Assessing variability in whole-farm environmental impact estimates using a partially-stochastic beef production model

Abstract Text:

Environmental impact (EI) studies often aim to identify resource use and greenhouse gas (GHG) emissions from an average production system without accounting for biological variability. These models are frequently used as means to compare EI between systems but they do not account for the variability expected in EI calculations. Our objectives were to develop a partially-stochastic model of beef EI and to use that model to examine implications of increased efficiency through improved calving rate. A whole-system model of beef production EI was adapted to account for the variability in land use, water use and GHG estimates. Variability in animal production parameters was not assessed. Three scenarios were tested: LOW (80% conception), CON (89% conception) or HIGH (100% conception). Projected changes in calving rate were compared with and without accounting for EI ranges. Reported state average crop yield and irrigation values were collected over a 20 yr period and used to represent variability in yield and irrigation estimates. Equations for CH₄ and N₂O were varied by their reported confidence bounds. Land use was expressed in m²/kg hot carcass weight beef (HCWb), water use was in L/kg HCWb and GHG were calculated as CO₂-equivalents (CO₂e)/kg HCWb

Table 1. Means and ranges of EI metrics across efficiency scenarios

Scenario	Land Use (m2/kg HCWb)	Water Use (L/kg HCWb)	GHG (CO2e/kg HCWb)
LOW	82 (50.1 – 246.9)	258.1 (226.9 – 328.1)	21.5 (10.4 – 34.3)
CON	75 (45.9 – 224.5)	265.8 (217.7 – 315.0)	19.9 (9.8 – 31.7)
HIGH	68 (42.0 – 203.3)	258.1 (211.1 – 306.0)	18.5 (9.1 – 29.3)

The LOW scenario (Table 1) had the greatest EI while the HIGH scenario had the lowest. This is in agreement with current literature relating efficiency to EI. When the variability around each environmental estimate was accounted for (Table 1), the ranges described by the model overlapped considerably for all levels of operation efficiency. As a percentage of the mean, crop yield variability resulted in land use estimates with an error bound of about 200% while GHG and water use varied by about 100%. Variability associated with EI estimates was typically greater than the projected differences between treatments simulated. The variability may be consistent with true biological variability thus before accurate assessments of on-farm management to improve EI can be conducted, better methods to understand and account for the causes of this variability must be developed. 

Keywords:

Environmental Impact; Variability; Beef Production