The climate impact and land use of cultivated meat: Evaluating agricultural feedstock production

Our study uses data on agricultural crops in Germany and feedstuff analyses to determine amino acid patterns and glucose contents [91, 92]. The INRAE-CIRAD-AFZ Feed Tables—a comprehensive reference for feed composition and nutritional values developed by French research institutions INRAE, CIRAD, and AFZ—are used for glucose quantity specifications [93]. Glucose content is derived from starch and sugar, however, the database does not provide detailed information on the monosaccharide composition of these fractions. For energy and oil crops, starch was considered primarily as D-glucose polymers, allowing near-complete conversion to glucose during hydrolysis. For grain legumes, only half of the starch and sugar fraction was con- sidered convertible to glucose due to non-starch polysaccharides and oligosaccharides. Based on Humbird (2021) [37], an overall amino acid yield of 70% after extraction from crops was assumed. Due to feed analyses capturing glutamine and asparagine with glutamic acid and aspartic acid, these amino acids were considered collectively (Table 3). To quantify nutrient delivery, we used average yields from 2017–2021 calculated by the Bavarian State Research Center for Agriculture [82]. Nutrient yield per hectare was calculated by multiplying crop nutrient contents (%) with per-hectare yields (t FM/ha), considering dry matter contents of the feedstuffs [83, 94]. The energy demand for CM production was sourced from Vergeer et al. (2021) [40], Mat- tick et al. (2015) [50], and Tuomisto et al. (2014) [95]. Information on solar panel efficiency

came from Bo¨hm et al. (2022) [96]. Crop productivity data was obtained from the Bavarian State Research Center for Agriculture (LfL, 2023) [82]. Finally, we assume that all raw materials, including soy, are produced on-site by the farm, which is set at 100 hectares. The linear programming model then balances the nutrient requirement (kg/kg CM) with the nutrient delivery (kg/ha) for CM production. A specific correction is applied that takes into account that the LfL data relates to 14% or 9% moisture (equivalent to 91% or 86% dry matter). This results in an adjustment of hectare yields based on the dry matter content of each analyzed feed material. Here is an example, in the case of soybeans. Dry matter content per analysis: Yield of 3.17 t/ha at 14% and dry matter following feed analysis of 89.5% results in a calcula- tion of 3.17 t/ha x 0.86/0.895, or approximately 3.05 t/ha. This corrected yield per hectare is then multiplied by nutrient content percentages, the nutri- ent extraction factor, and a conversion factor of 1,000 kg/t. This calculation gives the nutrient delivery in pure nutrient kilograms per hectare for each crop type. Table 3 presents the nutrient requirements for CM. The nutrient delivery is directly correlated with the nutrient requirement, expressed in kg/kg of CM, ensuring a consistent comparison of values in the analysis. Table 2. Crop rotation restrictions assumed in the linear programming model. Crop category Category limit (%) Individual crop restrictions (%) Grain legumes � 20% (without soy)

Soybean � 50%, lupine � 15%, field bean � 15%, pea � 15% � 50% (with soy) Summer crops � 67.77% Grain maize � 50%, sunflower � 20%, potato � 20% Winter crops � 67.77% Wheat � 33.33%, rye � 50%, triticale � 33.33%, winter barley � 33.33%, oats � 25%, rapeseed � 20% Foliage crops � 25% Soybean, lupine, field bean, forage pea, rapeseed, sunflower Rapeseed + Sunflower � 33.33% Rapeseed, sunflower Cereals � 67.77% Wheat, rye, triticale, barley, oats https://doi.org/10.1371/journal.pone.0316480.t002 PLOS ONE Evaluating agricultural feedstock for cultivated meat PLOS ONE | https://doi.org/10.1371/journal.pone.0316480 January 22, 2025 10 / 27