXtra, respectively. Alfalfa lignin concentration differences were not detected through acid detergent-lignin analysis of harvested bales. While the ranking of lignin concentrations, digestibilities, and growth rates followed prevailing logic, alfalfa variety treatment effects were not detected. Experimental designs with greater sensitivity should be implemented in the future.

Effects of degrees of grass competition on spreading of Kura clover

AbstractKura clover (Trifolium ambiguum M. Bieb.) is known to develop an extensive rhizome system when grown in monoculture. However, there is no information about the effects that competition from companion grasses could have on its rhizome development. This study was conducted to determine the difference in rhizome development of kura clover planted in grass, grass plus nitrogen (grass + N) fertilizer, or killed‐grass plots, and to evaluate the effect of defoliation height on kura clover rhizome development. Using 1‐year‐old smooth bromegrass (Bromus inermis Leyss) and Kentucky bluegrass (Poa pratensis L.) swards, experiment 1 and 2 commenced in August 1991 and May 1992, respectively, at the University of Wisconsin Arlington Agricultural Research Station. Ten kura clover entries (populations started from seeds or from rhizomes started from clones) were transplanted into each sub‐plot. Two cutting height sub‐plots were superimposed on each of the three grass sward types. In each Exp., two plants from the same seed or rhizome source represented the experimental unit. Individual plants spread up to 1 m in diameter over a 2‐year period after transplanting when there was no grass competition. For both swards, the mean spread of kura clover by rhizome growth was highest in killed‐grass plots (85 cm diameter), least in the grass + N plots (29 cm diameter) and intermediate (57 cm diameter) in the grass only plots. The defoliation height (4 cm or 10 cm) of the swards did not significantly affect the spread of kura clover regardless of grass status (killed or + or −N) in either smooth bromegrass or Kentucky bluegrass. We conclude that reducing the competition from grasses is a management tool that can be used to maximize the colonizing ability of kura clover.

Improving the mapping of crop types in the Midwestern U.S. by fusing Landsat and MODIS satellite data

Mapping crop types is of great importance for assessing agricultural production, land-use patterns, and the environmental effects of agriculture. Indeed, both radiometric and spatial resolution of Landsat’s sensors images are optimized for cropland monitoring. However, accurate mapping of crop types requires frequent cloud-free images during the growing season, which are often not available, and this raises the question of whether Landsat data can be combined with data from other satellites. Here, our goal is to evaluate to what degree fusing Landsat with MODIS Nadir Bidirectional Reflectance Distribution Function (BRDF)-Adjusted Reflectance (NBAR) data can improve crop-type classification. Choosing either one or two images from all cloud-free Landsat observations available for the Arlington Agricultural Research Station area in Wisconsin from 2010 to 2014, we generated 87 combinations of images, and used each combination as input into the Spatial and Temporal Adaptive Reflectance Fusion Model (STARFM) algorithm to predict Landsat-like images at the nominal dates of each 8-day MODIS NBAR product. Both the original Landsat and STARFM-predicted images were then classified with a support vector machine (SVM), and we compared the classification errors of three scenarios: 1) classifying the one or two original Landsat images of each combination only, 2) classifying the one or two original Landsat images plus all STARFM-predicted images, and 3) classifying the one or two original Landsat images together with STARFM-predicted images for key dates. Our results indicated that using two Landsat images as the input of STARFM did not significantly improve the STARFM predictions compared to using only one, and predictions using Landsat images between July and August as input were most accurate. Including all STARFM-predicted images together with the Landsat images significantly increased average classification error by 4% points (from 21% to 25%) compared to using only Landsat images. However, incorporating only STARFM-predicted images for key dates decreased average classification error by 2% points (from 21% to 19%) compared to using only Landsat images. In particular, if only a single Landsat image was available, adding STARFM predictions for key dates significantly decreased the average classification error by 4 percentage points from 30% to 26% (p<0.05). We conclude that adding STARFM-predicted images can be effective for improving crop-type classification when only limited Landsat observations are available, but carefully selecting images from a full set of STARFM predictions is crucial. We developed an approach to identify the optimal subsets of all STARFM predictions, which gives an alternative method of feature selection for future research.

Agronomic Management and Fungicide Effects on Oat Yield and Quality

ABSTRACTOat (Avena sativa L.) is an important crop in the U.S. The primary uses for oat in Wisconsin are for establishing alfalfa (Medicago sativa L.) underseedings, livestock forage, grain, and straw production, and for human nutrition. Therefore, the quantitative and qualitative characteristics of oat are of great importance. The objective of this study was to detect the effects of variety, seeding rate, seed treatment, and foliar fungicide on total grain yield, groat proportion, test weight, plant height, lodging severity, and β‐glucan content using data from two separate trials which were established from 2011 to 2013 at the University of Wisconsin Arlington Agricultural Research Station. The results from the two 3‐yr studies highlight the importance of selecting an appropriate oat variety with superior yield and quality characteristics. The use of a high seeding rate had no effect on any of the examined characteristics. However, treated seeds with Rancona Crest exhibited increased groat yield compared to untreated seeds by 187 kg ha−1. An important finding of this study is that the use of a foliar fungicide increased total and groat yields, and reduced lodging severity without impacting β‐glucan content. The results of this study indicate that the use of a superior variety such as ‘Badger’ and ‘BetaGene’, use of Rancona Crest seed treatment, and use of Headline foliar fungicide can increase oat yields in Wisconsin without reducing the crops’ quality characteristics.

Influence of groundwater on plant water use and productivity: Development of an integrated ecosystem – Variably saturated soil water flow model

Plant physiology influences the energy and water balance of the soil–plant–atmosphere continuum. However, impacts of soil water dynamics on plants in shallow groundwater environments are not completely understood, partially due to the limited ability of current models to simulate groundwater–vegetation interactions. In this study, we analyzed the influence of groundwater-induced soil temperature change on plant phenology, and the impact of variable depth to the water table on the net primary productivity (NPP), evapotranspiration and stomatal response, by integrating an advanced dynamic agroecosystem model (Agro-IBIS) and a variably saturated soil water flow model (Hydrus-1D) into a single framework. The model is first evaluated using field observations of soil moisture and temperature as well as annual NPP and weekly LAI measurements collected from three replicated maize plots at the Arlington Agricultural Research Station near Arlington, Wisconsin, USA. Comparisons showed reasonable agreement for each dataset without site-specific prior calibration. We then simulated the influence of groundwater on plant physiological responses as well as the energy, carbon, and water balance at the land surface. The model sensitivity analyses indicated that physiological functions of plants are sensitive to water table depth, and the aridity of a particular production site. For example, shallow groundwater limits water stress during dry years, helping to mitigate decreased NPP associated with water deficits. However, if the water table is persistently too close to the surface during the growing season, photosynthesis is negatively affected through oxygen stress on roots regardless of the aridity. To further explore factors influencing plant physiology other than oxygen stress, we designed simulations without oxygen stress effects. Results showed that under shallow groundwater conditions: (1) higher leaf level relative humidity causes higher water use efficiency because of a lower vapor pressure deficit between the leaf and atmosphere; (2) due to delayed corn plant emergence caused by cooler springtime soil temperatures reduces NPP. Our results suggest that models designed to more mechanistically simulate groundwater–vegetation interactions may lead to a more realistic representation of feedbacks between plant phenology, soil moisture, temperature, anoxia, NPP and ET. However, until critical data are collected to assess simulated feedbacks and advance our understanding of groundwater–vegetation interactions, model confidence will likely remain somewhat limited.

Tillage, Crop Rotation, and Hybrid Effects on Residue and Corn Anthracnose Occurrence in Wisconsin.

Corn anthracnose (Colletotrichum graminicola) is an important disease of field corn (Zea mays). Two phases, leaf blight and stalk rot, can reduce yield through either premature leaf senescence or reduced grain harvest due to stalk lodging. Corn residue is an important source of primary inoculum and is increased through cultural practices such as no-tillage and continuous corn cropping, which are common practices in Wisconsin. Field studies conducted at the Arlington Agricultural Research Station (ARS) and the West Madison ARS showed that the incidence and severity of anthracnose leaf blight were higher in continuous-corn crop rotations than in soybean-corn rotations (91% higher incidence, 24 to 78% higher severity). Anthracnose stalk rot was marginally affected by tillage in 2008 (P = 0.09), with higher incidence in chisel-plowed treatments. There was a positive association between spring residue cover and anthracnose leaf blight but no association was found between residue and stalk rot. No association was found between anthracnose leaf blight and stalk rot. There was a negative association between anthracnose leaf blight and yield but not between anthracnose stalk rot and yield. Managing residue levels through crop rotation would help to reduce anthracnose leaf blight but further work is needed to elucidate factors that lead to stalk lodging prior to harvest.

Forage Quality Management of Kura Clover in Binary Mixtures with Kentucky Bluegrass, Orchardgrass, or Smooth Bromegrass

Kura clover (Trifolium ambiguum M. Bieb.) is a potentially useful perennial legume because of its excellent nutritive value and persistence under environmental extremes. However, information about forage quality of kura clover - grass mixtures adapted to the North-Central USA is limited. Objectives of this research were to determine forage nutritional value of kura clover-grass mixtures under different harvest frequency and cutting height regimes. 'Rhizo' kura clover was grown alone and in binary mixtures with 'Park' Kentucky bluegrass (Poa pratensis L.), 'Comet' orchardgrass (Dactylis glomerata L.), and 'Badger' smooth bromegrass (Bromus inermis Leyss.) at the Arlington Agricultural Research Station located near Madison, WI. Three harvest frequencies (, , or annually) and two cutting heights (4- or 10-cm) were imposed on each binary mixture and on kura clover grown alone. Higher nutritive value was observed in the binary mixtures with more frequent harvest and lower cutting height. Averaged over 3 years and all harvest frequency and cutting height treatments, the nutritive value of the Kentucky bluegrass and smooth bromegrass mixtures was superior to that of the orchardgrass mixture ( NDF and CP in the Kentucky bluegrass mixture; NDF and CP in the smooth bromegrass mixture; NDF and CP in the orchardgrass mixture). All of the mixtures and harvest management systems evaluated in this study produced forage with quality equivalent to grade one alfalfa hay and suitable for highproducing livestock, even though the highest quality was observed in the Kentucky bluegrass mixture with harvesting at the shorter cutting height.

Effects of Seed Moisture Content, Cooking Time, and Chamber Temperature on Nuña Bean (Phaseolus vulgaris L.) Popping

Nuña beans are a type of common bean (Phaseolus vulgaris L.) native to the Andean region of South America that possess the unusual property of popping; however, little is known regarding postharvest environmental effects on popping. Seed of a photoperiod-insensitive, temperate-adapted nuña bean breeding line, ‘PB24’, was produced at the Arlington Agricultural Research Station, Arlington, WI, and evaluated in a hot air popper. The experimental design was a factorial with three levels of popping time (60, 90, and 120 s), five levels of chamber temperature (101, 146, 208, 244, and 268 °C), and eight levels of seed moisture (2.5%, 3.2%, 5.2%, 6.6%, 8.3%, 12.0%, 15.3%, and 20%). Percentage of popped seed, sufficiently expanded to shed the seedcoat, was calculated. A curvilinear decrease in popping percentage was observed with increasing seed moisture content. In contrast, a curvilinear increase in popping percentage was observed with increasing chamber temperature and popping time. Larger mean squares were observed for main effects and first-order interactions associated with seed moisture content and chamber temperature compared with popping time. A combination of seed moisture below 5%, popping chamber temperature of 244 °C, and popping time of 90 s resulted in popping percentages greater than 90%.

Effects of Stall Design and Rubber Alley Mats on Cow Behavior in Freestall Barns

The objectives of these studies were to evaluate cow preference between different stall designs and effects of rubber flooring on stall use and time at the feed bunk. Both studies were conducted at the University of Wisconsin-Madison Arlington Agricultural Research Station. In the stall design study, cows could choose among seven different stall base types, each available with a U-loop and 112-cm (45-in) neck rail or a new Y2K® stall partition (Artex Fabricators, Ltd., Langley, British Columbia, Canada) and 125-cm (50-in) neck rail, without a front mounting bar. Stall status was recorded four times per day, 2 d/wk. Data collection began January 29, 2003 for a 1-mo period. Cows had a 1-mo acclimation period to the new stalls, and data were collected for another 1-mo period. Cows were observed lying significantly more (P<0.0001) often (number of times) in stalls after installation of higher neck rails and Y2K® loops. Data collection for the rubber alley mat study began August 20, 2003 for a 1-mo period. Cows had a 2-wk acclimation period to the new flooring, and data were collected for another 1-mo period. Feed alley observations were taken at 15-min intervals, 1 d/wk. Stall use observations were taken four times per day, 1 d/wk. Cows spent similar time lying in stalls, but significantly less (P<0.001) time standing in stalls when concrete alleys were covered with rubber alley mats. Cows spent more (P<0.03) time at the feed bunk after alley mats were installed, but the increase was not greater than that observed in the control pen.

SENSITIVITY AND UNCERTAINTY ANALYSES OF CROP YIELDS AND SOIL ORGANIC CARBON SIMULATED WITH EPIC

Modeling biophysical processes is a complex endeavor because of large data requirements and uncertainty inmodel parameters. Model predictions should incorporate, when possible, analyses of their uncertainty and sensitivity. Thestudy incorporated uncertainty analysis on EPIC (Environmental Policy Impact Calculator) predictions of corn (Zea maysL.) yield and soil organic carbon (SOC) using generalized likelihood uncertainty estimation (GLUE). An automatic parameteroptimization procedure was developed at the conclusion of sensitivity analysis, which was conducted using the extendedFourier amplitude sensitivity test (FAST). The analyses were based on an experimental field under 34-year continuous cornwith five N treatments at the Arlington Agricultural Research Station in Wisconsin. The observed average annual yields pertreatment during 1958 to 1991 fell well within the 90% confidence interval (CI) of the annually averaged predictions. Thewidth of the 90% CI bands of predicted average yields ranged from 0.31 to 1.6 Mg ha-1. The predicted means per treatmentover simulations were 3.26 to 6.37 Mg ha-1, with observations from 3.28 to 6.4 Mg ha-1. The predicted means of yearly yieldover simulations were 1.77 to 9.22 Mg ha-1, with observations from 1.35 to 10.22 Mg ha-1. The 90% confidence width forpredicted yearly SOC in the top 0.2 m soil was 285 to 625 g C m-2, while predicted means were 5122 to 6564 g C m-2 andobservations were 5645 to 6733 g C m-2. The optimal parameter set identified through the automatic parameter optimizationprocedure gave an R2 of 0.96 for average corn yield predictions and 0.89 for yearly SOC. EPIC was dependable, from astatistical point of view, in predicting average yield and SOC dynamics.

Use of Impact Testing to Predict Softness, Cow Preference, and Hardening Over Time of Stall Bases

The objective of this study was to assess the softness and durability of commercially available free-stall bases, and to determine the relationship of stall base softness to cow preference. Clegg impact values were recorded at the University of Wisconsin–Madison Arlington Agricultural Research Station on June 19, 2002, and again on July 24, 2003. The Clegg Impact Soil Tester (model 95051, Lafayette Instruments, Lafayette, IN) with a 20-kg hammer was used in this study. The impact of the hammer on the free-stall base results in a digital display based on peak deceleration of the hammer's impact with the free-stall base in tens of gravities (CIV/H). The CIV/H value, as measured by the Clegg Impact hammer, is based on peak deceleration of the 20-kg hammer's impact with the surface, from a height of 30cm. Clegg impact measures were highly correlated with cow preference measurements. This relationship suggests that Clegg impact measures of compressibility were good indicators for predicting stall-base acceptance. A cork mattress, 4 foam mattresses, 4 rubber mattresses, 4 rubber mats, and a waterbed were evaluated in this study. Foam-based mattresses lost cushioning ability faster than rubber mattresses or rubber mats. Clegg impact values increased over the 13-mo time period for most stall base types, which indicated a tendency of stall bases to harden.

Parameters for Predicting Giant Foxtail Cohort Effect on Soybean Yield Loss

Quantifying the effect of cohort emergence time on weed–crop interference is critical for development of accurate crop yield loss models. Therefore, research was conducted to assess late‐season weed shoot biomass, weed density and cohort emergence time, weed relative leaf area, and weed relative volume as predictors of soybean [Glycine max (L.) Merr.] yield loss. Field experiments were conducted in 1999 and 2000 at the University of Wisconsin Arlington Agricultural Research Station near Arlington, WI. The main‐plot effect was giant foxtail (Setaria spp.) cohort emergence time at VE, VC, V1, and V3 soybean. The subplot effect was giant foxtail density at 0, 4, 16, 36, and 64 plants m−2. Linear and nonlinear regression equations were used to describe soybean yield loss based on measurements of giant foxtail shoot biomass, volume, leaf area, and density and cohort emergence time. Based on weed density and cohort emergence time, coefficient estimates used to describe soybean yield loss were similar between years. Based on weed relative leaf area, coefficient estimates for the giant foxtail damage coefficient (qL) and maximum soybean yield loss (mL) differed between years. Based on weed relative volume, coefficient estimates for the giant foxtail damage coefficient (qV) and maximum soybean yield loss (mV) were similar between years. Soybean yield loss increased linearly as giant foxtail shoot biomass increased. The stability of the density and cohort emergence time coefficients and the weed relative volume coefficients indicated these parameters may prove more useful that weed relative leaf area for predicting soybean yield loss.

First Report of Mycoleptodiscus terrestris Infecting Forage Legumes and Soybeans in Wisconsin.

Soilborne plant pathogens are regarded as important causes of failures of newly established and mature stands of forage legumes and reduced yield of soybean (Glycine max (L.) Merr.) in the North Central Region of the U.S. Birdsfoot trefoil (Lotus corniculatus L.) plants exhibited decaying roots and stems in yield trials in 1994 and 1995 at the Arlington Agricultural Research Station, Arlington, WI. After first harvest, over 40% of the plants failed to produce regrowth or regrowth was extremely slow. Dark brown lesions were evident on root and crown tissue and occasionally spread to newly initiated stem tissue. A fungus resembling Mycoleptodiscus terrestris was isolated from birdsfoot trefoil by placing surface-disinfested plant tissue on potato dextrose agar (PDA) amended with 10 ppm chlorotetracycline or tetracycline and 100 μl liter-1 of tergitol. Isolates of M. terrestris produced dark green mycelium and black scerotia on PDA. These same isolates also produced conidia with characteristics similar to those described by Ostazeski (1) for M. terrestris when cultures were grown on gamma ray-sterilized chrysanthemum leaf pieces positioned on the surface of 2% water agar on petri dishes. The setaed conidia were generally two-celled and oval, and measured 24 to 30 × 4.5 to 7.4 μm. Isolates were also obtained from soybean plants that expressed a premature decline symptom at the R6 to R7 growth stage. Soybean plants in two commercial fields initially expressed curled leaves followed by defoliation. Lower stems expressed a gray-tan discoloration of the cortical tissue. Alfalfa (Medicago sativa L.), red clover (Trifolium pratense L.), birdsfoot trefoil, and soybean were inoculated with isolates of M. terrestris. All crop species expressed symptoms and M. terrestris was recovered from symptomatic tissue. M. terrestris isolates have also been baited from other soils from southern Wisconsin with red clover, birdsfoot trefoil, and soybean. Although recognized in states south of Wisconsin, M. terrestris has not been implicated previously in poor health of forage legumes and soybeans in Wisconsin. This report provides evidence that M. terrestris inhabits agricultural soils farther north than previously recognized.

Use of Labeled and Experimental Insecticides for Control of European Corn Borer and Corn Ears Worm on Sweet Corn, 1996

Abstract ‘Jubilee’ sweet corn was planted at the Arlington Agricultural Research Station, Arlington, WI at 23,000 seeds per acre on 24 Jun. Each treatment was 6 rows wide and 60 ft long, arranged in a RCB design and replicated four times. The sprays were applied 21, 26 Aug and 3, 10 Sep. All treatments were applied with a high clearance sprayer equipped with D2/23 TeeJet disc-core nozzles and delivered 10.3 gallons of finished spray per acre at 40 psi when driven at 2.5 MPH. Plots were evaluated 18 Sep by removing 25 ears per treatment and recording the number of damaged ears and of the larvae present in different developmental stages.

Use of Labeled and Experimental Insecticides for Control of European Corn Borer and Corn Ear-Worm on Sweet Corn, 1996

Abstract ‘Jubilee’ sweet corn was planted at the Arlington Agricultural Research Station, Arlington, WI at 23,000 seeds per acre on 24 Jun. Each treatment was 6 rows wide and 60 ft long, arranged in a RCB design and replicated four times. The sprays were applied 21, 26 Aug and 3, 10 Sep. All treatments were applied with a high clearance sprayer equipped with D2/23 TeeJet disc-core nozzles and delivered 10.3 gallons of finished spray per acre at 40 psi when driven at 2.5 MPH. Plots were evaluated 18 Sep by removing 25 ears per treatment and recording the numbers of damaged ears and of the larvae present in different developmental stages.

Yield and Quality of Soybean Forage as Affected by Cultivar and Management Practices

AbstractSoybean [Glycine max (L.) Merr.] has potential for use as an alternative forage crop, however, little is known about the effects of cultural practices on forage yield and quality. A study was conducted to evaluate the effects of cultivar, row spacing, plant density, and harvest maturity on the yield and quality of soybean forage. The study was conducted at the Univ. of Wisconsin Arlington Agricultural Research Station, Arlington, WI on a Plano silt‐loam soil (fine‐silty, mixed, mesic, Typic Argiudoll) in 1987 and 1989. The cultivars Corsoy 79, Pella, and Williams 82 were grown at 20‐ and 76‐cm row spacings at planting rates of 280 000 and 890 000 seeds ha−1 and were harvested at the R1, R3, R5, and R7 stages of development. Harvest maturity had the greatest effect on soybean forage yield and quality of the management practices evaluated. The yield of soybean forage increased from 2.4 Mg ha−1 when harvested at R1 to 7.4 Mg ha−1 when harvested at R7, but quality declined between stages R1 and R5 then increased from R5 to R7 as pods developed and seeds filled. Late maturing cultivars (Maturity Group III vs. Maturity Group II) produced greater forage yields but lower quality forage when harvested at the same stage of development. The 20‐cm row spacing produced 1.2 Mg ha−1 more forage than the 76‐cm row spacing, but crude protein concentration was 8 g kg−1 less. The results of this experiment indicate that soybean can produce forage similar in quality to alfalfa and that management practices typically used for grain production are suitable for forage production.