Effect of water stress on foliar reflectance of three forest tree species.

In this dissertation the effect of moisture stress on the foliar reflectance of red pine (Pinus resinosa Ait.), sugar maple (Acer saccharum Marsh), and red oak (Quercus rubra L.), was studied. The study area was located near Ann Arbor, Michigan. The field work was conducted between May and September 1986. Nine trees, three from each species, were subjected to three treatments. One tree was covered near the base with plastic material to prevent rainfall from reaching the root zone. A second tree was irrigated and left to receive rainfall, while a third tree was used as a control receiving only rainfall. Moisture stress was calculated using Zahner's model, which calculates depletion of water in soils of different textures. The empirical method of Thornthwaite was used to calculate potential evapotranspiration. Then moisture stress was calculated as the difference between potential evapotranspiration and depleted water or actual evapotranspiration calculated by Zahner's model. Leaf water potential was measured using a nitrogen pressure apparatus. The obtained xylem tension readings were used to test the the utility of Zahner's model to estimate moisture stress. Leaf water content was also calculated using the gravimetric method. Soil water content at 60 and 100 cm depths was monitored throughout the study period. The foliar reflectance of the three species was measured using a Beckman DK-2A spectrophotometer in the wavelength range 500 to 2600 nanometers. The relationships between stress, calculated by Zahner's model and measured by the hydrostatic pressure cell, and foliar reflectance were studied. Results suggest that: (1) Zahner's model worked reasonably well in predicting water stress. (2) Water stress results calculated by Zahner's water deficit method and xylem tension of the hydrostatic pressure cell method were highly correlated. Zahner's model stress calculations showed higher correlation with foliar reflectance than the pressure cell method in the red pine and sugar maple trees. (3) Water stress was significantly correlated with foliar reflectance in the atmospheric windows between 1300 and 2400 nm, especially in the wavelengths 1550 to 1650, and 2100 to 2300 nm. (4) The significant relationship between water stress calculated by Zahner's model and foliar reflectance increases the possibility of predicting reflectance at the date of acquisition of the remotely sensed data by removing the effect of moisture stress.Ph.D.Biological SciencesEarth SciencesEcologyRemote sensingUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/128252/2/8907023.pd