Bermudagrass Response to Leaching Fractions, Irrigation Salinity, and Soil Types
Reuse of wastewater as an irrigation source for turfgrass is becoming a more viable and appealing option in arid environments where competition for good-quality water is increasing. The objective of this research was to determine the impact that varied leaching fractions, irrigation salinity, and soil types had on root growth and distribution, and fractional water uptake of bermudagrass [Cynodon dactylon (L.) Pers.]. Bermudagrass was grown for a 2-yr period in large columns packed with three different soil types (sandy loam, silt loam, and clay). Saline water was synthesized and applied at three different salinity levels (electrical conductivities of 1.5,3.0,and 6.0 dS m−1). Irrigations were applied 3 d wk−1 at a rate beyond measured evapotranspiration (ET) to establish three different leaching fractions (0.09, 0.18, and 0.27). The soil salinity (ECe), soil solution chloride (CI−), root density, and volumetric water contents were measured in soil cores taken with depth and time. Dry matter of weekly grass clippings was measured and recorded throughout the 2-yr period. Plant water status was monitored by measuring canopy temperatures and leaf xylem water potentials. Results indicated that bermudagrass was very tolerant to the range of salinity-leaching conditions imposed. However, differences were noted by treatments, with the sandy soil showing as much as a 25% yield decrement at the highest salinity level. Salinity of the irrigation water (ECI), rather than soil salinity (ECe), was more highly correlated with most of the soil-plant-water relationships observed. Root length density was best described by a hyperbolic function. Only limited success was found in correlating root length density with fractional water uptake. In addition, poor correlations were found between soil salinity with depth and fractional water uptake. These findings indicate that the ability to predict water uptake based on root distribution and/or soil salinity would be poor and that great error might occur in using such an approach in predictive models.
Copyright © . .