The research team, led by G. Osthoff of the University of the Free State with collaborators from North-West University and the University of Pretoria, set out with three objectives: to describe quantified blood metabolites in healthy giraffes and southern white rhinoceroses, to test for sex-associated differences within each species, and to compare the two species using matrix-matched samples. The choice of animals was deliberate. Giraffes are foregut-fermenting ruminant browsers, while white rhinoceroses are hindgut-fermenting grazers, meaning their blood chemistry could plausibly differ because of species physiology, diet, and the microbial fermentation that occurs at opposite ends of the digestive tract.

The rhinoceros dataset was substantial: 45 apparently healthy, free-ranging white rhinoceroses, 31 females and 14 males, were sampled during routine management procedures on a single semi-extensive wildlife property in the Northern Cape Province over five days in May 2019. Each animal was immobilised with etorphine, hyaluronidase and azaperone, and blood was drawn from the posterior auricular vein into EDTA tubes, centrifuged within two hours, and stored at minus 80 degrees Celsius. The giraffe data came from 24 valid sampling occasions involving 19 animals at Rooipoort and Sandveld Nature Reserves, with serum collected in 2017 and EDTA plasma from six animals in 2018.

The analytical pipeline was exacting. Samples were filtered through 10 kDa membranes to remove proteins, mixed with a phosphate buffer in deuterium oxide containing an internal standard, and run on a Bruker Avance III HD 500 MHz spectrometer. Metabolites were annotated against pure-compound spectral libraries supplemented with two-dimensional J-resolved and COSY spectra, and identities were graded according to the Metabolomics Standards Initiative, with Level 1 assignments confirmed against reference standards. Concentrations were reported in micromoles per litre, and blank cells where no resonance was discernible were treated as non-detections rather than zeros, a distinction the authors stress is critical for honest interpretation.

The statistics were equally careful. Concentrations were log2-transformed and tested with Welch’s t-test, which does not assume equal variances, followed by Benjamini-Hochberg false-discovery correction across each family of metabolites. Effect sizes were expressed as ratios of geometric means with Hedges’ g confidence intervals, and robustness was probed with equal-variance t-tests, Mann-Whitney U tests, Monte Carlo permutation tests with 50,000 label permutations, and leave-one-out analyses that removed each animal in turn. Notably, the team explicitly declined to use PLS-DA, OPLS-DA, VIP-based feature selection or pathway enrichment as inferential evidence, positioning the study as a model of statistical restraint in a field often criticised for overinterpretation.

The first headline result concerned sex. In the design-compatible giraffe comparison, restricted to 2017 Rooipoort serum from 10 females and six males, no metabolite differed significantly even before correction for multiple testing; the smallest p value, for 3-hydroxyisobutyric acid, was just 0.070. The authors caution that this null result should not be read as proof that the sexes are metabolically equivalent, given the small sample and the absence of a predefined equivalence margin, but it aligns with growing evidence that sex effects are not universal across wildlife metabolomes.

The rhinoceros sex comparison told a more nuanced story. Nine metabolites met the primary false-discovery criterion, with females showing higher N-acetylglucosamine, creatinine, alanine, cholic acid, glycocholic acid, acetoacetic acid, isoleucine and valine, and lower allantoin than males. Yet only N-acetylglucosamine survived every sensitivity analysis, remaining significant under Welch, permutation and rank-based testing and in all 45 leave-one-out runs. Alanine and creatinine showed moderate support, while the remaining six findings were threshold-sensitive, hovering near the significance boundary. The result is a textbook demonstration of why multiplicity control and transparent robustness categories matter: without them, six fragile findings might have been reported alongside one genuinely solid one.

The interspecies comparison produced the study’s most dramatic pattern. Comparing EDTA plasma from six verified 2018 giraffes with all 45 rhinoceroses, 36 of 50 directly comparable metabolites differed significantly, and 28 formed a conservative core supported by every statistical test and every leave-one-giraffe-out analysis. Giraffe plasma contained markedly more pyruvic acid, glycine, allantoin and trimethylamine, and far less methanol, N-acetylglucosamine, acetic acid, 3-hydroxyisovaleric acid and methylamine. Most intriguingly, a coherent pattern emerged among branched-chain amino-acid-related compounds: giraffes showed lower leucine, 2-oxoisocaproic acid, 3-methyl-2-oxovaleric acid, 3-hydroxyisobutyric acid and 3-hydroxyisovaleric acid, consistent with a genuine difference in circulating branched-chain amino-acid chemistry between the species. Lower acetate and propionate in giraffes is biologically compatible with the contrast between foregut and hindgut fermentation, though the study did not measure feed intake or fermentation products directly.

The authors are admirably candid about the limits of their data. Species was perfectly confounded with NMR acquisition batch, since giraffe spectra were acquired in 2018 and rhinoceros spectra in 2021 with no shared quality-control material, so not every difference can be attributed to biology. Collection year was completely confounded with serum versus plasma in the giraffe dataset, precluding any inferential temporal comparison. The six-animal giraffe plasma group, while shown by leave-one-out analysis not to be driven by a single outlier, is small, and the concentrations are offered as exploratory baseline data rather than formal veterinary reference intervals, which would require dedicated reference-interval methodology. Adjustment for sex did not materially change the conservative core, ruling out the different sex ratios as an explanation for the broad pattern.

Why does this matter beyond the curiosity factor? Blood metabolomics is increasingly used in conservation medicine to detect disease, nutritional stress and environmental exposure in endangered species, from iron storage disease in Sumatran rhinoceroses to mitochondrial dysfunction in black rhinoceroses. Reliable baseline data from healthy animals are the foundation on which such diagnostic applications must be built, and they are scarce for wildlife. This study provides carefully curated concentration data for two keystone African herbivores, along with a methodological template, predefined contrasts, multiplicity correction, permutation testing and leave-one-out stability checks, that future wildlife metabolomics studies would do well to follow. The authors call for prospective work sampling both species with the same specimen matrix, balanced sexes, contemporaneous acquisition and shared quality controls. Until then, the branched-chain amino-acid signature separating the world’s tallest ruminant from its second-largest land mammal stands as an intriguing candidate for targeted investigation into how gut architecture writes itself into the blood.

Subject of Research: Comparative 1H-NMR blood metabolite profiling of giraffes and white rhinoceroses

Article Title: Comparative 1H-NMR profiling of plasma and serum metabolites in giraffes and white rhinoceroses

Article References: Osthoff, G., Mason, S., Schmidt, L., Tordiffe, A., & Deacon, F. (2026). Comparative 1H-NMR profiling of plasma and serum metabolites in giraffes and white rhinoceroses. Metabolomics, 22(5), Article 159. https://doi.org/10.1007/s11306-026-02535-0

Image Credits: AI Generated

DOI: 10.1007/s11306-026-02535-0

Keywords: metabolomics, giraffe, white rhinoceros, NMR spectroscopy, blood plasma, serum, branched-chain amino acids, wildlife physiology, conservation medicine, foregut fermentation, hindgut fermentation, South Africa