Background/Question/Methods Paradoxically, litter decomposition often is positively related to its initial concentration of nitrogen, yet externally supplied nitrogen often has little or even negative effects on decomposition rates. Using long-term resource additions to decomposing litter, I tested three alternative hypotheses to explain this apparent inconsistency. I hypothesized that 1) organic forms of nitrogen stimulate decomposition more than do inorganic forms, 2) readily available carbon stimulates decomposition more than does nitrogen, or 3) nutrients other than nitrogen (e.g., phosphorus) stimulate decomposition more than does nitrogen. To resolve the paradox, the latter two hypotheses require that the concentrations of labile carbon or other nutrients, respectively, commonly are positively related to nitrogen concentrations in litter. To distinguish among these hypotheses, I decomposed oak leaf litter and oak leaves of varying initial nitrogen concentrations in a resource addition experiment that was replicated in three forested stands at the Cedar Creek Ecosystem Science Reserve, Minnesota for four years. Resource addition treatments included externally supplied inorganic nitrogen, organic nitrogen, labile carbon, inorganic nitrogen plus labile carbon, and a mix of nutrients other than nitrogen. Unamended plots served as controls.

Results/Conclusions As expected, decomposition at all sites was strongly positively related to initial nitrogen concentrations, with green leaves decomposing twice as fast as senesced litter, and litter or leaves with higher nitrogen concentrations decomposing more quickly than litter or leaves with lower nitrogen concentrations, respectively. In contrast to my expectations, addition of both inorganic and organic forms of nitrogen depressed rates of decomposition relative to unamended controls plots. Addition of labile carbon had neutral effects on decomposition, while addition of nutrients other than nitrogen (phosphorus, potassium, calcium, plus micronutrients) reduced rates of decomposition to the lowest rates of all of the treatments. Treatment effects were consistent among sites. Taken together, the results suggest that none of the proposed hypotheses can explain the inconsistent effects of litter versus externally supplied nitrogen on litter decomposition, so the paradox remains unresolved.