Recent studies provide compelling evidence that ecosystem nutrient cycling and soil carbon (C) storage differ according to tree associations with arbuscular versus ectomycorrhizal fungi. Our objective was to investigate if mycorrhizal association mediates the response of soil C and nitrogen (N) cycling to N fertilization in a diverse tropical forest. We utilized a fertilization experiment initiated in 2007 in a lower montane tropical forest in Fortuna, Panama. In each control or N-fertilized plot, we sampled surface (0-10 cm) soil underneath two tree species associated with AM fungi, Micropholis melinonia and Eschweilera panamensis, and one species associated with EM fungi, Oreomunnea mexicana. We performed extracellular hydrolytic enzyme activity assays, for β-glucosidase (BG), x-xylosidase (XYL), N-acetylglucosaminidase (NAG), leucine aminopeptidase (LAP), and acid phosphatase (AP). We also separated the soil samples into the free light fraction (fLF), occluded light fraction (oLF), and heavy fraction (HF). We analyzed the elemental composition of the bulk soil and each fraction, and used bulk density measurements to estimate soil organic C (OC) content. We also measured live and dead fine root biomass. We separately performed an in situ 15N tracer experiment to measure gross rates of N mineralization, nitrification, and dissimilatory nitrate (NO3-) reduction to ammonium (DNRA).
Results/Conclusions
Fertilization marginally increased total soil OC content under all species (p=0.13). Total soil OC content did not differ by species, but O. mexicana had more OC in the fLF and less OC in the HF compared to the two AM species (p<0.05 for both). Live fine root biomass did not differ by treatment or species, but dead fine root biomass decreased with fertilization (p=0.05). BG, XYL, and NAG activities were greatest for O. mexicana (p<0.05), but this was likely driven by greater labile C availability under O. mexicana because specific activities (per g fLF-OC) did not differ by species. Only AP activity responded to fertilization (p=0.07) and with a decline. O. mexicana had the greatest gross N mineralization rates, but gross nitrification did not differ by treatment or species. DNRA rates were greatest under E. panamensis (p=0.01) and also increased with fertilization (p=0.02); this tracked differences in soil NO3-, which were greatest for E. panamensis (p<0.001) and increased with fertilization (p=0.004). Our results remarkably showed distinct tree species effects on soil C and N pools and fluxes in a diverse tropical forest. However, responses to N fertilization were similar among all species, suggesting that mycorrhizal association did not mediate them.