Living on the edge: modeling climate change impacts on sub-humid forests growing in semi-arid environments
- Autores: Daniel Nadal Sala
- Directores de la Tesis: Carlos A. Gracia Alonso (dir. tes.), Santiago Sabaté Jorba (codir. tes.)
- Lectura: En la Universitat de Barcelona ( España ) en 2019
- Idioma: español
- Tribunal Calificador de la Tesis: Emilia Gutiérrez Merino (presid.), Nadine Rühr (secret.), Thomas Rötzer (voc.)
- Programa de doctorado: Programa de Doctorado en Ecología, Ciencias Ambientales y Fisiología Vegetal por la Universidad de Barcelona
- Materias:
- Enlaces
- Tesis en acceso abierto en: TESEO
- Resumen
Semi-arid environments are zones where annual precipitation is less than a half of annual potential evapotranspiration, yet water availability is high enough to allow tree growth. Climate change is expected to have a major impact on forests growing at those regions. Rising atmospheric [CO2] (Ca) is expected to increase forest productivity. However, this fertilizing effect may be partially offset by an increase in water stress, either by reductions in water availability or by increases in atmospheric evaporative demand. Additionally, species-specific responses to climate change may further promote invasive tree species expansion.
GOTILWA+ process-based model was used to project the performance of sub-humid forests growing in semi-arid conditions under climate change. However, a carpenter is just as good as the least sharpened of his tools. So, firstly it was developed and tested the RheaG Weather Generator Algorithm, a first-order Markov transition matrix-based WGA, in order to assure the ability to generate statistically robust meteorological time-series. Then, Bayesian inverse modeling was applied in order to calibrate GOTILWA+ model from “in situ” observations from two different forest stands, both occupied by water-demanding tree species growing surrounded by semi-arid conditions.
Firstly, combined effects of increased vapor pressure deficit (D), increased Ca and decreased water availability in an S.W. Australian Eucalyptus salinga Sm. plantation were evaluated. Increasing Ca up to 700 ppm alone was projected to increase E. saligna productivity up to a 33%, and forest carbon stock up to a ~60%. However, combined reductions in water availability and D increases offset part of this fertilizing effect, down to 13% and 35%, respectively. Furthermore, limitations on forest productivity due to D increases were projected to occur in a magnitude similar than productivity reductions due to reduced soil water availability. Afterwards, in a N.E. Iberian Mediterranean riparian forest where black locust (Robinia pseudoacacia L.) is outcompeting three autochthonous deciduous tree species, sap flow observations were used to calibrate GOTILWA+ model for black locust and European Ash tree (Fraxinus excelsior L.). Field observations suggested that black locust success was explained by its facultative phreatophytic behavior, as well as an increased water use efficiency in stem growth, when compared with co-occurring autochthonous tree species. GOTILWA+ projections, including regionalized climate change scenarios, suggested that under global warming black locust productivity and growth would be further enhanced than its native counterpart, the European ash. The reasons are an increase on daily productivity as Ca increases, and an enlargement of its vegetative period as temperature rises.
As conclusions, the invasive black locust growth performance is expected to be favored by global warming in Mediterranean riparian forests. On the other hand, E. saligna responses to climate change will strongly depend on the balance between the beneficial effects of increasing Ca and physiological limitations due to water stress increase. At stand level, results highlight the importance of accounting for the water available for the trees at the whole soil column, and not only at the superficial soil layers, a challenging issue that is often not resolved in simulation models. On the other hand, properly accounting for vapor pressure deficit changes is of a major importance when projecting forest responses to climate change, as it will strongly determine stand changes in productivity and water use efficiency. This thesis highlights the importance of training simulation models from field observations, not only to describe ecophysiological processes, but also to obtain the most likely set of parameters given the observations.
