Date of Award

Spring 2019

Project Type


Program or Major

Earth and Environmental Sciences

Degree Name

Doctor of Philosophy

First Advisor

Heidi Asbjornsen

Second Advisor

Thomas G Pypker

Third Advisor

Julian A Licata


Forecast scenarios predict an increase in the demand of alternative sources of energy during the coming decades, such as woody biomass crops (WBC). WBC have the potential to become a major challenge for the next generation of researchers, policymakers and land managers. However, the current rationale for promoting plant-based over petroleum-based energy sources emphasizes the benefits of reduced carbon dioxide and other emissions, while giving less attention to potential impacts to water resources.

It is well documented in the scientific literature that trees use large amounts of water for metabolic needs. Water use at the tree and ecosystem level has always been of scientific interest, however, the potential impact of water use in bioenergy plantations is often considered a “possible environmental impact”. Thus, understanding the ecological implications of water use in WBC is essential for their sustainable development.

The general goal of my research was to assess potential ecohydrological impacts associated with the production of biomass for bioenergy from aspen (Populus tremuloides Mich.) in Wisconsin, USA, and eucalyptus (Eucalyptus grandis) plantations in Entre Rios, Argentina. My doctoral research was part of a large international interdisciplinary NSF-PIRE research project that examined the impacts, barriers and opportunities related to bioenergy production across the Americas (USA, Mexico, Brazil, Argentina).

We selected plantation ages within the most common rotation cycles for each species. In Wisconsin, we studied three sites, a 10- and a 24-year-old (YO) coppice plantations, and a reference 34 YO mature forest. In Argentina, we studied two 1 YO plantations one at high and one at regular density, a 4 YO, a 10 YO plantation, and a reference grassland. This was a unique study to determine annual water use based on a combination of tree-level measurements of water use using sap flow sensors, and deterministic models of potential evapotranspiration.

We validated two sap flow methods (heat dissipation and heat ratio), and validated a third method (maximum heat ratio) that is capable of measuring with precision high and low sap flux densities (Fd, cm3 cm-2 cm1). According to the results from the validation studies, we were able to estimate tree-level water use within a 7% error margin (estimated as the difference between observed and estimated sap flow in L h-1) using heat ratio and maximum heat ratio methods, without generating species-specific parameters. However, using the heat dissipation method, the average estimation error without species-specific parameters was -53%, and improved to 5% once species-specific parameters were generated. Validating the maximum heat ratio method, allowed us to estimate Fd in young trees, which are often excluded from chronosequence studies due to their high Fd. Our estimates of Fd at different plantation ages, allowed us to extrapolate from the tree to the site level, using real tree-level response to various environmental variables.

Our analysis of P. tremuloides and E. grandis offered contrasting results. In P. tremuloides plantations, water use at the site level generally increased with age, even when site density decreased over time (from approx. 6500 to 1900 from 10 to 34 YO). We observed that young plantations (10 YO) used 80% of the annual water early in the growing season, compared to a 45% for the same period in the Mid-aged and Mature plantations. Site effects, specifically soil type and the resulting soil saturation (S), had a significant effect on T estimates. After modeling the effects of S, creating two artificial scenarios (e.g. limited and non-limited S) the 10 YO site showed the highest sensibility to changes in S, while the 34 YO mature site was the least affected. Average stand transpiration by site considering the effects of S increased with age, which supported our hypothesis regarding the relationship between stand age and stand T. However, the relationship between stand T was not constant across seasons, which according to our results might be caused by a higher hydraulic stress observed in the 10YO site, compared to the other two sites.

On eucalyptus plantations, when the density remained constant, stand T decreased with plantation age, reaching maximum water use rates at around 4 YO and declining afterwards for the remainder of the typical 15 YO rotation cycles. Due to similar site conditions in terms of soil characteristics, we did not observe a strong site effect. Our experimental site at high density (e.g., double of a regular site) showed an increase in average site T of 50%, and both 1 YO sites presented an opportunistic pattern in water use, increasing when soil water was available, but decreasing significantly when soil moisture was limiting. Finally, in both plantations we observed that reference evapotranspiration, estimated with the Penman-Monteith equation, was a poor predictor of water use in young plantations. We associate these results to the seasonal patterns of water use in young plantations.

Within the context of bioenergy production, our results provide ample evidence for the importance of water use in bioenergy plantations in the early stages of feedstock production. We also show that plantation density in fast and slow growing species, impacts the way trees respond to water availability in the soil.