Tiller Development in Salt-Stressed Wheat

Wheat (Triticum aestivum L. emend. Thell.) grain yields are highly dependent upon the number of spike-bearing tillers produced per plant. Salinity, drought, and other environmental stresses can greatly affect the development and viability of tillers. We determined the effects of soil salinity on the occurrence and rate of tiller development and the incidence of tiller abortion in spring wheat cultivars, Anza and Yecora Rojo. Plants were grown in Pachappa fine sandy loam soil (mixed, thermic, Mollic Haploxeralf) in outdoor lysimeters. Three salinity treatments were imposed by irrigating with waters containing equal weights of NaCl and CaCl2 (electrical conductivities of = 1,12, or 18 dS m-l). Salinity significantly decreased the number of primary and secondary tillers in both cukivars. Soil water salinities 27.5 dS m-l (mean electrical conductivity of the soil water in the rootzone during tiller development, &w) eliminated most of the secondary tillers and greatly reduced the number of TO, T3, and T4 tillers. However, the percentage of tillers producing spikes actually increased at i& up to 8 dS m-l. Higher salinities reduced the percentage of tillers with spikes, but not as much as the reduction in tillers. Tiller and spike production per plant decreased about 0.13 to 0.15 organs for each unit increase in i&w. Of all the potential tillers these cukivars can produce, the prhnary tillers on Leaves 1 and 2 (i.e., Tl and T2) were the least susceptible to salt stress, partly because they emerged before salinity builds up during the irrigation season. Adjusting planting densities to increase the number of anticipated spike-bearing culms per unit area could help to maintain yields on salt-affected soils. GR A I N YIELD in wheat is highly dependent upon the number of spike-bearing tillers produced by each plant (Power and Alessi, 1978; Nerson, 1980). The number of productive tillers depends on the environmental conditions present during tiller bud initiation and subsequent development stages. Environmental stress during tiller emergence can inhibit their formation and, at later stages, can cause their abortion. Numerous studies have shown that tiller appearance, abortion, or both are affected by limited soil water (Sionit et al., 1980; Schonfeld et al., 1989; Davidson and Chevalier, 1990), high temperature (Friend, 1965; Ishag and Taha, 1974; Rawson, 1971; Thome and Wood, 1987), low temperature (Huci and Baker, 1990), nutrient deficiencies (Black and Siddoway, 1977; Power and Alessi, 1978; Masle, 1985), shading (Fischer, 1975; McMaster et al., 1987), and salt stress (Maas and Grieve, 1990; Nicolas et al., 1993). Soil salinity decreases grain yield of wheat more when plants are stressed prior to booting than when stressed later (Maas and Poss, 1989). The yield component affected most by salt stress is the number of spikes produced per plant (Maas and Grieve, 1990). Neither of these previous studies, however, provided information about the effects of salinity on individual tiller development and abortion. Knowledge of how salt stress affects USDA-ARS, U.S. Salinity Lab., 4500 Glenwood Drive, Riverside, CA 92501. Contribution from the U. S. Salinity Lab., USDA-ARS, Riverside, CA. Received 20 Dec 1993. *Corresponding author (!ao3lcrivussl). Published in Crop Sci. 34:1594-1603 (1994). the production of spike-bearing tillers would be invaluable in increasing the salt tolerance of wheat and improving crop management practices. For example, breeding genotypes with fewer, but less vulnerable tillers, as suggested by Jones and Kirby (1977), and increasing planting density to offset the loss of tillers (Francois et al., 1994, unpublished data), could substantially increase yields on salt-affected soils. The potential for increasing wheat yields by restricting tillering was demonstrated in field trials by Islam and Sedgley (1981). Data that relate tiller productivity to soil salinity levels are also essential for development of growth simulation models that predict response to salt stress. This experiment was conducted to determine the occurrence and rate of tiller development and the incidence of tiller abortion of two wheat cultivars grown at three levels of soil salinity. A logistic regression model was developed that estimates the number of spike-bearing tillers as a function of soil water salinity. MATERIALS AND METHODS Experimental Conditions The experiment was conducted in outdoor lysimeters (3.0 by 3.0 by 1.5 m deep) at the U.S. Salinity Laboratory, Riverside, CA. The lysimeters contained Pachappa fine sandy loam. Triple superphosphate was mixed into the top 0.25 m of soil at 73 kg P ha-’ prior to planting. All irrigation water contained 0.6 mol me3 Ca(NO& and 1.0 mol mm3 KN03 to ensure adequate N and K fertility throughout the experiment. Two hard red spring wheat cultivars, Anza and Yecora Rojo, were planted in level beds in the center 2.4by 2.4-m area of each lysimeter on 11 Jan. 1989 and 1990. Each lysimeter contained eight rows of each cultivar. Rows were spaced 0.15 m apart, with the seeds placed 40 mm apart within the row. Plant population at this planting density was 167 plants m-‘. Sowing depth was approximately 1.5 cm. The planted area was surrounded by a wooden barrier extending 2.5 cm above and 12.7 cm below the soil surface to minimize runoff of ponded irrigation water. The experiment consisted of three salinity treatments replicated three times in a split-plot arrangement of a randomized complete block design, with salinity as main plots and cultivars as subplots. To facilitate germination, 25 mm of low-salinity water (electrical conductivity = 0.9 dS m-l) was applied to each lysimeter after sowing. Differential salination was initiated on 30 Jan. 1989 and 23 Jan. 1990 by applying irrigation water containing equal weights of NaCl and CaC12. Seedlings for 1989 and 1990 were at the twoand one-leaf stages, respectively. The average electrical conductivities of the three saline-irrigation waters (Ki,) were 0.9, 11.3, and 17.5 dS m-’ in 1989 and 0.8, 12.2, and 17.7 dS m-’ in 1990. All lysimeters were irrigated every 7 to 10 d to keep the soil matric potential of the control treatment above 85 J kg-’ Abbreviations: DAS, days after sowing; ail, electrical conductivity of irrigation water expressed in deciSiemens per meter (dS m-l); K~, electrical conductivity of soil water; i&v, depth-averaged electrical conductivity of soil water in the rootzone; iisw, depthand time-averaged electrical conductivity of soil water in the rootzone during tiller development. MAAS ET AL.: TILLER DEVELOPMENT IN SALT-STRESSED WHEAT 1595 at the 0.25-m depth. Between 25 and 50 mm of water was applied at each irrigation. The total amount of water applied to each lysimeter between sowing and harvest was 690 mm in 1989 and 580 mm in 1990. Lysimeters were protected from rainfall by covering them with clear plastic sheeting on rainy days. A neutron probe and tensiometers were wsed to monitor soil matric potential and to guide irrigation frequency. Soil water contents were measured before and after most irrigations at depths of 25, 45, 75, and 105 cm and at two locations within each lysimeter. Plant Growth Measurements Soil water salinity (K,) was determined by extracting soil solutions with porous ceramic suction tubes buried 25, 45, 75, and 105 cm below the soil surface at two locations within each lysimeter. Soil solutions were extracted from these soil depths after most irrigations and timeand depth-averaged K, was calculated from measurements of electrical conductivity. Tiller growth and development were monitored three times each week on 10 randomly selected plants of each cultivar in each lysimeter for a total of 30 plants per treatment. The plants were chosen from the third, fourth, and fifth rows in from edge of each lysimeter to minimize border effects. Main shoots, tillers, and all leaves were identified and tagged as they emerged. Tiller designations follow the numbering system of Klepper et al. (1982). Leaf and ligule appearance dates and blade length were determined on all culms. The plants were harvested at maturity and the culms with spikes were identified and counted. Statistical Analyses Standard meteorological measurements were made with a The relationships between average number of primary tillers, Class I agrometeorological station adjacent to the lysimeters. spikes, or both, and the Km-Stress levels were analyzed with Air temperatures were measured about 2 m above the soil linear regression models. All secondary tiller and spikesurface. Hourly temperatures were integrated over the 24-h production data were analyzed with non-parametric analysis of period and the mean daily temperatures were summed to give variance tests (Kruskal-WaIlis tests) for significant differences cumulative thermal time, expressed as c”C d. The minimum between control and saline treatments. The regression relating air temperatures were always above 0°C except for 2 or 3 h tillers to stress level was divided by the regression relating on Days 46 and 47 in 1990. spikes to stress level to provide a predictor of the percentage Soil Depth High -0-25 cm -45cm A_ a75cm YR m + 105cm