Energy supplements for beef heifers on cool season pastures - a database analysis

ABSTRACT: Pooled data analysis is an analytical method that combines results from multiple studies. This technique provides a more robust estimate of the effects of an investigation. We performed a database analysis from seventeen experiments developed at Federal University of Santa Maria, Rio Grande do Sul state, Brazil, between 1999 and 2017 to characterize individual performance per area and stocking rate with or without supplementation of replacement heifers grazing winter pastures. Data were separated into two groups: with and without energy supplement provision, and into five subgroups based on supplement levels. Heifers from both groups were maintained under similar forage biomass and leaf blade allowance. Statistical analyses were run on R software using a ‘meta’ package. Supplement supply increased average daily gain and gain of body condition scores by 11.1% and 20.0%, respectively. Supplement levels higher than 1.2% of body weight resulted in higher weight gain per area, with the stocking rate increasing with higher supplement levels.


INTRODUCTION
Breeding systems require selection and contention of beef heifers to maintain herd size and productivity, ensuring livestock sustainability (HENLEY et al., 2021).In Southern Brazil, grasslands are the main forage source for herd; however, their production decreases during the winter season (MEZZALIRA et al., 2012), affecting the energy and protein intake essential to meet animal requirements (BERETTA et al., 2000, TITTONELL et al., 2016).During feed scarcity season, improving animal energy balance is possible by establishing cultivated grasses adapted to climate conditions and tolerant to grazing that can extend grazing period, increasing nutritional forage value and decreasing seasonal variations (VENDRAMINI et al., 2006, VENDRAMINI & MORIEL, 2020).Furthermore, the use of supplementation can enhance the growth and reproductive performance of heifers (MARTIN et al., 2007), making it a useful tool to achieve ideal body weight and improve reproductive results (MULLINIKS et al., 2013).
Cool-season pastures are used to reduce negative effects of low temperatures on forage quality and productivity by establishing high nutritional Ciência Rural, v.53, n.8, 2023.Martini et al. pastures with frost-tolerant species (SALGADO et al., 2013).However, these pastures can limit animal performance through their heterogeneity along the productive cycle (PARIS et al., 2012).In this sense, supplementation can be used as a strategy to intensify rearing of heifers by providing nutrients that are not available in pastures, extending grazing season, and thus optimizing forage use, reducing reproductive cycles, and increasing animal performance (DIXON & STOCKDALE, 1999;BARBERO et al., 2015).Infield grazing experiments are of primary importance to assess the effects of supplementation on animal performance on winter pastures, but they are expensive and time-consuming.One alternative approach is the use of meta-analysis combining results from different related studies and estimating the effects of treatments with higher precision, consequently providing useful information for future livestock practices with lower costs and higher financial incomes (LOVATTO et al., 2007, RODRIGUES & ZIEGELMANN, 2010).
In this context, our objective was to examine the effects of supplementation on the performance of beef heifers and pasture-supporting capacity in winter pastures from the Southern Brazil region.We hypothesized that the use of supplements will increase: (1) average daily gain, (2) stocking rate, and consequently (3) average gain per area.The intent was to develop the first suite of information that would be useful for prediction of supplementation benefits on beef heifers' performance and pasture productivity for this region.

Experiments description and dataset construction
Datasets were constructed based on results from 17 experiments developed at Laboratório Pastos & Suplementos (Departamento de Zootecnia, Universidade Federal de Santa Maria) from 1999 to 2017 (Table 1).These experiments assessed 589 beef heifers (Angus breed and Charolais/ Brahman crossbreed), with initial age and corporal weight average of 8 months and 160.9 ± 22.6 kg, respectively.Established pastures were ryegrass (Lolium multiflorum Lam.) by itself or in a mixed consortium with black oat (Avena strigosa Schreb.),arrowleaf clover (Trifolium vesiculosum Savi), or red clover (Trifolium pratense L.).The average winter pasture use was 106 days, from May to September, and grazing management was continuous or rotational  stocking rate, with variable put-and-take animals to maintain forage mass and height of canopy according to experimental criteria.Two or three repetitions were done per area with three tester animals.The average of supplement provision was 0.8% (ranging from 0.15 to 1.5) of liveweight being offered daily at 0200 pm.Data were compiled on Microsoft ® Office Excel ® 2013 and separated into two groups with and without supplement, and into five subgroups according to daily quantity of offered supplement (Table 2).The average and standard deviation of variables were obtained from the raw data from each experiment.

Forage and animal variables
The selected variables from pasture attributes were: forage biomass (FB, kg DM ha-1), forage accumulation rate (FAR, kg DM ha-1 day-1), forage allowance (FA, kg DM per 100 kg BW), leaf blade allowance (LBA, kg DM per 100 kg BW), and canopy height (H, cm).Additionally, we included the following forage variables obtained by grazing simulation: crude protein (CP, %), neutral detergent fiber (NDF, %), and organic matter digestibility (OMD, %).Variables related to animal performance were: average daily gain (ADG, kg BW day-1), gain of body condition score (BCS), final body weight (FBW, kg), stocking rate (SR, kg BW ha-1), supplement conversion to body weight (SC, kg ha-1), and gain per area (GPA, kg BW ha-1 day-1).The GPA was obtained by average of SR divided by beef heifers' weight, multiplied by average daily gain of tester animals.The SC was obtained from supplement intake per hectare divided by the GPA difference between animals that received and did not receive supplements.

Statistical analyses
All statistical analyses for the metaanalysis were performed using R (R CORE TEAM, 2018) and the 'metacont' function within the package 'meta' (SCHWARZER, 2007), which produces both fixed-and random-effects estimates with continuous outcome data.The standardized mean difference (SMD) was used to obtain mean differences across groups, and selected experiment results were pooled using inverse variance weighting.The effect size on variable measure unit was obtained by multiplying the average of the standard deviation from animals that received supplement by the analysis standardized mean difference.The choice of model (fixed-effect or random-effect) was based on heterogeneity by I2 test (HIGGINS et al., 2003), which quantifies the impact of heterogeneity on meta-analysis through mathematical criteria independent of number of studies and treatment metric effect.Variable stocking rate (SR) was modelled as a function of supplement levels using the 'metareg' function from the 'meta' package, and variance estimates between studies were done using the restricted maximum likelihood (REML) method.Supplement conversion (SC) was analyzed by regression analysis according to supplement levels and its model was chosen based on coefficients (linear, quadratic, and cubic) significance using Student's t-test with α = 0.05 as the probability limit for rejection of null hypothesis.These analyses were made using the 'lm' function and were plotted using the 'ggplot2' package (WICKHAM, 2016).

RESULTS AND DISCUSSION
Based on heterogeneity analysis, the fixedeffect model was chosen for the FB, FAR, FA, LBA, H, CP, NDF, and OMD variables, while the randomeffect model was used for the ADG, BCS, FBW, GPA, and SR variables.
Variables ADG, BCS, and FBW did not show differences among subgroups; however, the  use of supplement, independent of level, increased individual performance (Table 3).Heifers that received supplements had an ADG of 1.0 ± 0.2 kg day -1 , which was 11.1% higher than that of heifers that did not receive supplements.Heifers that received supplements had ADG, BCS, and FBW 11.1%, 20%, and 5.3%, higher than heifers fed only with cool-season pastures, respectively (Table 3).
The FB correlates directly with available forage to animals, being considered one of the most relevant and utilized factors in grazing management (CONFORTIN et al., 2013).The average of forage biomass in our study was within the intended values for analyzed experiments (Table 3).According to ROMAN et al. (2007), the ideal values of FB for maximum animal performance in temperate climate zones range from 1100 to 1800 kg ha -1 of DM.For ryegrass pastures, canopy height should be maintained from 10 to 15 cm to optimize biomass fluxes and provide conditions for pasture growth that will allow animals to have higher forage intake and better performance results (PONTES et al., 2004).In this sense, when heifers were fed solely with cool-season pastures, FA was 11.5% higher than for supplemented heifers; however, despite this difference, both had higher values (3.4 and 3.1, respectively, for nonand supplemented heifers) than the 3% estimated by the National Research Council (NRC, 2000).BARGO et al. (2003) suggested that appropriate values for animals fed solely with pastures range from 3 up to 5 times more than that estimated for dry matter intake, and 2.5 times higher when animals receive supplements.Furthermore, FA values from experiments analyzed in the present study were within the range indicated by GRAMINHO et al. (2019), from 6 to 12 kg of DM per 100 kg of BW for ryegrass management, without jeopardizing foliar tissue fluxes and efficiency of pasture use.The sum of the average supplement intake and forage allowance totaled 10 kg of DM per 100 kg of BW, similar to the forage allowance of heifers maintained only on pastures.Additionally, it was not inferior to NRC (2000) estimates and probably was not a limiting factor of forage intake.
Chemical composition and digestibility are the main factors that influence pasture quality (SOLLENBERGER & CHERNEY, 1995).Beef heifers require 13.5% of crude protein for high animal performance (NRC, 2000); our study successfully exceeded this value.Content of NDF has an inverse relation with forage intake, with values ranging from 55 to 60% not limiting intake, according to VAN SOEST (1994).However, in our study, values were below this range and were thus considered as intake limiters.The average OMD was within the 65-70% range indicated by POPPI et al. (1994) for high digestibility diets and in these cases, voluntary intake is restricted by metabolic mechanisms, such as animal capacity to use absorbed nutrients.According to DIXON & STOCKDALE (1999), digestibility has a linear correlation with NDF, being higher in forages that have lower NDF and higher protein content.Therefore, our results for CP, NDF, and OMD characterized cool-season pastures as having high nutritional quality for heifers.
The highest ADG observed could be explained by the supplement additive effect, which increased dry matter intake and, consequently, provided higher amounts of energy to animals.Furthermore, this result can be linked with diet equilibrium provided by supplement use, which is a degradable carbohydrate source for rumen that optimizes volatile fatty acids and propionic acid production, hence increasing glycose availability for muscular, uterine, and fatty tissues storage (NOVIANDI et al., 2014).Supplementation of heifers on ryegrass pastures increases ADG and anticipates reproductive system development of 13-month-old heifers (GONZALEZ et al., 2016).BCS of higher supplemented animals was determined by weight gain composition by the end of the grazing period.Animals that fed on pastures with high protein content and received energy supplements tended to accumulate more fat faster than animals maintained without supplements.High protein and energy relations in consumed nutrients have potential to alter animal BCS (POPPI & MCLENNAN, 1995).In a study of the development of beef heifers, SILVA et al. (2018) showed that the main factors altering conception rate of 14-months-old heifers were BCS at the beginning and the end of the reproductive season.Furthermore, weight gain intensification was necessary to increase the nutrient levels of the diets of animals, aiming to reach a BSC of 4.0 ± 0.1, which had a higher conception rate.Heifers that received supplementation reached 59.2% of the 450 kg of mature weight.According to LARDNER et al. (2014), heifers did not have their reproductive performance affected and became more productive by reaching 55% of BW for their first mating when compared to heifers raised to reach 62% of BW, which are nutritionally more demanding, thus increasing financial investment.This weight change is linked to genetic modifications that aim to decrease the age of heifers' puberty (FUNSTON et al., 2012).
In accordance with our results, PÖTTER et al. (2010a) reported similar relations between SR and supplement levels that were due to the effect

Figure 1 -
Figure 1 -Bubble chart graphics: Y-axis represents the effect size using the standardized mean difference (SMD) method from meta-regression analysis.Size of dots represents analysis participation of each study.

Figure 3 -
Figure 3 -Dashed vertical line represents standardized mean difference of supplemented animals.The size of squares represents participation weight of each study on analysis and horizontal line indicates standard deviation of studies.

Table 3 -
Additional values in average daily gain (ADG, kg day -1 ), gain of body condition score (BSC, scores), and final body weight (FBW, kg) of heifers managed on pastures receiving supplements from Pastos & Suplementos Laboratory database.Mean difference between heifers supplemented or not in variable unit of measure; * % heterogeneity between experiments measured by I2 statistic; ** probability for statistic difference between groups calculated by inverse variance weighting; *** probability for statistic difference between subgroups calculated by inverse variance weighting.