Laboratory and in situ measurements of soil methane consumption in a moist forest area of central Panama indicate that the conversion of forests to agricultural lands diminishes the soil sink for atmospheric methane. Rates of microbial methane consumption in agricultural soils were one fourth those of undisturbed forest soils. This reduction in soil methane consumption may partially account for past and future increases in atmospheric methane concentrations. INTRODUCTION Atmospheric methane concentrations have been increasing at a rate of approximately 1% per year for at least 30 years and have more than doubled since the 1700s [Craig and Chou, 1982; Rasmussen and Khalil, 1984; Rinsland et al., 1985; Stephens, 1985; Pearman et al., 1986; Blake and Rowland, 1988]. The cause of the increase in methane concentration is a matter of societal concern becausemethane is a greenhouse gas. Among the biologically active greenhouse gases, methane is second only to carbon dioxide in its potential effect on global temperature [Dickinson and Cicerone, 1986]. Explanations for the rise in methane concentration include increasing methane sources such as rice paddies, cattle, termites, landfills, and biomass burning [Bolle et al., 1986; Cicerone and Oremland, 1988]. Alternatively, it has been suggested that decreasing the major methane sink, atmospheric reaction with hydroxyl radical, accounts for a part of the concentration increase [Khalil and Rasmussen, 1985]. An additional sink for atmospheric methane is microbial consumption of methane in soils. Estimates indicate that this process presently consumes as much as 5 to 10% of the annual methane emission, but little is known of how this sink may have changed over time [Seiler, 1984; Bolle et al., 1986; Cicerone and Oremland, 1988]. Global estimates of the oxidation of atmospheric methane by soils are based on few field data. Most of the existing published measurements have been made in relatively pristine forest and savanna habitats [Harriss et al., 1982; Keller et al., 1983, 1986; Seller et al., 1984]. We studied

Abstract. Nitrous oxide production was measured in intact cores taken from active pasture and old-growth forest Inceptisols in the Atlantic Lowlands of Costa Rica. Following additions of aqueous KNO, or glucose, or the two combined amendments, the cores were incubated in the laboratory to determine if N,O production rates were either N-limited or C-limited in the two land use types. Differences in rates of denitrification (N,O + N, production) among amended forest and pasture soils were determined by addition of 10% GH,.The forest soils were relatively insensitive to all amendment additions, including the acetylene block. Forest N,O production rates among the treatments did not differ from the controls, and were consistently lower than those of'the pasture soils. With the addition of glucose plus nitrate to the forest soils, production of N,O was three times greater than the controls, although this increase was not statistically significant. On the other hand, the pasture soils were definitely nitrogen-limited since N,O production rates were increased substantially beyond controls by all the amendments which contained nitrate, despite the very low N level (5 mg N kg-' soil) relative to typical fertilizer applications. With respect to the nitrate plus glucose plus acetylene treatment, denitrification was high in the pasture soils; N,O production in the presence of C,H, was 150% of the rate of N,O production measured in the absence of the acetylene block. The results are discussed in relation to the effects of agricultural land use practices and subsequent impacts of disturbance on N,O retease.