Methane cycling in upland soils of the Peruvian Andes and Amazon

Significant discrepancies exist in global estimates of the atmospheric methane (CH4) budget. This is particularly true for tropical South America where bottom-up approaches, rooted in field observation, tend to under estimate atmospheric observations. As such, a better understanding of soil environments, which are capable of acting as both source and sink for atmospheric CH4, is required. Soil-atmosphere CH4 exchange is fundamentally determined by the balance between strictly anaerobic methanogenic and aerobic methanotrophic microbial processes. For this reason, CH4 emissions are typically associated with anoxic wetland soils, whilst, oxic upland soils are thought to uptake CH4 from the atmosphere. However, there is increasing evidence that upland soils may act as sources of CH4 through methanogenic activity within cryptic wetlands or anoxic microsites. This thesis aims to: document soil-atmosphere CH4 fluxes in poorly represented tropical upland and montane ecosystems, investigate controls on CH4 flux with a focus on soil oxygen (O2) concentration and investigate relationships between methanogenic and methanotrophic processes under oxic conditions. These aims are addressed in three chapters focusing on lowland terra firme, premontane and montane forests and montane humid puna grasslands and wetlands along an Amazonian to Andean transect spanning ~ 3300 m of elevation in southeastern Peru. In the lowland rainforest intensive seasonal field campaigns and laboratory incubations were conducted on higher porosity ultisol and lower porosity inceptisol soils. Mean (s.e.) net CH4 fluxes for dry and wet seasons were, respectively, -1.59 (0.06) and - 1.39 (0.07) mg CH4-C m−2 d−1 for the ultisol and -0.95 (0.06) and -0.41 (0.10) mg CH4-C m−2 d−1 for the inceptisol. Greater uptake rates in the ultisol than the inceptisol were best explained by lower water-filled pore space (WFPS). Similarly, WFPS best explained between season variation in net CH4 flux from the inceptisol, whilst, we were unable to explain the smaller variations observed for the ultisol. Methanogenic processes were active in both the ultisol and inceptisol soils despite oxic conditions. In the premontane and montane forests, long-term monthly field measurements were conducted over two and a half years in premontane, lower montane and upper montane settings. Mean (s.e.) net CH4 fluxes for aggregated dry and wet season months were, respectively, -0.20 (0.15) and -0.08 (0.13) mg CH4-C m−2 d−1 for the premontane forest, -1.12 (0.13) and -0.97 (0.11) mg CH4-C m−2 d−1 for the lower montane forest and -1.55 (0.13) and -1.04 (0.11) mg CH4-C m−2 d−1 for the upper montane forest. Increased uptake with elevation was best explained by decreases in WFPS. Significant variation in net CH4 flux between seasons, driven by variation in WFPS, was only identified for the upper montane forest.