“Plant-based is more efficient (thermodynamics / trophic levels)”

Feed-to-food conversion

Evidence quality: High for the direction of conversion loss; moderate–high for its magnitude (Cassidy et al., 2013; Shepon et al., 2016). Animals retain only a fraction of feed energy and protein in edible products because feed also supports maintenance, movement, heat production, reproduction, and non-edible tissues (Shepon et al., 2016). Empirical food-system studies quantify this difference rather than relying only on a trophic-level rule (Cassidy et al., 2013; Shepon et al., 2016). A global crop-allocation analysis estimated that 36% of calories produced by crops were used as animal feed and that 12% of those feed calories ultimately contributed to human diets as meat and other animal products (Cassidy et al., 2013).

A detailed US analysis estimated feed-to-food conversion for dairy, beef, poultry, pork, and eggs (Shepon et al., 2016). Averaged across those categories, caloric and protein conversion efficiencies were 7–8%; beef was lowest at 3% for both measures, while other categories were higher (Shepon et al., 2016). These estimates support the stub's claim that feeding edible crops to livestock produces substantial calorie and protein losses, while also showing that the magnitude differs by animal product and production system (Shepon et al., 2016).

Opportunity cost and food availability

Efficiency can be expressed as nutritionally comparable food produced from a fixed area of cropland (Shepon et al., 2018). A US scenario analysis compared animal products with plant replacements selected to minimize cropland use (Shepon et al., 2018). It estimated opportunity losses of 96% for beef, 90% for pork, 75% for dairy, 50% for poultry, and 40% for eggs, and calculated that complete replacement of the modeled animal products could provide enough food for 350 million additional people under the study's nutritional and consumption assumptions (Shepon et al., 2018). This is a modeled opportunity-cost result, not a forecast that land reallocation or food distribution would occur automatically (Shepon et al., 2018).

At global scale, Cassidy and colleagues estimated that directing crops to human food rather than feed and biofuel could increase available food calories by as much as 70%, theoretically sufficient for about four billion additional people under the study's allocation assumptions (Cassidy et al., 2013). Both studies concern potential food availability; neither shows that biophysical availability alone resolves income, access, distribution, or dietary-quality constraints (Cassidy et al., 2013; Shepon et al., 2018).

Land and other environmental inputs

Life-cycle evidence is consistent with the conversion findings (Poore and Nemecek, 2018). A synthesis covering 38,700 farms and five environmental indicators found up to 50-fold variation among producers of the same food, but reported that the lowest-impact animal products typically exceeded vegetable substitutes (Poore and Nemecek, 2018). The result supports the stub's claim that land and resource use per unit of food is typically higher for animal products, while the large within-product variation cautions against treating every farm as identical (Poore and Nemecek, 2018).

The role of non-arable land and by-products

Not all livestock feed competes directly with human food (Mottet et al., 2017). A global feed analysis estimated that 86% of livestock feed dry matter consisted of materials not currently eaten by humans, including grass, crop residues, and processing by-products; it also estimated that livestock consumed one-third of global cereal production and used about 2.5 billion hectares of land (Mottet et al., 2017). Livestock can therefore convert some low-opportunity-cost biomass from non-arable land into food, while systems using human-edible feed retain the measured conversion losses (Mottet et al., 2017).