Coping with spectral interferences when measuring water stable isotopes of vegetables

Abstract Rationale Laser‐based analyzers are widely used in ecohydrology to analyze plant water isotopic compositions (δ 18 O and δ 2 H). The suitability of three different water extraction and isotope equilibration techniques was compared. We examined whether co‐extracted volatile organic compounds (VOCs) affect laser‐based isotope measurements and used the instrument's spectral parameters to post‐correct for interfering VOCs. Methods Cryogenic vacuum extraction, vapor headspace equilibration in bags, and vapor equilibration in situ probes were used to extract liquid water or water vapor for laser‐based isotope analysis (cavity ring‐down spectrometry, CRDS). Isotope data were calibrated by standards for each method separately. Spectral parameters of the instrument, appropriate to identify spectral interferences with MeOH and CH 4 , were identified and used for post‐correction. Differences between the three methods and between the origins of the vegetables were identified by statistical tests. Results VOCs were found in various amounts for the three different methods. They were co‐extracted or co‐equilibrated during the different extraction or equilibration methods. Correlation coefficients of isotope data and “CH 4 ” (spectral parameter) were 0.99 or better; however, slopes for δ 18 O were similar on different instrument types but different for δ 2 H. Our correction approach improved results and inter‐comparability of the methods considerably without knowing the chemical composition of the plant sap. Conclusions All three methods were sensitive enough to distinguish and resolve differences in natural abundance. Data quality was improved by the “CH 4 correction” approach but could probably be optimized by a plant species–specific correction. Standardized tools for contaminant removal or post‐correction applications from manufacturers, in particular for vapor‐mode analysis, are still needed.