Eco-hydrology of drylands: Lessons from 10 years of field experiments in the South American Dry Chaco rangelands

MAGLIANO, PN; CHAPPELL, NA

Londres

Congreso; 13º British Hydrological Society meeting; 2018

British Hydrological Society meeting

Drylands cover about 41% of Earth?s land surface and are home to more than 38% of the total global population of 7 billion. Ecological and hydrological processes are particularly tightly coupled in water-limited environments, or drylands. In these environments, a positive-feedback or self-reinforcing mechanism links water and vegetation. In other words, how water is redistributed and where it becomes concentrated are important determinants of vegetation patterns; and conversely, vegetation patterns also directly modifying the nature of runoff. These relations highlight the importance of assessing the vertical components of water balance (one dimension), but also the horizontal redistribution of water (two dimensions. Understanding these complex interactions in water-limited systems affected by land use/cover change, and on a climate change scenario,is the focus of the emerging field of ecohydrology. The South American Dry Chaco, covering ~1 million km2, is one ofEarth?s largest and flattest semiaridwoodland-dominated ecosystems, and is rapidlybeing converted to pastures and croplands (Figure 1). The southern edge of Dry Chaco (Arid Chaco; 400 mm/yr) is dominated by extensive livestock production systems which are very important in the whole economy of the region. Two main challenges that livestock productionfaces in this dry region are the lack of fresh surface orgroundwater sources for drinking supply, and the lowfraction of the NPP that is allocated to forage (asopposed to non-forage woody tissues) in the rangelands. While the lack of fresh wateris resolved by runoff harvesting in human-made impoundment, the availabilityof forage here as well as elsewhere depends on theredistribution of surface water via runoff/runon tovegetation patches. Both issues are highly dependent on horizontalwater redistribution at the patch to landscape scales and canbe affected intentionally or non-intentionally by themanagement of livestock grazing and associated tramplingand soil compaction. In this study, we present the results of 10 years of field observations and experiments, in combination with satellite and modeling analysis, onthese livestock production systems of Arid Chaco. We measured the components of one dimension water balance (rainfall, interception, stemflow, throughfall, runoff, infiltration, evaporation, transpiration and deep drainage), and surface runoff/runon redistribution process at different spatial scales (from 1m2 to 10 ha).The main results that we found were, first, rainfall event size distribution was very inequality, specifically, the 10% of the larger events explained 44% of total rainfall of the year (this relation was strong, tested in different dry sites and years). Second, evapotranspiration (sum of interception, evaporation and transpiration) represents ~98% of total rainfall inputs. Third, the replacement of native forest by pastures generated lower canopy interception, but higher runoff and evaporation soil; deep drainage was negligible in both cover types. In line with this, pastures presented lower soil saturated hydraulic conductivity, flatter surface micro-topography and lower heterogeneity of the water of the soil profile than native dry forests. Fourth, the most abundant woody species of Arid Chaco (Larreadivaricata) presented 13% of total rainfall channelized as stemflow, this percentage is the double than similar species measured in other dry sites in the world. Fifth, the intensity of rainfall inputs (mm/h) had a critical ecohydrological relevance, that is not wellcharacterized by the more commonly used single metricof event size, at explaining runoff/runon redistribution patterns. Sparsely vegetated patches (1m2)infiltratedrelatively more water than densely vegetated patchesfor lower intensity events, whereas more denselyvegetated ones infiltrated relatively more water thansparsely vegetated ones for higher intensity events. Finally, the small fraction of runoff of the system at the landscape scale (10ha) was enough to fill the man-made impoundments that farmers build to harvest water to livestock supply. This last point, especially to understand the temporal variability of rainfall-runoff processes and water storage in the impoundments is critical for improving livestock production and/or to maintain the same production in dry years. These results provide critical information to better understanding the hydrological functioning of drylands, which will be particularly necessary in coming years of climate change scenarios and global land use intensification.