Search

With current observations and future projections of more intense and frequent droughts in the tropics, understanding the impact that extensive dry periods may have on tree and ecosystem-level transpiration and concurrent carbon uptake has become increasingly important. Here, we investigate paired soil and tree water extraction dynamics in an old-growth upland forest in central Amazonia during the 2018 dry season. Tree water use was assessed via radial patterns of sap flow in eight dominant canopy trees, each a different species with a range in diameter, height, and wood density. Paired multi-sensor soil moisture probes used to quantify volumetric water content dynamics and soil water extraction within the upper 100 cm were installed adjacent to six of those trees. To link depth-specific water extraction patterns to root distribution, fine root biomass was assessed through the soil profile to 235 cm. To scale tree water use to the plot level (stand transpiration), basal area was measured for all trees within a 5 m radius around each soil moisture probe. The sensitivity of tree transpiration to reduced precipitation varied by tree, with some increasing and some decreasing in water use during the dry period. Tree-level water use scaled with sapwood area, from 11 to 190 L per day. Stand level water use, based on multiple plots encompassing sap flow and adjacent trees, varied from ∼1.7 to 3.3 mm per day, increasing linearly with plot basal area. Soil water extraction was dependent on root biomass, which was dense at the surface (i.e., 45% in the upper 5 cm) and declined dramatically with depth. As the dry season progressed and the upper soil dried, soil water extraction shifted to deeper levels and model projections suggest that much of the water used during the month-long dry-down could be extracted from the upper 2–3 m. Results indicate variation in rates of soil water extraction across the research area and, temporally, through the soil profile. These results provide key information on whole-tree contributions to transpiration by canopy trees as water availability changes. In addition, information on simultaneous stand level dynamics of soil water extraction that can inform mechanistic models that project tropical forest response to drought. [ABSTRACT FROM AUTHOR]

• Fe accumulates in decomposing litter within skid trail ruts. • Fine root biomass higher in uppermost 5 cm of skid trail ruts than forest controls. • Higher base saturation and pH in skid trails when compared to control. • Elevated levels of exchangeable Mn found throughout skid trails. • Soil compaction persists in skid trails 28 years after logging in a clay Ferralsol. The use of heavy machinery in logging operations leads to soil compaction, especially in highly used skid trails. In Amazonia, recovery of topsoil physical properties has been reported in skid trails. So, the present study sought to determine if this was a full recovery by investigating additional attributes, such as soil chemical properties and fine root biomass (FRB). In addition, forest floor litter was investigated to present a more complete view of soil recovery dynamics. To accomplish this, litter above sampling locations was collected, as well as soil samples below this in two depths. These depths represented the recovered topsoil layer (0–5 cm) and the still compacted soil layer (5–10 cm) below. Three areas were compared: control, middle of the skid trails, and skid trail ruts. Dry mass, C%, N% were determined for litter and FRB. Foliar nutrients (P, K, Ca, Mg, Fe, Zn, Mn) were also determined. Soil chemical properties such as pH, cation exchange capacity, base saturation, aluminum saturation, soil organic C and nutrients (total N, available P, exchangeable Ca, Mg, K, Na, Fe, Zn, Mn) were also assessed. The primary objective was to determine if there was a full recovery of properties evaluated in the topsoil, but also to determine if the compacted soil below acted as an effective hardpan that potentially exerted influence on the recovered topsoil. In the skid trail ruts, results revealed no significant differences with the controls in the recovered topsoil for FRB and all but one soil nutrient, Mn which was elevated. Additionally, the base saturation and volumetric water content of the ruts topsoil was significantly higher than the controls indicating a potential hardpan condition below. So, even though a topsoil may be considered fully recovered, the compacted soil below may still exert influence in the form of increased soil moisture and nutrient accumulation. Therefore, the key finding of this research is that for the first time a recovered topsoil in skid trails was demonstrated to be affected and influenced by the compacted soil below. This influence in the very clayey soils of Amazonia was observed after nearly-three decades. [ABSTRACT FROM AUTHOR]