Methane production and nitrogen balance of dairy heifers grazing palisade grass cv. Marandu alone or with forage peanut

Abstract

Livestock production systems are an essential agribusiness activity in Brazil, but a critical challenge of Brazilian farmers is to maintain the equilibrium of the ecosystem, using herbage resources efficiently with a minimum impact on the environment. Nitrogen (N) fertilization and the inclusion of forage legumes into tropical grass pastures are management strategies which increase the productivity and nutritive value of pastures and may also affect methane (CH₄) production by ruminants. The objective of this study was to examine the effects of either fertilizing palisade grass pastures with N or including the forage peanut (Arachis pintoi) into grass pastures on enteric CH₄ emission, microbial protein production in the rumen via purine derivatives in the urine, and N balance. Twenty-one nonlactating crossbred dairy heifers were used in a completely randomized design with 3 treatments. The treatments consisted of pastures of palisade grass without N fertilization (control), fertilized with urea (fertilized), and palisade grass mixed with forage peanut (mixed). Seven animals (replications) were used to evaluate dry matter intake, digestibility, CH₄ emission, urea, purine derivatives, and volume of urine, and N ingestion and excretion. Four paddocks (replications) were used to measure herbage mass; morphological, botanical, and chemical composition of herbage; and herbage allowance. The CH₄ emissions were determined using the sulfur hexafluoride (SF₆) tracer gas technique. The efficiency of N utilization (ENU) was calculated using the N balance data. Crude protein (CP) concentration of herbage increased with fertilization or legumes inclusion (P < 0.0001) while neutral detergent fiber (NDF) concentration decreased (P = 0.0355). The leaf allowance was higher in the fertilized treatment (P = 0.0294). Only uric acid excretion increased with N fertilization (P = 0.0204). The ENU was not affected by fertilized or mixed compared to control and averaged 55% (P = 0.8945). The enteric CH₄ production was similar between treatments and averaged 129 g/d (P = 0.3989). We concluded that the changes in chemical composition of herbage provided by N fertilization or the inclusion of the legume showed no reduction in enteric CH₄ emissions, but the ENU was more significant than previous studies with palisade grass, suggesting that different management strategies might alter the ENU under grazing conditions.

Introduction

Livestock production system is an essential agribusiness activity in Brazil as pastures occupy the largest area of the national territory. Thus, higher yields and productivity of livestock can help to improve the economic sustainability of the country. The most significant challenge of cattle production in Brazil is to maintain the equilibrium of the ecosystem, using herbage resources efficiently with a minimum impact on the environment (Cardoso et al., 2016). Although tropical grasses most used in extensive systems are relatively well adapted to low soil fertility, insufficient nutrients, especially nitrogen (N), coupled with inadequate pasture management can lead to low herbage production and, consequently, lower animal performance (Alencar et al., 2018). A common practice in pasture production systems which can increase the productivity and protein content of tropical herbages is N fertilization, but the excessive use of N fertilizers can have significant environmental impacts, such as the emission of nitrous oxide (N₂O) and ammonia (NH₃). A possible solution is the inclusion of forage legumes into pastures.

Tropical grasses have high fiber content which can result in high production of enteric methane (CH₄) (Archimède et al., 2018). In the rumen environment, CH₄ is produced anaerobically by methanogenic microorganisms, and its production is modulated mainly by the presence of free carbon dioxide (CO₂) and hydrogen (H₂), occurring the reduction of these gases to CH₄ and water (Beauchemin et al., 2008; Tedeschi and Fox, 2018). It is a process directly related to the characteristics of the animal diet, including the level of intake, nutritive value of available herbage and digestibility of ingested mass, genetics, animal management, and conditions of ruminal fermentation (Hristov et al., 2015).

In general, reductions in CH₄ emissions can be achieved by strategies such as improving the composition and quality of pastures and diet by reducing the cell wall content and increasing crude protein (CP) and soluble carbohydrates concentrations of herbage (Berndt and Tomkins, 2013). Also a reduction in CH₄ emissions can be achieved by improving animal genetics and diet supplementation (Cardoso et al., 2016).

Faced with the energy losses as enteric CH₄, increasing the efficiency of N utilization (ENU) by ruminants is a problem for many cattle ranchers because it results in excessive N losses, becoming environmentally and economically challenging. According to Russell et al. (1992), the reduction of N losses by feces and urine is possible by improving the efficiency of the capture of degraded N in the rumen through the synthesis of microbial protein. Furthermore, the adequate supply of specific amino acids to the ruminal bacteria can increase the fermentation of fiber carbohydrates (Tedeschi and Fox, 2018).

Regarding the availability of nutrients to the plants, the use of N fertilizers is an efficient means of increase herbage mass (McRoberts et al., 2018), since N is the most limiting element to the plant (Boddey et al., 2004) and has a positive correlation with leaf growth and quality, increasing the nutritive value of herbage (Andrade et al., 2016).

An alternative to reduce environmental impacts caused by the excessive use of chemicals is the introduction of legumes into the pasture (Muir et al., 2017, 2018). The practice induces indirect effects related to the biological N fixation (BNF) (Traill et al., 2018), reducing the emissions of CO₂ and N₂O to the atmosphere and the fossil energy used in the production of N fertilizers (Jensen et al., 2012). In addition, studies have shown an increase in herbage quality in mixed pastures, due to the lower fiber content, higher passage rate, and, in some cases, the presence of condensed tannins in forage legumes, which modify the ruminal fermentation, resulting in reduced CH₄ emissions by ruminants (Archimède et al., 2011; Tedeschi et al., 2014; Tedeschi and Fox, 2018).

Both N fertilization and BNF by legumes are important practices to improve development of pastures and may also affect CH₄ production (Andrade et al., 2016), in order to reduce the negative contribution of livestock to global warming and increase agronomic efficiency and animal performance. However, is necessary to quantify the production of enteric CH₄, especially when in mixed pastures with legumes or using N fertilizers.

Forage peanut (Arachis pintoi) may be the most promising tropical legumes for mixed pastures with tropical grasses, especially those on dairy farms. We hypothesize that mixed pastures of palisade grass and forage peanut may reduce enteric CH₄ emissions when compared to N fertilization or palisade grass pastures without N fertilizer and may increase the ENU by crossbred dairy heifers. Therefore, the objectives of this study were to investigate the potential use of forage peanut compared to N fertilization of palisade grass by 1) quantifying enteric CH₄ emission of crossbred dairy heifers; 2) estimating N balance and ENU by the animals, and 3) evaluating herbage mass, intake, nutrient digestibility, and chemical composition.

Materials and Methods

The experiment was approved by the UNESP Council of Animal Experimentation and Animal Use, Campus Jaboticabal, under protocol 10356/14. The experiment was conducted at Forage and Grasslands sector of the Sao Paulo State University “Julio de Mesquita Filho” (UNESP), in Jaboticabal, SP, Brazil (21°15′22″S latitude, 48°18′58″, 77 W longitude and 595 m elevation). The climate, according to the Köeppen system, is Aw type (tropical, characterized by dry winters). According to Cardoso et al., (2019a), the average annual temperature in the city of Jaboticabal-SP is 22.7 °C, with a minimum of 17.1 °C and a maximum of 30.1 °C, and average annual precipitation of 1,207 mm in the period from 1971 to 2010. The experimental pastures were palisade grass (Urochloa brizantha R. D. Webster cv. Marandu) alone or mixed with forage peanut (A. pintoi cv. Amarillo). The experiment was conducted in 2 growing seasons (2016/2017 and 2017/2018), in which 4 evaluations of enteric CH₄ emissions and other variables were performed in the grazing cycles between February and March 2017 and March and April 2018.

Experimental Design

The experiment was realized in a completely randomized design with 3 treatments and 7 replications (animals) for evaluation of CH₄ emission, dry matter intake (DMI), digestibility (DMD), urine parameters, and N balance and 4 replications (paddocks) for the variables of mass and chemical composition of herbage. The treatments consisted of pastures of palisade grass without an N source (control); pastures of palisade grass fertilized with N (150 kg N/ha/yr) (fertilized); and pastures of palisade grass mixed with forage peanut (mixed). The total experimental area was 4 hectares (ha) of which 2.56 ha was reserve area and 1.44 ha for treatments and replications. This area was divided into 12 paddocks, subdivided into 3 units each, with an approximate area of 400 m². The reserve paddocks of palisade grass were used for the animals when they were not in the experimental period.

Animals and Pastures Management

In the first year, 21 crossbred dairy heifers, nonlactating, 24-mo-old and with an average weight of 300 ± 5.9 kg were used, and in the second year, 26 heifers with the same genetic pattern, with an average weight of 270 ± 3.9 kg. The grazing method adopted was intermittent stocking, with a variable stocking rate, according to the mob stocking technique (Allen et al., 2011). The grazing started when palisade grass reached 95% light interception (LI), measured using the LI-COR canopy analyzer (LI-COR, Lincoln, NE). Stocking rate was adjusted based on the mean weight of heifers to achieve 15 cm of residue height, with just 1 d of occupation.

Maintenance fertilization of all treatments with P and K was determined according to the results of chemical analyses, carried out in July 2016 and July 2017, following the recommendation of Technical Bulletin 100 (Van Raij (1997). In the fertilized palisade grass treatment, 150 kg N/ha/yr were provided by urea (45% N), fractionated in 3 applications (November 10, 2016; January 10, 2017; and March 10, 2017 in the first year and November 14, 2017; January 15, 2018; and March 25, 2018 in the second year).

Enteric Methane Emissions

Enteric CH₄ emission of grazing animals was quantified by the sulfur hexafluoride (SF₆) tracer gas technique, according to the manual of Global Research Alliance for Greenhouse Gases on Agriculture (Berndt et al., 2014). A 10-d adaptation period was used for the halter and the PVC yoke use. A calibrated SF₆ permeation tube with average of 89.4 μg/h release rate was introduced into the rumen/reticulum of each animal 7 d before the first collection. After the adaptation period, a collector-storage PVC yoke with 60 mm, class 20 pipe was fitted on each animal.

A blank collector set (yoke + halter) was installed in the paddocks of each treatment to measure the background CH₄ concentration. Enteric CH₄ production was calculated as follow the Eq. 1:

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{Q}\mathrm{C}{\mathrm{H}}_4=\frac{\mathrm{Q}\mathrm{S}{\mathrm{F}}_6\times ([\mathrm{C}{\mathrm{H}}_4]\mathrm{y}-[\mathrm{C}{\mathrm{H}}_4]\mathrm{b})}{[\mathrm{S}{\mathrm{F}}_6]}}\end{array}}$$ (1)

where QCH₄ corresponds to the CH₄ emission rate; QSF₆ to the known rate of SF₆ emission; [CH₄]y to the CH₄ concentration (ppm) of the collecting yoke; [CH₄]b to the CH₄ concentration of the blank yoke; and [SF₆] to the concentration of SF₆ (ppt) of the blank yoke.

Enteric CH₄ collections were performed over 6 sequential days at 24-h intervals. The concentrations of CH₄ and SF₆ were determined on Shimadzu GC2014 gas chromatograph (Kyoto, Japan). From the primary data, CH₄ emissions were expressed as follows: g CH₄/d, g CH₄/d/kg^0.75 of body weight (BW), the percentage of digestible energy, and the CH₄ conversion rate (Ym). For the calculation of Ym, the Eq. 2 of Blaxter and Clapperton (1965) corrected by Wilkerson et al. (1995) was used, considering 0.05565 MJ/g CH₄ and the energy obtained in the hand-plucked samples (EH), in order to estimate the digestible energy from the digestive percentage of the gross energy (Ym).

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{Y}\mathrm{m}(\mathrm{\%})=\frac{(\mathrm{C}{\mathrm{H}}_4\times 0.05565)}{\mathrm{E}\mathrm{H}}\times 100}\end{array}}$$ (2)

Animal Urine Production and Nitrogen Balance

Spot samples of urine were collected from each animal to quantify N concentration and others urine compounds, following the methodology described by Chizzotti et al. (2008). The urine was collected from all the animals during 5 d in the paddocks at approximately 6 am, with the aid of an adapted collector. From the 5 collections, composite samples were obtained for each animal per grazing cycle.

Urine was filtered and 10 mL aliquots were immediately diluted with 40 mL of 0.072 M sulfuric acid (H₂SO₄) to prevent bacterial degradation of the purine derivatives (PD) and the precipitation of uric acid (UA). Samples were stored at −15 °C until analyses.

Concentrations of PD, such as UA and allantoin (ALLA), were determined using commercial Analisa kits by the enzymatic-colorimetric methodology (Cat. 451/MS80022230065) and Fujihara et al. (1987), described by Chen and Gomes (1992), respectively.

The total excretion of PD was calculated by summing the amounts of ALLA and UA excreted in the urine, expressed in mmol/d. Absorbed purines (Pabs, mmol/d) were calculated from the excretion of PD (Eq. 3). Then, the ruminal synthesis of nitrogen compounds (Nmic, g/d) was calculated as a function of Pabs, using Eq. 4 (Chen and Gomes, 1992). The ENU was calculated following the recommendations by Detmann et al. (2014) (Eq. 5).

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{P}\mathrm{D}(\mathrm{m}\mathrm{m}\mathrm{o}\mathrm{l}/\mathrm{d})=0.84\times \mathrm{P}\mathrm{a}\mathrm{b}\mathrm{s}+0.236\times \mathrm{B}\mathrm{W}{}^{0.75}}\end{array}}$$ (3)

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{N}\mathrm{m}\mathrm{i}\mathrm{c}(\mathrm{g}/\mathrm{d})=\frac{70\times \mathrm{P}\mathrm{a}\mathrm{b}\mathrm{s}}{0.93\times 0.137\times 1,000}}\end{array}}$$ (4)

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{E}\mathrm{N}\mathrm{U}=\frac{\mathrm{N}\mathrm{b}\mathrm{a}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{c}\mathrm{e}(\mathrm{g}/\mathrm{d})}{\mathrm{N}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{d}(\mathrm{g}/\mathrm{d})}}\end{array}}$$ (5)

where 70 corresponds to the N concentration in the purines (mg N/mol); 0.134 to the ratio N purine:total N in the bacteria (Barbosa et al., 2011); and 0.83 to the digestibility of microbial purines. The N balance was calculated from the sum of the N excretions in the feces and the urine, subtracted from the N ingestion; and Ning corresponds to the ingested N (g/d).

Another aliquot of concentrated urine was used to estimate the total N concentration by DUMAS method (LECO CN analyser) and the urea concentration through commercial Analisa kits by the colorimetric-enzymatic methodology (Cat. 427/MS80022230063). Creatinine concentration in the spot sample was also determined through commercial Analisa kits using the colorimetric methodology with alkaline picrate (Cat. 435/MS80022230066), and subsequently used to estimate the urinary volume. The daily creatinine excretion (CE, mg/mg) was estimated by the following Eq. 6 proposed by Chizzotti et al. (2008):

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{C}\mathrm{E}(\mathrm{m}\mathrm{g}/\mathrm{m}\mathrm{g})=37.88\times \mathrm{B}{\mathrm{W}}^{0.9315}}\end{array}}$$ (6)

N excreted via urine was obtained by multiplying the urine volume by urinary N concentration (Chizzotti et al., 2008).

Mass and Herbage Allowance

Herbage mass was quantified by measuring 16 points of height of palisade grass randomly with a graduated ruler before and after grazing, with the subsequent collection of 2 representative samples of the average height of the paddock, by cutting at 5 cm from the soil of all herbage present in a 0.25-m² metal frame (Barthram, 1985). Samples were separated into 2 subsamples, one for the determination of the dry matter (DM) concentration and the other for the morphological composition, fractionating them in leaf, stem + sheath, and dead material. After drying in an oven with forced air circulation at 55 °C for 72 h, total DM and morphological components were obtained to estimate the total herbage mass (HM) (Gimenes et al., 2011). Herbage allowance (kg DM/100 kg BW) was calculated as HM divided by the animal stocking rate and multiplied by the percentage of leaves to obtain the leaf allowance.

Chemical Composition of Herbage

Hand-plucked samples were collected to evaluate chemical composition of herbage at each replication and in all grazing cycles, totalizing 16 sample units per treatment (Halls, 1954). After drying and grinding, samples were submitted to chemical analysis, following the methodologies of AOAC (2012). Chemical components analyzed were as follows: total N (crude protein; CP) by the Kjeldahl method; neutral detergent fiber (NDF), and acid detergent fiber (ADF) by the ANKOM fiber analyzer method; lignin by the acid hydrolysis method; ash; ether extract (EE) by the Goldfish method; and gross energy through the complete combustion of the samples using an automatic bomb calorimeter. Total carbohydrates (TC), nonfibrous carbohydrates (NFC), and total digestible nutrients (TDN) were calculated according to NRC (2001).

Intake and Digestibility

Herbage intake by animals was estimated using chromium oxide (Cr₂O₃) as an external indicator. For this assay, 10 g of Cr₂O₃/d were given for 10 d to each animal via esophagus, of which the first 7 d for adaptation and the last 3 d for feces collection (Hopper et al., 1978) at preestablished schedules (15, 11, and 7 h), resulting in a composite sample for each animal, following the methodology of Le Du and Penning (1982). Fecal recovery of Cr₂O₃ was determined following the methodology of Williams et al. (1962). From these data, fecal excretion (FE) was calculated through the equation proposed by Le Du and Penning (1982) (Eq. 7).

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{F}\mathrm{E}(\mathrm{g}/\mathrm{d})=\frac{\mathrm{C}{\mathrm{r}}_2{\mathrm{O}}_3\mathrm{p}\mathrm{r}\mathrm{o}\mathrm{v}\mathrm{i}\mathrm{d}\mathrm{e}\mathrm{d}(\mathrm{g}/\mathrm{d})}{\mathrm{C}{\mathrm{r}}_2{\mathrm{O}}_3\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}(\mathrm{g}/\mathrm{k}\mathrm{g}\mathrm{D}\mathrm{M})}}\end{array}}$$ (7)

Herbage intake and digestibility of dry matter were estimated based on fecal production data using indigestible neutral detergent fiber (iNDF) as the internal marker. To quantify NDF concentration, hand-plucked and feces samples were adequately conditioned in ANKOM F-57 filter bags and arranged in the rumen of fistulated animals for in situ incubation for 240 h, according to the methodology of Van Soest et al. (2000). After removal from the rumen, the bags were washed until thoroughly bleached and dried in an oven with forced circulation at 55 °C for 72 h.

The bags were submitted to neutral detergent extraction in ANKOM fiber analyzer, and the entire drying procedure was repeated for iNDF quantification, according to the methodology of Van Soest (1994). From these values, DMI and DMD were calculated, following Eqs. 8 and 9:

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{D}\mathrm{M}\mathrm{I}(\mathrm{k}\mathrm{g}\mathrm{D}\mathrm{M}/\mathrm{d})=\frac{\mathrm{F}\mathrm{E}\times \mathrm{C}\mathrm{F}\mathrm{F}}{\mathrm{C}\mathrm{F}\mathrm{F}\mathrm{O}}}\end{array}}$$ (8)

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{D}\mathrm{M}\mathrm{D}=1-\frac{\mathrm{C}\mathrm{I}\mathrm{F}\mathrm{O}}{\mathrm{C}\mathrm{I}\mathrm{F}}}\end{array}}$$ (9)

where FE = fecal excretion (kg/d); CFF = concentration of the iNDF indicator in the feces (kg/kg); CFFO = concentration of the iNDF indicator in the herbage (kg/kg).

Statistical Analyses

Data were analyzed for the homoscedasticity and normality of the residues, using the Box-Cox and Cramer-von Mises tests, respectively, both using Nortest package of R program, version 3.5.2 (Team, 2018). The analysis of variance (ANOVA) was conducted using a completely randomized design procedure. For variables related to the animals, such as DMI, DMD, CH₄ emission, urea and PD (ALLA and UA) of urine, urinary volume, and N ingestion and excretion, the animal was the experimental unit. For variables related to the herbage, such as herbage mass, morphological, botanical, and chemical composition of herbage, and leaf and herbage allowance, the paddock was the experimental unit. In some cases, especially for CH₄ emission (g/kg DM), CH₄ conversion rate, ALLA, urinary N, N excreted in the urine, and efficiency of microbial protein synthesis (EMPS), homoscedasticity was not verified, then appropriate transformation was applied to maintain homoscedasticity.

All variables were analyzed as repeated measures and, the statistical model included herbage management, period, and interaction management × period as fixed effects, whereas paddock or animal was the random effect. The best covariance structure used for repeated-measures analyses was chosen as the one that achieved the lowest corrected Akaike and Bayesian information. Data were analyzed using the LME procedure of R (package NLME, R Core Team). Significant effects for treatment were declared at P < 0.05. When a significant effect was found for herbage management, the means of the 3 treatments were compared by the Tukey-HSD test at 5% probability. When period was significant, orthogonal polynomial contrasts were performed to assess the effect of N dose on the variables.

Results

Enteric Methane Emissions

Nitrogen fertilization and the inclusion of forage peanut in pastures did not affect DMI and daily enteric CH₄ emissions by crossbred heifers (P > 0.05) (Table 1). Emissions of enteric CH₄ estimated by DM unit consumed also did not show differences among treatments (P > 0.05) (Table 1). Since no difference was found in the animal’s BW and metabolic weight, the emission of enteric CH₄ expressed in g/kg BW also did not differ between treatments (P > 0.05). There was also no difference in Ym values (P > 0.05) (Table 1).

Table 1.

Enteric methane by dairy heifers raised in palisade grass pastures fertilized with urea (150 kg N/ha/yr, fertilized), mixed with forage peanut (mixed) and without N fertilization (control)¹

Item2	Treatments			SEM	P-value
	Fertilized	Mixed	Control
Body weight, kg	332	329	316	3.92	0.2133
Metabolic weight, kg0.75	77.65	77.24	74.88	0.69	0.2180
DMI, kg DM per animal per day	8.97	9.35	8.46	0.35	0.5715
CH4 emission, g per animal per day	140	132	115	6.90	0.3989
CH4 emission, g/kg DM	16.22	15.79	16.42	0.97	0.6092
CH4 emission, g/kg MW	1.83	1.57	1.83	0.08	0.4018
CH4 conversion rate, Ym; %	4.94	4.81	4.99	0.29	0.6082

1Average 2-yr evaluation (n = 28).

²DMI = dry matter intake; DM = dry matter.

Purine Derivatives

There was no treatment effect on ALLA excretion by the heifers (P > 0.05) (Table 2). On the other hand, N fertilization promoted an increase in UA excretion by the animals of this treatment (P < 0.05) (Table 2). The PD parameter had no treatment effect in this study (P > 0.05) (Table 2). In contrast, the relationship between allantoin and purine derivatives (ALLA:PD) was influenced by the treatments, being higher in the control compared to the mixed, which in turn was higher than the fertilized (P < 0.05) (Table 2). No treatment effect was observed on the Pabs excretion and urine volume (P > 0.05) (Table 2).

Table 2.

Purine derivatives of dairy heifers raised in palisade grass pastures fertilized with urea (150 kg N/ha/yr, fertilized), mixed with forage peanut (mixed) and without N fertilization (control)¹

Items2	Treatments			SEM	P-value
	Fertilized	Mixed	Control
ALLA, mmol/d	82.59	76.64	76.64	10.47	0.8302
UA, mmol/d	14.05a	9.54ab	7.76b	1.27	0.0204
PD, mmol/d	96.64	86.18	84.40	10.83	0.3711
ALLA:PD	0.85c	0.89b	0.91a	0.08	<0.001
Pabs, mmol/d	66.99	74.93	79.68	10.66	0.6007
Urinary volume, L/d	16.10	13.80	13.08	1.61	0.2409
Urinary N, g/L	2.38	2.21	2.46	0.12	0.7096

1Average 2-yr evaluation (n = 28).

²ALLA = allantoin; UA = uric acid; PD = purine derivatives; ALLA:PD = allantoin and purine derivatives relation; Pabs = absorbed purines.

a-cMeans without a common superscript differ at P < 0.05.

The Efficiency of Nitrogen Utilization

Dairy heifers grazing palisade grass pastures fertilized with N, mixed with forage peanut or without N fertilization presented the same daily intake of N from herbage (P > 0.05) (Table 3). There was no influence of N fertilization or introduction of forage peanut on the amount of N excreted in feces and urine (P > 0.05) (Table 3). Although N fertilization increased the CP concentration, there was no influence of treatment on urea excretion by urine (P > 0.05) (Table 3). There were no differences in the ENU and Nmic among treatments (P > 0.05) (Table 3).

Table 3.

Efficiency N utilization by dairy heifers raised in palisade grass pastures fertilized with urea (150 kg N/ha/yr, fertilized), mixed with forage peanut (mixed) and without N fertilization (control)¹

Items2	Treatments			SEM	P-value
	Fertilized	Mixed	Control
Nitrogen balance
N ingested, g/d	119	131	120	6	0.5498
N feces, g/d	52	59	52	2	0.3037
N urine, g/d	40	40	33	4	0.5612
ENU, %	56	54	55	2	0.8945
Urea, mg/dL	392	324	290	37	0.3312
Nmic, g/d	63	49	63	9	0.2786
EMPS, g/d	392	305	292	55	0.1758

1Average 2-yr evaluation (n = 28).

²ENU = efficiency of nitrogen utilization; Pmic = microbial protein; Nmic = synthesis of nitrogen compounds; EMPS = efficiency of microbial protein synthesis (g Pmic/kg TDN).

Mass and Herbage Allowance

There was no treatment effect on herbage mass (HM) (P > 0.05) (Table 4). Herbage allowance was also not affected by N fertilization and legume inclusion (P > 0.05) (Table 4). Pastures of palisade grass fertilized with N, mixed with forage peanut and without N fertilization showed a similar morphological composition (P > 0.05) (Table 4). Leaf allowance was higher in fertilized (P < 0.05) than in the mixed, both of which did not differ from the control (Table 4).

Table 4.

Herbage mass and chemical composition of hand-plucked samples of palisade grass fertilized with urea (150 kg N/ha/yr, fertilized), mixed with forage peanut (mixed) and without N fertilization (control)¹

Item2	Treatments			SEM	P-value
	Fertilized	Mixed	Control
Herbage characters
Herbage mass, kg/ha	4980	3690	4820	390	0.0702
Leaves, g/kg	329	309	360	16	0.3985
Stem and sheath, g/kg	302	263	255	18	0.5402
Dead material, g/kg	371	271	364	28	0.1207
Herbage allowance1	7.55	6.52	7.55	0.7	0.5852
Leaf allowance1	3.27a	1.94b	2.72ab	0.2	0.0294
Chemical composition
OM, g/kg DM	897a	883b	893ab	3	0.0466
Ash, g/kg DM	101b	117a	107ab	3	0.0327
CP, g/kg DM	112a	85b	83b	5	<0.0001
TC, g/kg DM	756b	794a	792a	7	0.0002
NFC, g/kg DM	169	178	156	8	0.4958
NDF, g/kg DM	589b	601ab	618a	6	0.0355
Hemicellulose, g/kg DM	300b	305ab	311a	4	0.0093
ADF, g/kg DM	296	302	307	4	0.2749
Cellulose, g/kg DM	259b	268b	285a	4	0.0010
Lignin, g/kg DM	25	26	27	4	0.9529
EE, g/kg DM	19	17	19	0.4	0.143
DMD, g/kg DM	620	567	604	15	0.3860
Energy, MJ/kg DM	189a	180b	180b	0.6	<0.0001
TDN, g/kg DM	575	518	553	15	0.3540

¹Average 2-yr evaluation (n = 16).

²BW = body weight; OM = organic matter; CP = crude protein; NDF = neutral detergent fiber; ADF = acid detergent fiber; EE = ether extract; DMD = dry matter digestibility; TC = total carbohydrates; NFC = nonfibrous carbohydrates; TDN = total digestible nutrients.a,bMeans followed by the same letters in the line did not differ by Tukey HSD test (P < 0.05).

Chemical Composition of Herbage

The OM concentration was higher (P < 0.05) in the fertilized compared to the mixed, both of which did not differ from the control (Table 4). The ash concentration was inversely proportional to that observed for OM due to the influence of N fertilization, being lower (P < 0.05) in this treatment than in the mixed and both similar to the control (Table 4). The N fertilization increased herbage CP concentration up to 26% and was different from the other treatments (P < 0.05) (Table 4). Total carbohydrates concentration was lower (P < 0.05) in the fertilized in relation to mixed and control. The amount of NFC and TDN showed that there was no influence of N fertilization or legume inclusion (Table 4). Nitrogen fertilization also affected the herbage NDF and hemicellulose concentrations, which had lower mean values (P < 0.05) compared to control, both of which similar to the mixed (Table 4). But there was no influence (P > 0.05) of N fertilization or on the introduction of legumes into the ADF and lignin concentrations. On the other hand, cellulose concentration was higher (P < 0.05) in the control compared to fertilized and mixed (Table 4). The amount of lipids, represented by the EE concentration, did not present a difference between treatments (P > 0.05). There was no difference in DMD between treatments (P > 0.05) (Table 4). The application of N in the pasture increased the herbage energy concentration (P < 0.05), but there was no difference in TDN between treatments (P > 0.05) (Table 4).

Discussion

Enteric Methane Emissions

Although there was a greater leaf allowance provided by N fertilization, DMI by heifers was similar in the 3 conditions (Table 1), showing that the adjustment of the stocking rate was correctly managed, without limiting the animal intake.

Since 1930, as observed by Kriss, and recently confirmed by Niu et al. (2018), it is known that there is a strong effect of DMI on the enteric CH₄ emissions. However, values similar to this study were reported by Barbero et al. (2015) in tropical pastures; and higher values were recorded by Primavesi et al. (2004), of 181.0 g CH₄ per animal per day by dairy crossbred heifers kept in pastures of palisade grass without N fertilization, and of 227.0 g CH₄ per animal per day in pastures of Tobiata grass (Megathyrsus maximus) fertilized. Also, Andrade et al. (2016) found higher means of enteric CH₄ emissions by steers grazing dwarf elephant grass with or without forage peanut, 146 and 180 g CH₄ per animal per day, respectively.

In experiments conducted in Canada, McCaughey et al. (1999) concluded that improving pasture quality by the inclusion of legumes reduced enteric CH₄ emission by 10%, due to the reduced fiber content and increased digestible energy and CP in the diet. Reduction in fiber and an increase in herbage CP concentration were also observed in this study (Table 4), as well as a reduction of 17% in CH₄ production (Table 1), but not differ significantly from treatments of the grass in monoculture.

In addition to the lower fiber content of legumes, there are other factors inherent to these herbages that can reduce methanogenesis, such as higher passage rate and, in some cases, the presence of condensed tannins (Beauchemin et al., 2008). In this context, many studies have been conducted in recent times potential mitigation effect of tannins on the enteric CH₄ emissions, due to its positive effect on ruminal fermentation. Waghorn et al. (2002) demonstrated that the tannin present in the legume Lotus pedunculatus was responsible for the reduction of up to 16% when used by sheep.

The hand-plucked samples of mixed pastures presented 14.2% (±3.6) of forage peanut in the DM. According to Beauchemin et al. (2008), the minimum intake of 0.8% of condensed tannin in DM is required for reduction in enteric CH₄ emission. Based on the study of Paulino et al. (2012), who reported that forage peanut has approximately 1.7% of condensed tannin, the metabolite supplement for the heifers of this study was approximately 0.25% in the ingested DM and thus, insufficient to cause a reduction in CH₄ emission. The desirable amount of forage peanut to achieve the minimum of condensed tannin to observe effects would be 40% in the herbage mass (DM basis).

It is important to mention that the seeding of forage peanut was done simultaneously to the palisade grass in November 2014, and in the 2 subsequent years, the planting of forage peanut seedlings in the pastures was carried out in order to increase the legume stand in the mixed. Nevertheless, the low proportion of legumes (8% of sward) may be partly justified by the slow establishment of the plant which resulted in low intake of forage peanut by the animals (Cook et al., 1994; Alencar et al., 2018; Sanchez et al., 2019).

Therefore, both strategies of N fertilization, which increased herbage CP and decreased NDF concentrations, and the introduction of forage peanut into tropical grass pastures, did not mitigate enteric CH₄ emissions. However, the results indicate a tendency to reduce CH₄ emissions that could become significant with a higher proportion of legume in the pasture. Despite this, the values found here are within the critical limit suggested by Crutzen et al. (1986) from 54 kg CH₄ per animal per year for cattle in pasture systems in Brazil, equivalent to approximately 148 g CH₄ per animal per day, and are also below the IPCC (2006) reference average for pasture dairy cows in Latin America, from 63 kg CH₄ per animal per day, or 173 CH₄ g CH₄ per animal per day.

According to Reynolds et al. (2011), the higher DMI increases CH₄ emissions per animal, but at the same time improves production efficiency, resulting in lower emissions per unit of product (meat or milk) or production cycle. However, the absence of difference in DMI and, consequently, in the emission of CH₄ per unit of DM consumed, this could not be verified in this study.

Methane conversion rate (Ym) results were below the range proposed by USEPA (2000), between 5.5% and 6.5% in North America and Eastern Europe and still below the IPCC (2006) range of between 6.5% and 7.5% for cattle under tropical conditions.

Although not verified in the present study, Blaxter and Clapperton (1965) argue that the improvement in diet quality provided by the introduction of legumes in the pastures tends to increase the voluntary intake of the animal, which promotes more efficient post-ruminal digestion, thus reducing the energy loss of the diet converted to CH₄.

Purine Derivatives

According to Yu et al. (2002), the excretions of ALLA, UA, xanthine, and hypoxanthine, the latter 2 being considered detrimental in cattle (Chen and Gomes, 1992), may be affected by factors such as CP concentration and dietary energy, DMI, BW, presence of feed additives, and animal species. In this study, however, there was no treatment effect on ALLA excretion by crossbred heifers. The results also differ from those described in the literature that indicate that the level of ALLA in cows is not constant and depends on the physiological stage and quality of the diet (Johnson et al., 1998). However, in this study, even with changes in CP and NDF concentrations of herbage promoted by N fertilization and introduction of legumes into the system, there was no interference in excreted ALLA concentrations.

According to Chen and Gomes (1992), PD originates from 2 sources: Pabs in the small intestine, and endogenous, which are released from the metabolism of nucleic acids. In this sense, as the Pabs are catabolized and proportionally recovered in the urine as PD. No treatment effect was also observed in the excretion of PD.

Urinary volume, estimated based on the creatinine concentration in the urine, was also not affected by N fertilization and legume inclusion. The values were similar to those found by Cardoso et al. (2019b) of 12.4 L/d at the same experimental site. The creatinine excretion and, consequently, urinary volume, is relatively constant as a function of BW and is barely affected by dietary factors (Barbosa et al., 2006). However, this behavior does not agree with Oliveira et al. (2016), that urinary volume is positively correlated with CP content of the diet. In this study, even with the highest CP concentration in the fertilized, did not affect urine volume by heifers, which was only slightly higher to others.

The Efficiency of Nitrogen Utilization

Since no differences were found in DMI under different conditions, the N ingested by heifers was also similar in treatments.

The absence of differences in N excreted in feces by heifers among treatments suggests that N retention was similar in different experimental conditions, agreeing with Atkinson et al. (2010), that retention of N is, for the most part, due to reduced excretion of N. The N excreted through urine depends on the supply of energy, in order to promote maximum use of CP ingested by the animal (Detmann et al., 2014).

The parameter ENU by ruminants indicates how much N available is being used by rumen microorganisms (Detmann et al., 2014). Since there was no difference in excreted and ingested N, ENU was similar among the treatments. The results infer that all conditions provided adequate amounts of available N and digestible OM for the production of microbial protein (Chizzotti et al., 2015). Therefore, there was a balance between ingested and excreted N, resulting in similar N balance (Da Silva et al., 2016).

According to Detmann et al. (2014), ENU by cattle on tropical pastures is very low, approximately 11.6%. In this study, ENU was considerably higher, and this is due to the pasture management in order to have a high proportion of leaves and consequently an adequate supplement of digestible CP and OM. According to Dijkstra et al. (2011), the overall average ENU is around 25%, with a range between 15% and 40% for ruminants in general.

Urea is the main form of elimination of nitrogenous compounds from mammals, especially ruminants, excreting through urine and correlates with N concentrations in blood plasma and N intake (Van Soest, 1994), that is, the levels of urea and CP of the diet. In this context, several studies described in the literature reported an increase in urea excretion as a function of diet CP concentration, suggesting that the reabsorption of this component is not constant and there is greater retention of urea under low diet CP intakes, and higher excretion under high intakes (Rennó et al., 2008). However, in this study there was no influence of treatment on urea excretion through urine, suggesting that there are no differences between the losses of N and, consequently protein, for heifers kept in palisade grass pastures fertilized, not fertilized or mixed with legumes.

In addition to DMI, amino acids, especially peptides, stimulate the production of microorganisms that develop in rapidly degradable energy sources, the reverse occurs when the energy substrates are slowly fermented (NRC, 2001). The absence of difference in urea excretion was similar to that found by Kropp et al. (1977), who reported that Nmic production was relatively constant regardless of the urea level addition in the diet of steers, which in this study was evidenced by the increase in CP concentration in fertilized.

Mass and Herbage Allowance

According to Boddey et al. (2004), N is the nutrient that most limits pastures productivity in the extensively managed pastures in Brazil. At the same time N increases HM accumulation by increasing the tillering rate, emission of new leaves, and favoring plant growth. However, the results showed that the N supply provided by both N fertilization and BNF was not enough to promote a significant difference in HM when compared to the control (Table 4). It is important to mention that the average temperature during the months when samples were taken in the 2 experimental years was 24.5 °C, while the average rainfall was 77.8 mm.

Grazing efficiency was close to the 50% goal because the average grazing height was 28 cm. The absence of effects on herbage allowance may be associated with the mob stocking technique management adopted in the research (Allen et al., 2011), in order to maintain the same stocking rate and DMI in the 3 conditions. However, a difference was observed in the leaf allowance (Table 4).

With regard the higher supply of leaves presented in the fertilized, it can be justified by the fact that the N from the fertilization provided greater N absorption and nutritional conditions favorable to the emergence and development of tillers. According to Casagrande et al. (2011), the greater supply of leaves is important in grazing systems because it better reflects the quality of herbage available to the animals. Moreover, there is a higher nutritive value and preference of the animals for this morphological component in relation to the stems.

Chemical Composition of Herbage

The CP concentration found in the mixed suggests that the supply of N provided by BNF was not enough to increase the emission of new leaves compared to the effect of N fertilization, contributing to lower CP concentrations. In addition, the CP values obtained in the palisade grass are within the range estimated for tropical herbages, between 5.2% and 12.8%, according to the meta-analysis published by Detmann et al. (2014), but below the mean (15% CP) observed at an experimental site close the study area in the last 16 yr (Ruggieri et al., 2019). It is important to highlight that the pasture management has a strong effect on CP and N fractions in tropical grasses.

The reduction of NDF concentration as a function of N fertilization corroborates with Benett et al. (2008) and Dupas et al. (2010) studies, who reported a linear reduction of NDF with the increase of applied N dose. The mean NDF concentration found in mixed is higher than 50%, verified by Carulla et al. (1991) in a mixed of B. dictyoneura with A. pintoi. However, if NDF concentration is lower than the 60% Poppi and Vega (1997) suggest that it can limit the herbage intake by grazing animals. McRoberts et al. (2018) explained that N fertilization can reduce plant NDF contents by stimulating the growth of new tissues, the DM of which contains less structural carbohydrates (hemicellulose, cellulose) and lignin.

The absence of fertilization and introduction of legumes influence on ADF concentrations corroborates with Rogers et al. (1996), who found that the application of N promotes minimal effects on the ADF concentrations, and with the statement of Karadavut et al. (2017) that the use of N fertilizers does not always cause changes in the fibrous fractions of plants.

Energy concentration of herbage was higher in the fertilized pastures, due to the higher CP and EE concentrations in this treatment. The results of TDN differ from Benett et al. (2008), who verified a considerable increase of TDN concentration with the increase of N doses, which mean was 567.2 g/kg DM. However, the averages presented in the conditions of this study coincide with that stipulated by Van Soest (1994), considering approximately 55% of TDN for tropical herbage plants, being able to be altered according to the climatic conditions, soil, development stage, and plants management.

Conclusions

Palisade grass pastures fertilized with 150 kg N/ha/yr or mixed with forage peanut did not reduce enteric CH₄ emission by crossbred dairy heifers, even with herbage composition changes. Dairy heifers raised in palisade grass pastures fertilized present higher UA excretion in the urine and lower ALLA:PD. However, there were no differences in ENU and efficiency of Nmic. Nitrogen fertilization increased the nutritive value of herbage, by raising the CP and reducing the NDF concentrations, and also provided a greater leaf allowance. Palisade grass pastures fertilized with 150 kg N/ha/yr, mixed with forage peanut or without N fertilization managed at the height equivalent to 95% of LI and at a grazing efficiency of 50%, with the stocking rate adjusted based on the herbage allowance, provides adequate conditions to not limit the animal intake and to avoid the degradation of the pastures.