Greenhouse Gas Emissions From Slash-And-Burn And Alternative Land Uses At The Benchmark Sites

“Best bet” Land-use Systems

Thematic reports

Carbon Sequestration And Trace Gas Emissions

 

Unique id: IDANQXXB

Source file: D:\Projects\ASB\ASB Country and Thematic reports\Climate Change WG Report\phase2final999.xml

 

Authors: C. A. Palm, P. L. Woomer, J. Alegre, L. Arevalo, C. Castilla, D. G. Cordeiro, B. Feigl, K. Hairiah, J. Kotto-Same, R. Lasco, , A. Mendes, A. Moukam, D. Murdiyarso, R. Njomgang, W. J. Parton, A. Ricse, V. Rodrigues, S. M. Sitompul, M. van Noordwijk

 

--------------------------------------------------------------------------

Deforestation and subsequent land use also result in emissions of methane and nitrous oxide, two other greenhouse gases.  Methane is the second most important greenhouse gas in terms of amounts and effect in the atmosphere.  Most upland, well-drained soils serve as a net sink of methane through consumption of methane by methanotrophic microorganisms in the soil.  There is increasing evidence that the size of this sink diminishes with increasing land-use intensity. Conversion to pastures in the humid tropics can actually result in a net source of methane from the soil (Steudler et al., 1996; Keller et al., 1997), through the process of methanogenesis. Factors that affect methane consumption or production in soils include bulk density, water-filled pore space, and nitrogen fertilization.  Currently there are insufficient data from the tropics that provide a mechanistic understanding and prediction of the net flux of methane from soils with changes in land use.

 

Tropical forest soils are also reputed to be a major source of nitrous oxide (Keller et al., 1997).  Nitrous oxide emissions can result from the processes of nitrification and denitrification (Firestone and Davidson, 1989) and are affected by N fertilization, land conversion, soil compaction and water logging.  Early data from tropical land-use conversion indicated a large flux of nitrous oxide from areas converted to pastures (Luizao et al, 1989). More recent information, however, suggests that this increase in flux is temporary, and the nitrous oxide fluxes may eventually be less than that of the nearby undisturbed forest (Keller and Reiners, 1993; Erickson and Keller, 1997).  Nitrogen fertilization seems to be the largest management factor affecting emissions (Davidson, et al., 1996; Erickson and Keller, 1997). Most nitrous oxide measurements in the tropics have been taken from undisturbed forest systems or pastures; few measurements have been taken from areas converted to alternative land uses.  A goal of the ASB work during Phase 2 was to sample and compare trace gas fluxes from the natural forests and alternative land uses in the benchmark areas and to identify factors influencing these fluxes.

 

2.1 Results of rainy season-dry season greenhouse gas monitoring at the benchmark sites.

Progress in measuring gas emissions was severely hampered by the lapse in ASB funding in 1995-6.  During that period, however, considerable time and effort were allocated to devise gas sampling, storage and analysis methods that would assure standardization among the sites.  The protocol revisions are in accordance with consultations held with experts in the field of trace gas emissions who are associated with the global change research community through the IPCC (International Panel on Climate Change) and GCTE (Global Change in Terrestrial Ecosystems of IGBP) programs.  This association of ASB with the global change community is essential if ASB results are to have impact at the larger scale and be incorporated in global change models.

 

Measurements of CO2, N2O, and CH4 fluxes were made in Brazil, Cameroon, Indonesia, and Perú (work in Perú was funded by ASB-DANIDA) during Phase 2.  The measurements were taken in the same locations as those for carbon stocks and biodiversity.  Measurements were taken during the rainy season and dry season in Indonesia and Perú but only once, during the rainy season, in Brazil and Cameroon.

 

The data obtained at the different benchmark areas proved to be extremely variable.  In fact, there were no significant differences at any particular site among the different land-use systems, including the natural forests, in terms of nitrous oxide flux or methane consumption. (The data for each of the sites can be found in the ASB country reports).  In addition, it is difficult to draw any conclusions from cross-site comparisons, which indicate that the fluxes depended, perhaps, more on soil and climatic conditions of the different sites than on land-use management.  For example, N2O fluxes from the different systems in Perú and Cameroon averaged 57 to 80 ug N m-2 hr-1, respectively, and was an order of magnitude higher than the average for Indonesia. Conversely, the methane consumption rates in Indonesia averaged 22 ug Cm-2 hr-1,almost four times that measured for Perú and Cameroon. The data from Brazil were obtained during an extremely wet period in which the soils were essentially saturated at all sites, and the data were affected by those conditions.  There was a net methane flux from most of the system and nitrous oxide fluxes were two orders of magnitude higher than observed in all other sites.  We caution that these data should not bias interpretations because of the extraordinary climatic conditions during sampling.

 

2.2 Results from intensive monthly sampling of greenhouse gas fluxes

Following the first year of monitoring, a conclusion was drawn that reliable estimates or comparisons of gas fluxes cannot be made from only one or two time measurements due to the number and variability of factors affecting the gas fluxes.  A decision was made to sample more intensively in a few, well-characterized locations.  Two sites, Indonesia and Perú, were chosen for intensive monthly sampling.  These sites were selected because of well-defined land-uses and the capacity of the laboratories at the sites to monitor the soil variables that affect gas fluxes.  Similar land-use categories are being monitored in the two benchmark areas, representing forests, complex agroforests, fallows, tree plantations, crops, and grasslands. Permanent bases have been established for each chamber so that measurements are taken from the same place each time, reducing the variability and also the disturbance caused by placing the chambers each time.  Ancillary soil measurements taken at each sampling time include extractable nitrate and ammonium, N mineralization, nitrification, bulk density, moisture content, texture, pH, and CO2 evolution.

 

Though the monthly samplings have not been completed for a year, the data collected for six months in an experiment in Yurimaguas, Perú are presented to show the trends. The experiment compares six different land-use systems.  Five of the treatments were established 14 years ago by slashing and burning a 12 year-old forest fallow. Those treatments are 1) shifting cultivation, which entailed one year of  cropping and abandonment to forest fallow; the fallow is currently 13 years old; 2) high-input cropping with a maize-soybean rotation and tillage, fertilization, and liming; 3) low-input cropping with an upland rice-cowpea-mucuna rotation; 4) a multistrata agroforestry system; and 5) a peach palm plantation.  These treatments are all compared to 6) the original forest fallow that is currently 16 years old.  The monthly N2O fluxes range from 5 to 25 ug N m-2 hr-1 (Figure 6a).  Although there are a few months in which there are significant differences among treatments (the high-input cropping treatment is higher following fertilization) the overall mean, 12 ug N m-2 hr-1, for the six months does not indicate any significant differences among the treatments (Figure 6b).  These fluxes are 75 to 85% smaller than those measured in Perú in the initial monitoring using removable bases and are similar to those currently being measured in other areas on acid, infertile soils in the humid tropics.

 

Methane fluxes do, however, show differences among treatments, with the high-input cropping system producing a net efflux of methane to the atmosphere in four of the six months (Figure 7a).  Date for the other treatments indicates methane consumption, but the size of this sink decreases with increasing land-use intensity.  The largest sink, 25 to 30 ug C m-2 hr-1, is found in the two forest fallows and the multistrata agroforestry system, and the smallest, 20 ug C m-2 hr-1, is in monoculture peach palm and low-input cropping systems (Figure 7b).  These differences in CH4 flux are related primarily to increased bulk density and resulting water-filled pore space in the cropping systems and peach palm plantation.  These methane consumption rates are similar to those reported from the first year of monitoring in Indonesia.

 

Data obtained from the monthly measurements in Perú and Indonesia will be used to test a gas flux model for the tropics.  A new version of the CENTURY model (NGAS-CENTURY) has been developed to simulate trace gas production. The NGAS-CENTURY version uses a daily time step and can simulate daily CH4 consumption, and N2O, NOx and N2 gas fluxes. Once NGAS-CENTURY has been validated and tested for the benchmark sites, it can then be used to predict trace gas fluxes from a variety of environments and land management systems that cannot possible be accomplished through intensive field sampling due to the expense and time involved.

 

2.3 Net radiative forcing of greenhouse gas emissions from slash-and-burn and alternative land-use systems

The net effect, or net radiative forcing of the fluxes of the different greenhouse gas, CO2, N2O, and CH4, or the relative contribution of the individual gases can be compared based on the global warming potential (GWP) of the different gases. Various conversions factor are used for GWP; an example from Lal et al., (1998) is that the GWP is 1 for CO2, 21 for CH4 and 310 for N2O.  In other words, an equivalent weight of nitrous oxide has a much stronger effect on radiative forcing than the same weight of carbon dioxide or methane.  Based on the GWP of the different gases, a calculation from preliminary data in Indonesia compared the overall effect of release of CO2 from deforestation and the resorption of C from the vegetation of the replacement land-use system (based on the time-averaged C) and the release of N2O and sink of CH4 over a 25-year time course (Tomich et al., 1998).  They found that the initial loss of CO2 through deforestation is by far the largest factor in terms of net radiative forcing.  This may be different in intensive systems where N fertilizers are applied; we do not yet have estimates from such systems in ASB, but other studies (Davidson, et al., 1996; Erickson and Keller, 1997) would indicate this to be the case.