There is an exogenous SSP specific component for the livestock density

To represent this healthy U.S. diet in GLOBIOM, we performed a series of additional conversions. First, we determined the allocation of GLOBIOM items across Calculator product groups based on how commodities are currently allocated across each product group. For example, the majority of barley is used for making alcohols and the remaining 9% is consumed as cereals, and about 21% of corn that is consumed by humans is consumed as a cereal, whereas the remaining 79% is used for making corn-based sugars . We then calculated healthy diet “shifters” for each Calculator product group by dividing the healthy diet kcal by the baseline diet kcal. A “shifter”, as we define it here, is a constant multiplier that allows conversion between scenario values. Food product group shifters allow for the creation of a healthy U.S. diet scenario using any baseline diet kcal values . We then combined the healthy diet shifters with the GLOBIOM-to-Calculator product group allocations to calculate diet shifters for GLOBIOM items . These shifters were used directly in GLOBIOM to create the Healthy US, Healthy World, and Sustainability scenarios. Unlike in the Calculator,dutch bucket for tomatoes shifters were applied to the demand curve in GLOBIOM, since final human consumption is determined endogenously .

This means that dietary changes between Calculator and GLOBIOM may not be identical, though they are highly similar .Using the US FABLE calculator, we developed two sets of yield shifters. The “BAU yields” shifters apply 2000–2015 yield growth trends in the U.S. to simulation years 2000–2050 . The Higher “U.S. Yields” shifters increase the growth rate by 200% between 2015 and 2050 if the annual positive rate is lower than 1%/year, 80% if the annual rate is higher than 1%/year, and turns negative historic growth rates into positive growth rates , but only applying these adjusted growth rates on the yields after 2020 . For example, if crop yields were declining at a rate of 0.5% per year historically, and the yield in 2020 was 4.5 tons/ha, we changed this in the higher “U.S. Yields” scenario to increase at a rate of 0.5%/ year starting in 2020 and thus the yield would be 4.6 tons/ ha in 2025. GLOBIOM endogenously adjusts yields based on cropping mix and management . Thus, GLOBIOM yields are a combination of our exogenously applied yield increases and GLOBIOM endogenous adjustments. As a result, the yields between GLOBIOM and the Calculator may not match exactly .We used the Calculator to explore the exogenous effects of changes in livestock productivity parameters. These livestock changes may represent technological innovations or management systems shifts .

The Calculator uses the historic USDA growth rate from 2010 to 2020 linearly extrapolated out to 2050 in the “BAU productivity” livestock productivity scenario, and we increased this growth rate by 20% in the “High productivity” livestock productivity scenario . For ruminant grazing density, the “Constant density” scenario uses the same ruminant density from 2010 to 2050, and reduces this by 6% by 2050 in the “Declining density” scenario. Though these changes were applied to all grazing livestock in the U.S., the vast majority of grazing livestock is cattle, thus, these changes effectively alter beef productivity and grazing density. We conducted these sensitivity analyses in the Calculator, because livestock productivity in GLOBIOM is a more complex combination of endogenous and exogenous factors than for crop yields. The amount of livestock product per unit of land area depends on the average feed conversion ratio and the grass yield. The grass yield is exogenous and can change over time under different climate scenarios , whereas the average feed conversion ratio is endogenous as the production system composition is endogenous.We constructed two main scenarios—Baseline and Sustainability—in both GLOBIOM and the US FABLE Calculator.

The values of all variables chosen for the Sustainability scenario are expected to favor sustainable outcomes . For the Baseline scenarios , we assume no change from the current average U.S. diet, SSP2-Middle-of-the-Road diets for ROW, and SSP2 baseline yields. GLOBIOM model runs generated scenario-specific values that were used as inputs in the Calculator for yields, livestock productivity, ruminant density, imports, and exports . In GLOBIOM, we ran five additional scenarios that isolate the roles of U.S. diets, ROW diets, and crop productivity assumptions to examine every combination of input assumptions. We could not replicate these scenarios in the Calculator because it cannot represent global demand and production. We simulated alternative crop productivity futures in GLOBIOM. As described above, yields are a function of both endogenous decisions and exogenous productivity growth rates that vary between business-as-usual and high yield scenarios . In this analysis we vary the latter to represent a range in expected technological change from business-as-usual to optimistic growth. The remaining five scenarios are as follows : 2: healthy U.S. diets and healthy ROW diets , 3: healthy U.S. diets and high U.S. yields ; 4: healthy U.S. diets only ; 5: healthy ROW diets only ; and 6: high U.S. yields only . For the Calculator sensitivity analysis , we used the A: sustainability scenario assumptions, but changed either the livestock productivity to be higher than in GLOBIOM or ruminant density to be lower than in GLOBIOM . Because of inherent differences in underlying data, model infrastructure, and system boundaries between the FABLE Calculator and GLOBIOM approaches, we report only the difference and percentage change from each model’s BAU scenario.

Simulated diets and crop yields reflect scenario adjustments. Scenario adjustments to commodity demand curves and crop yields in GLOBIOM resulted in both production and consumption changes . That is, instead of perfect alignment with assumed diet and productivity assumptions in a domestic-only LCA or mass balance approach, simulated diets and yields in GLOBIOM reflect endogenous prices and supply-side adjustments that cause variation in crop yields . GLOBIOM uses commodity-specific demand curves for representing human demand. Thus, applying the healthy U.S. diet shifters essentially shifts the entire demand curve, as opposed to the final demand for each commodity, which is determined by both the demand and supply curves and market dynamics. As a result, applying the same set of shifters to both the Calculator and GLOBIOM does not necessarily ensure the same percentage change in final per capita consumption across all items. However, demand curve adjustments to reflect healthy diets in particular, led to expected changes in consumption across all food product groups . Results indicate that the final per capita consumption in GLOBIOM very closely resembles that of the Calculator . Similarly, simulated yields from GLOBIOM vary across scenarios due to market adjustments in the U.S. and the rest of the world, illustrating the sensitivity of the U.S. production system to global market forces .Pastureland declines significantly while cropland contracts slightly in response to healthier U.S. diets . Healthier diets in the rest of the world and increases in U.S. crop yields only modestly reduce cropland in the U.S.,blueberry grow pot but significantly reduce cropland in the rest of the world. In the Sustainability scenarios, domestic land used for livestock forage and grazing decline by 37 mil ha in the US FABLE Calculator and 28 mil ha in GLOBIOM scenarios if the average American diet resembled the Healthy-style DGA diet by 2050 . These declines in pastureland are far more dramatic than for cropland, which declines by only 3.9 mil ha in the Calculator scenarios and 2–3.3 mil ha in GLOBIOM scenarios due to healthier U.S. diets . Percentage reductions in both cropland and pastureland are similar across the two models, with the reductions in the Calculator about 1% and 10% points greater, respectively. Both models assume that reductions of pastureland and cropland result in a commensurate increase in natural lands. Across GLOBIOM scenarios, healthy U.S. diets have the greatest single impact on pastureland changes . Pastureland use is more sensitive to dietary changes than cropland due to the low relative land use efficiency of beef production . Increasing crop productivity has no discernible impact on pastureland use. Simultaneously shifting to healthy diets in the rest of the world and the U.S. only negligibly changes pastureland requirements in the U.S. by 1–1.8 mil ha relative to only shifting to healthy diets in the U.S. , because most beef produced in the U.S. is domestically, as opposed to being exported.

Thus, U.S. pastureland should be most responsive to changes in domestic beef consumption. Correspondingly, shifting just the ROW to healthier diets actually slightly increases pastureland over the baseline by 1.8 mil ha despite a slight reduction in U.S. beef production, likely a result of decreased beef land use efficiency . As the ROW demand for U.S. beef declines, these small decreases in U.S. production result in beef land use efficiency reduction. For cropland, there is a similar spread of about 1 mil ha across the GLOBIOM scenarios that adopt healthier diets in the U.S., the ROW, or both. The greatest declines in cropland are observed with healthier diets. Increased crop productivity in the U.S. only reduces domestic cropland by less than 200,000 ha , due to increased production and exports from the greater global comparative advantage of U.S. crop commodities . As a result, higher U.S. yields alone cause a 7.5 mil ha decrease in croplands globally by 2050 relative to the baseline, which is partially offset by a 1.9 mil ha increase in grassland globally . Annual domestic GHG emissions decrease due to shifts to healthier diets in the U.S. and declines are primarily driven by livestock methane emissions reductions and land sequestration. As a result of a healthier U.S. diet, annual CO2e emissions from the agriculture, forestry, and other land use sectors reduces by 176–197 MT for GLOBIOM scenarios and 187 MT for the Calculator Sustainability scenario compared to the baseline by 2050. Livestock methane emissions drive the majority of total reductions for GLOBIOM scenarios, whereas land use change emissions drive the majority of reductions for the Calculator. Most of the land use change emissions reductions are due to increases in forest sequestration from natural regeneration on former cropland and pastureland . As with land use changes, domestic emissions reductions by 2050 show minor differences in U.S. emissions between GLOBIOM scenarios that vary yields and diets in the ROW. These differences are only apparent for cropland-related emissions—crop and soils non-CO2. Increasing yields alone has little to no effect on total emissions, since any additional sequestration from land use change is negated by increased crop and soil non-CO2 emissions due to more intensive farming practices. In particular, N2O emissions from fertilizer use increases as fertilizer intensity expands with higher yields; higher fertilizer application and associated input costs are exogenously required to increase to achieve higher exogenous yield growth rates. Healthy ROW results in near-term total emissions reductions of 50–60 MT CO2e/year, but these diminish to less than 10 MT CO2e/year by 2050. We do not find evidence of international leakage when the U.S. shifts to healthier diets . In fact, we find slight declines in global emissions in the Healthy U.S. and U.S. Yields scenarios . We do find that Healthy ROW alone reduces global AFOLU emissions by 26% despite having minor impacts on U.S. emissions . The future trajectory of beef productivity and ruminant density provide bounds for the range of possible land use and emissions impacts from healthy U.S. diets . To explore the role of technology improvements in the cattle industry and changes in production system or intensification in response to changes in demand, we use the Calculator to run a range of sensitivity scenarios that exogenously alter beef productivity and ruminant density of cattle. We find that beef productivity decreases significantly in response to lower domestic beef consumption and production after 2020 in the healthier U.S. diet GLOBIOM scenarios . Productivity increases and then decreases after 2030 or 2040 for the scenarios that maintain the current average U.S. diet . The business-as-usual beef productivity trajectory in the U.S., based on USDA data from the last 20 years is comparable with GLOBIOM baseline until 2040, when the productivity growth in GLOBIOM starts to level of due to market conditions. The sensitivity that increases this BAU livestock productivity growth rate causes beef production to exceed that of all GLOBIOM scenarios . These productivity increases result in the greatest pastureland reduction by 2050 across all scenarios. Emissions follow similar trends with increases in beef productivity resulting in significantly greater total emissions reductions compared to baseline or 90–127 MT/year greater reductions than the Calculator Sustainability scenario that uses GLOBIOM productivity assumptions .