Catchments in the Cerrado e Caatinga — more arid biomes — were prone to present a losing water condition that deteriorates the imbalance between water availability and demand in these regions (Gesualdoet al. , 2021). This situation is exacerbated by the long-term conversion of natural vegetation to different agricultural land uses (e.g., sugarcane, soybean, and corn), which have been responsible for more than half of the national grain production (Spera, 2017) mainly in the Cerrado. Furthermore, the increase in irrigated areas for food production in the Cerrado and Caatinga biomes has been recently related to an increase in evapotranspiration and baseflow reduction in these regions (Oliveira et al ., 2020; Lucas et al ., 2021). Therefore, quantifying the effective catchment area is key to better understand synergies and trade-offs of land use changes and the increase in irrigated areas. The observed losing water condition suggests a strong inter-catchment hydrological dependency among the catchments in these biomes as a substantial part of precipitation contributes to the subsurface flow.
Our ECI results were positively correlated with mean slope and mean elevation corroborating the gaining water condition found in 72% of the catchments located in the Atlantic Forest biome, which presents the highest mean elevation and slope (Almagro et al., 2020). Indeed, low and high slopes were associated with smaller and larger effective areas, respectively (Figure 5b). Different from Liu et al. (2020), we did not note high variability of the effective area — i.e., either positive or negative ECI — in flatter regions (Figure 5d). On the other hand, ECI only presented low variability with increasing elevation. For instance, the catchment in the Pantanal biome, characterized by a complex hydrological dynamic, presented a substantial deviation between its effective and topographic area, indicating a losing water condition. Thus, the observed losing water condition corroborates the characteristics of a flat lowland area, where rivers flood the plains and feed an intricate seasonal drainage system (Ivory, McGlue, Spera, Silva, & Bergier, 2019). Additionally, the topographical catchment delineation is more susceptible to errors in complex topographies since the effect of Digital Elevation Models accuracy is not well understood, leading to a mismatch between topographic and effective areas (Zandbergen et al., 2011).
Catchments with a well-defined precipitation seasonality were associated with gaining water conditions as in the southern part of the Cerrado and the entire Atlantic Forest biome. Nonetheless, the precipitation seasonality index has some limitations when applied in Brazil due to its climate characteristics. Low thermal amplitude characterizes the climate in the north and northeast regions so that it is difficult to determine whether precipitation occurs during the summer, the winter, or throughout the year. The WTD and HAND had little influence on ECI probably due to the low spatial resolution of the available data. Carrying out large-scale studies involving groundwater is still challenging since high-resolution products at large scales are scarce (Gleeson, Cuthbert, Ferguson, & Perrone, 2020). Besides, monitoring the groundwater table is associated with high levels of uncertainties, frequently limited to developed regions (Fan, Li, & Miguez-Macho, 2013). Similarly, the soil texture data used in this study also have a low spatial resolution (250 m), which may have compromised the RFA performance in identifying its influence on ECI estimates (Supplement S3). Even though this attribute presents the lowest influence, it plays an important role in the catchment’s water cycle and groundwater flow in terms of soil water storage and percolation.
Implications for the understanding of hydrological connectivity and potential advances in water resources management
In this paper, we assessed the concept of interconnected catchments through the investigation of their effective areas. The hydrological connectivity was inferred from the flow-process perspective, defined by Bracken et al. (2013) as the understanding of runoff patterns and processes on hillslopes. From this perspective, we spatially assumed the connection between catchments and their influencing attributes. Based on the ECI results, we can state that catchments are sub-superficially interconnected. Although the scientific community agrees on inter-catchment connectivity, this is still a recent concept and little explored by water management decision-makers.
Incorporating the knowledge of the effective area and its influencing factors in the water resources management, the groundwater boundaries and processes may be reasonably considered. These processes are often neglected when the topographic area is solely used as a management unit. Therefore, the inclusion of the magnitude of the effective area would improve the comprehension of more reliable hydrological processes on a catchment scale. Besides, understanding the deviation between topographic and effective areas copes with the lack of clear and detailed information about aquifer properties and limits (Hirata, Kirchheim, & Manganelli, 2020), which are important to have integrated water management (Samani, 2021).
The ECI is a relevant tool for tackling water vulnerabilities and inequalities in allocating and managing water. In the semiarid region, Brazilians are exposed to high levels of water insecurity and inequality exacerbated by recurrent and prolonged droughts (Gesualdo et al. , 2021). By knowing the magnitude of the deviation between the topographic and effective areas, the influencing climatic and physiographic attributes, and its underlying hydrological processes, decision-makers can assemble groups of nearby connected catchments. Based on this, water resources management would be settled into a pooling of catchments — a combination of the interests and needs of a group of catchments — considering their interconnection. The exclusive use of topographic delineation is a limiting factor although the water resources management practice is already designed for a large group of catchments (e.g., as in the São Francisco River Basin (one of the 12 hydrographic units in the country).
Efforts on water-related preservation and conservation have been made in contributing surface areas upstream reservoirs and pumping points even though these are often much smaller or larger than the actual contributing areas. Therefore, advances in understanding and identifying effective areas play a key role in synergistically managing watershed services related to water yield and provision. For instance, it would imply a greater environmental responsibility by water-gaining catchments benefiting from water ecosystem services provided by water-losing catchments. We emphasize the paramount importance of integrating hydrological processes and water ecosystem services relevant for catchment management as alterations in flow processes impact water provision and vice-versa (Grizzetti, Lanzanova, Liquete, Reynaud, & Cardoso, 2016). Thus, managing a pool of catchments can increase synergies and lessen trade-offs of water transfer processes. Sustainable inter-basin water transfer is an alternative for addressing the imbalance in water availability and demand mainly in the Caatinga and Cerrado biomes (Gesualdo et al., 2021). Hence, advancing the understanding of hydrological connectivity between semiarid and arid catchments better copes with water scarcity. In this context, the ECI guides possible solutions to a question encompassed by Guswa et al. (2014): “What parcel of land is the highest priority for conservation?”. The effective area deviation can support decision-makers in identifying catchments with the highest priority for conservation and best management practices implementation.
Considering the inter-catchment connectivity contributes to investigating the extent of groundwater pollution and projecting efficient water use in activities such as agriculture. In the agriculture sector, the quantification of effective catchment areas would allow decision-makers to strategically manage the increase in water-fed irrigation areas in Caatinga and Cerrado biomes and understand how this disturbs the water balance in these regions. Although the land cover was not one of the most important attributes to identifying the effective catchment area, it has a major role in unveiling the mechanisms of movement and storage of water at a catchment scale. Nevertheless, anthropogenic impacts on water fluxes are still poorly understood (Neupane & Kumar, 2015) such as land use and land cover changes and inter-basin transfers.
There are future research opportunities for addressing surface and groundwater integration, such as: