Interaction of soil health indicators to different regenerative farming practices on mineral soils

Abstract

Regenerative farming is a common practice adopted to enhance soil health. Routinely, cheaply measured soil health indicators can be used to predict unknown soil health indicators, which makes analysis simple and fast. The objective of this study was to determine the interaction and relationship between soil health indicators and the effects of regenerative farming practices on soil health indicators. We conducted a comprehensive on-farm study across Florida, measuring 11 soil health indicators for 592 soil samples that were collected from the surface 15 cm of soil on 13 experimental sites. Sampled fields were conventionally managed or subjected to either cover cropping, organic amendment application, or their combination, and fallow fields as control plots. We tested the Pearson correlation coefficient between 11 soil health indicators and further tested the types of relationships between each indicator using regression analysis. Strong positive correlations (r ≥ + .5 to 1.0) were obtained, especially between organic matter (OM) with maximum water holding capacity (MWHC), cation exchange capacity (CEC), soil protein (SP), and total Kjeldahl nitrogen (TKN); MWHC with CEC, SP, and TKN; TKN with CEC and SP; total phosphorus (TP) with Mehlich-3 phosphorus (M3P). A strong negative correlation (r ≥ –.5 to –1.0) was observed between bulk density (BD) with OM, MWHC, CEC, and TKN. No significant change in soil health indicators was observed in control plots except for active carbon (AC) and CEC. Organic amendments and a combination of organic amendments with cover crops were effective in improving soil health indicators. However, cover crops alone had no effect on soil health indicators except for CEC and M3K.

Abbreviations
• AC
• active carbon
• BD
• bulk density
• C
• carbon
• CEC
• cation exchange capacity
• M3K
• Mehlich-3 potassium
• M3P
• Mehlich-3 phosphorus
• MWHC
• maximum water holding capacity
• OM
• organic matter
• SP
• soil protein
• TKN
• total Kjeldahl nitrogen
• TP
• total phosphorus.

INTRODUCTION

Soil health, often used interchangeably with soil quality is defined as “the continued capacity of soil to function as a vital living ecosystem that sustains plants, animals, and humans”. Soil health is assessed by measuring the soil properties that provide clues on how well the soil can function. These measurable soil properties are also known as soil health indicators which include physical, chemical, and biological markers used to diagnose a soil. Some of the common soil health indicators include soil pH, bulk density (BD), maximum water holding capacity (MWHC), organic matter (OM), active carbon (AC), cation exchange capacity (CEC), soil protein (SP), total Kjeldahl nitrogen (TKN), total phosphorus (TP), total potassium (TK), extractable Mehlich-3 phosphorus (M3P) and Mehlich-3 potassium (M3K) (Moebius-Clune et al., 2016). While these indicators can independently affect soil health, they can interact with each other to either enhance or suppress overall soil health (Bhadha et al., 2017; Libohova et al., 2018; Olorunfemi et al., 2016). Increase in OM can result in increased MWHC and lower BD (Bhadha et al., 2017; Libohov et al., 2018). Prediction of soil health indicators using one or more measured soil indicators (referred to as “Pedotransfer”) has been reported in previous studies (Krogh et al., 2000; McBratney et al., 2002). Olorunfemi et al. (2016) reported a successful prediction of CEC and MWHC from basic soil physical and chemical properties including pH and soil organic carbon (OC). Researchers around the world attempt to predict CEC from clay and organic carbon (Bell & van Keulen, 1995; Drake & Motto, 1982; Yuan et al., 1967).

Continuous monocropping, intense agricultural operations, removal of natural vegetation, and soil erosion are main factors that can lead to decline in soil health and crop yields. In recent years, regenerative farming practices including shifting from monoculture to crop rotation, incorporation of the cover crop during fallow season, and addition of soil amendments has increased, and this has shown to potentially enhance soil health (Ashworth et al., 2018; Jian et al., 2020). Cover crops add crop residues to the soil which improve the soil physical, chemical, and biological properties. Crop residue which is typically composed of OM mineralizes over time and improves soil aggregation, increases available WHC, and lowers BD (Ashworth et al., 2018; Blanco-Canqui & Lal, 2009; Steward et al., 2018). Planting of leguminous cover crops as cover crop or crop rotation may result in greater soil N due to N₂ fixation. Our previous study reported that growing cover crops during fallow periods could be a viable option for Florida's growers since it showed beneficial effects on soil health (Bhadha et al., 2021). The addition of organic amendments to soil is also a common practice for promoting soil health and nutrient cycling. Soil organic amendments have a wide range of benefits that include increased OM and WHC, reduced BD, improved crop yields and quality (Cogger et al., 2008; Favoino & Hogg, 2008). In this study, two types of common locally derived organic wastes have been selected (bagasse and horse bedding). Bagasse as one of the sugarcane industry by-products that generated and accumulated in the mills, is a fibrous organic material left after the extraction of sugar juice from sugarcane. In Florida, it is typically utilized as a fuel for sugar mills, but there is still a surplus amount of bagasse that needs disposal. Horse bedding that obtained from local equestrian facilities in Florida, also produces more than 200,000 t of bedding material annually. Application of these two locally generated organic materials have the potential to improve soil health in Florida while reducing the wastes.

Core Ideas

• Organic amendments, and the combination of organic amendments and cover crops had a positive effect on soil health indicators.
• Cover crops alone were not effective in enhancing most soil health indicators.
• Strong positive and negative correlations were observed between soil health indicators.

Soils of Florida are typically sandy, with low OM (<4%) and nutrient content (Bhadha et al., 2017). Nutrient deficiencies, poor physical attributes, and low water storage capacity are major farming constraints associated with these soil types. A conventional cropping system on mineral soils in Florida relies on a large amount of external nutrients, which have the potentials to be lost via leaching (Xu et al., 2019). In this study, cover cropping, organic amendment (bagasse) application, cover crops combined with organic amendment (horse bedding) have been selected as three types of regenerative farming practices. The seeding rates of cover crops and application rates of organic amendments were based on the recommendations from local farm managers and growers. Previous studies highlighted the benefits of cover crops, organic amendments, and the combined effect of cover crop and organic amendments at the farm scale in each individual study within Florida (Bhadha et al., 2021; Xu et al., 2019; Xu et al., 2021), but comparing all these farming practices across Florida (at the regional level) is relatively unknown. Also, the interaction and relationship between soil health indicators within the region is limited. The purpose of this study was to: (a) determine the interaction and their relationship between soil health indicators and (b) evaluate the effects of regenerative farming practices on soil health indicators. Results from this study provides valuable insights for growers and researchers about the potential of including cover crops, organic amendments, and a combination of cover crop and organic amendments in a farming system. The information on interaction and relationship between soil health indicators will help make indirect prediction of unknown soil health indicators from a limited number of known measurements.

MATERIALS AND METHODS Site description and treatment used

Sites selected included four different soil orders of varying inherent properties, farm types, and regenerative farming practices (Table 1). The criteria used to select experimental sites in this study include, (a) regenerative farming practices (either cover crops, addition of organic amendments, and the combination of cover crops and organic amendments); (b) geographical location (different counties across Florida, as shown in Figure 1); (c) farm size and type [farm size ranged from large commercial production 81 ha to small organic growers 2-4 ha); farm types including sugarcane (Saccharum officinarum L.), sweetcorn (Zea mays L.), and vegetables]; (d) growers’ willingness (the collaborative growers expressed strong interest in performing regenerative farming practice on their farms). Thus, in total, 13 agricultural farms across Florida were identified for this study. The climate of Florida ranges from humid subtropical to tropical with an average annual temperature 20.8°C and an average annual precipitation 1,242 mm (WorldClimate Data, http://www.worldclimate.com/climate/us/florida). Since this is a comprehensive on-farm study with multiple collaborative growers across the state, each farm followed its unique management practices from 2017 to 2021. The management practices included (a) cover crops, (b) organic amendments (bagasse), and (c) cover crops mixed with organic amendment (horse bedding) (Table 1). Different cultivars of cover crops were grown on the farms in rotation with cash crops, either in summer, winter, or both. Cover crops used during this study includes alfalfa (Medicago sativa L.), buckwheat, mustard seeds [Brassica nigra (L.) Koch], oat (Avena sativa L.), pea (Pisum sativum L.), barley (Hordeum vulgare L.), cereal rye (Secale cereale L.), crimson clover (Trifolium incarnatum L.), faba bean (Vicia faba L.), flax (Linum usitatissimum L.), lentil (Lens culinaris Medik.), oat, pea, Triticale (Triticosecale), radish (Raphanus sativus L.), rye, safflower (Carthamus tinctorius L.), vetch (Vicia spp.), wheat (Triticum spp.), white clover (Trifolium repens L.), buckwheat (Fagopyrum esculentum Moench), cowpea [Vigna unguiculata (L.) Walp.]/sunn hemp (Crotalaria juncea L.), cowpea/sudan grass (Sorghum × drummondii).

TABLE 1 Description of farm type, location of farm, treatment applied, and soil samples

Site	Farm type	Soil order	Location	Treatment	Treatment performing dates	Total soil sample	Samples from control plota	Samples from treated plot
1	vegetables	Ultisols	Walton County	alfalfa (Medicago sativa), buckwheat, mustard seed (Brassica nigra), oat (Avena sativa), pea (Pisum sativum) barley (Hordeum vulgare), cereal rye (Secale cereale), crimson clover (Trifolium incarnatum), faba bean (Vicia faba), flax (Linum usitatissimum), lentil (Lens culinaris), oat, pea, Triticale (×Triticosecale), radish (Raphanus sativus), rye, safflower (Carthamus tinctorius), vetch (Vicia spp.), wheat (Triticum spp.), white clover (Trifolium repens)	Summer: May–August 2018 Winter: December 2018–April 2019	14	2	12
2	pasture	Ultisols	Columbia County	oat & rye mix	Winter: December 2018–April 2019	12	0	12
3	peanut	Ultisols	Columbia County	oat	Winter: December 2018–April 2019	12	0	12
4	hay	Ultisols	Columbia County	oat	Winter: December 2018–April 2019	12	0	12
5	vegetables, fruits, flowers	Entisols	Alachua County	Summer: Buckwheat, cowpea & sunn hemp mix	Summer: May–August 2018	12	0	12
6	vegetables	Alfisols	Palm Beach County	sunn hemp	Summer: May–August 2018	12	0	12
7	organic vegetables	Alfisols	Palm Beach County	cowpea, cowpea & sudan grass mix; cowpea & sunn hemp mix	Summer: May–August 2018 Summer: May–August 2019	32	8	24
8	vegetables	Alfisols	Palm Beach County	cowpea & sunn hemp mix; cowpea & sunn hemp mix	Summer: May–August 2018 Summer: May–August 2019	24	8	16
9	sugarcane	Spodosols	Hendry County	sunn hemp	Summer: May–August 2018	14	2	12
10	sugarcane	Spodosols	Hendry County	sunn hemp	Summer: May–August 2019	14	2	12
11	sugarcane	Spodosols	Hendry County	sorghum, sudangrass, and sunn hemp	Summer: May–August 2017	56	10	46
12	sugarcane	Entisols	Hendry County	5-cm bagasse (7,700 kg ha−1) 10-cm bagasse (15,400 kg ha−1) 10-cm bagasse + extra N (58 kg N ha−1)	May 2017	288	54	234
13	sugarcane	Entisols	Hendry County	horse bedding & cowpea (HB+CC)	horse bedding: Feb 2017 cowpea: May–August 2017	90	0	90

^aControl plot for cover cropping was left fallow; control plot for organic amendment (Bagasse) were maintained similar to the treatment plots except that the treatments were not added; and cover crops mixed with organic amendments (HB + CC) had no control plots.

Organic amendment used in this study was bagasse. Bagasse incorporated into the top 15 cm of soil included: a 5-cm layer of bagasse at a rate of 7,700 kg ha⁻¹; a 10-cm layer of bagasse at a rate of 15,400 kg ha⁻¹; and a 10-cm layer of bagasse with extra N using ammonium nitrate at a rate of 58 kg N ha⁻¹. The extra N was applied to account for the initial immobilization of N that could occur. The timeline for the management practice is shown in Figure 2.

The farm that followed mixed management (referred to this study as HB + CC) practice applied horse bedding before planting the cover crop. Horse bedding was obtained from local equestrian facilities in Palm Beach County of Florida. Horse bedding was incorporated into the top 15 cm of soil and 12 wk after the application of horse bedding, cover crop (cowpea, Vigna unguiculata L.) was planted at approximately 34 kg ha⁻¹ rate. The basic physiochemical properties of bagasse and horse bedding is presented in Table 2.

TABLE 2 Physiochemical properties of bagasse and horse bedding

		Bagasse
Physiochemical properties	Unit	Concentration	Horse bedding
pH		3.79 ± 0.01	7.37 ± 0.17
Bulk density	g cm−3	0.11 ± 0.001	0.19 ± 0.04
Organic matter	%	95.0	94.68 ± 1.59
P	mg kg−1	0.263 ± 0.008	1,160 ± 638
K	mg kg−1	0.997 ± 0.043	9,098 ± 1421
Ca	mg kg−1	2.101 ± 0.012	4,673 ± 1,033
Fe	mg kg−1	0.326 ± 0.008	105.7 ± 46.78
Al	mg kg−1	0.217 ± 0.005	126.5 ± 31.69

Soil sampling

Since this study was conducted with multiple collaborative growers across the state, each farm followed its own experimental design. For example, the farm that applied bagasse as organic amendments, was performed a completely randomized design with three replications. Efforts were made to ensure sampling of site level replication when available; however, this was not possible for all the experimental sites, such as in study site 1, 9, and 10, only one composite soil sample was collected from the fallow fields (control plots) at each sampling.

Baseline soil samples were collected from all fields before the application of different treatments (referred to in this study as pre-, which represented initial soil conditions). The second soil samples (referred to in this study as post-) were collected after the cover crops were cut and tilled into the soil surface prior to cash crop being planted from farms that practiced cover crop and cover crop mixed organic amendments (HB + CC). Farms that applied bagasse as organic amendments, after treatment soil samples were collected three times. First post-harvest soil sampling was done after sugarcane harvest (18 mo after bagasse application), second soil samples were collected after first ratoon harvest (36 mo after bagasse application), and third soil samples were collected after final harvest of sugarcane (48 mo after bagasse application). A minimum of five composite samples were collected from the top 15 cm of surface soil from each field for each sampling. A composite sample comprised of mixing 10 samples collected along a transect from an individual field. Total numbers of soil samples collected from each farm depended on the farm size. Soil samples were collected in 3.78 L (1-gallon) Ziploc bags and transported to the Soil, Water, and Nutrient Management Laboratory at Everglades Research and Education Center in Belle Glade, FL, for analyses.

Analytical methods

All pre- and post- soil samples were air-dried, passed through 2-mm sieve, and analyzed for soil health indicators, including soil pH, BD, MWHC, OM, CEC, AC, SP, M3P, M3K, TP, and TKN. Soil pH was determined with a 1:2 soil/water ratio using Accumet AB250 pH meter. Bulk density was calculated by dividing soil mass in a fixed core volume. Maximum water holding capacity was determined by the modified method described in Jenkinson and Powlson (1976) based on saturation procedure. Organic matter content was calculated based on the loss on ignition (LOI) method at 550 °C. Cation exchange capacity was estimated using the ammonium acetate method (Sumner & Miller, 1996). Active carbon (C) was determined based on K permanganate (KMnO₄) oxidizable C using 0.02 M KMnO₄ for mineral soils, in which approximately 2.5 g of soil was reacted with 20 ml of 0.02 M KMnO₄ for exactly 2 min, filtered, and the supernatant solution was then analyzed using Thermo Scientific Genesys 30 spectrophotometer at 550 nm (Schindelbeck et al., 2016). Soil protein was determined by using a sodium citrate extraction method (Schindelbeck et al., 2016) under autoclaving with high temperature and pressure. The extracted protein was quantified by using the Thermo pierce colorimetric bicinchoninic acid assay (BCA) as calibrated against protein standards of known concentration. The color development was read by using Thermo Scientific Genesys 30 spectrophotometer at 550 nm. Soil available P and K were determined using Mehlich-3 extraction method and then analyzed using Agilent 5110 inductively coupled plasma-optical emission spectrometer (ICP-OES) (Santa Clara, CA). Total P was determined by ashing samples for at least 5 h (not to exceed 16 h) at 550 °C in a muffle furnace followed by extraction with 6 M HCl and analyzed using ICP-OES. The TKN was determined using the digestion method followed by colorimetric determination (EPA method 351.2).

Statistical analysis

A dataset of n = 592 was used for computing statistical analysis. Soil pH and BD followed the normal distribution. However, other soil health indicators including OM, MWHC, CEC, SP, TP, M3P, M3K, and TKN were not normally distributed. Log₁₀ transformation was successful in obtaining a normal distribution (Figure 3). The statistical analysis was performed on the log-transformed data in order to meet the assumptions of linear mixed models, but the results demonstrated in figures (except Figure 3) and tables present nontransformed values. Pearson correlation coefficient was calculated to test the level of interaction between soil health indicators. Based on the Pearson correlation coefficient results the strength of association were classified into three groups: weak (r ≥ ± .01 to .3), medium (r ≥ ± .3 to .5), and strong (r ≥ ± .5 to 1.0). Parameters that had a larger strength of association were further tested for the types of interaction using regression analysis. Effects of different agriculture practices on soil health indicators were analyzed using the generalized linear mixed models (GLIMMIX) method (SAS version 9.4, SAS Institute Inc.). Means were separated using Tukey–Kramer multiple-comparison procedure when the F test was significant at p ≤ .05.

RESULTS AND DISCUSSION Correlation analyses

Pearson correlation coefficients were computed for every pair of soil health indicator (Table 3). Strong positive correlations (r ≥ + .5 to 1.0) were obtained, especially between OM with MHWC, CEC, SP, and TKN; MWHC with ECE SP, and TKN; TKN with CEC and SP; TP with M3P. Strong negative correlation (r ≥ –.5 to −1.0) was observed between BD with OM, MWHC, CEC, and TKN (Table 3). Those pair of soil health indicators that were strongly correlated were further tested for their types of relationship. Pairs of soil health indicator that had r² ≥ .5 were OM and BD, OM and MWHC, OM and TKN, MWHC and BD, and TP and M3P (Figure 4). 

TABLE 3 Pearson correlation coefficients for soil health indicators (n = 592)

Indicator	pH	BD	OM	MWHC	AC	CEC	SP	TP	M3P	M3K	TKN
pH	1.0000
BD	–0.3217	1.0000
OM	0.4539	–0.6989	1.0000
MWHC	0.4141	–0.7187	0.7591	1.0000
AC	0.3488	–0.2818	0.4085	0.3648	1.0000
CEC	0.2651	–0.5519	0.6251	0.5357	0.4412	1.0000
SP	0.2056	–0.4242	0.7114	0.5103	0.4192	0.4647	1.0000
TP	0.3755	–0.3186	0.4375	0.3336	0.1050	0.1773	0.4647	1.0000
M3P	–0.0023	–0.0009	–0.0278	–0.0149	–0.0476	–0.0816	–0.1439	0.6327	1.0000
M3K	0.2402	–0.2893	0.3945	0.2600	0.4307	0.2965	0.3207	0.3210	0.1424	1.0000
TKN	0.4842	–0.6032	0.8141	0.7280	0.4818	0.5810	0.6046	0.3620	–0.0805	0.3283	1.0000

Note. AC, active carbon; BD, bulk density; CEC, cation exchange capacity; M3K, Mehlich-3 potassium; M3P, Mehlich-3 phosphorus; MWHC, maximum water holding capacity; OM, organic matter; SP, soil protein; TKN, total Kjeldahl nitrogen; TP, total phosphorus.

The results of correlation analysis revealed a significant and positive relationship between OM and MHWC (Table 4). For every 1% increase in OM accounts for 2.9% increase in MHWC. Previous studies have also shown a similar trend of increase in MWHC with OM. Bhadha et al., 2017 reported that every 1% increase in OM resulted in a 2.3 and 1.3% increase in soil MWHC for sand and lime rock (calcareous bedrock), respectively. Other research shows that a 1% increase in OM can increase MWHC up to 5% depending on soil texture (Emerson et al., 1994). It was observed that soil with higher OM demonstrated higher CEC. Results indicate that every 1% increase in OM equates to 1.1% increase in CEC. This finding is also in agreement with Olorunfemi et al., 2016, where CEC showed a positive correlation with soil pH, soil organic C, OM, MWHC, and clay content. Soil samples with higher CEC values were found to have a high level of OM and soil pH (Fasinmirin & Olorufemi, 2012; Vogelman et al., 2010). Organic matter in soil is a major source of negative electrostatic sites which creates a strong correlation between CEC and OM in the soil (Olorunfemi et al., 2018). Soil protein is an effective indicator of soil health and considered a proxy for available organic N (Hurisso et al., 2018). Results indicate 1% increase of OM accounts for 42.2% increase in SP. Plant residues are the source of much of the OM and protein is one of the components of OM (Moebius-Clune et al., 2016). An increase in OM increased the SP. Soil protein can be increased by adding organic amendments and cover cropping whereas protein tends to decrease with increasing soil disturbance (Moebius-Clune et al., 2016). The SP content is organically bound N which influences the soil ability to store N and make it available by mineralization during the growing season (Hurisso et al., 2018; Roberts & Jones, 2008; Schindelbeck et al., 2016). Nitrogen mineralization is an important indicator of soil fertility and soil health. Since SP is the organically bound N in the soil, increased SP in the soil will increase soil N. In addition, SP content is also associated with soil aggregation and affects the MWHC (Moebius-Clune et, al., 2016). Active C promotes microbial growth which influences the soil health and quality parameters such as MWHC, CEC, soil structures, and aggregation (Moebius-Clune et, al., 2016). Results showed strong correlation between SP and TKN. Soil proteins represent the large pool of organically bound N in soil OM that can be mineralized by soil microbial communities (Hurisso et al., 2018; Roberts & Jones, 2008; Schindelbeck et al., 2016).

TABLE 4 Relationship between soil health indicators with coefficient of determination

Relationship between soil health indicators	Equation	R2
BD-OM	y = −14.123x + 22.676	.51
BD-MWHV	y = −57.228x + 127.21	.52
BD-CEC	y = 22.142x + 35.844	.27
BD-TKN	y = −4532.3x + 7273.2	.37
OM-MWHC	y = 2.9822x + 39.687	.55
OM-CEC	y = 1.1001x + 2.2319	.26
OM-SP	y = 42.235x + 202.95	.50
OM-TKN	y = 294.51x + 105.51	.61
MWHC-CEC	y = 0.272x – 7.3638	.26
MWHC-SP	y = 7.9154x – 33.879	.28
MWHC-TKN	y = 65.111x – 2062.2	.48
CEC-TKN	y = 533.89x0.4498	.27
SP-TKN	y = 3.6771x – 63.089	.34
TP-M3P	y = −0.219x + 43.499	.70

Note. AC, active carbon; BD, bulk density; CEC, cation exchange capacity; M3K, Mehlich-3 potassium; M3P, Mehlich-3 phosphorus; MWHC, maximum water holding capacity; OM, organic matter; SP, soil protein; TKN, total Kjeldahl nitrogen; TP, total phosphorus.

Soil OM had a strong negative correlation with BD. Based on our data, every 1% increase in OM decrease bulk density by 14.1%. Soil OM has a lower particle density than mineral soils, which reduces the overall BD. Also, OM enhances soil aggregation and structure, which lowers BD. Bulk density is also negatively correlated with MWHC, CEC, and TKN. As the BD increased MWHC decreased. At higher BD the larger pores are lost and there are less pores to potentially hold water. Libohova et al. (2018) reported a weak and negative correlation between BD and MWHC.

Effects of regenerative farming practices on soil health indicators

Soil health indicators were not significantly different between pre- and post- measurement for control plots except for AC and CEC (Figure 5). Cation exchange capacity and M3K were the only parameters that were decreased in post-harvest soils compared to pre-planting of cover crop (Figure 5). In HB+CC treatment BD decreased, whereas OM, MWHC, TP, and M3P content increased after post-harvest (Figure 5).

5 FIGURE

Effects of regenerative farming practices on soil health indicators. (a) soil pH, (b) bulk density, (c) organic matter, (d) maximum water holding capacity, (e) active C, (f) cation exchange capacity, (g) soil protein, (h) total P, (i) Mehlich-3 P, (j) Mehlich-3 K, and (k) total Kjeldahl N. HB + CC, horse bedding + cover crop. ^aMeans sharing a common letter within each treatment are not significantly different at p = .05 significance level

Pre: soil sample collected before treatment application.

Post 1: Cover crop treatment– after incorporation cover crop into the soil; HB+CC– after first sugarcane harvest; Bagasse– after first sugarcane harvest (18 mo after bagasse application).

Post 2: Bagasse– after first ratoon harvest (36 mo after bagasse application).

Post 3: Bagasse– after second ratoon harvest (48 mo after bagasse application).

Application of bagasse decreased soil pH, AC, and CEC over a 3-yr period. The effects of bagasse were short termed for indicators such as BD, which decreased at 18 mo after application but were not significantly different at 36 and 48 mo after application compared to pretreatment whereas TKN increased at 18 mo after application compared to pretreatment, 36 mo, and 48 mo after application of bagasse. Maximum water holding capacity and M3P increased after bagasse application. Soil protein and M3K decreased at 18 mo after measurement but increased at 36 mo after measurement compared to pretreatment.

The soil pH in this study ranges from neutral to slightly alkaline (6.63–7.5). In Florida, the alkaline soil may be due to the presence of limestone bedrock comprising of Ca- and Mg-based minerals (Bhadha et al., 2018). Bagasse was the only treatment that showed significant changes in the soil pH. Application of bagasse lowered the soil pH because the pH of bagasse added was acidic (3.79 ± 0.01). The significant decrease in BD at 18 mo after bagasse application and HB+CC treatment may be due to the low BD of bagasse (0.1 ± 0.001 g cm⁻³) and horse bedding (0.19 ± 0.04 g cm⁻³). However, at 36 and 48 mo, BD showed no significant changes after bagasse application, which may be due to the decomposition of bagasse. Results suggest bagasse and mixture of horse bedding and the cover crop could do a better job in decreasing BD in a short period than cover cropping alone. The lowering of BD improves soil air and water movement, alleviating soil compaction which promotes root growth. Increase in OM after HB+CC treatment may be due to the addition of OM from horse bedding (94% OM) and residue from cover crops. Increasing soil OM is key to improve soil health (Doran & Zeiss, 2000). Organic matter binds soil particles into aggregates which improves soil structure, reduces compaction, and increases infiltration rates. Organic matter also serves as a reservoir for nutrients and water in the soil as well as enhances microbial biodiversity and activity. Also, there was a significant increase in MWHC after bagasse and HB+CC treatment. The increase in MWHC may be due to increased soil OM. Water holding capacity of soil could be more than double if the OM of sandy soil is increased from 0.55 to 5% (Hudson, 1994). Bhadha et al. (2017) reported on mineral soil MWHC could be increased by adopting farming practices that increase soil OM.

The CEC significantly decreased after cover cropping and bagasse application.

Interestingly, we observed no significant change in CEC after HB+CC treatment although it had increased OM content. The soil with higher OM and clay particles demonstrated higher CEC values (Manrique et al., 1991; Martel et al., 1978). Fasinmirin and Olorunfemi (2012) and Vogelmann et al. (2010) reported that soil with a high level of OM and pH had higher CEC. The CEC of soil also depends on soil pH, types of clay, and amount and source of OM (Rashidi & Seilsepour, 2008). The soils of farms used in this study had sandy texture with coarse particles and low OM. Active C is the measure of the fraction of soil C pool that can be easily mineralized by microbial communities. Active C is one of the most important soil health indicators that are useful in the early detection of changes in crop and regenerative farming practices (Schindelbeck et al., 2016). Roper et al. (2017) reported an increase in AC with reduced-tillage and addition of organic amendments. However, we observed a decrease in AC after bagasse application which may be because of seasonal fluctuations. Active C was reported as the most sensitive indicator since it was easily affected by fluctuation in environmental conditions (Sahoo et al., 2019).

We observed a decrease in SP after 18 mo but an increase in SP after 36 mo of bagasse application. The increase in SP after 36 mo may be due to higher SP present in bagasse which required some time to become available for the plant. Although SP was decreased TKN was increased after 18 mo of bagasse application. Total P was increased for bagasse treatment. Increase in P availability will be beneficial to the field crop. Melich-3 P increased for bagasse and HB+CC treatment, which is a good indicator of bioavailable P in mineral soils. Melich-3 K decreased for cover cropping and at 18 mo after bagasse application. The decrease in M3K after cover crop may be due to K uptake by cover crops and potential K leaching. Potassium has a high mobility in sandy soils, especially under the humid climate of Florida, K leaching is common. However, no significant change in M3K was observed for mixed treatment. 

CONCLUSIONS

Based on the results, strong, medium, and weak correlations were observed between the measured 11 soil health indicators. Specifically, strong positive correlations (r > + .5 to 1.0) were obtained, especially between OM with MWHC, CEC, SP, and TKN; MWHC with CEC, SP, and TKN; TKN with CEC and SP; TP with M3P. A strong negative correlation (r > –.5 to −1.0) was observed between BD with OM, MWHC, CEC, and TKN. These results suggest possibilities for simplifying comprehensive soil health assessment and predicting the unknown soil properties based on the measured indicators.

In general, the addition of organic amendments and combination of organic amendments with cover cropping could be a viable option for Florida's growers since these treatments showed beneficial effects on soil health. However, cover cropping alone was not effective in enhancing most of the soil health indicators such as decreasing CEC and M3K. This could be because the short duration of this study, assessing only a single season of cover crops. Subsequent years of cover cropping could result in additional positive effects as documented by mid-western Long-Term Agroecosystem Research (LTAR) studies (Baffaut et al., 2020; Conway et al., 2020). Further research is warranted in addressing the economics of conducting these practices and a sustainable mechanism to utilize locally available amendments. Locally derived organic amendments, such as bagasse and horse bedding, can be potentially incorporated in mineral soils to improve soil health in Florida, particularly in combination with cover crops during the fallow period.

ACKNOWLEDGMENTS

This research was supported in part by the USDA National Institute of Food and Agriculture - Hatch Project FLA-ERC-005552 and the Florida Specialty Block Grant Program (Agreement no. 026705). The authors would like to thank the many collaborative growers involved in this study including Patrick Troy, Kim Yates, Tripp Whidden, Derek Orsinego, Margaret Duriez, Noah Shitama, Libby Schmidt, and Jeff Willis. Mr. Salvador Galindo and Martino Trotta for assistance with field and lab work.

AUTHOR CONTRIBUTIONS

Nan Xu: Data curation; Formal analysis; Writing – original draft. Abul Rabbany: Data curation; Methodology; Validation; Writing – review & editing. Jay Capasso: Methodology; Writing – review & editing. Kevin Korus: Methodology; Writing – review & editing. Stewart Swanson: Investigation, Methodology, Writing-review & editing; Jehangir H. Bhadha: Conceptualization, Funding acquisition, Investigation, Project administration, Supervision, Visualization, Writing-original draft.

CONFLICT OF INTEREST

Authors declare no conflict of interest.