DOC Under-litter and Litter leachate
The mean UL DOC concentration was 44.26 (± 15.99) mg.L-1 for the studied hillslope. The variation between the events was significant. However, there was a large reduction in the determination coefficients compared to that observed under-canopy (TF and CL), with a total loss of model significances related to the variables on antecedent humidity and characteristics of the rainfall events (table 4). The 10-days AcR was the variable that most approached the significance of the linear model for UL, maintaining the inverse proportionality trend.
About UL data, there is an even bigger reduction in the coefficients of determination for the variables of antecedent humidity and characteristics of rainfall events. Such observation indicates no influence of the dilution and pre-wash effects on DOC leaching from litter.
The proximity to a significant correlation observed between UL data and 10-days AcR should be residual, reflecting part of the correlation observed in TF, once the TF concentrations are start-values for the DOC through-litter increase. However, it is noteworthy that litter is the main DOC source in the forest leaching process and, thus, the UL field values show less anterior compartment dependence (TF DOC) than was observed in the BP-TF transition.
The comparison between different hillslope positions showed high significance on DOC spatial variations. The VBT presented the lowest mean DOC concentration for UL and LL for all events, except for LL at EV7 (figure 7). The SMT tended to intermediate concentrations and both mid-hillslope areas (UHS and LHS) achieved the highest concentrations. The similarity in the spatial variation between the results obtained at the different events reinforces the confidence in the average result obtained.
Analyzing the DOC results in view of the spatial variability of litter stocks, it is evident that both mid-hillslope areas (UHS and LHS), which with the highest DOC concentrations, have similar litter stocks between itself and intermediate litter mass compared to SMT and VBT. In VBT rainwater percolates thin stocks with little accumulated organic matter, resulting in lower DOC concentrations in the leachate. On the other hand, the thicker litter stock at SMT did not materialize in higher DOC concentrations, demonstrating some limiting factor to leaching compared to what occurs in mid-hillslope areas.
Since the areas (hillslope positions) produce similar litterfall, the difference between the litter stocks reflects differentiated decomposition rates. Therefore, DOC concentrations are related to decomposition rates analogously to that observed for litter stocks, with more intense leaching in areas with intermediate decomposition rates. This observation indicates the nonlinearity of the relationship between DOC concentrations with litter stocks and their decomposition rates, suggesting a curve trend (Second Degree Polynomial) that best demonstrates an “optimal” condition for DOC leaching in this environment under moderate stocks and decomposition rates (figure 8).
It’s important to state that we do not intend to expose a standard curve representing a probabilistic model, as it would require a larger range of positions to be significant. Our intention is only to emphasize the nonlinearity trend under the conditions of the analyzed environment, believing that such a general idea may contribute to analyzes in similar environments. The high coefficient of determination value found for the curve in DOC related to the litter stocks, besides being favored by the low number of areas, reflects the fact that the stocks have similar accumulated litter mass in UHS and LHS that produced similar DOC concentrations, which was an interesting result.