Journal list menu

Volume 5, Issue 2 e20240
ORIGINAL RESEARCH ARTICLE
Open Access

Impacts of winter grazing on soil health in southeastern cropping systems

Hayley Crowell

Corresponding Author

Hayley Crowell

Dep. of Crop, Soil, and Environmental Sciences, Auburn Univ., 201 Funchess Hall, Auburn, AL, 36849 USA

Correspondence

Hayley Crowell, Dep. of Crop, Soil and Environmental Sciences, Auburn Univ., 201 Funchess Hall, Auburn, AL, 36849, USA.

Email: [email protected]

Contribution: Data curation, Formal analysis, Writing - original draft, Writing - review & editing

Search for more papers by this author
Audrey V. Gamble

Audrey V. Gamble

Dep. of Crop, Soil, and Environmental Sciences, Auburn Univ., 201 Funchess Hall, Auburn, AL, 36849 USA

Contribution: Conceptualization, Funding acquisition, ​Investigation, Methodology, Project administration, Resources, Supervision, Writing - review & editing

Search for more papers by this author
Yucheng Feng

Yucheng Feng

Dep. of Crop, Soil, and Environmental Sciences, Auburn Univ., 201 Funchess Hall, Auburn, AL, 36849 USA

Contribution: Conceptualization, Methodology, Writing - review & editing

Search for more papers by this author
Kipling Balkcom

Kipling Balkcom

USDA-ARS, National Soil Dynamics Lab., 411 South Donahue Drive, Auburn, AL, 36832 USA

Contribution: Data curation, Methodology, Writing - review & editing

Search for more papers by this author
Anna Yang

Anna Yang

College of Life Sciences, Anhui Normal Univ., Wuhu, Anhui Province, P.R. China

Contribution: Data curation, Methodology, Validation

Search for more papers by this author
First published: 13 April 2022
Citations: 6

Assigned to Associate Editor Benjamin Wherley.

Abstract

Coastal Plain soils are often characterized by low soil organic carbon (SOC) as a result of natural and anthropogenic factors. A rotation of cotton (Gossypium hirsutum L.) and peanut (Arachis hypogaea L.) under conventional tillage is typical in this region, but an opportunity to increase SOC and improve soil health by incorporating winter grazing of cover crops exists for producers with row crop and livestock operations. A study to assess winter grazing impacts on soil health was established in the U.S. Coastal Plain. Three cattle removal dates and an ungrazed control were evaluated for impacts on selected soil health indicators: SOC, permanganate oxidizable carbon (POXC), water stable aggregates (WSA), penetration resistance (PR), microbial biomass C (MBC), and arbuscular mycorrhizal fungi (AMF) colonization rates. After two years, MBC was highest in the control treatments, likely due to greater cover crop biomass on the soil surface at termination. No differences were observed between treatments for SOC, POXC, WSA, PR, or AMF. Increased cotton lint yield was observed in control and mid-February treatments in 2019, likely due to greater cover crop residues on the soil surface following grazing, which may have conserved soil moisture during the growing season. Peanut yields were unaffected by treatments in 2020. Lack of differences in soil health indicators suggests that integrating winter-grazing livestock does not negatively nor positively impact selected dynamic soil properties in the short-term, but more time under grazing treatments is needed to thoroughly evaluate how winter-grazing livestock impacts soil health and crop yield.

Abbreviations

  • AMF
  • arbuscular mycorrhizal fungi
  • AUCC.I
  • area under the curve for cone index
  • MBC
  • microbial biomass C
  • POXC
  • permanganate oxidizable carbon
  • PR
  • penetration resistance
  • SOC
  • soil organic carbon
  • SOM
  • soil organic matter
  • WSA
  • water stable aggregates.
  • 1 INTRODUCTION

    Soils in the Southeast United States tend to be highly weathered, easily erodible, carbon-depleted, and low in water-holding capacity (Shaw et al., 2010; Simoes et al., 2009). For many years, conventional agricultural practices, such as intensive tillage, insufficient addition of organic residues, and monocropping have degraded soils across the region. Furthermore, the hot, humid climate in the Southeast causes accelerated breakdown of organic matter, which often makes it challenging to improve soil health (Balkcom et al., 2013). The combination of intensive agricultural practices, soil type, and climate has resulted in lower soil organic matter (SOM) levels than in other regions of the US, and crops produced in the Southeast may yield less than their economic potential as a result (Sainju et al., 2007; Siri-Prieto et al., 2007). Soil organic matter is viewed as the most significant single indicator of soil health because it is integrally tied to many other chemical, physical, and biological properties of soil (Reeves, 1997; Weil et al., 2003). Concerns regarding long-term sustainability of row crop production in the Southeast have led to a renewed interest in soil conservation practices. Implementing soil conservation practices can sustainably intensify land use and improve agroecosystem health. For example, conservation tillage and cover cropping are known to improve soil health by building SOM inputs into the soil, increasing above and belowground biodiversity, improving soil structure, and improving water-holding capacity (Rosa-Schleich et al., 2019; Schipanski et al., 2014; Lehmann & Kleber, 2015).

    Soil health can be described as soil properties that determine its capacity to produce economic goods and services and regulate the environment (Lal, 1993). Many soils around the globe are severely degraded as a result of natural and anthropogenic factors, such as erosion, insufficient residue returns, and intensive tillage. Certain soil characteristics, such as porosity, C content, and microbial activity, are crucial to soil health and agricultural productivity. Such chemical, physical, and biological properties can serve as indicators of soil health. Although soils of the Southeast are not as naturally fertile as those of other regions, they can be highly productive with conservation practices, including conservation tillage, crop rotations, and cover cropping (Balkcom et al., 2013). Thanks to plentiful research, the use of conservation systems has increased dramatically across the country and in the Southeast. In the last five years, the average hectarage of cover crops planted in the United States has increased by nearly 40% (CTIC, 2020). However, limited research exists on the impacts that conservation systems have on soil health while further diversifying land-use with grazing.

    Since the Neolithic age, crop–livestock systems have been used that integrate row production with animal husbandry (Caravalho et al., 2010). Integrated crop–livestock systems provide opportunities for increased plant diversity, nutrient cycling, maximization of farmland use, and economic gains (Ribeiro et al., 2019). Depending on climate and crop type, integrated crop–livestock systems can take many different forms. In the Southeast, mild winters allow growers to incorporate grazing livestock on winter cover crops between summer cash crop growing seasons. Many row crop producers in the Southeast also have cow–calf or stocker livestock operations, in which much of the operating cost comes from winter feeding. Using an integrated crop–livestock system to maximize forage availability is a strategy that can reduce winter feed cost (Han et al., 2018). The use of cropland year-round has potential to increase crop productivity through improved soil health and economic efficiency, thereby improving livelihoods of producers in the Southeast. While some producers in the Southeast already incorporate grazing of winter annual forages into their row crop production systems, these forage crops are typically not managed to build SOM or improve soil health. For example, forage crops are intensively grazed at the end of the growing season to terminate the crop, leaving minimal crop residue on the soil surface at cash crop plating. Research is needed to identify best management practices for an integrated crop–livestock system to optimize both soil health and economic gains.

    Recent research has shown that adding grazing livestock to a production system can have benefits to agroecosystem health such as enhanced nutrient cycling, breaking weed or pest cycles, and increased organic matter additions from manure (George et al., 2013; Salton et al., 2014). However, thorough research on how to manage integrated crop–livestock systems to improve soil health in the Southeast could lead to greater adoption by producers. Therefore, our objectives of this study were (a) to determine effects of cover crop grazing on soil health indicators in a cotton (Gossypium hirsutum L.)–peanut (Arachis hypogaea L.) rotation under conservation tillage and (b) to determine dates at which cattle should be removed from grazed cover crops to optimize soil health benefits.

    Core Ideas

    • Physical and chemical soil health indicators were unaffected by grazing period.
    • Microbial biomass C was greatest in an ungrazed control treatment.
    • Cotton lint yield was improved for ungrazed and shorter grazing periods after a year.

    2 MATERIALS AND METHODS

    2.1 Site description and experimental design

    A research trial was established in the fall of 2018 at the Wiregrass Research and Extension Center in Headland, AL (31°30′ N, 85°17′ W) on a soil classified as a Dothan fine sandy loam (fine-loamy, kaolinitic, thermic Plinthic Kandiudults). The site had been previously managed (>8 yr) in a peanut (Arachis hypogaea L.)–rye/oat (Secale cereale L./Avena sativa L.)–pearl millet [Pennisetum glaucum (L.) R. Br.] rotation under conventional tillage. A 7.3-ha field was divided into twelve 0.61-ha paddocks with each paddock receiving one of four treatments. Treatments included: (a) mid-February cattle removal, (b) mid-March cattle removal, (c) mid-April cattle removal, and (d) an ungrazed control. The four cover crop grazing treatments were organized in a randomized complete block design and replicated three times.

    Stocker cattle of approximately seven to eight months in age in January 2019 were used to graze the experimental treatments under continuous grazing to maintain forage allowance of 1-kg dry matter per 1-kg body weight and forage height of approximately 15 cm. Grazing began in January, and all cattle were removed from each designated treatment according to each cattle removal date assigned to that paddock. Cattle had ad libitum access to water and high-magnesium mineral during the grazing period. Cattle had no access to shade due to moderate seasonal temperatures.

    Prior to initiation of the experiment in 2018, the area was disked, subsoiled, and field cultivated before planting cover crop treatments. A winter cover crop mixture of ‘FL401’ rye , ‘Cosaque’ oat, ‘AU Sunrise’ crimson clover (Trifolium incarnatum L.), and ‘T-raptor’ brassica hybrid [Brassica rapa × napus]; was planted. Seeding rates varied between grazing and control plots based on guidelines for managing cover crops versus winter grazing according to Alabama Cooperative Extension System recommendations. Seeding rates for the ungrazed control plots were 33.6 kg ha−1 rye, 33.6 kg ha−1 oat, 16.8 kg ha−1 crimson clover, and 3.36 kg ha−1 brassica hybrid. Seeding rates for the grazed plots were 50.4 kg ha−1 rye, 50.4 kg ha−1 oat, 18.8 kg ha−1 crimson clover, and 3.36 kg ha−1 brassica hybrid. Cover crops were planted using a Great Plains 1205 no-till drill (Great Plains Manufacturing Inc.) with 19-cm row spacing and terminated approximately two weeks prior to cotton and peanut planting in 2019 and 2020, respectively. Field operation dates are shown in Table 1.

    TABLE 1. Dates of field operations at the Wiregrass Research and Extension Center in Headland, Alabama for Year 1 (October 2018–October 2019) and Year 2 (November 2019–October 2020)
    Operation Year 1 Year 2
    Cover crop planting 29 Oct. 4 Nov.
    Grazing begins 11 Jan. 13 Jan.
    Mid-February cattle removal treatment 15 Feb. 12 Feb.
    Mid-March cattle removal treatment 15 Mar. 9 Mar.
    Mid-April cattle removal treatment 5 Apr. 9 Apr.
    Cover crop termination 18 Apr. 14 Apr.
    Cash crop plantinga 30 Apr. 4 May
    Soil Sampling 14 May 1 May
    Cash crop harvest 12 Oct. 1 Oct.
    • a Cash crops in 2019 and 2020 differed: cotton was planted in 2019 and peanut was planted in 2020.

    In all treatments, cover crops were chemically terminated with a tank mix of glyphosate at 1.25 kg ai ha−1 and pendimethalin at 1.6 kg ai ha−1 for burndown and rolled approximately one month after the last cattle removal treatment. Phosphorus, potassium, and pH were maintained according to soil test recommendations from the Alabama Agricultural Experiment Station (Mitchell, 2012). Nitrogen application varied between control plots and grazed plots because of different management recommendations for cover crops and for winter grazing according to Alabama Cooperative Extension System recommendations. Grazed plots received 67.2 kg ha−1 N application and control plots received 26.9 kg ha−1 N application. In 2019, ‘Deltapine 1518′ cotton was planted with a 91.44-cm row spacing following cover crop termination. In 2020, ‘GA-06G’ peanut was planted with a 91.44-cm row spacing following cover crop termination. Conservation tillage was achieved with a KMC rip/strip (Kelley Manufacturing Company) noninversion subsoiling implement. Pesticides were applied according to Alabama Cooperative Extension System recommendations (Majumdar et al., 2020; Smith et al., 2020).

    2.2 Field data collection

    Soil samples were collected using bucket augers following cover crop termination in spring of 2019 and 2020. Soil samples to be used for lab analyses of permanganate oxidizable carbon (POXC), water stable aggregates (WSA), and soil organic carbon (SOC), were collected from depth intervals of 0–5, 5–10, 10–15, and 15–30 cm two weeks after cover crop termination in 2019 and 2020. Due to variability in soil texture, two Global Positioning System (GPS)-marked locations were chosen for soil sampling within each paddock and are referred to as sampling locations. The same Global Positioning System locations within each paddock were used for soil sampling in both 2019 and 2020. Ten sub-samples at each sampling location were combined to form a composite sample for laboratory analysis. A composite of ten samples from the 0–15 cm depth at each sampling location was collected and transported to the laboratory on ice for analysis of microbial biomass carbon (MBC).

    Cover crop biomass sampling occurred within one week prior to cover crop termination. A composite of four random 0.25-m2 subsamples were taken at each Global Positioning System-marked sampling location. Biomass samples were oven-dried for at least 48 h at 60 °C and weighed to obtain dry cover crop biomass.

    Root samples for arbuscular mycorrhizal fungi (AMF) colonization rates were collected during the cash crop growing period in 2019 and 2020. Cotton roots were sampled at the fourth true leaf stage, and peanut roots were sampled approximately 60 d after planting. Sampling was conducted by uprooting five to ten plants from each Global Positioning System-marked sampling location. Plants were extracted by wetting the soil and placing a trowel approximately 30 cm into the soil adjacent to the selected plant. The trowel was then used to gently extract the plant and attached root system. Plants were placed in sealable plastic bags and transported to the laboratory on ice. Roots were removed from the plants and gently rinsed in water for two minutes. Lateral roots were cut from the taproot and carefully placed in tightly sealed glass vials containing 0.5 M formalin–acetic acid–alcohol. Vials were kept in a refrigerator at approximately 4 °C until assessing AMF colonization rates.

    2.3 Lab analysis

    All soils were sieved to 2 mm for soil chemical analyses. A portion of the soils was finely ground using a coffee grinder and analyzed for total carbon via dry combustion with a CN LECO 2000 analyzer (LECO Corp.) in 2019 and via combustion on a CHNS/SIR elemental analyzer (vario EL cube, Elementar) in 2020. No soil samples contained inorganic carbon; therefore, total carbon was used as a measurement of SOC.

    Soil samples were analyzed for POXC as described in Weil et al. (2003). A 2.5-g air-dried soil sample was placed in a 50-ml centrifuge tube with 18 ml of deionized water and 2 ml of 0.2 M potassium permanganate (KMnO4) stock solution. Centrifuge tubes were placed in a shaker and shaken at 240 oscillations per minute for two minutes and then removed from the shaker and placed in a dark area for ten minutes to settle. The oxidation of active C results in the conversion of Mn (VII) to Mn (II). After, 0.5 ml of supernatant from each centrifuge tube was added to a new centrifuge tube with 49.5 ml of DI water for a 100-fold dilution of supernatant, and a 0.2-ml aliquot was transferred to a 96-well microplate. The absorbance of the samples and standards was read at 550 nm on a spectrophotometric microplate reader (Biotek MQX200). Concentration of POXC was calculated as described by Weil et al. (2003) and Culman et al. (2012).

    Water stable aggregates were measured through methods described in Kemper and Rosenau (1986). A subset of each air-dried soil sample was sieved to only include 1- to 2-mm sized aggregates. A 4-g sample of the 1- to 2-mm aggregates were weighed into cup-like sieves of 24 mesh cm−1 wire and rewetted to near field capacity using a household humidifier. The rewetted samples were then uniformly raised and lowered into preweighed individual metal containers of deionized water at a rate of 35 times min−1 for three minutes. New preweighed metal containers were filled with diluted sodium hexametaphosphate (NaPO3)6 and NaOH dispersal solution, and the aggregate samples were raised and lowered into the solution at 35 times min−1 until all remaining aggregates dispersed. All metal containers were then dried at 105 °C overnight or until constant weight was reached. All metal containers were weighed again. The weight of soil remaining in the metal containers was compared with the percent of stable aggregates, with corrections for the dispersal solution.

    Soil penetration resistance (PR) was measured in situ at each sampling location each year approximately one month after cash crop planting using a tractor-mounted hydraulic, five-probe penetrometer to obtain cone-index values described by Raper et al. (1999). The tractor was positioned so the center penetrometer rod was in the cash crop row. Two penetrometer rods on either side were located 22.5 and 45 cm away from the cash crop row to include both trafficked and nontrafficked rows. Cone-index values were recorded to 50 cm in the soil profile. Area under the curve for cone index (AUCC.I) values were calculated by averaging cone index values across all row positions and depths to simplify soil strength analysis. Soil moisture data were collected at the time of PR data collection at depths of 0–15 and 15–30 cm using soil probes. Area under the curve for cone index was calculated following the equation below, in which I represents the row position, CIi represents the average cone index value according to row position, di represents the distance between individual row position measurements, and k represents the total number of row positions (Balkcom et al., 2016):
    AU C C . I = i = 1 k 1 = [ C I i + 1 + C I i ] d i 2 \begin{equation*} {\rm{AU}}{{\rm{C}}_{\rm{C.I}}}\; = \;\mathop \sum \limits_{i = 1}^{k - 1} = \frac{{[{\rm{C}}{{\rm{I}}_{\left( {i + 1} \right)}} + {\rm{C}}{{\rm{I}}_i}]{d_i}}}{2} \end{equation*}

    Soil samples were analyzed for MBC using the chloroform fumigation-incubation method described in Jenkinson and Powlson (1976). Field-moist soil samples were sieved to 4 mm, and sub-samples were dried in an oven at 105 °C for at least 48 hr to determine soil moisture content. Triplicate moist soil samples (25 g on a dry weight basis) and six blanks were weighed into individual preweighed 150-ml beakers. Deionized water was added to bring the soil samples to 50% of their water-holding capacity, then each sample was placed in a 1-L mason jar that contained 1.5 ml of water to maintain 100% humidity in the jar. All mason jars were closed and remained in the dark at room temperature (approximately 25 °C) for five days. Two replicates of the samples were then fumigated with ethanol-free chloroform at room temperature for 24 hr in a desiccator while the third sample served as the unfumigated control and was used for soil respiration determination. Sodium hydroxide (0.5 M NaOH) was used to trap CO2 produced. Following a 10-d incubation in the dark at room temperature (25 °C) for 10 d, 0.25 M HCl was used to titrate the unreacted NaOH to determine CO2–C released from the soil samples during the incubation period. Soil MBC was calculated using a conversion factor of 0.41 (Voroney & Paul, 1984).

    Root samples were analyzed for AMF colonization rates using acid fuchsin staining as described in Berch and Kendrick (1982). Roots were rinsed three times with water and placed in new, clean test tubes containing 10% KOH solution. Test tubes were placed in a 90 °C hot water bath for approximately 90 min for dissociation. Once dissociated, roots were rinsed with water and immersed in lactic acid for three minutes to neutralize the 10% KOH. Roots were then transferred to a clean 55 by 75 mm microscope slide (Fisher Scientific) and stained with 0.5% acid fuchsin. Each slide was then heated, and, once cooled, roots were transferred to a clean microscope slide where lactic acid glycerol was added to remove the stain from roots. Once acid fuchsin staining appeared to be removed, two to three root segments approximately 5 cm in length were arranged on the center of the slide parallel to one another. Roots were then checked under an optical microscope at 160X magnification for percent colonization according to methods in McGonigle et al. (1990) (Figure 1). Moving in one direction at a constant distance, fifty eyesight views were obtained. Each eyesight view and percent colonization was calculated.

    Details are in the caption following the image
    Image of cotton root with arbuscular mycorrhizal fungi (AMF) infection under optical microscope (160X magnification) showing (a) cotton root xylem, (b) cotton root cortical cell, (c) AMF hyphae, and (d) AMF vesicles

    2.4 Statistical analysis

    Data were analyzed using mixed models in SAS® PROC GLIMMIX with Kenward-Roger degrees of freedom. Year, treatment, depth, and their interactions were considered fixed effects. Random effects were replication within year and treatment within replication within year. Repeated measures for depth were accounted for using the autoregressive [(AR)1] covariance structure. Mean separations were performed using Tukey's honestly significant difference test (α = .1). Relationships among soil health indicators were determined using Pearson's correlation. Correlations are described as weak (R < .30), moderate (R = .31 to .70), and strong (R > .70).

    3 RESULTS AND DISCUSSION

    3.1 Cover crop biomass

    Cover crop biomass at termination did not exhibit a treatment-year interaction but was affected by year and treatment independently (Table 2). Mean dry weight biomass production was 4,384 kg ha−1 in 2019 and 3,378 kg ha−1 in 2020. Mean dry weight biomass production for the ungrazed control, mid-February cattle removal, mid-March cattle removal, and mid-April cattle removal was 7,130, 4,052, 2,575, and 1,768 kg ha−1, respectively, across years. The ungrazed control treatment produced 75 to 300% greater biomass compared with other cattle removal treatments. This was expected as cover crops in the ungrazed control were not consumed by livestock. The mid-February cattle removal treatment had 57 and 130% higher cover crop biomass, respectively, than the mid-March and mid-April treatments, which were not significantly different from each other. Grazing livestock were removed from the paddock earlier in the mid-February treatments than in the mid-March and mid-April treatments, which allowed for more biomass to develop.

    TABLE 2. Summary of analysis of variance (ANOVA) for biomass, soil organic carbon (SOC), permanganate oxidizable carbon (POXC), water stable aggregates (WSA), area under the curve for cone index (AUCC.I), microbial biomass carbon (MBC), and microbial respiration (CO2–C) in response to grazing treatment, depth, year and their interactions
    ANOVA, Pr > F
    Source of variance df Biomass SOC POXC WSA AUCC.I MBC CO2–C
    Treatment (T) 3 <.0001 .1600 .1438 .1042 .1654 .0170 .2171
    Depth (D) 3 <.0001 .0204 <.0001
    Year (Y) 1 .0303 .6904 .0084 <.0001 .1654 .8034 .5791
    T × D 9 .6261 .9030 .3897
    T × Y 3 .8554 .8985 .6753 .7775 .8439 .3543 .5622
    D × Y 3 .9532 .1040 .4351
    T × D × Y 9 .4771 .4718 .9399

    3.2 Soil organic carbon

    Soil organic carbon was not affected by treatment or year and did not exhibit interactions between treatment, year, or depth (Table 2). The lack of differences in SOC between treatments may be attributed to varied soil type, climate, and duration of the experiment as SOC is integrally tied to many other chemical, physical, and biological properties of soil, climate, and time (Reeves, 1997; Weil et al., 2003). In Coastal Plain soils, the surface horizon texture varies from loamy sand to sandy loam (Causarano et al., 2008). Coarser soils, like those in the Coastal Plain, tend to accumulate SOM very slowly (Hassink & Whitmore, 1997). Conversely, soils with a higher clay content are better able to preserve SOM, since metal oxides and aluminosilicate clays form stable complexes that protect SOM (Torn et al., 1997). The decomposition of SOM is further accelerated for coarse-textured soils examined in this study by the warm, humid climate of the Southeast. Therefore, SOM increases will likely occur after additional time under current conservation management practices.

    Depth was the only significant factor affecting SOC (Table 2), and SOC was stratified by depth class. Soil organic carbon decreased as depth increased (Figure 2a). In the top 5 cm, SOC averaged 8.35 g kg−1 while the 5–10, 10–15, and 15–30 cm depths averaged 7.67, 6.93, and 6.26 g kg−1 of SOC, respectively. Using the assumption that approximately 58% of SOM mass is made up of SOC (Stockmann et al., 2013), the measured SOC values equate to 1.44, 1.32, 1.20, and 1.08% SOM at 0–5, 5–10, 10–15, and 15–30 cm depths, respectively. The greater SOC at shallower depths is likely linked to biomass present on the soil surface, as above and belowground plant residues are the primary sources of SOC (Lorenz & Lal, 2005). Increased SOC at the soil surface has been observed in other studies conducted on Coastal Plain soils under conservation tillage in the Southeast (Balkcom et al., 2013). Greater amounts of SOC near the soil surface is considered a positive quality, as SOM present at the soil surface is important for soil aggregation, which prevents erosion and promotes water infiltration into the soil (Franzluebbers, 2002). In no-till systems, SOC is stratified with depth while in conventional tillage systems, SOC is more evenly distributed (Motta et al., 2002).

    Details are in the caption following the image
    (a) Soil organic carbon (SOC) and (b) permanganate oxidizable carbon (POXC) by depth at Wiregrass Research and Extension Center. Depths with the same letter do not differ within the soil health indicator (α = .1)

    3.3 Permanganate oxidizable C

    Permanganate oxidizable C was not affected by treatment and did not exhibit any interactions with year or depth, but year and depth each affected POXC independently (Table 2). While there was no difference in POXC between cattle removal treatments, the lack of difference indicates there was little negative or positive response to length of cattle grazing. While this may change over time, the lack of response may indicate that utilizing an integrated crop–livestock system does not negatively impact soil health in the short-term for producers in the Southeast who are looking to use land year-round for economic benefits.

    Soil organic carbon and POXC had a weak, significant correlation at the 0–30 cm depths (Figure 3). While POXC is a portion of the active SOM pool that makes up between 33 and 50% of SOM, the weak correlation may be due to soil texture and other environmental factors. The warm, humid climate of the Southeast and the coarsely-textured soils of the Coastal Plain, which cycle C more rapidly than soils of other regions, may be a factor in the weak correlation between POXC and SOC. Active C fractions like POXC are more quickly degraded than the total SOC pool by microorganisms. This effect may be greater in coarse-textured soils compared with fine-textured soils, since metal oxides and aluminosilicate clays that form stable complexes with SOM are present in lower concentrations. Lucas and Weil (2012) conducted research in croplands of the mid-Atlantic region of the United States and found that POXC was not predictive of crop responses (i.e., stover, grain, and biomass yields) in coarse-textured soil (i.e. sand and loamy sand soils) but was predictive of crop responses in finer-textured soils (i.e., silt loam and channery loam/silt loam soils of the Piedmont region).

    Details are in the caption following the image
    Correlation between soil organic carbon (SOC) and permanganate oxidizable carbon (POXC) from 0–30 cm depth based on Pearson's correlation coefficient (R) using 192 observations (N)

    Permanganate oxidizable C was affected by depth (Table 2). As depth increased, POXC decreased numerically (Figure 2b). The shallowest depth, 0–5 cm, had 20.6% greater POXC than the 15–30 cm depth (Figure 2b). Greater POXC in the top 5 cm of soil is likely due to accumulation of crop residues and manure that contribute readily decomposable C sources at the soil surface. With depth, POXC was less stratified than SOC over both years of this research (Figure 1).

    Weak correlations between SOC and POXC closer to the surface were observed in this study. The 0–5 (Figure 4a), 5–10 (Figure 4b), and 10–15 cm (Figure 4c) depths all had significant correlations with SOC, and the 5–10 cm (Figure 4b) depth had the strongest correlation. The 15–30 cm depth showed no correlation (Figure 4d). These observations support a suggestion that the relationship between SOC and POXC varies with depth (Wang et al., 2017). As aforementioned, active forms of soil C make up a portion of the SOC, so a correlation between POXC and SOC would be expected. However, because active C fractions are more readily available for use as food and energy for microbial organisms, these fractions would likely be depleted and unable to extend to deeper depths in the soil profile.

    Details are in the caption following the image
    Correlation between soil organic carbon (SOC) and permanganate oxidizable carbon (POXC) at (a) 0–5 cm depth, (b) 5–10 cm depth, (c) 10–15 cm depth, and (d) 15–30 cm depths based on Pearson's correlation coefficient (R) with 48 observations (N)

    Year affected POXC (Table 2). A 29% decrease in POXC occurred from 2019 to 2020. Timing of sampling could be a factor that affected the difference in POXC between 2019 and 2020. In 2019, cover crops were terminated on April 18th, and soil samples were collected on May 14th; in 2020, cover crops were terminated on April 14th, and soil samples were collected on May 1st (Table 1). The two-week difference between the elapsed time from cover crop termination to soil sampling between 2019 and 2020 may have resulted in less decomposition of biomass residue by microorganisms at the time of soil sampling in 2020, thereby impacting POXC concentrations. Similarly, cover crop biomass may have affected the decrease in POXC, since biomass was 22% lower in 2020 compared with 2019 across treatments. Permanganate oxidizable C and other labile C fractions act as fuel for soil microbes, influencing nutrient cycling and other biologically related soil properties (Weil et al., 2003). Active forms of C are also known to be easily affected by environmental conditions, so differences in soil temperature and moisture between years may have also contributed to temporal differences in POXC.

    3.4 Water stable aggregates

    Grazing time and its interactions with year and depth did not affect WSA, but year and depth independently affected WSA (Table 2). The lack of influence that cattle removal treatments had on WSA may indicate that this type of integrated crop–livestock system has minimal impacts on aggregate stability for Coastal Plain soils in early years of adoption. While research on the impact grazing animals have on soil physical properties is limited, the research available indicates minimal influence on soil physical properties like WSA. For example, Franzluebbers and Stuedemann (2008) found that grazing had little effect on aggregate stability over four years on sandy loam and sandy clay loam soils in Georgia between ungrazed and grazed cover crops with cattle consuming approximately 90% of available forage. Their study demonstrated that grazing had little effect on aggregate stability even when 90% of available forage was removed by grazing (Franzluebbers & Stuedemann, 2008). Carvalho et al. (2010) found that integrating livestock into a cropping system with moderate grazing intensities (i.e., maintaining 20 and 40 cm forage height) over seven years resulted in increased soil aggregation compared with intensively grazed treatments (i.e., maintaining 10-cm forage height). Carvalho et al. (2010) maintained moderate grazing intensity treatments to forage heights greater than were maintained in the current study and had more time under this management, which may explain the differences observed in aggregate stability compared with the current study. In the current study, greater residues in control and mid-February treatments did not contribute to increased aggregate stability. Additionally, WSAs were not correlated to SOC content. More time under current management may be needed to see changes in WSA or a relationship between SOC inputs and WSA.

    3.5 Penetration resistance

    Penetration resistance, which was evaluated using an approach to represent soil strength across five row positions (i.e., distances from the center of the cash crop row) for the 0–50 cm depth, was not influenced by treatment or year and did not exhibit a year/treatment interaction. However, the ungrazed control treatment had the lowest PR numerically in 2019 and in 2020 (Table 3). Intrinsic soil properties such as texture and moisture that result from changes in depth may have influenced soil strength more than cattle removal treatments (Balkcom et al., 2016; Williams & Weil, 2004), or the use of noninversion subsoiling following grazing may have reversed the potential negative impacts cattle grazing may have had on soil strength, as intended. A lack of consistency in PR over time under an integrated crop–livestock system was observed in a study by Tracy and Zhang (2008), in which PR measurements showed no consistent trend in 2004 or 2005 on silty, clay loams, suggesting that winter grazing may compact soils relative to ungrazed plots with cover crops in some years. Further, Franzluebbers and Stuedemann (2015) found little evidence of compaction due to winter grazing on sandy loam and sandy clay loam soils, proposing that integrated crop–livestock systems can be used on Ultisols in the Southeast without negatively impacting soil compaction. More data over time may provide greater insight as to how grazing livestock and different grazing lengths impact soil strength. Maintaining or increasing SOC may help prevent compaction when combined with practices like noninversion subsoiling, which was used in this study. Soil strength is known to decrease with increased soil moisture (Gerard et al., 1961). However, there was no difference between treatments for soil moisture readings collected at the time of sampling (data not shown), indicating that the lack of treatment differences in AUCC.I values were not influenced by soil moisture at sampling.

    TABLE 3. Means according to cattle removal treatment for soil organic carbon (SOC), permanganate oxidizable carbon (POXC), and water stable aggregates (WSA) across depths area under the curve for cone index (AUCC.I), microbial biomass carbon (MBC), and microbial respiration (CO2–C) by treatment across 2019 and 2020 at Wiregrass Research and Extension Center
    Treatment
    Soil health indicator Control Mid-Feb. Mid-Mar. Mid-Apr.
    SOC (g kg−1) 7.71 7.16 7.36 6.96
    POXC (mg kg−1) 337 289 295 310
    WSA (%) 93.24 92.42 90.93 91.73
    AUCC.I 149 181 169 189
    MBC (μg g−1) 249.3 a 190.9 b 194.3 ab 193.8 b
    CO2—C (μg g−1) 106.8 84.7 85.8 82.8
    • Note. Values followed by the same letter in a row are not significantly different between treatments according to Tukey's honestly significant difference at α = .1.

    Contour graphs provide a visual representation of soil strength profiles collected with the multiprobe penetrometer according to soil depth and distance from row middle (Figure 5). While there was no overall difference in PR measurements between treatments or years, some visual observations can be made based off contour graphs. Increased force was required to push the multiprobe penetrometer through the soil below approximately 40-cm depth at all distances from the row middle in mid-March and mid-April treatments in 2019 (Figure 5). For the center penetrometer position, the effect of subsoiling is clearly demonstrated within each treatment to a depth of 30 cm.

    Details are in the caption following the image
    Area under the curve for cone index (AUCC.I) to represent penetration resistance (PR) between treatments by depth and distance from row middle at Wiregrass Research and Extension Center in 2019 and 2020

    3.6 Microbial biomass CC

    Microbial biomass-C was not affected by year and did not exhibit a year/treatment interaction; however, MBC was influenced by treatment (Table 2). The ungrazed control treatment had approximately 30% greater MBC than the mid-February and mid-April cattle removal treatments (Table 3). The MBC in the mid-March cattle removal treatment was not different from any other treatment. Greater MBC in the control may be attributed to greater crop residues remaining on the soil surface compared with other treatments. The control treatment produced the greatest amount of biomass on the surface, which provides a C food source for microorganisms. Higher aboveground growth may also lead to greater root exudation below ground, thus enhancing C-containing substrates in the soil. Additionally, biomass may have provided protection from soil moisture loss, which may have contributed to more favorable conditions for biological activity. Franzluebbers and Studemann (2015) observed that grazing livestock resulted in a minimally positive impact on MBC in sandy loam soils in the Southeast and that using no-till was more important to preserving biologically active soil C following pasture termination. The lack of negative impacts on biological activity over the seven-year study supported the recommendation to use winter-grazing livestock in the Southeast United States (Franzluebbers & Studemann, 2015). George et al. (2013) saw greater MBC in grazed, nonirrigated plots compared with grazed, irrigated plots and nongrazed plots with and without irrigation. George et al. (2013) suggested various processes that could have increased MBC in grazed, nonirrigated plots, such as increased root growth due to grazing (Gao et al., 2008), increased root growth in dry areas to capture growth due to species abundance (Derner et al., 2006), and increased soil moisture as a result of the labile C additions from manure that may have preserved soil moisture levels similar to that of irrigated plots. Caravalho et al. (2010) saw an increase in MBC as grazing intensity increased on winter-grazed oat and ryegrass which was partially attributed to increased labile organic carbon additions from livestock manure. Contradictory results of how MBC is impacted by grazing livestock in an integrated crop–livestock system may be due to differences in methods used to quantify MBC (i.e., chloroform fumigation-incubation method, chloroform fumigation-extraction method), soil texture, climate, and stocking rate. More time under this management system may provide greater insights to how MBC is influenced by grazing time.

    Microbial biomass C had a significant (P = .0029), moderate (R = .42127) correlation with SOC for the 0–15 cm depth. Soil MBC is a portion of the SOC pool that is composed of living organisms, including bacteria and fungi; therefore, a positive correlation between MBC and SOC is expected. A similar correlation was seen in a long-term tillage study on sandy clay loams and clay soils in Spain where MBC had a moderate correlation (R = .352) with SOC (Melero et al., 2009). The positive correlation between SOC and MBC and the effect that grazing time had on MBC in the first two years of this study adds to the evidence that MBC serves as a soil health indicator that is more responsive to changes in management.

    3.7 Microbial respiration

    No differences were observed for microbial respiration (CO2–C) between treatments in 2019 or 2020, nor was a year-treatment effect observed (Table 2). In 2019, CO2–C ranged from 90.63 to 99.75 μg g−1 soil. In 2020, CO2–C ranged from 74.96 to 113.83 μg g−1 soil. A meta-analysis conducted by Zhao et al. (2017) found that heavy grazing significantly decreased the size of the microbial community, whereas light to moderate grazing increased the size of the microbial community. Additionally, Zhao et al (2017) found that microbial community size and soil respiration were positively correlated. The size of the microbial community plays an important role in the mechanisms involved in soil C dynamics under varying grazing intensities, so more information on the microbial community and more time may be needed to understand how various grazing lengths impact microbial respiration in the Southeast.

    3.8 Arbuscular mycorrhizal fungi colonization rates

    Arbuscular mycorrhizal fungi colonization rates were not influenced by grazing treatments in 2019 (P = .8082) or 2020 (P = .3739). In 2019, AMF colonization rates ranged from 72 to 74%. In 2020, AMF colonization rates ranged from 67 to 73%. Years were analyzed separately because different cash crops were planted each year (i.e., cotton in 2019 and peanut in 2020). Research on AMF colonization rates and integrated crop–livestock systems is extremely limited, especially in the Southeast. Davinic et al. (2013) found that two types of integrated crop–livestock systems that integrated perennial and annual crops in rotation with cotton increased AMF abundances in cotton on a Texas clay loam soil compared with continuous monoculture cotton without grazing, which differs from the current study that planted the same cover crop species in each treatment. Arbuscular mycorrhizal fungi play an important role in water and nutrient availability for plants, as well as soil aggregation due to their release of glomalin-related proteins that are known to improve soil aggregates. Cosentino et al. (2006) found that fungal biomass was correlated to WSA, but no correlation between AMF colonization rates and WSA were seen in this study. Because of different crops in rotation, more time under this integrated crop–livestock system may be needed to compare colonization rates and distinguish differences. Arbuscular mycorrhizal fungi colonization rates were not correlated to any other soil health indicator in this study.

    3.9 Yield

    Grazing time affected cotton lint yield in 2019. Cotton lint yield of the ungrazed control and mid-February cattle removal treatments were greater than the mid-April and mid-March treatments (Table 4). Cotton lint yield's coefficients of variation for mid-March and mid-April cattle removal treatments (40.4 and 42.1, respectively) were also greater than those of the control and mid-February cattle removal treatments (34.8 and 32.8, respectively; data not shown). Greater cotton lint yield and lower yield variability may be related to greater cover crop residue remaining on the soil surface that minimized the soil moisture loss during the growing season. The lower yields in the mid-March and mid-April treatments could also be linked to compaction observed at the 35 to 50 cm depths (Figure 2). Grazing time did not affect peanut yield in 2020, but peanut yield decreased numerically with increasing grazing time (Table 4).

    TABLE 4. Cotton lint yield in 2019 and peanut yield in 2020 between treatments at Wiregrass Research and Extension Center
    Treatment 2019 cotton yield 2020 peanut yield
    kg ha−1
    Control 1,810 a 6,130
    Mid-February 1,799 a 5,911
    Mid-March 1,575 b 5,723
    Mid-April 1,609 b 5,706
    • Note. Values followed by the same letter in a row within each year are not significantly different between treatments according to Tukey's honestly significant difference at α = .1.

    There is conflicting research on the effects that grazing livestock have on cash crop yield. For example, in a four-year study evaluating winter grazing treatments in a cotton production system, Schomberg et al. (2014) observed that cotton yield was numerically lower in an ungrazed compared with grazed treatment, but the difference was not significant. Franzluebbers and Stuedemann (2014) found that winter grazing did not impact yield of corn or soybean on an integrated crop–livestock system on Ultisols in Georgia. Negative impacts on yield have been reported in several studies of integrated crop–livestock systems that used high grazing intensities and stocking rates, but due to differences in climate, soil type, vegetation, and many other factors, it can be challenging to generalize the impacts that an integrated crop–livestock system has on yield. The noninversion subsoiling practices used in this study helped to mitigate compaction that may have been caused by cattle. More research is needed to properly evaluate how the type of integrated crop–livestock system used in this study affects cotton and peanut yield. Future data from the upcoming years may provide this information.

    4 CONCLUSION

    After the first two years of integrating winter-grazing livestock into a cotton–peanut rotation, there were few negative impacts on selected soil health indicators. Of the indicators analyzed in this study, MBC was the only one to exhibit a response to grazing treatment. Soils in ungrazed cover crops had greater MBC than some grazed treatments, likely due to greater biomass present on the soil surface. The ungrazed treatment and treatments with short grazing periods resulted in increased residue on the soil surface at the time of cover crop termination. Higher biomass levels may have contributed to increased cotton lint yields, possibly by limiting soil moisture loss during the cash crop growing season. Physical soil health indicators were unaffected by the presence of livestock and length of grazing, indicating that grazing did not negatively impact soil structure or compaction at the stocking rates used in the current study. There was no effect of grazing time on SOC storage, which was expected since SOC is known to take several years to build in coarse-textured soil and the hot, humid climate of the Southeast. Although POXC has been reported to be more responsive to changes in management than SOC, POXC did not show a response to cattle removal treatments in the current study. The lack of treatment differences indicates that integrating winter-grazing livestock does not negatively impact soil health in row crop production systems of the Southeast in early years of adoption. However, more time under this system is required to thoroughly evaluate how winter-grazing livestock impacts long-term soil health and yield.

    ACKNOWLEDGMENTS

    We gratefully acknowledge Anna Johnson, Jessie Williford, Noah Hull, and Taibo Liang for their contributions to lab and field analyses. Financial support was provided by the Alabama Natural Resources Conservation Service and the National Institute of Food and Agriculture.

      AUTHOR CONTRIBUTIONS

      Hayley Crowell: Data curation; Formal analysis; Writing – original draft; Writing – review & editing. Audrey V Gamble: Conceptualization; Funding acquisition; Investigation; Methodology; Project administration; Resources; Supervision; Writing – review & editing. Yucheng Feng: Conceptualization; Methodology; Writing – review & editing. Kipling Balkcom: Data curation; Methodology; Writing – review & editing. Anna Yang: Data curation; Methodology; Validation.

      CONFLICT OF INTEREST

      The authors declare no conflict of interest.