Large quantities of organic carbon are trapped in histosols which can be either a source or a sink of methane, nitrous oxide and carbon dioxide. Intrinsic soil characteristics, changes in soil use and management, in addition to climate factors, influence intensively in the flow of these gases, but artificial drainage is one of the practices that most affect their behavior. This research was carried out in order to evaluate the influence of drainage in the flow of methane (CH4), nitrous oxide (N2O) and carbon dioxide (CO2), considering variations in groundwater level, in the soil gravimetric moisture, as well as air temperature and rainfall. An experiment was carried out over a two year period in an Typic Haplosaprists within an experimental farm located in Ponta Grossa, PR, a town situated in the second plateau of that state. Under natural conditions, and with full hydromorphic, there was an emission of CH4, from 238 to 5365 μg of CH4 m-2 h-1. In the drained area, the removal of water has determined the consumption of liquid methane between 39 and 242 μg of CH4 m-2 h- 1 . Net consumption of N2O in soil under natural conditions varied from -0,13 to -31,2 μg of N2O m-2 h-1. However, in the drained area, the removal of water has determined a net emission in rates that may vary between 0,19 to 55,7 μg of N2O m-2 h-1. In drained soil the increase in the number of days after drainage has determined the increase of the emission of N2O. Under natural conditions, consumption of nitrous oxide was observed due to the excessive moisture and low values for redox potential. Under natural conditions emission rates of carbon dioxide ranged from 244 to 870 mg of CO2 m-2 h-1. In the drained area the same flow has varied from 238 to 1018 mg of CO2 m-2 h-1. Although there was no significant difference in the rate of carbon dioxide emission in both soil, the effect of accumulated drainage resulted in an increase of 17,9 Mg C- CO2 eq ha-1 yr-1 in the global warming potential overall, due to the loss of 37,8 Mg C from the stock of soil carbon, up to a meter deep, in just two years. The flow of methane was significantly affected by air temperature over time in natural histosols only. In histosol drained the temperature effect was secondary, once the decreased moisture was the factor that determined the consumption of CH4. Air temperature had a major effect on the flow of nitrous oxide only in natural histosol. On the other hand, rainfall had a notorious effect over the flow of N2O in drained histosol. Rainfall had no major effect on CH4 flow for both, natural and drained soil. In the first, the rainfall did not increased the high natural moisture of the soil and in the second, the rainfall water was rapidly removed due to the high soil permeability. In the case of CO2, air temperature and the rainfall intensively has influenced the flows in both soils. Although the drainage has transformed drained histosol in a methane sink, this soil started releasing nitrous oxi de and large quantities of carbon dioxide. Given that histosols located in hydrophilous grassland perform important functions for the ecosystem, such as acting as a carbon sink, retaining water and contributes in the decrease of the quantities of nitrate, could possibly change it into N2, so they should not be drained but effectively preserved. The capacity of histosol, under natural conditions, of trapping 581 Mg ha-1 of organic carbon and 8,6 million liter of water per hectare, at one meter depth, justifies entirely that they must not be drained.
