Fatty acids and secondary metabolites can predict grass-finished beef and supplemental cattle feeds

Concerns about environmental and human health impacts of beef production have prompted interest in agroecological, rotational grazing systems on biodiverse pastures. Prior research shows grass-finished beef can differ markedly from grain-finished beef, including lower total fat, higher CLA, more omega-3 PUFA, and a lower n-6:n-3 ratio. However, nutrient profiles vary widely among grass-finished products, and less is known about secondary metabolites (phytochemicals) that may transfer from forages to meat. Many producers must supplement with conserved forages (hay, baleage) or other feeds (e.g., soybean hulls) due to seasonal forage limitations. This study investigated whether profiles of fatty acids and secondary metabolites in beef differ when grass-finished cattle on pasture are supplemented with hay, baleage, or soybean hulls, or finished in confinement on baleage plus soybean hulls, and whether such profiles can authenticate beef production systems and supplemental feeding practices.

Previous comparisons of grass- vs grain-finished beef primarily focused on macronutrients, fatty acids, and vitamins, showing higher omega-3 PUFA, CLA, and vitamin E in grass-finished beef, and lower n-6:n-3 ratios. Nonetheless, grass-finished beef can vary substantially by producer (e.g., reported n-6:n-3 ratios ranging from ~1.8:1 to 28.3:1), suggesting production and supplemental feeding practices influence outcomes. Phytochemicals from diverse forages can transfer to meat and may confer antioxidant/anti-inflammatory properties. Conserved forages (hay, baleage) often exhibit reduced PUFA and phytochemicals compared to fresh pasture, and ensiling alters phenolic profiles, potentially increasing certain phenolic acids while degrading polyphenols. Soybean hulls, a fiber-rich by-product sometimes used despite not being permitted under some grass-fed standards, have variable phytochemicals and may alter rumen conditions affecting biohydrogenation and metabolite transfer. There is growing interest in analytical authentication of grassland-origin products using metabolomics, lipidomics, vitamins, and FA ratios (e.g., n-6:n-3).

Design and animals: Two-year study (2020–2021) at Michigan State University’s Kellogg Biological Station (Hickory Corners, MI). Each year, 60 Simmental–Angus influenced steers (~387 ± 47 kg) were randomly allocated (after weight stratification) to one of four diets with three replicate groups (pens/paddocks) of five animals per diet: (1) GHAY: pasture access plus 4.5 kg/head/day hay (control); (2) GBLG: pasture access plus 4.5 kg/day baleage; (3) GSH: pasture access plus 4.5 kg/day soybean hulls; (4) BLGSH: confinement, ad lib baleage plus 4.5 kg/day soybean hulls. Pasture comprised alfalfa, red and white clover, orchard grass, and endophyte-free tall fescue. Hay contained alfalfa, orchard grass, tall fescue; baleage contained alfalfa and orchard grass. Rotational strip-grazing occurred three times weekly. Over two years, sample losses resulted in n = 115 total animals. Sample collection: Feed and pasture samples were collected in late July of each year. Beef longissimus lumborum samples were collected at slaughter in November (18–20 months of age) and stored at −80 °C. Feed analyses: Proximate composition (DM, ash, CP, NDF, ADF, gross energy) via standard AOAC and related methods; chlorophyll A/B quantification by UV-Vis spectrophotometry after acetone extraction; total phenolics by Folin–Ciocalteu (mg GAE/g). Fatty acid (FA) extraction used microwave-assisted extraction; FAME prepared by transesterification; GC–MS (PerkinElmer 680/600S, HP-88 column) quantified FA using internal/external standards and EI fragmentation; CLA isomers identified by reference standards. Beef analyses: FA profiles quantified by GC–MS as above (multiple injection conditions including splitless). Targeted metabolomics performed on beef and feed using UHPLC–MS/MS (SCIEX 7500 Triple Quad, Shimadzu Nexera LC-40; Kinetex F5 column) in positive/negative ESI with MRM; samples processed with methanol extraction, protein precipitation, SPE (C18), and normalized using isotopically labeled internal standards (QReSS). Quantification (mg/100 g) for compounds with external standards; others reported in arbitrary units. Vitamin E (alpha-tocopherol) quantified by HPLC-UV after hexane extraction; minerals by ICP–MS; lipid oxidation measured by TBARS (MDA-equivalent) assay. Statistics and multivariate analysis: Power analysis targeted 15 heads/treatment/year (α=0.05, β=0.80, CV=0.10). Linear mixed models assessed diet, year, and interaction (fixed effects), with random effect pen nested within diet; Tukey post hoc; Dunn–Šidák correction; p<0.05 significant. Metabolomics visualization via MetaboAnalyst 5.0: PCA; random forest (500 trees; 7 predictors; OOB error for validation). Ranked heatmap of top 50 compounds. Pathway analysis used KEGG Bos taurus with hypergeometric enrichment and topology by relative betweenness centrality. Data were mass-normalized and log-transformed.

• Feed and beef FA profiles: No significant differences in total SFA or MUFA among beef groups. GHAY beef had significantly higher total PUFA (p=0.004), driven by n-3 PUFA (p<0.001), including higher EPA (C20:5 n-3; p<0.001), DPA (C22:5 n-3; p<0.001), and DHA (C22:6 n-3; p=0.004). The n-6:n-3 ratio differed by diet (p<0.001), lowest in GHAY and highest in GSH.
• Metabolites: Of 94 measured, pyridoxine (vitamin B6), alpha-tocopherol (vitamin E), hippuric acid, and gallic acid differed between diets (all p<0.05). GHAY showed higher pyridoxine and alpha-tocopherol versus BLGSH; hippuric acid was about two-fold higher in GHAY than the other groups; gallic acid was lower in GHAY than GBLG, GSH, and BLGSH.
• Multivariate classification: PCA indicated partial overlap among groups (PC1 explained 24.9% variance). Random forest achieved overall predictive accuracy of 73% with class errors: GHAY 0% (100% accuracy), BLGSH 3% (97% accuracy), GBLG 50%, GSH 59% (41% accuracy). Top discriminators included vitamin E, n-6:n-3 ratio, TBARS, total n-3 PUFA, DPA, and EPA.
• Pathway analysis: Enrichment in TCA cycle, arginine biosynthesis, glyoxylate/dicarboxylate metabolism, arginine/proline metabolism, riboflavin metabolism, taurine metabolism, ketone biosynthesis/degradation, vitamin B biosynthesis, retinol metabolism, and terpenoid backbone biosynthesis.
• Interpretation: Conserved forages (hay, baleage) can be used without markedly compromising beef nutrient profiles; metabolomics and lipidomics show high potential to authenticate exclusively forage-based vs feed-based diets. Baleage and soybean hulls were associated with higher gallic acid derivatives in feed; hippuric acid patterns suggest higher overall phenolic intake in GHAY.

The study addressed whether supplemental feeds in grass-finished systems alter beef FA and secondary metabolites and whether these signatures can authenticate production practices. Despite relatively few significant univariate differences across 80 FA and 90 metabolites, multivariate models reliably identified GHAY and BLGSH beef with high accuracy, demonstrating strong authentication potential, particularly for exclusively pasture/hay-fed animals and confinement on conserved forages. Elevated vitamin E and n-3 PUFA, together with lower TBARS in pasture-based groups, align with established grass-finish benefits. The two-fold higher hippuric acid in GHAY indicates greater phenolic intake, consistent with diverse pasture plus hay supplementation, whereas baleage-associated phenolic acids (e.g., gallic acid) were reflected in beef from GBLG and GSH. Soybean hulls contributed less to phytochemical richness transfer, potentially due to fiber-driven rumen effects. The n-6:n-3 ratio and vitamin E emerged as robust discriminators, supporting their utility in authentication frameworks. Overall, the findings suggest producers can incorporate conserved forages strategically while maintaining favorable beef nutrient profiles and that metabolomics/lipidomics can underpin empirical verification of grass-finished claims.

Metabolomics and lipidomics effectively differentiated beef from finishing systems that varied by supplemental feeds, achieving 100% predictive accuracy for pasture plus hay (GHAY) and 97% for confinement on baleage plus soybean hulls (BLGSH). Key discriminators included vitamin E, TBARS, n-3 PUFA (EPA, DPA, DHA), and the n-6:n-3 ratio, along with select phytochemicals. Although between-diet differences were often modest, the combined datasets allowed reliable authentication of diet backgrounds. These results support the use of analytical controls (e.g., FA ratios, vitamin E, metabolite panels) to verify grassland-origin products and inform producers seeking to optimize beef nutrient density with conserved forages. Future work should assess human health impacts of consuming grass-finished beef with different supplements, test additional supplemental feeds (e.g., distiller’s grains), evaluate dose-response effects, and examine carcass fatness/marbling influences on authentication accuracy.

• Feed intake was not recorded; actual consumption of supplemental feeds by pasture groups is unknown, so precise diets cannot be confirmed.
• Botanical composition proportions of the diverse pasture were not quantified, limiting attribution of specific metabolites to plant species.
• Supplemental feed amounts (4.5 kg/head/day) may have been too low to elicit larger metabolic differences; dose-dependence was not explored.
• The confinement group was not a conventional high-grain feedlot; finishing on baleage plus soybean hulls (with an n-6:n-3 ratio ~1.7:1) may explain fewer differences relative to pasture groups compared with prior grain-based studies.
• Forage sampling strategy may not have fully captured phenolic exposure in GHAY; the metabolomics panel may have missed certain polyphenols.
• Meat is a complex matrix; not all metabolites may have been extracted or identified; field is evolving.
• Sampling constraints (e.g., COVID-19) limited feed collections to late July in both years.