<p>Trees can exchange methane (CH4) with the atmosphere through their stems. However, the magnitudes, patterns, drivers and origin of these emissions as well as the biogeochemical pathways that might result in net CH4 production or uptake are still poorly understood. One of the most important constraints is the limited information on the spatial and temporal variability of these emissions. Manual measurements are useful for measuring spatial variability of stem emissions (both within and between trees), but their low temporal frequency hinders our understanding of temporal patterns. In contrast, high-frequency measurements capture temporal variability, but instrumentation cost and complex technical logistics preclude high number of spatial replicates. In this study we combined manual and automated measurements of tree stem emissions in 18 different bitternut hickory trees (Carya cordiformis) in an upland forest during one growing season. Methane emissions were measured at two stem heights (75 and 150 cm) in three trees every 30 min, whereas the other 15 trees were measured once every two weeks at three different stem heights (50, 110 and 170 cm). Additionally, sap flow, soil temperature, soil water content, ground water level, and CH4 concentrations in the heartwood and in the soil profile were measured. Finally, we performed incubations of stem cores to test its potential for producing CH4. All trees were net sources of methane during the experiment, but some of them showed sporadic capture of CH4. High-frequency measurements revealed large temporal variability of stem emissions even within hours. Trees showed a seasonal trend of CH4 emissions partially explained by sap flow, soil moisture and temperature, but the pattern and the magnitudes were not consistent between and within trees. Even when a larger number of trees were studied (15 trees with manual measurements every two weeks), no consistent spatial pattern emerged among trees or with stem height, with emissions differing up to two orders of magnitude among trees. We found high CH4 concentrations in the heartwood of the trees (up to 75,000 ppm), no relevant concentrations in the soil profile (<6 ppm in all cases), and methanogenic capacity in all trees (stem cores were able to produce CH4 in laboratory incubations), supporting the interpretation that CH4 emitted by treestems was likely produced in the heartwood of the trees rather than being produced in soils and transported by the roots. Our results provide evidence on the potential origin of CH4 emitted by tree stems, but also indicate that the spatial and temporal patterns of stem emissions should be better described in order to assess the role of trees in local-to-global CH4 budgets.</p>
How to best address spatial and temporal variability of soil-derived nitrous oxide and methane emissions
