Reconstructing herbivore diets: a multivariate statistical approach to interpreting amino acid nitrogen isotope values.

Recent studies have demonstrated that there exists significant variability in amino acid (AA) δ15N values of terrestrial plants, discriminating among plant types (i.e., legume seeds, grasses, tree leaves) as well as tissues of the same plant. For the first time, we investigate the potential of the spacing between the δ15N values of different AAs to differentiate between plant types and thus elucidate their relative importance in herbivore diet. Using principal component analysis, we show that it is possible to distinguish among five plant categories-cereal grains, rachis, legume seeds, herbaceous plants, and woody plants-whose consumption has different implications for understanding herbivore ecology and management practices. Furthermore, we were able to correctly classify the herbaceous plant diet of modern cattle using AA δ15N values of their tooth dentine adjusted for trophic enrichment. The AA δ15N patterns of wild and domestic herbivores from archaeological sites seem to be consistent with diets comprised predominantly of herbaceous plants, but there is variation in AA δ15N values among individuals that may reflect differing inputs of other plant types. The variation in AA δ15N values does not necessarily reflect the variation in herbivore bulk collagen δ13C and δ15N values, indicating that AA δ15N values have the potential to provide additional insights into plant dietary sources compared to bulk tissue isotope values alone. Future work should focus on defining trophic enrichment factors for a wider range of terrestrial herbivores and expanding libraries of primary producer AA δ15N values.

Determination of Arginine δ15N Values in Plant and Animal Proteins by Gas Chromatography-Combustion-Isotope Ratio Mass Spectrometry.

Nitrogen (N) stable isotope techniques are widely used in ecology, archaeology, and forensic science to explore trophic relationships and provenances of organisms and materials, most widely using bulk δ15N values of whole organisms, tissues, or other materials. However, compound-specific isotope values can provide more diagnostic isotope "fingerprints" and specific information about metabolic processes. Existing techniques for nitrogen isotope analysis allow the determination of δ15N values of 14 amino acids (AAs), accounting for ca. 75% of plant protein and collagen N. The majority of remaining N is from arginine, comprising 16 and 14% of collagen and plant protein N, respectively. We therefore aimed to develop a method to detect arginine and determine its δ15N value (δ15NArg) by gas chromatography-combustion-isotope ratio mass spectrometry (GC-C-IRMS), to further contribute to the understanding of the metabolic routing of this important AA. We demonstrate that arginine, as its N-acetyl isopropyl ester, is amenable to GC analysis using a 15 m midpolarity DB-35 column, eluting with baseline resolution from other AAs. The recorded δ15N value by GC-C-IRMS was within the error of that of the underivatized compound determined by elemental analyzer-isotope ratio mass spectrometry (EA-IRMS). The newly developed GC-C-IRMS method was applied to modern plant protein and cattle collagen, enabling their δ15NArg values to be related to AA biosynthesis. Determination of archaeological cattle collagen δ15NArg values confirmed the suitability of this method to provide further insights into past diets and ecosystems. Bulk collagen δ15N value reconstruction including δ15NArg values better reflect the measured bulk values, as the isotopic ratio of 91% of collagen N can now be determined at the compound-specific level.

Compound-specific δ15N values express differences in amino acid metabolism in plants of varying lignin content.

Amino acid δ15N values of foliage of various plant taxa, grown at the experimental farm stations of North Wyke, UK and Bad Lauchstädt, Germany were determined by GC-C-IRMS. The difference between δ15N values of glutamate (Glx) and phenylalanine (Phe) were found to differ significantly between woody and herbaceous plants, with mean Δ15NGlx-Phe (i.e. δ15NPhe - δ15NGlx) values of -9.3 ±â€¯1.6‰ and -5.8 ±â€¯2.1‰, respectively. These differences in values are hypothesised to be due to the involvement of Phe in the phenylpropanoid pathway, by which lignin and other phenolic secondary metabolites are produced, leading to isotopic fractionation and enrichment of the remaining Phe pool available for protein biosynthesis. This results in the more negative Δ15NGlx-Phe values observed in woody plants relative to herbaceous plants, as the former are assumed to produce more lignin. To test this assumption, plant leaf tissue lignin concentrations were estimated by solid state 13C cross-polarisation, magic-angle-spinning (CPMAS) NMR spectroscopy for a subset of plants, which showed that tree foliage has a higher concentration of lignin (12.6 wt%) than herbaceous foliage (6.3 wt%). The correlation of lignin concentration with Δ15NGlx-Phe values demonstrates that the difference in these values with plant type is indeed due to differential production of lignin. The ability to estimate the lignin content of plants from amino acid δ15N values will, to give one example, allow refinement of estimates of herbivore diet in present and past ecosystems, enabling more accurate environmental niche modelling.