Simulated effects of temperature and precipitation change in several forest ecosystems

The Nutrient Cycling Model (NuCM) was used to investigate the effects of increased temperature (+4°C) and changing precipitation (increased and decreased) on biogeochemical cycling at six forest sites in the United States: a Picea rubens forest at Nolan Divide in the Great Smoky Mountains, North Carolina; mixed deciduous forests at Walker Branch, Tennessee and Coweeta, North Carolina; a Pinus taeda forest at Duke, North Carolina; a P. eliottii forest at Bradford, Florida; and a P. contorta/P. jeffreyii forest at Little Valley, Nevada. Simulations of increased temperature indicated increased evapotranspiration and reduced water flux. Simulations of changes in precipitation indicated disproportionately large variations in soil water flux because of the relative stability of evapotranspiration with changes in precipitation. Increased temperature caused N release from forest floors at all sites. At the N-saturated Nolan Divide site, this resulted in no change in N uptake or growth but increased soil solution A1 and NO₃/^- and increased N leaching losses. At the N-limited sites, the release of N from the forest floor caused increased growth, and, in some cases, increased NO₃/^- leaching as well, indicating that N released from the forest floor was not efficiently taken up by the vegetation. Increased precipitation caused increased growth, and decreased precipitation caused reduced growth in the N-limited sites because of changes in wet N deposition. Changes in precipitation had no effect on growth in the N-saturated Nolan Divide site, but did cause large changes in soil solution mineral acid anion and A1 concentrations. Increased precipitation caused long-term decreases in soil exchangeable base cations in most cases because of the disproportionately large effects on soil water flux; however, increased precipitation caused decreases in exchangeable base cations in cases where atmospheric deposition was a major source of base cations for the system. The simulation results illustrate the extreme complexity of the possible responses of nutrient cycling processes to climate change. By virtue of the fact that the NuCM model does not contain physiological algorithms, these simulations demonstrate that changes in temperature and precipitation can produce widely varying ecosystem-level responses through their effects on biogeochemical cycling processes alone and that generalizations about the relative importance of temperature versus precipitation changes are hazardous. (C) 2000 Elsevier Science B.V.