Soil carbon stocks in experimental mesocosms are dependent on the rate of labile carbon, nitrogen and phosphorus inputs to soils

The soil sink for carbon is important in regulating climate and soil fertility. The sink strength is dependent on the balance of soil carbon decomposition and formation. Variation in the rates of these processes to manipulations of resource availabilities under global change, such as elevated atmospheric carbon dioxide, is not explained by soil microbial theory. To investigate disparate responses of soil carbon dynamics in field investigations, to altered carbon, nitrogen and phosphorus availability, we couple fractionation, isotope and mesocosm techniques to quantify soil carbon decomposition and formation under different resource regimes. These regimes involve addition of multiple levels of carbon, nitrogen and phosphorus, alone and in combination. We hypothesize that: (i) there is no net effect of labile carbon input rate on soil carbon stocks because reductions in soil carbon decomposition are offset by reductions in soil carbon formation; (ii) with simultaneous nutrient addition soil carbon stocks will increase because nitrogen will inhibit further soil carbon decomposition, and mitigate reductions in soil carbon formation observed under elevated labile carbon availability alone; (iii) this increase in soil carbon stocks will be a product of greater formation and decreased decomposition of slower-cycling, mineral-associated, soil carbon, whereas less stable, particulate soil carbon will simply turnover faster (due to greater soil carbon decomposition and formation). In contrast to our predictions formation of soil carbon is positively correlated with labile carbon input rates. In addition, nutrient amendment does not interact with carbon amendment to affect total soil carbon contents. However, there are significant interactive effects when the formation and decomposition responses of different soil carbon fractions are considered. For nitrogen alone, its effects on soil carbon fractions follow our hypotheses. However, phosphorus amendment increases decomposition of the soil carbon fraction that constitutes a longer-term sink. Our results highlight the need for rhizodeposition, phosphorus and soil carbon fractions to be explicitly considered when interpreting potential soil organic carbon responses to altered resource availability. In the discussion, we make four recommendations for future investigations to improve our understanding of soil carbon responses to altered carbon, nitrogen and phosphorus availabilities.