Headwaters drive streamflow and lowland tracer export in a large‐scale humid tropical catchment

Abstract

Headwaters are generally assumed to contribute the majority of water to downstream users, but how much water, of what quality and where it is generated are rarely known in the humid tropics. Here, using monthly monitoring in the data scarce (2,370 km2) San Carlos catchment in northeastern Costa Rica, we determined runoff‐area relationships linked to geochemical and isotope tracers. We established 46 monitoring sites covering the full range of climatic, land use and geological gradients in the catchment. Regression and cluster analysis revealed unique spatial patterns and hydrologically functional landscape units. These units were used for seasonal and annual Bayesian tracer mixing models to assess spatial water source contributions to the outlet. Generally, the Bayesian mixing analysis showed that the chemical and isotopic imprint at the outlet is throughout the year dominated by the adjacent lowland catchments (68%) with much less tracer influence from the headwaters. However, the headwater catchments contributed the bulk of water and tracers to the outlet during the dry season (>50%) despite covering less than half of the total catchment area. Additionally, flow volumes seemed to be linearly scaled by area maintaining a link between the headwaters and the outlet particularly during high flows of the rainy season. Stable isotopes indicated mean recharge elevations above the mean catchment altitude, which further supports that headwaters were the primary source of downstream water. Our spatially detailed “snap‐shot” sampling enabled a viable alternative source of large‐scale hydrological process knowledge in the humid tropics with limited data availability.

A preliminary isotope‐based evapotranspiration partitioning approach for tropical Costa Rica

“…The average annual precipitation is approximately 4200 mm with average temperatures of around 26°C and high average relative humidity of approximately 84% (Zúñiga Mora, 2011). More details on the San Carlos catchment characteristics, monitoring, and sampling can be found in Birkel et al (2020).…”

“…In the San Carlos catchment as well as on the national scale, point isotope values were spatially interpolated using a spline interpolation method implemented in ArcGIS 10.5 at a pixel size of 100 m. This spatial interpolation enables the projection of continuous surfaces (isoscapes) from discrete data distributed over sampling points on a surface (Vargas et al, 2011; Villatoro et al, 2008). The national‐scale isoscapes of Sánchez‐Murillo and Birkel (2016) and the catchment‐scale isoscapes of Birkel et al (2020) together with the respective hydrological water balance (Figure 1c) form the basis of the water partition isotope model for T estimation explained below (Table 2). The long‐term mean annual availability of surface runoff (Q in mm) was calculated from the climatic water balance with mean annual precipitation P minus mean annual evapotranspiration ET.…”

“…Building on work by Sánchez‐Murillo and Birkel (2016) and Birkel et al (2020), who presented isoscapes and a basic water balance at the national and catchment scale, respectively, here, we aim at developing annual average T fluxes using isotope‐based mass balance models (e.g., Gibson et al, 2016; Jasechko et al, 2014) at the national scale of Costa Rica and on a monthly basis in the San Carlos catchment for comparison purposes indicating potential uncertainties due to different data sources. Linking the water partitioning to ecosystems for a potential impact assessment can be approached based on simple ecosystem indicators.…”

End member and Bayesian mixing models consistently indicate near‐surface flowpath dominance in a pristine humid tropical rainforest

“…Model frameworks such as MixSIAR use Bayesian hierarchical modelling techniques and compositional data transformation to assess the likelihood of source proportions (Stock et al, 2018). An increasing number of studies has used Bayesian theory to assess water sources to streams (Beria et al, 2020; Birkel et al, 2020; Duvert et al, 2020), but a systematic comparison of the relative performances of the EMMA and Bayesian approaches is still lacking (He et al, 2020).…”

A longer‐term perspective on soil moisture, groundwater and stream flow response to the 2018 drought in an experimental catchment in the Scottish Highlands

“…These vary from assessing sources of plant water uptake in droughts at the plot scale in Mediterranean environments (e.g. Barbeta et al, 2015) to assessment of groundwater inputs to larger catchments under low flow conditions in geographical regions as diverse at the Italian Alps (Chiogna et al, 2018), Costa Rica (Birkel et al, 2020) and upland areas in central India (Noble & Arzoo Ansari, 2019). In the BB, the trajectory to more stable, intermediate isotopic ratios was consistent with the dominance of groundwater inflows, which increase in the valley bottom (Scheliga et al, 2017).…”