Detailed assessment of isotope ratio infrared spectroscopy and isotope ratio mass spectrometry for the stable isotope analysis of plant and soil waters

As an alternative to isotope ratio mass spectrometry (IRMS), the isotope ratio infrared spectroscopy (IRIS) approach has the advantage of low cost, continuous measurement and the capacity for field-based application for the analysis of the stable isotopes of water. Recent studies have indicated that there are potential issues of organic contamination of the spectral signal in the IRIS method, resulting in incorrect results for leaf samples. To gain a more thorough understanding of the effects of sample type (e. g., leaf, root, stem and soil), sample species, sampling time and climatic condition (dry vs. wet) on water isotope estimates using IRIS, we collected soil samples and plant components from a number of major species at a fine temporal resolution (every 2 h for 24-48 h) across three locations with different climatic conditions in the Heihe River Basin, China. The hydrogen and oxygen isotopic compositions of the extracted water from these samples were measured using both an IRMS and an IRIS instrument. The results show that the mean discrepancies between the IRMS and IRIS approaches for delta O-18 and delta D, respectively, were: -5.6 parts per thousand and -75.7 parts per thousand for leaf water; -4.0 parts per thousand and -23.3 parts per thousand for stem water; -3.4 parts per thousand and -28.2 parts per thousand for root water; -0.5 parts per thousand and -6.7 parts per thousand for xylem water; -0.06 parts per thousand and -0.3 parts per thousand for xylem flow; and -0.1 parts per thousand and 0.3 parts per thousand for soil water. The order of the discrepancy was: leaf > stem approximate to root > xylem > xylem flow approximate to soil. In general, species of the same functional types (e. g., woody vs. herbaceous) within similar habitats showed similar deviations. For different functional types, the differences were large. Sampling at nighttime did not remove the observed deviations. Copyright (C) 2011 John Wiley & Sons, Ltd.