Abstract.The memory timescale that characterizes root-zone soil mois-

Stemflow (SF) has been recognized as an important process delivering water to 1 spatially localized areas of the forest floor. Based on almost six years of daily SF data from 2 nine teak trees in a Thai plantation, together with detailed above-canopy meteorological data 3 and daily leaf area index (LAI) data, this study sought to better understand how specific biotic 4 and abiotic factors affect both individual and stand-scale SF production from teak trees. 5 Specifically, the factors affecting stemflow funneling ratio (SFFR) at individual and stand 6 scales were analyzed and compared by means of boosted regression tree (BRT) analysis. The 7 largest variation of SFFR among individual teak trees was observed in the leafed state. For 8 trees taller than average, the BRT analysis revealed that vapor pressure deficit during rain was 9 the most influential factor affecting SFFR. Vapor pressure deficit had a negative influence on 10 SFFR, implying significant control of evaporative demand during rain. In contrast, for trees 11 shorter than the average, rain duration (R D ) was the most influential variable, having a positive 12 correlation with SFFR. The stand-scale BRT analysis for all nine teak trees demonstrated that 13 R D was the most influential factor affecting SFFR (exerting positive influence) but that an 14 array of other factors (both abiotic and biotic) played intricate and complex roles in governing 15 SFFR. The effect of LAI on SFFR was complicated and varied greatly among individual teak 16 trees. It is possible that spatially heterogeneous flowpaths of intercepted water inside the teak 17 canopy, which could be a product of the large-sized mature leaves of teak, may account for the 18 tree-to-tree variation in the responses of SFFR to changing LAI. Although our study focused 19 on teak trees, the demonstrated physically-based mechanistic explanations of stemflow 20 production could apply to other even-aged deciduous forests and monospecific plantations. 21 phenophase 23 2 65affect teak stemflow generation is necessary in order to draw a more general conclusion of the 66 characteristics of teak stemflow. 67It is well known that deciduous tree species produce larger stemflow quantities in the leafless 68 state than the leafed state (Helvey and Patric, 1965; Dolman, 1987; Neal et al., 1993; Park et 69 al., 2000; Herbst et al., 2008; Staelens et al., 2008; Siegert and Levia, 2014). Exceedingly little 70 work, however, has gone beyond the dichotomous comparison of stemflow production 71 4 between leafed and leafless states. In fact, to our knowledge, there has been no study 72 examining stemflow of deciduous tree species in relation to a detailed daily time series of leaf 73 area index (LAI, m 2 m -2 ). Given the importance of the presence and absence of leaves on 74 stemflow production, a more detailed analysis on any effects of daily LAI changes on 75 stemflow production is warranted. This is especially true for deciduous trees during the leafing 76 phenophase and trees species with large leaves, such ...

No abstract

Temporal variability in root zone soil moisture content (w) exhibits a Lorentzian spectrum with memory dictated by a damping term when forced with white‐noise precipitation. In the context of regional dimming, radiation and precipitation variability are needed to reproduce w trends prompting interest in how the w memory is altered by radiative forcing. A hierarchy of models that sequentially introduce the spectrum of precipitation, net radiation, and the effect of w on evaporative and drainage losses was used to analyze the spectrum of w at subtropical and temperate forested sites. Reproducing the w spectra at long time scales necessitated simultaneous precipitation and net radiation measurements depending on site conditions. The w memory inferred from observed w spectra was 25–38 days, larger than that determined from maximum wet evapotranspiration and field capacity. The w memory can be reasonably inferred from the Lorentzian spectrum when precipitation and evapotranspiration are in phase.

No abstract