Ecosystem functional responses such as soil CO 2 emissions are constrained by microclimate, available carbon (C) substrates and their effects upon microbial activity. In tropical forests, phosphorus (P) is often considered as a limiting factor for plant growth, but it is still not clear whether P constrains microbial CO 2 emissions from soils. In this study, we incubated seven tropical forest soils from Brazil and Puerto Rico with different nutrient addition treatments (no addition, Control; C, nitrogen (N) or P addition only; and combined C, N and P addition (CNP)). Cumulative soil CO 2 emissions were fit with a Gompertz model to estimate potential maximum cumulative soil CO 2 emission (C m ) and the rate of change of soil C decomposition (k). Quantitative polymerase chain reaction (qPCR) was conducted to quantify microbial biomass as bacteria and fungi. Results showed that P addition alone or in combination with C and N enhanced C m , whereas N addition usually reduced C m , and neither N nor P affected microbial biomass. Additions of CNP enhanced k, increased microbial abundances and altered fungal to bacterial ratios towards higher fungal abundance. Additions of CNP, however, tended to reduce C m for most soils when compared to C additions alone, suggesting that microbial growth associated with nutrient additions may have occurred at the expense of C decomposition. Overall, this study demonstrates that soil CO 2 emission is more limited by P than N in tropical forest soils and those effects were stronger in soils low in P.
Highlights• A laboratory incubation study was conducted with nitrogen, phosphorus or carbon addition to tropical forest soils. Soil CO 2 emission was fitted with a Gompertz model and soil microbial abundance was quantified using qPCR. Phosphorus addition increased model parameters C m and soil CO 2 emission, particularly in the Puerto Rico soils. Soil CO 2 emission was more limited by phosphorus than nitrogen in tropical forest soils.