Establishing rhizoma peanut–bahiagrass mixtures

Abstract

Rhizoma peanut (RP; Arachis glabrata Benth.) is a perennial legume with ability to persist and spread in mixtures with C₄ grasses. Most RP establishment studies focus on its use as a monoculture hay crop, and more information is needed on establishment strategies for RP–grass mixtures. The objectives of this study were to compare techniques for establishing RP–grass mixtures and quantify their effects on RP shoot emergence, frequency of occurrence (percentage of quadrats containing RP shoots), percentage ground cover, and RP shoot and root-rhizome biomass. Treatments were (a) RP planted in monoculture (MRP; control), (b) RP planted simultaneously with bahiagrass (Paspalum notatum Flügge; RP-BG), or (c) RP strip planted into existing bahiagrass pasture following removal of bahiagrass sod strips (Strip). There was no treatment or sampling date effect on RP shoot density (29.5 shoots m^–2) during 10 wk after spring planting. By October, RP ground cover was greater for Strip (43%) than MRP (32%) and for MRP than RP-BG (18%), and Strip and MRP had greater RP frequency than RP-BG (82, 72, and 58%, respectively). Strip and MRP (1.39 and 1.10 Mg ha^–1, respectively) had greater year-of-planting RP shoot mass than RP-BG (0.52 Mg ha^–1), but root-rhizome biomass was not affected by treatment. Measures of establishment success for Strip generally were equal or superior to MRP, and Strip is a viable option for establishing RP–grass mixtures. Weed competition reduced RP establishment success in RP-BG relative to the other treatments because herbicides tolerated by both bahiagrass seedlings and emerging RP shoots are unavailable.

Abbreviations
• MRP
• monoculture rhizoma peanut
• RP
• rhizoma peanut
• RP-BG
• rhizoma peanut–bahiagrass mixture planted simultaneously
• Strip
• strip planting of rhizoma peanut in existing bahiagrass

INTRODUCTION

Cow-calf operations are widespread in the southeastern United States and rely primarily on warm-season perennial grasses as a feed source. Though well adapted and generally reliable, grass monocultures depend upon N fertilizer for optimal performance, and often their nutritive value limits animal performance. Grass–legume mixtures may provide a variety of benefits for the environment (Dubeux et al., 2017; Sollenberger et al., 2019), and the generally superior nutritive value of legumes relative to C₄ grasses can increase animal performance and production efficiency (Boddey et al., 2004; Hernández-Garay et al., 2004). There are few warm-season perennial legumes adapted to the southeastern United States, but rhizoma peanut (RP; Arachis glabrata Benth.) has demonstrated persistence under a range of grazing management strategies (Ortega-S. et al., 1992).

Rhizoma peanut has been used primarily as a monoculture hay crop instead of pasture due to the high cost of vegetative establishment and greater economic return from sale of hay with superior nutritive value compared with animal products from grazed pasture. There are large areas of pasture grazed by livestock in the southern Gulf Coast region that could benefit from incorporation of RP if establishment methods were reliable and fit budget constraints. Recent research has documented feasibility of strip planting RP into portions of existing grass pastures to reduce cost (Castillo, Sollenberger, Blount, et al., 2013; Castillo, Sollenberger, Ferrell, et al., 2013; Mullenix et al., 2014). The vigorous rhizome system provides a mechanism for RP to spread from the planted strip into surrounding grass (Sollenberger et al., 2014). Significant attention to management for controlling grass competition is required when strip planting RP. Some producers have expressed interest in planting RP and the grass component simultaneously into a prepared seedbed; however, this approach has not been investigated.

The RP germplasm Ecoturf (Prine et al., 2010) is a relatively decumbent type that has been established successfully by strip planting into bahiagrass (Paspalum notatum Flügge) (Mullenix et al., 2014). It has demonstrated phenotypic plasticity under frequent and intense grazing by shortening stems and maintaining greater residual leaf mass after defoliation events, thus limiting its reliance on stored reserves for subsequent regrowth (Mullenix et al., 2016; Shepard et al., 2018). These characteristics increase the likelihood of it contributing successfully to grazed grass–legume mixtures.

More information is needed regarding strategies for successfully establishing RP–grass mixtures and for understanding how interplant competition affects establishment success. The objectives of this experiment were (a) to compare Ecoturf RP establishment after planting in monoculture, in binary mixture with bahiagrass (simultaneous planting of both species into a prepared seedbed), and in strips in existing bahiagrass swards, and (b) to determine the effect of varying levels of competition associated with different planting techniques on Ecoturf RP shoot emergence, frequency of occurrence, ground cover, and shoot and root-rhizome biomass accumulation.

Core Ideas

• More reliable and cost-efficient options are needed for establishing mixtures of rhizoma peanut and bahiagrass.
• When planted simultaneously with bahiagrass, competition hindered establishment of rhizoma peanut.
• Lack of herbicides tolerated by bahiagrass seedlings and rhizoma peanut shoots contributed to the poor performance.
• Strip planting rhizoma peanut into existing bahiagrass and use of herbicide in the strips was successful.
• Control of competition to rhizoma peanut shoots during establishment is critical to achieving successful stands.

MATERIALS AND METHODS Experimental site

The experiment was conducted at the University of Florida Beef Research Unit near Gainesville, FL (29°27′ N, 82°35′ W) during 2015 and 2016. A new set of plots was planted in an adjacent area each year. Prior to initiation of the experiment, the site was a long-term (>30 yr) ‘Pensacola’ bahiagrass pasture. Soils at the site are Pomona sands (sandy, siliceous, hyperthermic Aeric Haplaquods). Soil samples taken prior to initiation of the experiment were analyzed by the University of Florida Extension Soil Testing Laboratory. The soil pH was 6.1, and Mehlich-3 extractable P, K, and Mg were 32 (medium), 28 (low), and 67 mg kg⁻¹ (high), respectively. Fertilizer was applied according to soil test recommendations, with 15 kg P and 56 kg K ha⁻¹ applied in April 2015 and 2016. Also, 500 kg ha⁻¹ of calcitic lime was applied to the plots planted in 2016. The P source was triple superphosphate, and the K source was muriate of potash.

Treatments and experimental design

Treatments were three establishment methods: (a) Ecoturf RP monoculture planted into a prepared seedbed (MRP), (b) a binary mixture of Ecoturf RP and Pensacola bahiagrass planted into a prepared seedbed on the same day (RP-BG), and (c) a binary mixture of Ecoturf RP and Pensacola bahiagrass with RP strip planted into existing bahiagrass (Strip), as previously described by Castillo, Sollenberger, Blount, et al. (2013) and Mullenix et al. (2014). Four replicates of the three treatments were arranged in a randomized complete block design. Each experimental unit was 5 m wide and 10 m long.

Site preparation, planting, and plot management

Land preparation was initiated on 30 July 2014, when plots assigned to MRP and RP-BG treatments were sprayed with glyphosate herbicide at a rate of 3.4 kg a.i. ha⁻¹ to kill existing bahiagrass. Approximately 3 wk after spraying, these plots were tilled several times with a heavy cutting disk. A leveling disk was used subsequently at 2- to 3-wk intervals during the remainder of summer 2014 and monthly during fall and winter before planting occurred in 2015. Plots in an adjacent area were managed in the same manner during 2015 in preparation for their use when the study was repeated in 2016.

Preparation for planting RP in strips in existing bahiagrass was based on a successful approach used for ‘Florigraze’ RP by Castillo et al. (2014) and was initiated on 24 Mar. 2015 for the 2015 experiment and 9 Feb. 2016 for the 2016 experiment. The first step was removal of a 2-m-wide strip of bahiagrass sod through the center of the 5-m-wide plot using a commercial sod cutter, resulting in 2-m-wide by 10-m-long strips through the entire length of the plots. There was 1.5 m of intact bahiagrass remaining on either side of this strip. Strips were then tilled with a leveling disk prior to planting to create a prepared seedbed within the strip.

Immediately prior to planting, a tractor with a tool bar on which four hoes were mounted was used to create furrows for planting RP rhizomes. Furrows were 50 cm apart and approximately 8 cm deep. There were four furrows across the width of each strip in the Strip treatment and eight furrows across the 5-m width of the other treatments.

On 26 Mar. 2015 and 8 Apr. 2016, Ecoturf rhizomes were dug by a collaborating producer, transported to the planting site, and planted the same day. The planting rate was 1,250 kg ha⁻¹ of planted area, and rhizomes for each plot were weighed separately and evenly distributed in all furrows. Rhizomes were covered with soil immediately after planting. In RP-BG plots, after RP rhizomes were covered, bahiagrass seed was broadcast at a seeding rate of 25 kg ha⁻¹. A cultipacker was used to incorporate the seed and increase seed– or rhizome–soil contact. During the period from planting until the onset of summer rains in June of both years, rainfall was supplemented with irrigation to ensure that the weekly sum of the two equaled at least the 30-yr weekly average rainfall.

Method of weed control varied among treatments due to differences in species’ sensitivity to herbicides. The RP strip in the Strip treatment and entire MRP plots were sprayed with ammonium salt of imazapic (Impose, Makhteshim Agan of North America) at 0.07 kg a.i. ha⁻¹ plus 2,4-D amine (dimethylamine salt of 2,4-dichlorophenoxyacetic acid) (2,4-D amine Weed Killer, Universal Crop Protection Alliance) at 0.28 kg a.i. ha⁻¹ when grass or broadleaf weeds were 5–10 cm tall in order to control sedge (Cyperus sp.), broadleaf weeds, and bahiagrass seedlings. Spraying occurred on 15 May and 20 July 2015 for plots planted in the first year and on 3 May and 23 June 2016 for those planted in the second year. The RP strip in the Strip treatment was also treated with Fusilade (fluazifop-P-butyl) at 144 ml ha⁻¹ on 6 May and 20 July 2015 to control bahiagrass and occasional common bermudagrass [Cynodon dactylon (L.) Pers.]. The MRP plots were sprayed with Fusilade on 20 July 2015 and 3 May 2016 at the same rate as used for Strip. Weeds in the RP-BG plots could not be sprayed during the year of planting because bahiagrass seedlings are sensitive to all herbicides available for use in RP. Thus, these plots were mowed to 10 cm every 4–6 wk to minimize shading from weeds. Cutting height of the mower was taller than the height of RP in the canopy, so no RP herbage mass was removed during mowing.

Response variables Rhizoma peanut shoot density

Shoot density was determined after first-shoot emergence and biweekly thereafter. Shoot density measurements were initiated on 1 May 2015 and 4 May 2016, 5 and 4 wk after planting, respectively, and they ended 6 wk later when individual shoots could no longer be identified. Counts were made in four 30- by 100-cm quadrats in the interior of each plot as described by Aryal et al. (2020). Quadrats were placed with the long side perpendicular to the planted row, and each end of the quadrats was placed at the midpoint of the inter-row space (row spacing was 50 cm). Thus, each quadrat contained two 30-cm lengths of planted row. Quadrat locations were flagged so that the same area could be assessed at each counting date. Shoot density was expressed as the average of the four locations per plot in units of shoots per square meter.

Rhizoma peanut ground cover and frequency

Ground cover of RP was estimated visually by the same observer starting at the end of the shoot emergence period and every 28 d thereafter throughout the year of planting (July to October each year). A 1-m² (2 × 0.5 m) quadrat was placed at two marked locations in the center of the plots containing RP, and ground cover was estimated from the same area at each date. The 0.5-m side of the quadrat was oriented parallel to planted RP rows, and the quadrat encompassed four rows of RP. The quadrat was divided into a hundred 10- by 10-cm squares to facilitate estimations. At each sampling event, ground cover by RP was estimated in the same 20 squares per quadrat location and averaged to obtain the measure of RP ground cover for each experimental unit (Castillo, Sollenberger, Blount, et al., 2013; Mullenix et al., 2014). Frequency of RP (Castillo, Sollenberger, Blount, et al., 2013) occurrence (i.e., percentage of squares assessed per plot with at least one RP shoot present) was evaluated at the same time and in the same 10- by 10-cm squares per quadrat placement as percentage cover. Two quadrats were sampled per plot, and frequency was calculated as [(number of squares per plot containing at least one RP shoot)/40] × 100 (Castillo, Sollenberger, Blount, et al., 2013).

Rhizoma peanut shoot and root-rhizome mass

Rhizoma peanut shoot and root-rhizome mass were measured at the end of the growing season of the year of planting, on 27 Oct. 2015 for the 2015 planting and on 20 Oct. 2016 for the 2016 planting. These responses were also measured at the end of growing season of the year after planting, on 20 Oct. 2016 for the 2015 planting and on 26 Oct. 2017 for the 2016 planting. Quadrats of 30 × 100 cm were placed at two representative locations per plot. The long side of the quadrat was placed perpendicular to the planted RP rows and was positioned so that each end was at the mid-point of a row and there were two RP rows included within the quadrat.

All shoot and root-rhizome biomass and soil were removed to a 20-cm depth. Shoot material was removed by clipping to soil level. Subsequently, soil was removed from belowground plant biomass by washing with water over a 2-mm mesh screen. Shoot and root-rhizome fractions were separated into RP, grass, and other. Rhizoma peanut rhizomes are characterized as axonomorphic, typically composed of primary (conical shape, thickening from branching points) and secondary (<3-mm diameter, cylindrical) rhizomes (Krapovickas & Gregory, 1994; Saldivar et al., 1992). Bahiagrass rhizomes were identifiable by their scaly appearance, short internodes, and horizontal growth from which leaves grow vertically (Houck, 2009; Sampaio & Beaty, 1976). Separated fractions were dried at 60 °C until constant weight and weighed. Only mass of RP fractions is reported because no bahiagrass was planted in the Strip or MRP treatments, and plants arising from the soil-seed reserve were controlled in the year of planting using herbicides, as described earlier. Mass of RP fractions was reported in megagrams per hectare of land area planted to RP (i.e., the entire plot for MRP and RP-BG), and the planted strip for the Strip treatment.

Statistical analysis

Data were analyzed using R Software (R Core Team, 2016). A mixed model was used for analysis with the package “nlme” (Pinheiro et al., 2016). Treatments and sampling date were fixed effects. Sampling date was considered a repeated measure for RP shoot emergence, ground cover, and frequency. Year and block were considered random effects. Least squares means were analyzed with the package “lsmeans” (Lenth, 2016). Main effects were considered significant when P ≤ .05, and interactions were discussed when P ≤ .10.

RESULTS AND DISCUSSION Rhizoma peanut shoot density

There was no effect of treatment on RP shoot density (29.5 shoots m⁻²; P = 0.36) throughout the shoot emergence period. Thus, during the 6-wk period of shoot emergence there was no detectable negative effect of competing BG seedlings in the RP-BG treatment relative to MRP and Strip.

Strip planting of Florigraze RP into existing bahiagrass was evaluated by Castillo, Sollenberger, Blount, et al. (2013), Castillo, Sollenberger, Ferrell, et al. (2013), and Castillo et al. (2014), and this method was assessed for Ecoturf, Florigraze, ‘UF-Peace’, and the germplasm Arblick by Mullenix et al. (2014). The range in shoot density at the end of the shoot emergence period (32–41 shoots m⁻²) in the current study was considerably less than the 88 shoots m⁻² reported for Florigraze (Castillo et al., 2014) and 66 shoots m⁻² for Ecoturf (Mullenix et al., 2014), but it is similar to that for UF Peace (30 sprouts m⁻²; Mullenix et al., 2014). In most previous studies, treatments with greatest shoot emergence had the greatest RP percentage ground cover later in the growing season. An exception was a study of seedbed preparation effects on RP establishment (Castillo et al., 2014). In that experiment, shoot density soon after planting was greater for tilled vs. no-till systems, but RP ground cover was 50% greater for no-till vs. tillage at the end of the growing season, attributed in part to lesser weed pressure in the no-till system. The authors indicated that controlling competition to RP after shoot emergence had a greater effect on establishment success than seedbed preparation. This observation is supported by the results of the current study, as discussed below.

Rhizoma peanut ground cover

The treatment × sampling date interaction approached significance (P = .098), so treatments were compared within sampling dates across years (Figure 1). From the first sampling event in July through the last in October, RP cover increased from 5 to 32% for MRP, from 8 to 18% for RP-BG, and from 12 to 43% for Strip (Figure 1). Cover was greater for Strip than the other treatments at all dates, with treatments ranked Strip > MRP > RP-BG at the end of the growing season (Figure 1).

FIGURE 1. Ecoturf rhizoma peanut (RP) ground cover (percentage of ground covered by RP plants relative to total area sampled) response to treatment beginning at the end of the shoot emergence period in July 2015 and 2016, after planting in spring each year, and continuing through October of 2015 and 2016. Asterisks indicate significance of treatment comparisons (***, P ≤ .001; **, P ≤ .01). MRP = monoculture rhizoma peanut; RP-BG = rhizoma peanut–bahiagrass mixture planted simultaneously; Strip = rhizoma peanut strip planted into existing bahiagrass

Rhizoma peanut ground cover for Strip and MRP in the current study are comparable with the range of 30–40% reported for Ecoturf and Florigraze when strip planted into bahiagrass at the same location (Mullenix et al., 2014), as well as the results of Williams et al. (2008) and Interrante et al. (2011) for Florigraze without weed competition. The seasonal pattern of increasing ground cover for Ecoturf observed in the current study is also similar to that reported by Mullenix et al. (2014). Florigraze RP achieved approximately 25% cover by the end of the year of planting when weeds in a planted strip were controlled using imazapic or imazapic plus 2,4-D, but cover reached only 10% at the end of the growing season if mowing alone was used to control competition (Castillo et al., 2014). As noted earlier, they reported greater shoot density for a prepared seedbed vs. a no-till system, but ground cover at season end was greater for no-till, in part because of lesser weed pressure. This helps to explain observations in the current study. Specifically, RP shoot density was not negatively affected by planting simultaneously with bahiagrass, but the greater competition to establishing RP in the RP-BG mixture than in MRP, attributable to the lack of available herbicides for use in RP-BG, resulted in superior planting-year performance when RP was planted alone (i.e., in Strip or MRP).

Rhizoma peanut frequency

There were no interaction effects of treatments on RP frequency, with frequency being defined as percentage of squares is a subdivided quadrat having at least one RP shoot relative to total number of squares assessed (Castillo, Sollenberger, Blount, et al., 2013). However, RP frequency was affected by treatment and sampling date (P < .01). Averaged across sampling dates and years, Strip had greater RP frequency than MRP and RP-BG treatments (70, 55, and 48%, respectively), and frequency increased from 38% in July to 72% in October. For consistency with ground cover data already reported, RP frequency results were presented for each sampling date (Figure 2). The pattern of frequency response and ranking of treatments was very similar to that for cover. Frequency was affected by treatment at all sampling dates (P ≤ .03), with differences between Strip and MRP decreasing throughout the year of planting and differences between MRP and RP-BG increasing as the year progressed. Thus, the negative effects of bahiagrass and weed competition on RP frequency increased over time in RP-BG plots.

FIGURE 2. Ecoturf rhizoma peanut (RP) frequency (percentage of 40 squares assessed per plot with presence of at least one RP shoot) response to treatment starting at the end of the shoot emergence period in July 2015 and 2016, after planting in spring each year, and continuing through October of 2015 and 2016. Asterisks indicate significance of treatment comparisons (***, P ≤ .001; **, P ≤ .01; *, P ≤ .05). MRP = monoculture rhizoma peanut; RP-BG = rhizoma peanut–bahiagrass mixture seeded simultaneously; Strip = rhizoma peanut strip planted into existing bahiagrass

Previous research by Mullenix et al. (2014) showed very similar patterns of Ecoturf frequency response as was observed for MRP and Strip (i.e., increasing throughout the growing season to reach ∼80% by October). They found that decumbent RP entries like Ecoturf achieved relatively high frequency and ground cover by the end of the year of planting. Imazapic and imazapic + 2,4-D treatments, the latter identical to that applied to MRP and Strip in the current study, resulted in Florigraze RP frequency of 60–70% (Castillo et al., 2014). The mowing treatment (no herbicide) in that study is the most comparable with RP-BG in the current study, and it resulted in ∼40% frequency by the end of the growing season, somewhat less than the 58% that that we observed. They found that greater light penetration to the level of RP in the canopy resulted in greater RP cover and frequency (Castillo et al., 2014). Thus, lesser RP light interception due to shading from bahiagrass or weeds in the RP-BG plots likely reduced RP cover and frequency in that treatment relative to MRP and Strip.

Rhizoma peanut shoot and root-rhizome mass

In October of the year of planting, there was an effect of treatment on RP shoot and total biomass (P = .009 and .014, respectively), but there was no treatment effect on RP root-rhizome mass (P = .14) (Table 1). Strip and MRP resulted in a >100% increase in RP shoot mass compared with RP-BG (Table 1). Root-rhizome mass followed the same general treatment ranking, but there was greater variation for this response and no differences among treatments were detected. Total RP biomass (i.e., shoot plus root-rhizome) was greater for Strip than for RP-BG but there was no difference between Strip and MRP, and between MRP and RP-BG (Table 1). For all treatments, root-rhizome biomass contributed slightly more than 50% of total biomass (52–58%) by the end of the year of planting.

TABLE 1 Ecoturf rhizoma peanut (RP) shoot, root-rhizome, and total plant organic matter mass responses to treatment in October 2015 and 2016 following planting in late March 2015 and early April 2016

Treatment	Shoot	Root-rhizome	Total
	Mg ha−1
MRP	1.10a	1.17a	2.27ab
RP-BG	0.52b	0.73a	1.25b
Strip	1.39a	1.49a	2.88a
SE	0.15	0.22	0.35

Note. Data for the Strip treatment are expressed in Mg ha⁻¹ of planted area, and all data are means across two years and four replicates (n = 8). Means within a column followed by the same letter are not different (P > .05). MRP, monoculture rhizoma peanut; RP-BG, rhizoma peanut–bahiagrass mixture planted simultaneously; Strip, rhizoma peanut strip planted into existing bahiagrass.

By the end of the growing season of the year after planting, shoot biomass for MRP was greater than the other treatments (Table 2). Shoot mass was 304% greater for MRP than for RP-BG and 178% greater for Strip than RP-BG. Root-rhizome mass for MRP and Strip did not differ, but they were more than 180% greater than for RP-BG. Similarly, total biomass did not differ between MRP and Strip, but it was at least 179% greater for MRP and Strip compared with RP-BG. These means for root-rhizome biomass are quite similar to those reported in October following March planting of Ecoturf in monoculture (1.10 Mg ha⁻¹; Aryal et al., 2020) and are within the range of values (0.71–1.60 Mg ha⁻¹) reported in the same study for shoot biomass.

TABLE 2 Ecoturf rhizoma peanut (RP) shoot, root-rhizome, and total organic matter mass responses to treatment at the end of the growing season in the year after planting

Treatment	Shoot	Root-rhizome	Total
	Mg ha−1
MRP	1.86a	4.40a	6.25a
RP-BG	0.46c	1.20b	1.68b
Strip	1.28b	3.41a	4.69a
SE	0.16	0.51	0.66

Note. Data for the Strip treatment are expressed in Mg ha⁻¹ of planted area, and all data are means across two years and four replicates (n = 8). Means within a column followed by the same letter are not different (P > .05). MRP, monoculture rhizoma peanut; RP-BG, rhizoma peanut–bahiagrass mixture seeded simultaneously; Strip, rhizoma peanut strip planted into existing bahiagrass.

CONCLUSIONS

Planting bahiagrass (seed) simultaneously with Ecoturf RP (rhizomes) did not negatively affect Ecoturf shoot emergence, but RP ground cover and frequency of occurrence during the year of planting were less in RP-BG than Strip or MRP plots. This was most likely because of competition from bahiagrass and weeds in the RP-BG treatment. Only mowing could be used for weed control in the RP-BG plots, as bahiagrass seedlings are sensitive to all herbicides registered for use in RP. Greater competition from bahiagrass or weeds in RP-BG also reduced RP shoot and total biomass accumulation in the year of planting. In addition, shoot, root-rhizome, and total biomass accumulation were least for RP-BG through the end of the growing season of the year after planting. We conclude that Ecoturf RP establishes well when planted in pure stand in a prepared seedbed or when strip planted into bahiagrass because competition to establishing RP can be controlled using herbicides during the establishment period. In contrast, planting Ecoturf and bahiagrass simultaneously in a prepared seedbed is not a recommended practice for establishing RP–grass mixtures because herbicides are lacking that are tolerated by both emerging bahiagrass seedlings and RP shoots, and resultant RP establishment is poor.

AUTHOR CONTRIBUTIONS

Erin M. Shepard: Conceptualization; Formal analysis; Investigation; Methodology; Writing – original draft. Lynn E. Sollenberger: Conceptualization; Formal analysis; Investigation; Methodology; Project administration; Resources; Supervision; Writing – original draft; Writing – review & editing. Marta M. Kohmann: Investigation; Methodology; Writing – review & editing. Liliane Severino da Silva: Investigation; Methodology; Writing – review & editing. John F. Harling: Writing – review & editing. Jose C. B. Dubeux: Conceptualization; Writing – review & editing. João M. B. Vendramini: Conceptualization; Writing – review & editing.

CONFLICT OF INTEREST

The authors declare no conflict of interest.