Fig. 1. Regression plots of methane (CH4) fluxes against tree stem sampling position above the forest floor.
Top left, fluxes measured from Cuniã, Brazil, from two free-draining forested plots (20 × 30 m2) in the catchment draining to the Madeira River (n = 100) during 15–25 March 2013; top right, for mature trees stems of two common tree species: Heisteria concinna (orange) and Simarouba amara (yellow) in experimental litter manipulation treatments in a tropical forest on free-draining soil in Panama, Central America, between 18 and 27 November 2015 (24 trees); bottom right, ash and sycamore in a temperate deciduous woodland on free-draining soil in Wytham Woods, United Kingdom, between October 2015 and January 2016 (24 trees); bottom left, measurements from Scots pine and Norway spruce (18 trees) in the Skogaryd Research Catchment, near Vänersborg in southwestern Sweden. Grey bands denote 95% confidence intervals, which for Cuniã (large number of trees) falls in the regression line area. In Cuniã (Brazil) and Gigante (Panama) upland forests, CH4 uptake fluxes were largest at the highest sampling points at about 1.75–2 m above the forest floor (−56.5 ± 59.4 µg of CH4 m−2 h−1 and −46.7 ± 64.7 µg of CH4 m−2 h−1, respectively). For figure clarity, some of the largest values are not included, but the highest individual measured uptake values were −290 µg of CH4 m−2 h−1 in Cuniã and −488 µg of CH4 m−2 h−1 in Gigante. At Wytham, we found mean uptake at the highest sampling position (2 m) of −18.5 ± 20.4 µg of CH4 m−2 h−1 for ash and −14.0 ± 27.4 µg of CH4 m−2 h−1 for sycamore, with the highest recorded individual uptake fluxes being −54.8 µg of CH4 m−2 h−1 on ash and −142 µg of CH4 m−2 h−1 on sycamore. General information relating to these sites is presented in Extended Data Table 5.