Ecohydrology of a seasonal cloud forest in Dhofar

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Civil and Environmental Engineering, 2005.Includes bibliographical references (p. 203-211).The hydrology of a semiarid broadleaf forest in Dhofar (Oman) is investigated by performing a field experiment and a modeling study. Based on the results of the field experiment, the ecosystem in Dhofar is classified, for the first time, as a seasonal deciduous cloud forest. Owing to the seasonal cloud cover, the soil storage in this ecosystem is filled and emptied in such a fashion that allows trees to allocate most of the water to transpiration by their relatively deep roots. In addition to annual rainfall amounts of about 100 millimeters, more water is gained by trees directly through interception of cloud droplets (horizontal precipitation). Through- fall measured below the canopy is about twice the rainfall measured above the canopy. Stemflow contributes about one third of the total water received by the soil. Soil moisture observations are consistent with increased infiltration in proximity of the tree trunks. Sap flow and meteorological observations indicate that transpiration is suppressed during the wet season, which allows for deep infiltration of the water received at the surface. This soil water storage supports transpiration by the relatively deep-rooted trees for up to four month into the dry season. All these factors combine to shape and enhance the water conserving nature of the ecosystem, and to allow for tree dominance in an otherwise too dry environment. A dynamical vegetation model is used to investigate tree- grass competition in this arid region. The model simulations elucidate the role of clouds in reducing the available energy for transpiration during the wet season. Without representing this role in the model, the simulated soil water storage by the end of the wet season is not enough to allow for tree survival and dominance.(cont.) The same model is used to investigate the role of the soil in controlling relative performance of trees and grass in this ecosystem. A characteristic rooting depth that is optimal for tree growth is identified, function of soil type and climate. This is the smallest depth at which transpiration is maximized and all other water sinks such as surface runoff or drainage are minimized. The optimal rooting depth is deeper when the evaporative demand during the wet season is low, similar to conditions in Dhofar. Such conditions improve competitiveness of trees in the model simulations. A horizontal precipitation module is used to illustrate how the contribution of this process is markedly reduced over grass as compared to trees. When the horizontal precipitation module is coupled to the vegetation model, two stable states are simulated by the model. Equilibrium vegetation simulated by the model starting from forest (grass) initial conditions is dominated by trees (grass). Deforestation, in the model, reduces soil water input and hence would tend to inhibit re-emergence of trees as a dominant land cover. Implications of this feedback for the re-forestation efforts in Dhofar are discussed. The results of this study should provide a solid basis for sound environmental management of the ecosystem.by Anke Hildebrandt.Ph.D