Detecting Physiological Water Stress in Southern New Mexico Pecan Orchards using Remote Sensing, Poster Board #213

While leaf-level plant physiological parameters can provide useful information to better understand drought stress in pecans, extraordinary cost and long collection times, along with the necessity of leaf destruction, limit their use.  Remote sensing applications are considered a promising technique in detecting and up-scaling leaf-level physiological responses to large areas.  The objective of this study was to detect physiological changes in pecan trees exposed to soil moisture deficits using remotely sensed surface reflectance derived from Landsat-7 Enhanced Thematic Mapper (ETM) and Terra Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER).  The study was conducted simultaneously on two southern New Mexico mature pecan orchards in 2011.  Two treatments were applied to both orchards; well-watered and water deficit. The beginning of each irrigation cycle was considered the well-watered level, while water deficit was assumed at the end of the irrigation cycle. Irrigation cycle varied from 14 to 20 days. Irrigation cycles were synchronized with satellite overpasses.  Several reflectance indices were computed from the remotely sensed data including reflectance in individual bands, Band Ratios (BR), the Normalized Difference Vegetation Index (NDVI), the Soil Adjusted Vegetation Index (SAVI), and shortwave infrared (SWIR) moisture indices of Landsat-7 ETM+.  The empirical relationships between reflectance indices and in situ measurements of pecan foliar midday water potential were compared.  Green reflectance was related to water deficit and showed significant relationships (P = 0.049) from May to June (r² = 0.65).  SWIR_{(2.08-2.35 µm)} and BR SWIR_{(1.55-1.75 µm)}/SWIR_{(2.08-2.35 µm)} were also significantly related (P = 0.036 and 0.035 respectively) to foliar water deficit from June to September (r² = 0.67 and r² = 0.65, respectively).  On the other hand, SWIR moisture indices show significant relationships (P = 0.016) with foliar midday water potential from mid of September to mid of November (r² = 0.75).  NDVI and SAVI were not significantly related to foliar midday water potential of pecans.  Although ASTER data are still being analyzed, results illustrated that reflectance indices such as BR and SWIR moisture indices have potential as nondestructive proxies for detecting water deficit in pecan trees.