Growth and Water Relations of Kentucky Coffee Tree in Protective Shelters During Establishment

Growth and water relations of Kentucky coffee tree [Gymnocladus dioica (L.) K. Koch] whips in translucent tubelike shelters were investigated. In a container study, 1.2m-high shelters were placed over whips following transplanting, then diurnal microclimate, water relations, and water use were measured. Shelter air temperature and vapor pressure were substantially higher, and solar radiation was 70% lower, than ambient conditions. Sheltered trees responded with nearly three-times higher stomatrd conductance than nonsheltered trees. However, due to substantially lower boundary layer conductance created by the shelter, normalized water use was 40910 lower. In a second experiment, same-sized shelters were placed on whips following spring transplanting in the field. Predawn and midday leaf water potentials and midday stomatal conductance (g,) were monitored periodically through the season, and growth was measured in late summer. Midday gs was also much higher in field-grown trees with shelters than in those without. Sheltered trees in the field had four times greater terminal shoot elongation but 40% less stem diameter growth. Attenuated radiation in the shelters and lower specific leaf area of sheltered trees indicated shade acclimation. Shelters can improve height and reduce water loss during establishment in a field nursery, but they do not allow for sufficient trunk growth. Tree shelters, developed initially to reduce herbivory on newly planted tree seedlings, have been shown to improve elongation (Potter, 1988). Apparently, this is due to greenhouse-like conditions in the shelter, where interior temperatures and humidities are higher but transpiration rates from tree seedlings are sometimes lower (Burger et al., 1992). Tree shelters may benefit nursery field production by enhancing growth, reducing production time, and reducing water stress following field transplanting where routine irrigation is not available. Not all species perform equally well in shelters (Potter, 1988), and physiological responses to shelter microclimate are not completely known. Information on gas exchange and leaf temperature (t1) may provide insight as to why some species do not grow well inside shelters and suggest possible management options. The objectives of this study were to investigate microclimate changes and tree physiological response and growth in shelters for a nursery setting. Received for publication 6 July 1993. Accepted for publication 4 Feb. 1994. The cost of publishing this paper was defrayed in part by the payment of page charges. Under postal regulations, this paper therefore must be hereby marked advertisement solely to indicate this fact. 1Assistant Professor. Current address: Dept. of Plants, Soils, and Biometeorology, Utah State Univ., Logan, UT 84322. HORTSCIENCE, VOL. 29(7), JULY 1994 Materials and Methods Container experiment. Shelter microclimate and its effect on growth and physiology of Kentucky coffee tree was studied in a container experiment in 1991. Six whips (0.45 to 0.6 m tall) were potted with a 1 pinebark :1 peatmoss : 1 perlite mixture in 11-liter containers in mid-April and then placed in an east–west row on a gravel pad with full sun exposure. Trees were irrigated with drip emitters at 4 liters per tree every other day and were top-dressed with a 200-ppm N solution once a week. The above-ground portion of three trees was enclosed with 1.22-m-tall, 70to 100mm-diameter, corrugated-plastic, translucentbeige, tree shelter tubes (Tubex, St. Paul, Minn.) immediately after planting. When crown growth had reached the seasonal maximum following budset, the daytime pattern of tree water relations and shelter microclimate were measured on 31 July and 15 Aug. Total daily transpiration was determined on each date by enclosing the well-watered root ball and container with plastic, to eliminate soil evaporation, and then weighing the tree, including the container, the medium, and the shelter, before data collection in the morning and after cessation in the evening. A supporting post was attached to the shelters outside the enclosing plastic and also weighed with the plant. Predawn and midday water potential ( Ψ ) were measured on three leaflets per tree with a pressure chamber (model Arimad II; Kfar Charuv-Water Supply Accessories, Ramat Hagolan, Israel) immediately after excision. Dawn-to-dusk water relations were studied on both dates. Stomatal conductance (gs) and t1 were measured with a steady-state porometer (model 1600; LI-COR, Lincoln, Neb.) and infrared-thermometer (model 3100; Everest Interscience, Fullerton, Calif.), respectively. Conductance and t1 were measured through a small window cut into the south side of the shelter at mid-canopy level. After the window was opened, a single t1 measurement that integrated foliage temperature over the exposed area was taken immediately 0.2 m from the window. This test was followed quickly by gs measurements on three illuminated leaves per tree. Measurements were taken as rapidly as possible to avoid confounding effects of outside conditions on t1 and gs. The window was taped between measurements to prevent air movement into the shelter. Insufficient crown density, which permitted transmission of background radiation, precluded leaf temperature measurement of trees without shelters. After 15 Aug., foliage was harvested from each tree, and leaf area was measured with a leaf area meter (model 3 100; LI-COR). Dawn-to-dusk wet-bulb and air temperatures (ta) also were measured by drawing shelter atmosphere into a fan-aspirated thermocouple psychrometer (model 90023C; Atkins Technical, Gainesville, Fla.) through a closed cylinder inserted through a small hole into the shelter. Wind speed in the shelters also was measured periodically through the day with a hot wire anemometer (series 490; Kurtz Instruments, Carmel Valley, Calif.). The hole was sealed between measurements. Field experiment. Kentucky coffee tree whips were planted in a Hosmer silt loam (fine-silty, mixed-mesic, Typic Fragiudalf) in mid-Mar. 1991. The design was a randomized complete block plus or minus shelters with 10 single-tree replications. Same-sized shelters were placed over treatment seedlings, and the base was inserted 30 to 40 mm into the ground. Shelters were then firmly attached to supporting stakes placed on the north side of the tube. These trees received no supplemental irrigation, and the surrounding soil surface was clean-cultivated through the season. Global shortwave radiation and air temperature in the shelters were measured over 2 days in late Aug. 1991. A horizontally leveled pyranometer (model 200SA; LI-COR) was fixedto the wall immediately above the canopy inside one representative shelter. Because the limited spectral response of this pyranometer is calibrated to the daylight spectrum under full sun, values were corrected later to a star pyranometer (model 3020; Weathertronics-Qualimetrics, Sacramento, Calif.) with a broader spectral response in the blue and infrared wavebands. A thermistor also was inserted 40 mm into the shelter at canopy level to measure shelter air temperature and another thermistor was positioned outside the tube within a radiation shield 2m high to measure ambient temperature. The next day these measurements were repeated in a shelter without a tree to determine the effect of transpiration on air temperature in the shelter. Signal-wire holes were sealed to prevent air moving into the shelter. Output was logged