Explore
Communities in English
Advertise on Engormix

The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression

Published: April 3, 2023
By: Diego A. Martinez, Jordan T. Weil, Nawin Suesuttajit, Cole Umberson, Abdullah Scott and Craig N. Coon / Department of Poultry Science, University of Arkansas, Fayetteville, AR 72701, USA.
1. Introduction
Modern broilers show increasing growth rates, improved feed conversions, and increased breast meat size and yield, which play a pivotal role in the economics and sustainability of poultry production. Indeed, across the years, broilers have become leaner [1], and the industry aims to increase breast meat production due to its higher market price.
To achieve this aim, operation professionals need to be able to determine, ahead of time, what the final results will be, particularly the body weight (BW) and the breast yield of the flocks. However, appropriate indicators are needed, and it is necessary to more accurately determine how likely it is for the performance to correlate to processing traits.
Similarly, a correct cost-benefit balance between lowering the feed conversion ratio (FCR) and increasing breast meat production is considered by industry nutritionists to optimize the feeding cost and increase the market value of the broilers, respectively. To do so, dealing with the relationship between performance and processing yield traits is crucial.
Optimum production requires the best possible diet formulation strategies, so the industry will eventually move towards a productive energy system to more accurately take into account the effective use of nutrients to produce meat [2]. The Arkansas Net Energy system [3] has been developed by taking into account the body energy gain (net energy for gain) based on the protein and fat gains (determined by Dual Energy X-Ray Absorptiometry) multiplied by their corresponding caloric coefficients, and the fasting heat production (net energy for maintenance; determined with indirect calorimetry). This system has been shown to be well correlated with the performance of broilers [4], and accurate equations have been validated to predict the fasting heat production based on the body composition [5], thus reducing the dependence on respiratory chambers. However, a better understanding of the relationship between FCR, the dynamics of body composition, and breast meat production will support future modeling approaches to predict body composition and ease the determination of body energy gain. A better understanding of broiler performance and breast meat production will also support the development of strategies to attenuate meat quality issues [6].
Consequently, better quantifying the association between performance, body composition, and processing yield will support production professionals, nutritionists, and researchers to make decisions and better understand the underlying processes involved in productive and economic efficiency. Most studies on the relationships between these variables have focused on comparing genetic lines [7] or determining individual variations within the same genetic line [8,9]. Indeed, negative phenotypic and genotypic correlations have been reported between FCR and body weight gain (BWG) when the intra-genetic line variability was explored in individual birds [10,11]. Furthermore, it has been reported that a positive linear relationship exists between percent increases in the body protein content (g) and in the breast meat or thigh weights (g) as the bird grows [12]. However, as can be ascertained from the existing literature, no study has determined how the changes in one of these response variables as the result of applying multiple interventions correlate with changes in another. Modeling these relationships by applying a meta-regression multifactorial approach including studies that assessed diverse factors, may provide a broader point of view, clearer insight, and robust inferences. The present study aims to summarize broiler studies reported in the literature whose response variables included performance, body composition, and processing yield, regardless of the factor studied, to model the relationships between specific variables through meta-regression.
2. Materials and Methods
The meta-analysis was conducted and reported following the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) 2020 guidelines [13].
2.1. General Approach
A multifactorial approach [14,15] was applied to favor generalization, so the data used to build the models was not limited to a single factor tested but to any whose effects were measured in productive performance, body composition gain, and processing yield.
2.2. Eligibility Criteria
The scientific papers used fitted the following inclusion criteria: An study was included if it (1) was an actual experiment, (2) reported results in the three groups of variables (performance, body composition, processing yield) as the result of testing any experimental treatment using a single broiler genetic line, (3) the sample analyzed (i.e., to determine body composition) was the whole body (including feathers), and (4) reported a variability index per response variable.
2.3. Information Sources, Search Strategy, and Selection Process
Scientific papers were searched in Scopus [16] and Google Scholar [17] to identify papers published until December 2021 considering the following keywords in English: “{breast yield} AND broilers AND {body protein} OR {protein deposition} OR {protein gain}”. The following keywords in Portuguese were also used: “peito AND frango AND {proteína corporal} OR {deposição protéica} OR {ganho de proteína}”. Only articles in English and Portuguese were included. The documents gathered were examined. Those not related to the focus of the present review, thesis or dissertations, documents that were not papers published in peer-reviewed scientific journals, and duplicates were removed. One reviewer (D.A.M.) selected the papers by inspecting the titles, abstracts, and the whole body of the manuscripts and considering the criteria mentioned earlier. The reduced set of papers was approved by a second reviewer (C.N.C.).
2.4. Data Collection Process, Data Items, and Effect Measures
A database was built where each row consisted of a set of independent data and included all the data corresponding to a comparison of two treatments from the same experiment at the same age, obtaining as many datasets as non-redundant comparisons were observed. Data from different experimental periods within a paper were considered completely independent and conformed separate rows in the database [14]. Every row included identifying information about the experiment (reference, journal, year) and the treatments compared, reported treatment values in each response variable, number of replications, and variability values. Column sections included BWG (g/bird/d), FCR (accounted as one point for every g/g value [18]), body protein gain ratio (PGR; g/bird/d), body fat gain ratio (FGR; g/bird/d), protein-to-fat gain ratio (PFG), carcass yield (CAR; %), breast- (Pectoralis major + P. minor) to-legs (leg quarters) ratio (BLR), and market value (MKV; USD/bird; defined as the value of the whole breast meat plus the leg quarters value). One reviewer (D.A.M.) collected the data from each paper. For every response variable in each data row, the effect size was considered the difference between the means of the two treatments being compared [19]. Possible outlying values were not assessed or removed.
2.5. Synthesis Methods
Two synthesis stages applied in the study consisted of (A) grouping the effect sizes by response variable to determine their heterogeneity and (B) including the percent (%) changes of other response variables considered as moderators to determine the relationship between the changes in the predictor (i.e., moderator) and response variables. All studies providing data were considered eligible for each synthesis.
Some response variables were calculated based on the data reported in each study. The following calculations were applied:
PFG = PGR/FGR
where PFG, PGR, and FGR were all expressed in g/bird/d,
BLR = B/L,
where BLR was the breast-to-legs ratio, and B and L were the breast and leg quarters, respectively (weight or yield; both in the same unit), and
MKV = (BRE × BREP + LQU × LQUP) × BW
were BRE was the breast meat yield (%), LQU was the leg quarters yield (%), BREP and LQUP were the market prices ($/g) of the whole deboned breast (without wings) and leg quarters, respectively [20], and BW was expressed in g. If PGR was not reported but nitrogen retention, then the PGR was calculated as follows,
PGR = NR × 6.25,
where PGR was expressed in g/bird/g and NR was the nitrogen retention (g/bird/d). If FGR was not reported, but energy gain was available, FGR was calculated based on the caloric coefficients of body protein and fat by rearranging an equation previously reported [21] to clear the body fat content, as follows:
BF = (BE − 22.75 − (5.45 × BP))/8.95,
where, BF was the body fat (g/bird), BE was the body energy (kcal/bird), and BP was the body protein (g/bird). Then, the fat gain ratio was defined as:
FGR = BFf − BFi.
where FGR was the fat gain ratio (g/bird in a period), and BFf and BFi the final and initial body fat (g/bird) contents in the period, respectively. Then, Equation (5) was inserted into Equation (6), replacing both BFf and BFi , obtaining the following equation:
FGR = (EGR − 5.45 × PGR)/8.95,
where EGR was the energy gain ratio (kcal/bird/d), and both FGR and PGR were expressed in g/bird/d. If not reported, the standard deviation (SD) was calculated as follows:
SD = SEM × √ n
where SD was the standard deviation, SEM was the standard error of the mean, and n was the number of replications [22]. The SD of variables not reported in the primary studies (e.g., PFG and MKV) was calculated assuming that their coefficient of variation (CV; %) was the average of the CV values of those variables involved in their calculation and that:
SD = (CV × MV)/100,
where SD was the standard deviation, CV was the coefficient of variation (%), and MV was the mean value of the response variable.
The grouping of the effect sizes per response variable was visually displayed as forest plots and their relationship with the changes in moderator variables as bubble plots (datapoints drawn as circles with sizes proportional to their precision).
2.6. Statistical Analyses
2.6.1. Data Usefulness Determination: Basic Meta-Analyses (Random Models)
This study aimed to model changes in the effect sizes based on changes in moderator variables. Considering that the effect sizes resulted from applying multiple factors tested in the primary studies, overall effect sizes were reported only to inform the weighted mean effect size. The data was considered useful if heterogeneity, as an indicator of data diversity [23], was detected in the effect sizes across primary studies (Cochran Q test; p < 0.01 [24]). The heterogeneity was determined with the following random-effects model:
Yi = U + Ui + Ei
where yi was the observed effect size in the i-th data row (i: 1,. . . k) and also
Yi = 0iei
where µ was the mean true effect, θi was the unknown true effect in the i-th data row, ei was the intra-experimental sampling error in the i-th data row, ui was the inter-experimental deviation regarding the overall effect size in the i-th data row, k was the number of data rows, being that ui ∼ N(0,τ2) and ei ∼ N(0, vi), and both independently, where N denoted the normal distribution of the random inter-experimental deviation (u) and the intra-experimental sampling error (e), τ2 was the variability among the true effects in the different data rows, so that τ was the standard deviation of the true effect sizes across rows (pure heterogeneity); and vi was the approximately known sampling variance of the estimated effect size in the i-th data row. A weighted least squares method was applied where the weighing criterion was the inverse of the variance [25], as follows
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 1
where θw was the true weighted mean effect size; wi was the weighing factor applied for the i-th data row, θi was the true effect size in the i-th data row. Then, θw was the weighted average of the true effects (θi ) in the set of k data rows, with weights equal to the inverse of the corresponding variances, as follows
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 2
The heterogeneity was estimated with the REML (Restricted Maximum Likelihood) method [26] and evaluated with the Chi-square test (heterogeneous if p < 0.05 [27]). The normal distribution of model residuals was evaluated both with the Shapiro-Wilk test (non-normal if p < 0.05 [28]) and the confidence to rely on the Central Limit Theorem based on the sample size [29]. The statistic I 2 , defined as the portion (%) of the total variability attributed to pure heterogeneity, was also determined [14].
2.6.2. Modeling the Relationships between Variables: Meta-Regressions (Mixed Models)
Mixed-effects models were fitted to the data, considering the effect size was also a function of the level of a moderator variable (a co-variate), as follows
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 3
where yi was the observed effect size in the i-th data row (i: 1,. . . k), and also
Yi = 0iei
where θi was the unknown true effect in the i-th data row, ei was the intra-experimental sampling error in the i-th data row, β0 was the intercept, βj was the slope of the j-th moderator variable (j: 1,. . . p), xij was the value of the j-th moderator variable in the i-th data row, ui was the inter-experimental random deviation regarding the overall effect size in the i-th data row, ei was the intra-experimental sampling error in the i-th data row, and k was the number of data rows. Furthermore, in this case ui ∼ N(0, τ2), ei ∼ N(0, vi), and the mentioned weighted least squares method was applied. The estimate of the coefficient of the moderator variable (β1) was tested with the Chi-square test. The normal distribution of residuals was tested as mentioned earlier. When residuals were approximately normal, but the p-value was < 0.05, the data were re-analyzed to determine the probability associated with the slope of the moderator (β1) but applying a permutation test with 10,000 iterations [30]. The PR2 statistic (pseudo-R2 [31]), defined as the amount of variability in the effect sizes among data lines explained by the model, was calculated as follows [32]:
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 4
where PR2 was the pseudo-R 2 statistic, τˆ 2 RE was the total heterogeneity estimated through the random-effects model, and τˆ 2 ME was the residual heterogeneity estimated throuthe mixed-effects model. The 95% confidence intervals (95%CI) of moderator estimates were calculated.
All relationships were assessed considering simple linear meta-regressions and the models fully reported if the slopes were significant (p < 0.05) [27]. Complementarily, the type of genetic line (fast- or slow-growing birds) and an interaction term were also included as moderators for the sole purpose of conducting a parallelism test [33] and determining if the relationships between BWG and FCR and between FCR and PFG were equivalent in both types of birds. The relationships (slopes) were considered parallel (i.e., equivalent) if the interaction terms showed p values > 0.05 [33].
2.6.3. Bias Assessment
The risk of bias due to missing results was assessed combining the application of [30]: (1) Funnel plot-associated Trim-and-Fill method to identify a possible number of missing values and obtain an estimated set of them. (2) Egger’s test to determine the existence of bias (bias if p < 0.05) in the dataset. (3) Comparing the 95%CI of the overall effect sizes calculated with the original and adjusted datasets (including the Trim-and-Fill obtained values) to determine the existence of bias in the overall effect size. The overall effect size was considered biased if the following two criteria were met: (A) Estimates were statistically different (p < 0.05; both overall effect sizes fell outside the 95%CI of the counterpart [34]) and (B) the boundaries of the 95%CI had the same sign (positive or negative). Biases were considered influential if inferences were built based on the overall effect sizes.
2.7. Software
The data was analyzed using R programming language [35] with RStudio 1.1.456 interface [36]. The goodness-of-fit to the normal distribution was tested using the package stats [35], the meta-analysis procedures, including the Egger’s asymmetry test with Metafor 2.0-0 [25], and the graphs plotting with lattice 0.20-38 [35].
3. Results
3.1. Study Selection
Searches conducted provided 1087 records, from which 1032 were removed due to not pertaining to the focus of the study (899), through title inspection, not being scientific papers (112), or because of duplicates (21). Fifty-five papers were screened, from which the following were excluded: two papers presenting models but not treatment mean and variability values, one observational study, and 12 not reporting results in the three target groups of response variables. After reviewing in detail 40 papers, 26 were excluded because the chemical composition was determined for the carcass (12 papers) or meat (14 papers) and did not determine the chemical composition of the whole bird (Figure 1). The remaining 14 papers were included in the study (Table 1).
Table 1. Characteristics of the studies used to determine the relationships between performance, body composition gain, and processing yield in broilers.
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 5
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 6
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 7
Figure 1. PRISMA flowchart of the study search and selection process. * Not pertaining, documents not related to the focus of the study; Not papers, documents not actual papers published in scientific journals; Models, papers reporting mathematical models, but not mean values per treatment; Observational, observational studies; Carcass composition or Meat composition, chemical composition not determined in the whole bird but the carcass or meat, respectively.
3.2. Characteristics of Studies Included
A widely diverse database was built, including 78 data rows from the 14 scientific papers published between 1992 and 2019 in nine scientific journals. Fifteen rows included females and 57 male birds. Both fast-(two genetic lines; sampled between 21 and 49 days of age; 31 rows) and slow-(four genetic lines; sampled between 21 and 105 days of age; 47 rows) growing birds were included. The database included the following factors tested: Ambient temperature (9 rows), number of feeding phases (2 rows), feed restriction (1 row), source of dietary amino acids (intact protein or industrial amino acids; 6 rows), dietary methionine precursors (10 rows), organic trace minerals (8 rows), and dietary contents of energy (17 rows), protein (7 rows), and amino acids (18 rows).
The database included 3744 protein or N analyses to determine protein gain, 2336 analyses of fat or gross energy in bomb calorimeters to determine the fat gain, 1896 birds used to determine processing traits, and 1000 independent groups including a total of 22,256 birds to determine performance.
3.3. Data Usefulness
All assessed variables showed heterogeneity among rows (Table 2). The tau (τ) values found indicated that the treatments tested in the primary studies, whatever they were, produced effects for BWG, PGR, and FGR that typically varied from −8.4 to +8.4, −1.9 to +1.9, and −0.5 to +1.3 g/bird/d, respectively. The effects on PFG varied from −0.35 to +0.20, on FCR from −42 to +28 points, on CAR from −0.60 to +0.97 percent points, on BLR from −0.04 to +0.03, and on MKV from −0.14 to +0.01 $/bird. Consequently, the data in all tested variables were considered useful. Detailed effect sizes and their corresponding 95%CI are presented in Figures S1–S5. The BLR and MKV datasets and the BLR weighted mean effect sizes in the BLR showed bias (Table S1). These did not influence the data’s usefulness as the overall effect size was relevant only to report the weighted mean effect size, based on the general approach used in the study. Therefore, they were not considered influential.
Table 2. Heterogeneity in the effect sizes in each variable assessed in the study.
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 8
1 Expressed in the same units as the variable. 2 τ is the standard deviation of the true effect sizes across primary studies (pure heterogeneity) expressed in the units of the corresponding variable. I 2 is the portion (%) of the total variability attributed to pure heterogeneity among studies.
3.4. Relationships between Variables
Several relationships were explored between performance, body composition, and processing yield variables. The changes in BWG were related to the effect sizes in FCR (Figure 2; p < 0.001), and every percentage point increase in the BWG was associated with a reduction in FCR of 23.5 g less feed (95%CI: 19.4 to 27.5) per kg of body weight gain. This relationship was equivalent in slow- and fast-growing birds (parallel slopes; interaction p = 0.602; Table 3).
Table 3. Slow- and fast-growing genetic lines show equivalent relationships between the changes in growth rate and feed efficiency and between those in feed efficiency and body composition 1
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 9
1 FCR, feed conversion ratio; BWG, body weight gain; PFG, protein-to-fat gain ratio. PR2 , pseudo-R2 statistic. k, number of studies or data rows. * p values determined by a permutation test. p values > 0.05 denote a lack of interaction, parallel slopes, and equivalent relationships between primary predictors (change in BWG; change in FCR) and response variables (effect sizes in FCR and PFG).
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 10
Figure 2. Bubble plot of the relationship between growth rate and feed efficiency of broilers. The feed conversion ratio (FCR) changes are associated with changes in body weight gain (BWG). 95%CI = 95% confidence interval. PR2 = pseudo R2 . k = number of studies or data rows. Each circle is a primary study or data row, and its size is proportional to the relative weight given to calculate the regression, based on its precision. The continuous straight line is the prediction line, and the dashed curves are the boundaries of its 95% confidence interval.
The changes in BWG were not shown to be related to the effect sizes in PFG (Figure 3A; p < 0.170), but those changes in the FCR were (Figure 3B; p < 0.001), so the reduction in FCR was associated with reductions in the PFG. This relationship was consistent in slow- and fast-growing birds (parallel slopes; interaction p = 0.900; Table 3).
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 11
Figure 3. Bubble plot of the relationship between performance and body composition of broilers. The changes in body weight gain (A) are not associated with changes in body composition, but the changes in feed conversion ratio (B) are associated with. 95%CI = 95% confidence interval. PR2 = pseudo R 2 . k = number of studies or data rows. Each circle is a primary study or data row, and its size is proportional to the relative weight given to calculate the regression, based on its precision. The continuous straight lines are the prediction lines, and the dashed curves are the boundaries of their 95% confidence intervals.
The effect sizes in CAR showed no relationship with the changes in BWG, FCR, or PFG (Figure 4). In addition, the carcass yield showed no relationship with the carcass conformation, as measured by the BLR (Figure 5; p = 0.950).
The data showed that the changes in the body composition were highly related to those in the carcass conformation (i.e., breast-to-legs ratio) and the carcass market value (considering the breast and legs market prices; Figure 6). Every percent point increase in the PFG translates into a 0.001 increase in the BLR and represents 0.54 cents marginal value increase per carcass.
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 12
Figure 4. Bubble plot of the relationship between performance and body composition with carcass yield of broilers. The changes in the body weight gain (A), feed conversion ratio (B), or protein-to-fat gain ratio (C) show no association with the changes in carcass yield. PR2 = pseudo R2 . k = number of studies or data rows. Each circle is a primary study or data row, and its size is proportional to the relative weight given to calculate the regression, based on its precision. The continuous straight lines are the prediction lines, and the dashed curves are the boundaries of their 95% confidence intervals.
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 13
Figure 5. Bubble plot of the changes in carcass yield and those in carcass conformation, as measured by the breast-to-legs ratio, are not related. 95%CI = 95% confidence interval. PR2 = pseudo R2k = number of studies or data rows. Each circle is a primary study or data row, and its size is proportional to the relative weight given to calculate the regression, based on its precision. The continuous straight line is the prediction line, and the dashed curves are the 95% confidence interval boundaries.
The Relationship between Performance, Body Composition, and Processing Yield in Broilers: A Systematic Review and Meta-Regression - Image 14
Figure 6. Bubble plot of the changes in body composition are associated with the changes in carcass conformation (A) and market value (B) of the broilers. 95%CI = 95% confidence interval. PR2 = pseudo R . k = number of studies or data rows. Each circle is a primary study or data row, and its size is proportional to the relative weight given to calculate the regression, based on its precision. The continuous straight lines are the prediction lines, and the dashed curves are the boundaries of their 95% confidence intervals.
4. Discussion
The relationships between the changes in performance, body composition, and processing yield were evaluated. A database with 78 data rows was extracted from 14 primary studies testing multiple interventions (88% data rows related to nutrition). Broiler response was determined for the variables (performance, body composition, and processing yield) in the database.
In the present study, the heterogeneity of data was considered a formal verification of its usefulness to explore the influence of moderator variables on the effect sizes. Indeed, not only was the FCR (effect size SD = 35 FCR points), but most of the data used were highly diverse and heterogeneous (Table 2). The wide range of values in the tested variables increased the ability to build robust inferences from the models, so predictions are more likely to be interpolations. This can be considered a strength of the models in this study. Making statements outside the range of data used to develop the models (i.e., extrapolation) does not guarantee their accuracy and precision, reducing their validity [34].
The data demonstrates that regardless of the intervention applied to improve the performance, the higher the BWG, the better the feed efficiency (Figure 2). This finding demonstrates how important it is to ensure the broilers grow fast to support the sustainability of the poultry industry. The reduction of natural resources used (i.e., feed usage) as the FCR improves ultimately reduces the carbon footprint per kg of meat produced [2]. This consideration and the fact that both slow- and fast-growing broilers show the same relationship between BWG and FCR imply that producing slow-growing birds may not be sustainable. Furthermore, these operations need to ensure their birds grow as fast as possible to reduce their FCR and support their profitability. Indeed, a higher water footprint (L per kg of meat produced) for slow-growing birds can be inferred from the literature based on an equivalent water consumption per unit of weight gain [51] but a lower breast yield [52]. Finally, it may be contradictory to slow down the growth rate of the broilers to achieve a claimed better welfare (e.g., less skeletal integrity issues [53]) and then have to try to speed it up.
Interestingly, the data showed that the better the FCR, the lower the PFG (Figure 3), both in fast- and slow-growing birds (Table 3). This relationship does not imply that the better the FCR, the fatter the bird. It indicates that when the FCR improves as the result of a treatment, the fat gain (g/day) increases in a higher percentage than the protein gain (g/day). This relationship is consistent with the higher energy efficiency for fat gain versus protein gain [2]. The lower energy efficiency for protein gain has been associated with a higher heat production per unit of protein gain than fat gain in swine [54] and a higher fasting heat production per unit of actual body protein mass than fat mass in broilers due to protein turnover [5]. Furthermore, considering that the whole body fat content is lower than the protein content [29,55], an increased percentage of BWG from fat gain compared to protein gain is more likely to decrease the PFG.
The analyses showed that the effect sizes observed in the carcass yield are unrelated to the changes in performance, body composition (Figure 4), or even carcass conformation (Figure 5). The absence of a relationship between the carcass yield and the PFG in the study may be explained by the low variability in the carcass yield (effect size SD < 1 percent point) and the lower number of data rows available (k = 55; 30% lower sample size [56,57]. The carcass yield depends mainly on the relative weights of the carcass, viscera, and abdominal fat [53]. A higher PFG is expected to be associated with a lower abdominal fat content. It has been reported that the yield of the whole carcass is hardly affected by dietary factors [58]. However, the breast yield seems to be more sensitive to nutritional interventions [58] than the whole carcass yield. Indeed, it has been reported that the dietary supplementation of methionine produced effects (p < 0.05) that were more prominent in the abdominal fat yield (−13.8%) and the breast meat yield (+6.9%) than in the whole carcass yield (+0.8%) [59].
Finally, the observed association of the changes in body composition, as measured by the PFG, with the effect sizes in the carcass conformation agree with the higher protein content in the breast (64.6%) compared to the legs (50.8%) and the opposite relationship in the fat content (breast, 24.7%; legs, 31.7%) [37]. This condition explains the tight relationship between the body composition (PFG) and the marginal extra value of the carcass (MKV; Figure 6). As an example, a 40% change in the PFG (the difference between the mean positive and negative changes in the present study; +20% versus −20%) would represent a difference of $0.216 per bird, equivalent to $22 million of extra value in a broiler operation producing 100 million broilers/year. This finding supports that body composition may be used as a predictor of the economic value of broiler production.
Future research may explore the dynamics and relationships between the body protein and fat content and their ratio. The findings in the present study will support better modeling approaches to predict body energy retention, as calculated from the protein and fat gains and their corresponding caloric coefficients, and ultimately favor the determination of the net productive energy (Arkansas Net Energy). In addition, practical body composition measurements, such as the abdominal fat pad weight [60], may be explored as indicators of the final carcass conformation.
Limitations and Interpretation
Even though all the data utilized in the meta-regression analyses were from actual experiments, our findings do not establish formal cause-effect relationships but associations.
All data available fitting the required criteria were included in this study; however, the factors tested in the primary studies were mainly focused on nutrition. Furthermore, none of the studies included specific factors that may influence the relationships established herein, such as stress [61], gut microbiota changes [62], and coccidia, Clostridium perfringens, or feed ingredient-associated specific enteric challenges [18]. Therefore, the relationships established herein may change with new data.
One of the biggest challenges in mathematically modeling biological systems is that researchers need to balance three opposing criteria (precision, generality, and realism), where only two of them can be optimized at a time (the Levins’ Three Dimensions framework [63]). Usually, research in animal and poultry sciences is conducted under well-controlled conditions, favoring precision and realism at the expense of generality. Then, results differ when research conditions change (e.g., commercial conditions). In contrast, the general approach of the present study prioritized generalization and realism at the expense of precision [63]. The dataset included studies with extremely diverse characteristics, such as the year of publication (up to 30 years back), treatment factors tested, ages, genetics, dietary energy and amino acid contents, and both sexes. Even though these factors introduced variability in the data, some relationships were successfully characterized. Furthermore, those uncontrolled factors were random ones. Consequently, the relationships established in the present study are more generalizable than inferences from a single study under certain conditions. The influence of the varying conditions was intentionally not assessed to avoid misleading inferences based on insufficient statistical power. However, all effect sizes came from factors tested within the same genetic line; therefore, the generalizations in our study do not include relationships among genetic lines [64–66].
In addition, it should be considered that the actual performance, body composition, and carcass values were not used in the study but the changes that resulted from experimental treatments. Therefore, the influence of uncontrolled factors was minimized. This way, the differences in performance and body composition between sexes or across the years did not influence the calculations, but their relative sensitivity did. For example, the data showed that the relationship between the changes in body weight gain and feed conversion ratio is virtually the same in fast- and slow-growing birds, even though their genetics differ. Even though performance, lean mass, and breast meat yield have evolved with genetics across the years, it is unlikely that the relationship between the percent changes in body composition and the percent changes in processing yield as the result of an experimental treatment has done so.
Finally, the database available from the literature will increase if only two of the targeted areas are included (broiler performance, body composition, and processing yield) instead of including all three. If, for example, the same procedure is replicated but restricting the searches only to performance and body composition, more data will be found, and the influence of the genetics (i.e., year of publication), sex, and nutrient profile of the diets may be assessed
5. Conclusions
In conclusion, the results indicate that positive changes in body weight gain are associated with improvements in the feed conversion ratio and that the body composition predicts the carcass conformation and its market value. The results of the study on the association between body weight gain and the feed conversion ratio imply the importance of maximizing broiler weight gain to support the sustainability of broiler production.
     
This article was originally published in Animals 2022, 12, 2706. https://doi.org/10.3390/ ani12192706. This is an Open Access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

1. Cerrate, S.; Corzo, A. Lysine and energy trends in feeding modern commercial broilers. Int. J. Poult. Sci. 2019, 18, 28–38. [CrossRef]

2. Maharjan, P.; Martinez, D.A.; Weil, J.; Suesuttajit, N.; Umberson, C.; Mullenix, G.; Hilton, K.M.; Beitia, A.; Coon, C.N. Physiological growth trend of current meat broilers and dietary protein and energy management approaches for sustainable broiler production. Animal 2021, 15, 100284. [CrossRef] [PubMed]

3. Hilton, K.M. Development of Arkansas Net Energy Equation. Ph.D. Dissertation, University of Arkansas, Fayetteville, AR, USA, 2020. Available online: https://www.proquest.com/docview/2384839456?pq-origsite=gscholar&fromopenview=true (accessed on 8 July 2022).

4. Suesuttajit, N.; Weil, J.T.; Umberson, C.; Martinez, D.; Coon, C.N. A Comparison of apparent metabolizable energy, net energy, and productive energy (Ark NE) for 4-56d broiler performance studies. Anim.-Sci. Proc. 2022, 13, 316–318. [CrossRef]

5. Martinez, D.A.; Suesuttajit, N.; Hilton, K.; Weil, J.T.; Umberson, C.; Scott, A.; Coon, C.N. The fasting heat production of broilers is a function of their body composition. Anim.-Open Space 2022. accepted.

6. Maharjan, P.; Beitia, A.; Weil, J.; Suesuttajit, N.; Hilton, K.; Caldas, J.; Umberson, C.; Martinez, D.; Kong, B.; Owens, C.M.; et al. Woody breast myopathy broiler show age-dependent adaptive differential gene expression in Pectoralis major and altered in-vivo triglyceride kinetics in adipogenic tissues. Poult. Sci. 2021, 100, 101092. [CrossRef]

7. Nogueira, B.R.F.I.; Reis, M.P.; Carvalho, A.C.; Mendoza, E.A.C.; Oliveira, B.L.; Silva, V.A.; Bertechini, A.G. Performance, growth curves and carcass yield of four strains of broiler chicken. Braz. J. Poult. Sci. 2019, 21, 1–8. [CrossRef]

8. N’Dri, A.L.; Mignon-Grasteau, S.; Sellier, N.; Tixier-Boichard, M.; Beaumont, C. Genetic relationships between feed conversion ratio, growth curve and body composition in slow-growing chickens. Br. Poult. Sci. 2006, 47, 273–280. [CrossRef]

9. Wen, C.; Yan, W.; Zheng, J.; Ji, C.; Zhang, D.; Sun, C.; Yang, N. Feed efficiency measures and their relationships with production and meat quality traits in slower growing broilers. Poult. Sci. 2018, 97, 2356–2364. [CrossRef]

10. Yi, Z.; Li, X.; Luo, W.; Xu, Z.; Ji, C.; Zhang, Y.; Nie, Q.; Zhang, D.; Zhang, X. Feed conversion ratio, residual feed intake and cholecystokinin type A receptor gene polymorphisms are associated with feed intake and average daily gain in a Chinese local chicken population. J. Anim. Sci. Biotechnol. 2018, 9, 1–9. [CrossRef]

11. Chen, C.; Su, Z.; Li, Y.; Luan, P.; Wang, S.; Zhang, H.; Xiao, F.; Guo, H.; Cao, Z.; Li, H.; et al. Estimation of the genetic parameters of traits relevant to feed efficiency: Result from broiler lines divergent for high or low abdominal fat content. Poult. Sci. 2021, 100, 461–466. [CrossRef]

12. Danisman, R.; Gous, R.M. Effect of dietary protein on the allometric relationships between some carcass portions and body protein in three broiler strains. S. Afr. J. Anim. Sci. 2011, 41, 194–208. [CrossRef]

13. Page, M.J.; Moher, D.; Bossuyt, P.M.; Boutron, I.; Hoffmann, T.C.; Mulrow, C.D.; Shamseer, L.; Tetzlaff, J.M.; Akl, E.A.; Brennan, S.E.; et al. PRISMA 2020 explanation and elaboration: Updated guidance and exemplars for reporting systematic reviews. BMJ 2021, 372, n160. [CrossRef]

14. Martinez, D. Meta-Análisis de Estudios en Histomorfometría Intestinal, Energía Metabolizable y Comportamiento Productivo en Pollos de Carne. Ph.D. Dissertation, La Molina National Agrarian University (Universidad Nacional Agraria La Molina), Lima, Peru, 2021. Available online: http://repositorio.lamolina.edu.pe/handle/20.500.12996/5140 (accessed on 8 July 2022).

15. Martinez, D.; Ponce de Leon, C.; Vilchez, C. Quantitative relationships among intestinal histology, nutrient absorption and performance of broilers: A meta-regression approach. In Proceedings of the 2019 Poultry Science Association Annual Meeting, Montreal, QC, Canada, 15–18 July 2019; Volume 98, p. 84. Available online: https://poultryscience.org/files/galleries/2019-PSAAnnual_Meeting_Abstracts.pdf (accessed on 8 July 2022).

16. Scopus. Available online: https://www.scopus.com (accessed on 8 July 2022).

17. Google Scholar. Available online: https://scholar.google.com (accessed on 8 July 2022).

18. Martinez, D.A. Evaluacion de un producto a base de aceite esencial de oregano sobre la integridad intestinal, la capacidad de absorcion de nutrientes y el comportamiento productivo de pollos de carne. arXiv 2022, arXiv:2204.03728.

19. Gaibor, G.J.; Martinez, D.A.; Vilchez, C. Effect of selenium supplementation on productive performance in pigs: Meta-analysis. Rev. Investig. Vet. Peru 2020, 31, e17551. [CrossRef]

20. USDA (U.S. Department of Agriculture; Agricultural Marketing Service). Broiler Market News Report. Friday, 1 July 2022. Available online: https://mymarketnews.ams.usda.gov/filerepo/sites/default/files/2750/2022-07-01/606153/ams_2750_001 23_01.pdf (accessed on 8 July 2022)

21. Caldas, J.V.; Boonsinchai, N.; Wang, J.; England, J.A.; Coon, C.N. The dynamics of body composition and body energy content in broilers. Poult. Sci. 2019, 98, 866–877. [CrossRef]

22. Inca, J.S.; Martinez, D.A.; Vilchez, C. Validation of prediction equations of the egg characteristics in laying hens. Poult. Sci. J. 2022, 10, 71–82. [CrossRef]

23. Palmowski, A.; Buttgereit, T.; Palmowski, Y.; Nielsen, S.M.; Boers, M.; Christensen, R.; Buttgereit, F. Applicability of trials in rheumatoid arthritis and osteoarthritis: A systematic review and meta-analysis of trial populations showing adequate proportion of women, but underrepresentation of elderly people. Semin. Arthritis Rheum. 2019, 48, 983–989. [CrossRef]

24. Hedges, L.V.; Olkin, I. Parametric estimation of effect size from a series of experiments. In Statistical Methods for Meta-Analysis; Academic Press Inc.: Miami, FL, USA, 1985; pp. 107–145.

25. Viechtbauer, W. Conducting Meta-Analyses in R with the Metafor Package. J. Stat. Softw. 2010, 36, 1–48. [CrossRef]

26. Veroniki, A.A.; Jackson, D.; Viechtbauer, W.; Bender, R.; Bowden, J.; Knapp, G.; Kuss, O.; Higgins, J.P.T.; Langan, D.; Salanti, G. Methods to estimate the between-study variance and its uncertainty in meta-analysis. Res. Synth. Methods 2016, 7, 55–79. [CrossRef]

27. Bougouin, A.; Appuhamy, J.A.D.R.N.; Kebreab, E.; Dijkstra, J.; Kwakkel, R.P.; France, J. Effects of phytase supplementation on phosphorus retention in broilers and layers: A meta-analysis. Poult. Sci. 2014, 93, 1981–1992. [CrossRef]

28. Inca, J.S.; Martinez, D.A.; Vilchez, C. Phenotypic correlation between external and internal egg quality characteristics in 85-weekold laying hens. Int. J. Poult. Sci. 2020, 19, 346–355. [CrossRef]

29. Martinez, D.A.; Suesuttajit, N.; Weil, J.T.; Maharjan, P.; Beitia, A.; Hilton, K.; Umberson, C.; Scott, A.; Coon, C.N. Processing weights of chickens determined by Dual-Energy X-Ray Absorptiometry: 2. Developing prediction models. Anim.-Open Space 2022, 1, 100023. [CrossRef]

30. Martinez, D.A.; Ponce-de-Leon, C.L.; Vilchez, C. Meta-Analysis of commercial-scale trials as a means to improve decision-making processes in the poultry industry: A phytogenic feed additive case study. Int. J. Poult. Sci. 2020, 19, 513–523. [CrossRef]

31. Borenstein, M.; Hedges, L.V.; Higgins, J.P.T.; Rothstein, H.R. Meta-regression. In Introduction to Meta-Analysis; John Wiley & Sons, Ltd.: West Sussex, Shire, UK, 2009; pp. 187–203.

32. López-López, J.A.; Marín-Martínez, F.; Sánchez-Meca, J.; Van den Noortgate, W.; Viechtbauer, W. Estimation of the predictive power of the model in mixed-effects meta-regression: A simulation study. Br. J. Math. Stat. Psychol. 2014, 67, 30–48. [CrossRef

33. Martinez, D.A.; Weil, J.T.; Suesuttajit, N.; Beitia, A.; Maharjan, P.; Hilton, K.; Umberson, C.; Scott, A.; Coon, C.N. Processing weights of chickens determined by Dual-Energy X-Ray Absorptiometry: 3. Validation of prediction models. Anim.-Open Space 2022, 1, 100022. [CrossRef]

34. Ramsey, F.L.; Schafer, D.W. The Statistical Sleuth: A Course in Methods of Data Analysis, 3rd ed.; Brooks/Cole Cengage Learning: Boston, MA, USA, 2013.

35. R Core Team. R: A Language and Environment for Statistical Computing (Software); R Foundation for Statistical Computing: Vienna, Austria, 2018. Available online: http://www.R-project.org (accessed on 8 July 2022).

36. RStudio Team. RStudio: Integrated Development for R (Software); RStudio, Inc.: Boston, MA, USA, 2016. Available online: http://www.rstudio.com (accessed on 8 July 2022).

37. Filho, D.E.F.; Rosa, P.S.; Vieira, B.S.; Macari, M.; Furlan, R.L. Protein levels and environmental temperature effects on carcass characteristics, performance, and nitrogen excretion of broiler chickens from 7 to 21 days of age. Braz. J. Poult. Sci. 2005, 7, 247–253. [CrossRef]

38. Filho, D.E.F.; Rosa, P.S.; Figueiredo, D.F.; Dahlke, F.; Macari, M.; Furlan, R.L. Dietas de baixa proteína no desempenho de frangos criados em diferentes temperaturas. Pesqui. Agropecuária Bras. 2006, 41, 101–106. [CrossRef]

39. Hauschild, L.; Bueno, C.F.D.; Remus, A.; Gobi, J.P.; Isola, R.D.G.; Sakomura, N.K. Multiphase feeding program for broilers can replace traditional system. Sci. Agric. 2015, 72, 210–214. [CrossRef]

40. Liu, W.; Yuan, Y.; Sun, C.; Balasubramanian, B.; Zhao, Z.; An, L. Effects of dietary betaine on growth performance, digestive function, carcass traits, and meat quality in indigenous yellow-feathered broilers under long-term heat stress. Animals 2019, 9, 506. [CrossRef]

41. Mendonca, M.O.; Sakomura, N.K.; dos Santos, F.R.; Barbosa, N.A.A.; Fernandes, J.B.K.; Freitas, E.R. Metabolizable energy levels and energy:protein ratio for birds broilers of slow growth raised in free range system. Acta Sci. Anim. Sci. 2007, 29, 23–30. Available online: https://www.redalyc.org/articulo.oa?id=303126486004 (accessed on 8 July 2022).

42. Mendonca, M.O.; Sakomura, N.K.; dos Santos, F.R.; Freitas, E.R.; Fernandes, J.B.K.; Barbosa, N.A.A. Metabolizable energy levels for slow growth male broilers raised in semi confined system. Rev. Bras. Zootec. 2008, 37, 1433–1440. [CrossRef]

43. Nascimento, D.C.N.; Sakomura, N.K.; Siquira, J.C.; Pinheiro, S.R.F.; Fernandes, J.B.K.; Furlan, R.L. Digestible methionine + cystine requirements of ISA Label broilers reared in free-range system. Rev. Bras. Zootec. 2009, 38, 869–878. [CrossRef]

44. Nascimento, D.C.N.; Sakomura, N.K.; Siqueira, J.C.; Dourado, L.R.B.; Fernandes, J.B.K.; Malheiros, E.B. Digestible lysine requirements for ISA Label broilers reared in free-range system. Arq. Bras. Med. Veterinária Zootec. 2009, 61, 1128–1138. [CrossRef]

45. Oliveira, W.P.; Donzele, R.F.M.O.; Donzele, J.L.; Albino, L.F.T.; Antunes, M.V.L.; Campos, P.H.R.F.; Souza, M.F.; Pastore, S.M. Digestible lysine levels obtained by two methods of formulation of diets for 22-to-42-day-old broilers. Rev. Bras. Zootec. 2014, 43, 579–589. [CrossRef]

46. Perrault, N.; Leeson, S. Effect of environmental temperature, dietary energy, and feeding level on growth and carcass composition of male broiler chickens to 35 days of age. Can. J. Anim. Sci. 1992, 72, 695–702. [CrossRef]

47. Neto, M.A.T.; Takeara, P.; Toledo, A.L.; Kobashigawa, E.; Albuquerque, R.; Araujo, L.F. Níveis de lisina digestível para frangos de corte machos no período de 37 a 49 dias de idade. Rev. Bras. Zootec. 2009, 38, 508–514. [CrossRef]

48. Neto, M.A.T.; Kobashigawa, E.; Namazu, L.B.; Takearra, P.; Araujo, L.F.; Albuquerque, R. Lisina digestível e zinco orgânico para frangos de corte machos na fase de 22 a 42 dias de idade. Rev. Bras. Zootec. 2010, 39, 2460–2470. [CrossRef]

49. Xi, P.B.; Yi, G.F.; Lin, Y.C.; Zheng, C.T.; Jiang, Z.Y.; Vasquez-Anon, M.; Song, G.L.; Knight, C.D. Effect of methionine source and dietary crude protein level on growth performance, carcass traits and nutrient retention in Chinese color-feathered chicks. Asian-Australas. J. Anim. Sci. 2007, 20, 962–970. [CrossRef]

50. Yao, J.H.; Li, S.Q.; Zhong, L.L.; Huang, S.X.; Zhang, W.J.; Xi, H.B. The relative effectiveness of liquid methionine hydroxy analogue compared to dl-methionine in broilers. Asian-Australas. J. Anim. Sci. 2006, 19, 1026–1032. [CrossRef]

51. Deeb, N.; Cahaner, A. Genotype-by-environment interaction with broiler genotypes differing in growth rate. 3. Growth rate and water consumption of broiler progeny from weight-selected versus nonselected parents under normal and high ambient temperatures. Poult. Sci. 2002, 81, 293–301. [CrossRef]

52. Fanatico, A.C.; Pillai, P.B.; Cavitt, L.C.; Owens, C.M.; Emmert, J.L. Evaluation of slower-growing broiler genotypes grown with and without outdoor access: Growth performance and carcass yield. Poult. Sci. 2005, 84, 1321–1327. [CrossRef] [PubMed]

53. Uculmana, C.; Martinez-Patino, D.; Zea, O.; Vilchez, C. Effect of the calcium phosphorus ratio on bone characteristics, percent of ashes and skeletal integrity of broilers. Rev. Investig. Vet. Peru 2018, 29, 1268–1277. [CrossRef]

54. Reeds, P.J.; Fuller, M.F.; Cadenhead, A.; Lobley, G.E.; McDonald, J.D. Effects of changes in the intakes of protein and non-protein energy on whole-body protein turnover in growing pigs. Br. J. Nutr. 1981, 45, 539–546. [CrossRef] [PubMed]

55. Caldas, J.V.; Hilton, K.M.; Boonsinchai, N.; Mullenix, G.; England, J.A.; Coon, C.N. Maintenance energy requirements in modern broilers fed exogenous enzymes. Int. J. Poult. Sci. 2022, 21, 107–118. [CrossRef]

56. Martinez, D.A.; Suesuttajit, N.; Weil, J.T.; Maharjan, P.; Beitia, A.; Hilton, K.; Umberson, C.; Scott, A.; Coon, C.N. Processing weights of chickens determined by Dual-Energy X-Ray Absorptiometry: 1. Weight changes due to fasting, bleeding, and chilling. Anim.-Open Space 2022, 1, 100024. [CrossRef]

57. Quisirumbay-Gaibor, J.R.; Torres, R.; Yupanqui, J.; Martinez Patino-Patroni, D.; Vilchez Perales, C. Suplementación alimenticia de glutamina sobre el desempeño productivo en pollos de engorde. Siembra 2019, 6, 15–24. [CrossRef]

58. Ivey, F. Desenvolvimento e aplicação de modelos de crescimento para frangos de corte. In Proceedings of the I ACAV Interantional Simposium—Embrapa on Poultry Nutrition, Concordia, Santa Catarina, Brazil, 17–18 November 1999; pp. 22–35. Available online: https://www.infoteca.cnptia.embrapa.br/bitstream/doc/436985/1/documento56.pdf (accessed on 13 August 2022).

59. Schutte, J.B.; Pack, M. Sulfur amino acid requirement of broiler chicks from fourteen to thirty-eight days of age: Performance and carcass yield. Poult. Sci. 1995, 74, 480–487. [CrossRef]

60. Martinez, D.; Uculmana, C. Artichoke extract (Cynara scolymus L.): Experiences of use in animal production markets and opportunities for its production in Peru. Agroind. Sci. 2016, 6, 155–161. [CrossRef]

61. Lyte, J.M.; Shrestha, S.; Wagle, B.R.; Liyanage, R.; Martinez, D.A.; Donoghue, A.M.; Daniels, K.M.; Lyte, M. Serotonin modulates Campylobacter jejuni physiology and in vitro interaction with the gut epithelium. Poult. Sci. 2021, 100, 100944. [CrossRef]

62. Lyte, J.M.; Martinez, D.A.; Robinson, K.; Donoghue, A.M.; Daniels, K.M.; Lyte, M. A neurochemical biogeography of the broiler chicken intestinal tract. Poult. Sci. 2021, 101, 101671. [CrossRef]

63. Silverman, E. Methodological Investigations in Agent-Based Modelling; Springer International Publishing AG: Cham, Switzerland, 2018; pp. 61–81.

64. Maharjan, P.; Mullenix, G.; Hilton, K.; Beitia, A.; Weil, J.; Suesuttajit, N.; Martinez, D.; Umberson, C.; England, J.; Caldas, J.; et al. Effects of dietary amino acid levels and ambient temperature on mixed muscle protein turnover in Pectoralis major during finisher feeding period in two broiler lines. J. Anim. Physiol. Anim. Nutr. 2020, 104, 1351–1364. [CrossRef]

65. Maharjan, P.; Weil, J.; Beitia, A.; Suesuttajit, N.; Hilton, K.; Caldas, J.; Umberson, C.; Martinez, D.; Owens, C.M.; Coon, C. In vivo collagen and mixed muscle protein turnover in 2 meat-type broiler strains in relation to woody breast myopathy. Poult. Sci. 2020, 99, 5055–5064. [CrossRef] [PubMed]

66. Maharjan, P.; Hilton, K.M.; Mullenix, G.; Weil, J.; Beitia, A.; Suesuttajit, N.; Umberson, C.; Martinez, D.A.; Caldas, J.V.; Kalinowski, A.; et al. Effects of dietary energy levels on performance and carcass yield of 2 meat-type broiler lines housed in hot and cool ambient temperatures. Poult. Sci. 2021, 100, 100885. [CrossRef]

Related topics
Authors:
Diego Martinez Patiño-Patroni
University of Arkansas (USA)
Follow
Jordan Weil
University of Arkansas (USA)
Follow
Nawin Suesuttajit
University of Arkansas (USA)
Follow
Abdullah Scott
Follow
Craig Nelson Coon
University of Arkansas (USA)
Follow
Show more
Join to be able to comment.
Once you join Engormix, you will be able to participate in all content and forums.
* Required information
Would you like to discuss another topic? Create a new post to engage with experts in the community.
Create a post
Join Engormix and be part of the largest agribusiness social network in the world.
LoginRegister