Sumario: |
The impacts of forest conversion on runoff generation in the tropics have received much
interest, but scientific progress is still hampered by challenging fieldwork conditions and
limited knowledge about runoff mechanisms. Here, we assessed the runoff generation,
flow paths and water source dynamics of a pristine rainforest catchment in Costa Rica
using end member mixing analysis (EMMA) and a Bayesian mixing model (MixSIAR). Geochemical tracer data collected over a 4-week field campaign were combined with tritium
data used to assess potential deeper groundwater flow pathways to the perennial
stream. The streamflow composition was best captured using three end-members,
namely throughfall, shallow (5–15 cm) and deeper (15–50 cm) soil water. We estimated
the end-member contributions to the main stream and two tributaries using the two
mixing approaches and found good agreement between results obtained from EMMA
and MixSIAR. The system was overwhelmingly dominated by near-surface sources, with
little evidence for deeper and older groundwater as tritium-derived baseflow mean transit time was between 2.0 and 4.4 years. The shallow soil flow pathway dominated
streamflow contributions in the main stream (median 39% and 49% based on EMMA
and MixSIAR, respectively), followed by the deeper soil (32% and 31%) and throughfall
(25% and 19%). The two tributaries had even greater shallow soil water contributions relative to the main stream (83% and 74% for tributary A and 42% and 63% for tributary B).
Tributary B had no detectable deep soil water contribution, reflecting the morphology of
the hillslope (steeper slopes, shallower soils and lower vegetation density compared to
hillslope A). Despite the short sampling campaign and associated uncertainties, this study
allowed to thoroughly assess runoff generation mechanisms in a humid tropical catchment. Our results also provide a first comparison of two increasingly used mixing models
and suggest that EMMA and MixSIAR yield comparable estimates of water source partitioning in this tropical, volcanic rainforest environment.
|