2015

My PhD Thesis, entitled "Soil moisture dynamics in dehesas and its relationships with climate and vegetation", was supervised by Dra. Susanne Schnabel and recently defended in 2015, obtaining the highest level: cum laude per unanimity. The Thesis has also been awarded with the Extraordinary Award to the best Doctoral Thesis.

Summary

Water plays a critical role in almost all processes and natural cycles of the planet. The maintenance and equilibrium of agrosilvopastoral ecosystems with scattered tree cover and Mediterranean climate depends on factors that determine its configuration, the physical and environmental characteristics and human management. A prominent role is played by these factors in one of the key components of these agrosilvopastoral ecosystems: water.

This work investigates the soil moisture dynamic at different spatio-temporal scales in agrosilvopastoral ecosystems (dehesas), conditioned by the availability of water. The relationships between such dynamics and the factors involved, such as climate and vegetation, are also analyzed, as well as the response and sensibility of the pasture to soil water availability.

Soil water content was monitored continuously with a temporal resolution of 30 minutes and by means of capacitance sensors, mainly for the hydrological years 2010–2011 and 2011–2012. They were installed at 5, 10 and 15 cm, and 5 cm above the bedrock and depending on soil profile. This distribution along the soil profile is justified because soils are generally very shallow and most of the roots are concentrated in the upper layer. The sensors were gathered in 17 soil moisture stations (SMS) mainly in two contrasting situations characterized by different vegetation covers: under tree canopy and in open spaces or grasslands. Temporal scales ranged from minute to year, and longer climate series in combination with ecohydrological spatially distributed models were also considered and used. Spatial scales varied from soil profile (m2) to catchment watershed (≈1 km2). These works were carried out in 3 experimental farms of the Spanish region of Extremadura.

Results indicated that, generally, under tree cover, lower annual soil water content was observed than in grasslands. However, the combination of factors, such as meteorological variability and modifications caused by trees, could determine greater or lower soil water content under tree canopies than in grasslands. A homogenization of soil water content for different vegetation covers was observed for wetter periods, while differences in soil water content tended to increase in the drier ones.

The amount of rainfall reaching the soil may be modified temporally by different vegetation covers according to antecedent event conditions (from driest to wettest). The frequencies of re-wetting cycles seemed to be more relevant than either the duration or amount of precipitation in the interception of the different vegetation covers.

The annual wetting-drying cycle of the soil experienced different seasonal variability according to the amount and temporal distribution of rainfall. Matrix flow was observed as the dominant process during soil wetting cycles as opposed to preferential flow which was less frequent. However, if the total volume of water is considered, then preferential flows became the dominant process. Besides, the data mining technique Multivariate Adaptive Regression Spline (MARS) proved being useful to identify and rank the factors influencing flow types as well as modelling their occurrence.

The results highlighted the importance of the upper soil layer (first 15 cm) on ecohydrological processes as the main layer to supply water for pasture production. Herbaceous vegetation cover only presented a significant growth when this layer was able to satisfy the water demand in the appropriate period. Additionally, the surface layer proved to be the more sensitive to external factors, such as temperature changes.

The combined use of ecohydrological spatially distributed models and stochastic weather generators proved to bean effective tool for estimating trends in long-term pasture production and hydrologic conditions in semiarid rangeland areas characterized by high spatial and temporal variability of rainfall.

Keywords: soil moisture, ecohydrology, vegetation, climate, dehesa.