Response of electronic, printed-circuit wetness sensors was compared to visual observations of free water on processing-tomato leaflets during 13 dew-onset and 11 dew-dryoff events. Deployment angle and painting of the sensor surface significantly (P < 0.01) influenced the mean absolute time difference between observation of the first wet or dry leaflet at the top of the tomato canopy and the start of sensor response (kΩ) to dew onset or dryoff, respectively. Compass orientation of painted sensors deployed at 45° to horizontal had no significant effect on response to dew onset or dryoff. For sensors deployed at 45° during dew onset, mean absolute time difference between the first observed wet leaflet and the start of unpainted sensor response was 4.00 h, compared to 0.58 and 1.09 h for sensors with three and nine coats of paint, respectively. At deployment angles of 30 or 0°, paint coating had a lesser influence on time differences between visual observation and sensor response to dew onset. During dew dryoff, absolute time differences between visual confirmation of the first dry leaflet and the start of sensor response were ≤1.03 h for all sensors. Trends were similar when the visual observation criterion was 50% wet or dry leaflets during dew onset or dryoff, respectively, rather than first wet or dry leaflet. Standard deviation of sensor response during dew onset was generally larger for unpainted sensors than for sensors with three coats of paint, especially when deployed at a 45° angle. The apparent temperature of unpainted sensors at 0 or 30° deployment angles decreased much more rapidly during the period preceding dew onset than for painted sensors at the same deployment angles, whose apparent temperatures cooled at rates similar to those of tomato leaflets positioned at these angles. The results indicate that deployment angle can significantly affect accuracy and precision of dew-duration measurements by unpainted, but not painted, electronic wetness sensors.

Five cultivars of Fraxinus americana (white ash) and five of F. pennsylvanica (green ash) were graft-inoculated with three strains of ash yellows phytoplasmas at Ames, IA, and with thrsee other strains at Ithaca, NY. A sixth green ash cultivar was tested only in New York. Trees were allowed to grow in field plots for 3 years. Infection was detected via the DAPI (4′, 6-diamidino-2-phenylindole 2HCl) fluorescence test. Incidence of witches'-brooms on infected trees was greater on white ash than green ash and varied significantly among phytoplasma strain treatments at both locations. Volume growth of infected ash, averaged across cultivars over 2 years in Iowa and 3 years in New York, was 49 and 59%, respectively, as great as that of noninfected trees. Foliar greenness was reduced significantly by infection at both locations, and this reduction was positively correlated with growth reduction. Cultivars at each location varied significantly in growth of noninfected trees and in growth of diseased trees relative to that of nonin-fected trees (a measure of phytoplasma tolerance), but cultivar means for these variables in Iowa were not significantly correlated with those in New York. Green ash cvs. Bergeson, Dakota Centennial, and Patmore and white ash cv. Autumn Applause were above average in tolerance at both locations. Phytoplasma strains at each location varied significantly in aggressiveness as indicated by host growth suppression.

No abstract