Soil carbon isotopic proxies for determining the photosynthetic pathway of floral communities: a method inter-comparison

Abstract

C3 and C4 plants have a unique range of carbon isotope values as a result of their distinct photosynthetic processes. These differences are retained within leaf wax n-alkanes and bulk soil organic matter after entering the soil. As a result, soil organic matter and n-alkanes are used as isotopic tools for determining the proportion of C4 versus C3 vegetation cover for palaeoenvironmental reconstruction. However, unlike n-alkanes, soil organic matter is susceptible to isotopic enrichment due to decomposition, which affects the δ13C values. The validity of these two methods as accurate estimations for proportional C4 vegetation cover is yet to be tested in a comparative setting. In this study, these two methods were compared with ground vegetation surveys to determine whether they can be used interchangeably for palaeoenvironmental reconstruction. Surface soil samples were collected from 20 plots along a North to South transect through central Australia. These samples were analysed to determine their carbon isotopic composition, which was used to estimate the proportional C4 cover at each locality. I hypothesised that both soil organic matter and n-alkanes would accurately reflect spatial trends seen in the C3 and C4 community. I also hypothesised that high proportional C4 cover would correlate with high grass cover and low proportional C4 cover would correlate with high tree cover. However, proportional C4 cover would produce different relationships with climate variables depending on the method. Soil organic matter and n-alkane-derived proportional C4 cover were positively correlated with the vegetation survey data. Therefore, all three approaches provided the same general geographical trend. All three methods also produced similar relationships with proportional grass and tree cover and climate variables. These results demonstrate that either isotopic approach can be used to document large-scale geographic trends in proportional C4 cover without concern for variance associated with a particular method.