Understory and small trees contribute importantly to stemflow of a lower montane cloud forest

Stemflow (Sf) measurements in tropical rain and montane forests dominated by large trees rarely include the understory and small trees. In this study, contributions of lower (1- to 2-m height) and upper (>2-m height and <5-cm diameter at breast height [DBH]) woody understory, small trees (5 < DBH < 10 cm), and canopy trees (>10-cm DBH) to Sf per unit ground area (Sf(a)) of a Mexican lower montane cloud forest were quantified for 32 days with rainfall (P) during the 2014 wet season. Rainfall, stemflow yield (Sf(y)), vegetation height, density, and basal area were measured. Subsequently, stemflow funneling ratios (SFRs) were calculated, and three common methods to scale up Sf(y) from individual trees to the stand level (tree-Sf(y) correlation, P-Sf(y) correlation, and mean-Sf(y) extrapolation) were used to calculate Sf(a). Understory woody plants, small trees, and upper canopy trees represented 96%, 2%, and 2%, respectively, of the total density. Upper canopy trees had the lowest SFRs (1.6 +/- 0.5 Standard Error (SE) on average), although the lower understory had the highest (36.1 +/- 6.4). Small trees and upper understory presented similar SFRs (22.9 +/- 5.4 and 20.2 +/- 3.9, respectively). Different Sf scaling methods generally yielded similar results. Overall Sf(a) during the study period was 22.7 mm (4.5% of rainfall), to which the understory contributed 70.1% (15.9 mm), small trees 10.6%(2.4mm), and upper canopy trees 19.3%(4.4mm). Our results strongly suggest that for humid tropical forests with dense understory of woody plants and small trees, Sf of these groups should be measured to avoid an underestimation of overall Sf at the stand level.