Effects of Mortality and Growth Implant Use on Feedlot Net Returns and Greenhouse Gas Emissions

Abstract

This study uses established procedures to estimate the effects of changes in mortality and growth implant protocols on feedlot net returns (NRs). We then propose new methods for estimating concurrent impacts to feedlot greenhouse gas emissions intensity. Reducing mortality consistently increases NRs and reduces greenhouse gas emissions intensity in the feedlot regardless of sex or placement weight. Results indicate that use of two implants in the feedlot may increase NRs and reduce greenhouse gas emissions per pound of dressed beef produced, compared to just one growth implant.

1. Introduction

Sustainability has become a prevalent topic in agricultural research and policymaking, focusing on the intersection of environmental impacts, economic profitability, and social preferences. With air quality and greenhouse gas (GHG) emissions leading the conversation, major food industry firms have committed to reducing supply chain emissions and pledged to become carbon neutral within the next two decades (JBS Foods, 2021; McDonald’s, n.d.; Tyson Foods, 2021). Additionally, livestock have been identified as important contributors to methane and nitrous oxide emissions (U.S. EPA, 2024c). The United States leads the world in beef production, and beef cattle production consistently accounts for the largest share of total cash receipts for U.S. agricultural commodities (USDA ERS, 2023). The value of beef production in the United States brings the social and environmental impacts of these operations to the forefront of agricultural sustainability. Industry goals to reduce beef value chain emissions affect every segment of the supply chain, especially producers. How production practices and technologies simultaneously affect economic viability and environmental sustainability of feedlot operations is essential information for producers managing emissions. Further, assessing incremental effects of altering management strategies is important to ensure that feedlots remain profitable while supporting industry goals for emissions reduction. This study is designed specifically to address this issue.

While previous research has linked cattle performance to feedlot net returns (NRs) and GHG emissions, none have evaluated the concurrent economic and environmental effects of combined management practices (Brooks et al., 2011; Cooprider et al., 2011; Crawford et al., 2022; Irsik et al., 2006; Jones et al., 1996; Mark, Schroeder, and Jones, 2000; Stackhouse-Lawson et al., 2013; Swanson and West, 1963; Trapp and Cleveland, 1989). The primary objective of this study is to estimate how mortality and growth implant use affect economic returns and GHG emissions intensity of cattle feeding.

Practices that advance cattle performance, such as improved health outcomes or growth implants used in animal production to stimulate growth and increase efficiency, may improve both economic and environmental robustness of cattle feeding operations. Practices that increase NRs through improved cattle performance may also reduce GHG emissions intensity in cattle feeding. This analysis utilizes methods established in the literature to estimate the effects of changes in mortality and growth implant protocols on NRs and proposes methods for estimating concurrent impacts to feedlot GHG emissions per pound of dressed beef produced.

Meeting industry goals to reduce supply chain emissions requires an understanding of how operational decisions impact net GHG emissions. However, demonstrating potential emissions reductions may not be enough to incentivize producers to alter management strategies or implement new protocols. Investigating concurrent impacts to NRs and GHG emissions could incentivize producers to pursue emissions-reducing practices. This analysis illustrates combined economic and environmental impacts of mortality and growth implant strategies. Results reveal that reducing mortality and using growth implants can simultaneously increase feedlot NR and reduce GHG emissions per pound of dressed beef produced.

2. Background

The U.S. Environmental Protection Agency categorizes GHG emissions as Scope 1, Scope 2, and Scope 3 (U.S. EPA, 2024a; U.S. EPA, 2024b). Scopes 1 and 2 emissions are those produced directly and indirectly, respectively, from sources owned or controlled by the organization. Scope 3 emissions are the result of activities within the organization’s value chain that are not directly within the control of the organization ¹ (U.S. EPA, 2024a; U.S. EPA, 2024b). For beef processors and retailers, on-farm activities are Scope 3 emissions. Thus, industry goals to reduce beef value chain emissions inevitably affect each segment of the supply chain, including producers and stakeholders. The most notable GHGs in beef cattle production are carbon dioxide (CO₂), methane (CH₄), and nitrous oxide (N₂O). GHG emissions can be described as carbon dioxide equivalents (CO₂e), an aggregate measure of GHG emissions in terms of the 100-year global warming potential (GWP) of carbon dioxide (Lynch, 2019). Identifying environmentally friendly management practices that are economically feasible and measuring impacts of these practices are essential for ensuring the continuity of the beef supply chain.

The United States Roundtable for Sustainable Beef (USRSB) has developed industry-wide sustainability goals and identified metrics for monitoring continuous improvement in beef production. Further, the USRSB has set a goal to reduce feedlot GHG emissions by 10% per pound of dressed beef produced before 2030 (U.S. Roundtable for Sustainable Beef, 2022). Economic viability and optimized animal productivity are central to the mission of the USRSB. In feedlots, improved closeout performance and healthier animals are indicative of achieving desired outcomes (U.S. Roundtable for Sustainable Beef, 2022). Feeder cattle performance metrics, such as average feed conversion (AFC) and average daily gain (ADG), are highly dependent upon animal health and diet (Belasco, Cheng, and Schroeder, 2015; Irsik et al., 2006). Although feeder and fed cattle prices account for much of the variation in NRs to cattle feeding, research has long linked cattle performance to returns to management (Brooks et al., 2011; Irsik et al., 2006; Jones et al., 1996; Mark, Schroeder, and Jones, 2000; Swanson and West, 1963; Trapp and Cleveland, 1989). Researchers have also shown that more efficient cattle create fewer emissions (Broocks et al., 2017; Cooprider et al., 2011; Stackhouse-Lawson et al., 2013).

The implications of using growth implants are of particular interest in this study. Use of growth implants increases feed efficiency and ADG (Crawford et al., 2022; Smith and Johnson, 2020; Wileman et al., 2009). Additionally, growth promoting technologies may reduce feedlot GHG emissions per pound of dressed beef produced (Crawford et al., 2022; Stackhouse-Lawson et al., 2013; Stackhouse et al., 2012; Webb et al., 2020). In June 2023, the U.S. Food and Drug Administration implemented a clarification concerning the use of growth implants in beef cattle production systems. The clarification states that cattle may be administered only one growth implant per production phase unless the implant is specifically labelled for reimplantation (U.S. FDA, 2023a). As of August 2023, there were 20 commercial implants labelled for use in growing beef cattle fed in confinement for slaughter. However, only five implants were labelled for reimplantation at the time, and approved reimplantation programs do not allow for use of more than two growth implants in cattle fed in confinement for slaughter (U.S. FDA, 2023b). Thus, producers opting for use of multiple implants during the feeding phase are limited in the selection of implants available for use in a reimplantation program. Findings from this study have important implications for policymakers and producers regarding the number and potencies of growth implants used to manage beef cattle fed in confinement for slaughter.

3. Data and methods

This study utilizes proprietary pen-level closeout data for cattle from nine Midwestern feedlots from 2018 through 2021. The data are stratified by placement weight and sex, including 7,813 pens (960,658 head) of beef steers and heifers entering the feedlot between 400 and 1,000 lbs. Mean descriptive statistics are reported in Table 1. Summary statistics for data stratified by sex are reported in Appendix Table A.1. Weight categories are determined to evaluate lightweight (400–599 lbs.) and heavyweight (900–1000 lbs.) placed cattle. Steers placed at 400–599 lbs. and heifers placed at 900–1000 lbs. are excluded from the analysis due to insufficient observations ² .

Table 1. Mean descriptive statistics for pen-level operational feedlot data

Cattle purchase and sale prices are not available in the feedlot data, so average prices for the closeout period were used as proxies. Feeder cattle purchase prices, dressed carcass prices, and 5-area carcass premiums and discounts are obtained from the Livestock Marketing Information Center (LMIC, n.d.). Weekly corn prices for the corresponding states where cattle were fed serve as a proxy for feed prices, and feed prices used to calculate NRs are based on week of arrival (LMIC, n.d.). Annualized quarterly interest rates were used to estimate interest costs (Federal Reserve Bank of Kansas City, 2024). Yardage cost was assumed at $0.40/head/day (Dennis et al., 2020).

Cattle performance metrics vary modestly across placement weights, with mean ADG ranging from 2.97–3.34 lbs./head/day for heifers and 3.45–3.78 lbs./head/day for steers. AFC ranges from 6.21–6.57 lbs. of feed dry matter per lb. gain for heifers and 6.10–6.35 for steers. Average health costs per head are reported for each pen, however the allocation of health costs across preventative vs. treatment measures is unknown. Health costs and mortality are higher for lightweight cattle, which is expected due to increased health risk and time spent in the feedlot.

Implant potency protocol refers to the dose and combination of hormones in the implants used. Implant potency is defined as low (< 200 mg progesterone/testosterone), moderate (≥ 200 mg progesterone/testosterone or <200 mg trenbolone acetate), or high (≥ 200 mg trenbolone acetate). These potencies were determined following Beck et al. (2022) in addition to consultation with veterinary professionals. Low potency implants are rarely utilized by the feedlots represented in this data; thus, only cattle receiving moderate and high potency implants are included in this study (Appendix Table A.2). Most cattle received two implants in the feedlot, with the exception of heifers placed at 400–599 lbs., and a larger proportion of steers received high potency implants compared to heifers (Table 1).

Feedlot GHG emissions were estimated by UpLook v1.0, a commercially available GHG emissions modeling software, using IPCC guidelines. The Integrated Farm Systems Model (IFSM) was used to validate emission estimates used, using a random sample of 500 pens from the data. GHG emissions estimates reported in the data and those estimated using the IFSM are highly correlated (0.96), providing confidence in the use of pen-level feedlot GHG emissions estimates given in the data for our analysis.

Total GHG emissions are described as non-feedlot emissions (emissions from cow-calf/backgrounding phase of production) and feedlot emissions. Average feedlot GHG emissions (kg CO₂e) per head are reported in Figure 1 (Appendix Table A.3). Animals of the same sex placed at lighter-weights, naturally have larger feedlot GHG emissions due to longer time on feed and greater feed consumption compared to heavy-weight placements. For example, for heifers, feedlot GHG emissions per head decline by 11% in going from 400–599 lb. to 600–749 lb. placement weights and another 9% from 600–749 lb. to 750–899 lb. placement. Steers tend to have higher feedlot GHG emissions per head than heifers of similar placement weight.

Figure 1. Average Feedlot GHG Emissions (kg CO₂e) per head, by placement weight and sex. GHG emissions are estimated by Uplook v1.0.

Average feedlot GHG emissions (kg CO₂e) per pound of dressed beef produced are reported in Figure 2 (Appendix Table A.3). Consistent with per head values, feedlot GHG emissions intensity is higher for steers compared to heifers of similar placement weight when assessed on a per pound of dressed beef produced. For animals of the same sex, feedlot GHG emissions intensity decreases as placement weight increases. For example, for heifers, feedlot GHG emissions per pound of dressed beef produced decline by 15% in going from 400–599 lb. to 600–749 lb. placement weights and another 13% from 600–749 lb. to 750–899 lb. placement.

Figure 2. Average Feedlot GHG Emissions (kg CO2e) per pound of dressed beef produced, by placement weight and sex. GHG emissions are estimated by Uplook v1.0.

While it is well established that cattle feeding NRs vary across management and animal health protocols, the impacts of management practices on GHG emissions and other environmental factors are less documented. This study estimates the effects of mortality and growth implant protocol on AFC and ADG, as well as effects of cattle performance on feedlot GHG emissions intensity. Estimated AFC and ADG are used to calculate NRs and predict GHG emissions per pound of dressed beef produced for various management scenarios. Varying levels of mortality are used to illustrate changes in animal health management, and changes in the number of implants used and implant potency are used to illustrate changes in animal productivity management. This analysis is concerned with the potency of the final two implants received in the feeding phase. Moderate potency protocol indicates an animal received a moderate potency implant followed by a moderate or high potency implant. High potency protocol indicates an animal received a high potency implant followed by a moderate or high potency implant.

Irsik et al., (2006) estimated AFC and ADG models and assessed impacts of different treatments on performance of feedlot cattle. Here, the primary effects of interest are mortality, number of implants used, and potency of implants used. Building from the methods in Irsik et al., (2006), we model AFC and ADG as (individual pen subscripts i are omitted for simplicity):

(1) $$\eqalign{ & AFC = {\beta _0} + {\beta _1}ln\left( {CPW} \right) + {\beta _2}\left( {MORT \cdot 100} \right) + {\beta _3}HC + {\beta _4}{T_1} + {\beta _5}{T_3} + {\alpha _m} \cdot {P_m} + \sum\limits_{n = 2}^9 {{\gamma _n}} \cdot {F_n} \cr & \quad \quad \quad \;\; + \sum\limits_{j = 2}^4 {{\delta _j}} \cdot {D_j} + \varepsilon \cr} $$

(2) $$\eqalign{ & ADG = {\beta _0} + {\beta _1}ln\left( {CPW} \right) + {\beta _2}\left( {MORT \cdot 100} \right) + {\beta _3}HC + {\beta _4}{T_1} + {\beta _5}{T_3} + {\alpha _m} \cdot {P_m} + \sum\limits_{n = 2}^9 {{\gamma _n}} \cdot {F_n} \cr & \quad \quad \quad \;\; + \sum\limits_{j = 2}^4 {{\delta _j}} \cdot {D_j} + \varepsilon \cr} $$

Variables included in AFC, ADG, and NR equations are defined in Table 2. The natural logarithm of CPW, cattle placement weight, is used in the models following prior research (Belasco et al., 2009) and because we expect a nonlinear relationship between placement weight and these independent variables. Because most cattle receive two implants in the feedlot, we evaluate the effects of receiving one growth implant (T ₁ ) and three growth implants (T ₃ ) compared to receiving two growth implants (T ₂ = base). There is little variability in the implant potency protocol for heifers as nearly all received two implants. Implant potency protocol is not included in estimation of cattle performance metrics for heifers, due to multicollinearity between implant potency protocol and number of implants received. For steers, inclusion of implant potency protocol does not introduce multicollinearity and is thus included in model estimation. One limitation of this analysis is lack of genetic information for cattle included in the data. We include fixed effects to account for variation across feedlots, as well as placement quarter fixed effects to account for seasonality. Expected changes in AFC (Eq. 1) are used to predict changes in NRs. The methods of Dennis et al., (2020) are modified to calculate NRs per head as:

Table 2. Feeding net return variables

(3) $$NR=TR-FDRC-YC-FC-HC-IC$$

where total revenue (TR), feeder cattle cost (FDRC), yardage cost (YC), feeding cost (FC), and interest cost (IC) are defined as (HC, health costs are used directly as reported in the data set):

(4) $$TR = DP \cdot avgcarcwt \cdot \left( {1 - MORT - CULL) + (CULL \cdot CULLW \cdot CULLP} \right)$$

(5) $$FDRC = FRP \cdot CPW$$

(6) $$YC = 0.4 \cdot DOF$$

(7) $$FC = FEED \cdot (AFC \cdot (CSW \cdot (1 - MORT - CULL) - CPW))$$

(8) $$IC = \left( {0.5 \cdot \left( {YC + FC + HC} \right) + FDRC} \right) \cdot DOF \cdot \left( {{{IR} \over {365}}} \right)$$

Researchers have linked cattle performance to GHG emissions, and recent literature has established more efficient cattle produce fewer emissions (Cooprider et al., 2011; Crawford et al., 2022; Stackhouse-Lawson et al., 2013). We estimate feedlot GHG emissions per pound of dressed beef produced (FLGHG) as a function of AFC, ADG, and days on feed (DOF). Expected changes in AFC and ADG due to altered management practices (Eqs. 1–2) are used to predict changes in FLGHG. Feedlot fixed effects are included to account for differences in GHG emissions intensity across feedlots such as energy use.

(9) $$\textit{FLGHG}=\beta _{0}+ \beta _{1}DOF+\beta _{2}AFC+\beta _{3}ADG+\sum _{n=2}^{9}\gamma _{n}\cdot F_{n}+\varepsilon$$

4. Results and discussion

Analyses are conducted across four placement weight categories. ³ Average placement weight for the entire sample is approximately 750 lbs. Most observations (88%) fall into weight categories of 600–749 lbs. and 750–899 lbs. Regression results for each model are reported in the following sections. Robust standard errors are estimated to account for potential heteroskedasticity.

4.1. Feed conversion and average daily gain models

Tables 3 and 4 provide parameter estimates for the AFC (Eq. 1) and ADG (Eq. 2) models. Complete results including feedlot fixed effects and quarterly placement effects for AFC and ADG models are reported in Tables A.4. and A.5., respectively. The effect of placement weight on AFC is positive and statistically significant for heifers and steers placed at 600–749 lbs. and 750–899 lbs., indicating that cattle placed at lighter weights are more efficient than heavier placements over their time on feed. The effect of placement weight on ADG is greater for heifers placed at 400–599 lbs. compared to heifers placed at heavier weights; however, a greater proportion of these cattle received three or more implants compared to other groups of cattle. As expected, AFC increases with mortality, while ADG decreases. For heifers and steers placed at 750–899 lbs., a one percentage point increase in mortality (e.g., increase from 1% to 2%) is associated with a 0.09 lb. decrease in ADG.

Table 3. Results for AFC (feed conversion) model (lb. dry feed/lb. gain) ^a

^a Standard error in parentheses = *p < 0.10, **p < 0.05, ***p < 0.01.

Table 4. Results for ADG (average daily gain) model (lb. per day) ^a

^a Standard error in parentheses = *p < 0.10, **p < 0.05, ***p < 0.01.

Effects of health costs on AFC and ADG vary across sex and placement weight. For example, a one dollar increase in health cost per head is associated with a 0.001 lb. decrease in ADG for heifers placed at 600–749 lbs. However, a one dollar increase in health cost per head is associated with a 0.003 lb. increase in ADG for steers placed at 750–899 lbs.

Most cattle received two implants in the feedlot. However, the updated U.S. FDA ruling on use of growth implants in cattle production states that cattle may be administered only one growth implant per production phase unless the implant is specifically labelled for reimplantation (U.S. FDA, 2023a), and there are few approved reimplantation programs for cattle grown in confinement for slaughter (U.S. FDA, 2023b). Including effects of one implant versus two implants provides insight on potential impacts to cattle performance of the number of implants administered to feedlot cattle. AFC increases and ADG decreases for cattle receiving one implant compared to two implants, indicating that a second implant may improve performance of feedlot cattle. On average, steers placed at 600–749 lbs. or 750–899 lbs. receiving one implant gain 0.2 lbs. less per day compared to those receiving two implants. However, a third implant is not associated with an improvement in cattle performance. On average, steers placed at 600–749 lbs. receiving three or more implants gain 0.18 lbs. less per day compared to those receiving two implants.

Implant potency protocol is excluded from heifer performance estimation due to multicollinearity. Reduced potency of implants used in the feedlot is associated with a 0.07 lb. increase in ADG and 0.1 lb. decrease in AFC for steers placed between 600–749 lbs. However, reduced potency of implants is associated with a 0.13 decrease in ADG and an insignificant increase in AFC for steers placed between 750–899 lbs.

4.2. Greenhouse gas emissions model

Results for feedlot GHG emissions models (Eq. 9) are reported in Table 5. ⁴ To maintain confidentiality of third-party provided emissions predictive models, only signs and statistical significance of coefficients of interest can be reported. The models each explain 79% or more of variation in GHG across pens of cattle. Coefficients for days on feed (DOF), AFC, and ADG are all positive and statistically significant (p < 0.01) across placement weight and animal sex models.

Table 5. Sign of coefficients and goodness of fit measures for feedlot GHG emissions models ^a

^a Effects of DOF, AFC, and ADG are statistically significant at p < 0.01 across all placement weights. Coefficients are not reported to maintain confidentiality of third-party provided emissions predictive models.

Additional days on feed are associated with a net increase in feedlot emissions per pound of dressed beef produced. The effect of an additional day on feed increases with placement weight as heavier cattle consume more feed, increasing total GHG emissions per day. Higher feed conversion indicates reduced feed efficiency as heavier cattle generally require more feed to maintain weight. Therefore, positive effects of both AFC and ADG on feedlot GHG emissions intensity are as expected.

4.3. Effects of reducing mortality and using growth implants

Incremental effects are estimated to determine how changes in management practices impact cattle performance metrics, feedlot NRs, and GHG emissions intensity. A one-at-a-time (OAT) method is preferred for this analysis, as it allows for investigation of the effects of changing one independent variable and increases comparability of results across models. A drawback of using OAT is that it fails to consider other input shifts that may occur simultaneously with management practice changes. As a result, point-estimates of changes in variables of interest may be understated or overstated, depending on correlations.

A baseline is established by calculating NRs and estimating GHG emissions intensity for each pen, with dummy variables set to defaults. Point estimates for cattle performance metrics, net returns, and feedlot GHG emissions intensity are calculated using methods previously described in equations 1–9 for various changes in mortality and implant use. Mortality is shocked (25% and 50% decrease) to predict impacts on AFC and ADG, causing changes in NRs and GHG emissions intensity. Similarly, changes in growth implant programs are also evaluated. Since implant potency protocol is concerned with the final two implants received in the feedlot, and the default for number of implants received is two, effects of changing implant potency protocol are compared to the baseline. Health costs are adjusted for scenarios evaluating changes in implant protocol, accounting for cost differences between using one less (more) implant or using implants with reduced potency. ⁵ Additionally, DOF is adjusted to account for changes in ADG, assuming placement weight and sale weight remain constant.

Incremental effects of reducing mortality and growth implant protocol on NRs and GHG emissions intensity for heifers placed between 750–899 lbs. are reported in Figures 3 and 4. Incremental effects of reducing mortality for steers placed between 750–899 lbs. are reported in Figures 5 and 6. Effects for other placement weights are reported in the Appendix (Figs A.1–A.8). Reducing mortality simultaneously increases NRs and reduces feedlot GHG emissions intensity, regardless of sex or placement weight. Impacts of changing implant protocol varies between heifers and steers, and across placement weight.

Figure 3. Incremental effects of reducing mortality on feedlot net returns and GHG emissions per pound of dressed beef for heifers placed between 750–899 lbs.

Figure 4. Incremental effects of number of growth implants and implant potency on feedlot net returns and GHG emissions per pound of dressed beef for heifers placed between 750–899 lbs.

Figure 5. Incremental effects of reducing mortality on feedlot net returns and GHG emissions per pound of dressed beef for steers placed between 750–899 lbs.

Figure 6. Incremental effects of number of growth implants and implant potency on feedlot net returns and GHG emissions per pound of dressed beef for steers placed between 750–899 lbs.

For heifers placed between 750–899 lbs., reducing mortality by 25% and 50% is associated with a $4.64 and $9.37 increase in NRs per head, respectively. As mortality decreases by 25%, associated GHG emissions intensity decreases by approximately 0.003 kgCO₂e/lb. beef or 0.2%. Decreasing the number of implants from two to one is associated with a decrease in NRs of roughly $7.27 per head and an increase in GHG emissions intensity of approximately 0.14 kgCO₂e/lb. beef or 8%. These findings suggest that using a second implant reduces GHG emissions per pound of dressed beef produced. These findings are consistent with Crawford et al., (2022), which finds that although using growth implants may increase total GHG emissions per head, growth implants decrease GHG emissions per pound of hot carcass weight.

For steers placed between 750–899 lbs., reducing mortality by 25% and 50% is associated with a $5.59 and $11.31 increase in NRs per head, respectively. As mortality decreases by 25%, GHG emissions intensity decreases by approximately 0.004 kgCO₂e/lb. beef or 0.2%. Decreasing the number of implants from two to one is associated with a decrease in NRs of approximately $10.97 per head, and an increase in GHG emissions intensity of 0.1 kgCO₂e/lb. beef or 5%. Moderate potency protocol is associated with approximately a $2.36 decrease in NRs per head and a 0.02 kgCO₂e, or 1%, reduction in GHG emissions intensity, compared to high potency protocol.

5. Conclusions and policy implications

Beef industry sustainability goals may affect producer management decisions, which will in turn affect economic and environmental health of the beef value chain. Recent emphasis placed on reducing GHG emissions in the U.S. beef supply chain creates a need for information regarding emissions-reducing potential of practices enabling producers to remain profitable. The novelty of this analysis is quantification of economic together with environmental impacts of feedlot management strategies. Changes in GHG emissions per pound of dressed beef produced through implementation of various management practices have important implications for policymakers and stakeholders alike.

This study analyzes concurrent effects of mortality and use of growth implants in Midwestern feedlots. Reducing mortality simultaneously increases feedlot NRs and reduces GHG emissions per pound of dressed beef produced. Enhanced cattle health to decrease mortality is one way that GHG emissions intensity might be reduced through better identification, diagnosis, and treatment of sick or injured animals, or procurement strategies to prioritize healthy cattle upon arrival. However, added costs of greater emphasis on animal health must be assessed relative to savings from increased animal performance. For example, our findings illustrate that for heifers placed at 750–899 lbs., reducing mortality by 25% could improve returns to management and reduce emissions intensity if accomplishing this lower mortality costs less than $4.64 per head, on average. Enhancement of proactive health programs, such as for animal preconditioning or antimicrobial metaphylaxis, also may reduce mortality.

All observed associations between use of growth implants and cattle performance in this study are based on operational data. Decisions on the number and type of implants used are dependent upon factors that may not be included in this analysis. Impacts of changing growth implant programs vary between steers and heifers, and across placement weight, but using at least two implants, as compared to just one, increased NRs per head and reduced feedlot GHG emissions per pound of dressed beef produced. Using multiple moderate potency implants may further reduce feedlot GHG emissions intensity. The 2023 FDA ruling indicates that only five commercially available implants are currently labelled for reimplantation in cattle grown in confinement for slaughter (U.S. FDA, 2023b), and only two of the five implants are moderate potency. Increased regulation of growth technologies used in cattle production may limit producers’ abilities to reduce GHG emissions intensity in a cost-effective manner.

Despite a unique and detailed data set employed, this study certainly has limitations. The analysis completed here relies upon observational data. Although rich in numbers of observations and reflective of animal performance observed in commercial production settings, we estimate models using data with associated endogenous or unobserved management decisions. For example, a pen of cattle that may have had a different implant protocol from another pen that appear similar in the data, could have been different in ways we cannot observe. Likewise, we do not know if health costs are preventative or reactive to diagnosed health ailments. Additionally, cattle prices and costs of other inputs are not collected as part of the data. Coupled with the unobservable characteristics of these pens, this makes NRs our best estimates. These and related limitations of the data must be considered as implications are drawn. Additionally, the methods implemented here only investigate the changes in NRs and emissions intensity associated with mostly changing one input at a time. Future research investigating effects of management practices and feedlot GHG emissions intensity could include further variation of inputs. Utilizing methods which allow for modification of multiple inputs at once may allow for more detailed analysis, although adding to model complexity.