Growth Response of Major U.S. Cowpea Cultivars. I. Biomass Accumulation and Salt Tolerance

Over the last several years, there has been increasing interest in amending the soil using cover crops, especially in desert agriculture. One cover crop of interest in the desert Coachella Valley of California is cowpea [Vigna unguiculata (L.) Walp.]. Cowpea is particularly useful in that as an excellent cover crop, fi xing abundant amounts of nitrogen which can reduce fertilizer costs. However, soil salinity problems are of increasing concern in the Coachella Valley of California where the Colorado River water is a major source of irrigation water. Unfortunately, little information is available on the response of cowpea growth to salt stress. Thus, we investigated the growth response of 12 major cowpea cultivars (‘CB5’, ‘CB27’, ‘CB46’, ‘IT89KD-288’, ‘IT93K-503-1’, ‘Iron Clay’, ‘Speckled Purple Hall’, ‘UCR 134’, ‘UCR 671’, ‘UCR 730’, ‘8517’, and ‘7964’) to increasing salinity levels. The experiment was set up as a standard Split Plot design. Seven salinity levels ranging from 2.6 to 20.1 dS·m were constructed, based on Colorado River water salt composition, to have NaCl, CaCl2 and MgSO4 as the salinization salts. The osmotic potential ranged from –0.075 to –0.82 MPa. Salt stress began 7 days after planting by adding the salts into irrigating nutrient solution and ended after 5 consecutive days. The plants were harvested during fl owering period for biomass measurement (53 days after planting). Data analysis using SAS analysis of variance indicated that the salinity in the range between 2.6 and 20.1 dS·m signifi cantly reduced leaf area, leaf dry weight, stem dry weight and root dry weight (P ≤ 0.05). We applied the data to a salt-tolerance model, log(Y) = a 1 + a 2 X + a 3 X, where Y represents biomass, a 1 , a 2 and a 3 are empirical constants, and X represents salinity, and found that the model accounted for 99%, 97%, 96%, 99%, and 96% of salt effect for cowpea shoot, leaf area, leaf dry weight, stem dry weight and root dry weight, respectively. We also found signifi cant differences (P ≤ 0.05) of each biomass parameter among the 12 cultivars and obtained different sets of the empirical constants to quantitatively describe the response of each biomass parameter to salinity for individual cowpea cultivars. Since a signifi cant salt × cultivar interaction effect (P ≤ 0.05) was found on leaf area and leaf dry weight, we concluded that salt tolerance differences exist among the tested cultivars. Over the last several years, there has been increasing interest in amending the soil using cover crops, manures, composted yard wastes, and other organic matter. Covers crops are important in maintaining soil productivity and environmental quality (Hargrove and Fry, 1987; Powers, 1987). More recently, this interest has expanded to include desert agriculture (Guldan et al., 1997; Hutchinson and McGiffen, 2000). The desert valleys of southern California have become a profi cient producer of a large variety of vegetables (Mayberry et al., 1995). Although hot and arid, these low-desert valleys are well suited to growing high-value fruit and winter vegetables as well as other crops if properly irrigated. However, as pointed out by Hutchinson and McGiffen (2000), any cropping system employing the use of cover crops would need to fi t into the vegetable production timetable of desert agriculture. Because cowpea (Vigna unguiculata) is adapted to high temperature and drought (Ehlers and Hall, 1997; Hall and Patel, 1985), it has become one cover crop of some interest as it may be suitable for growth during summer months in the desert valleys of California when vegetables are not usually grown. Recently, Abdul-Baki, et al. (1998) suggested an approach whereby cowpea could be used as a second cover crop in date orchards by seeding them in late May after terminating Lana vetch. Recently, Hutchinson and McGiffen (2000) indicated that cowpea may also be useful as a mulch for weed control in the low desert. In desert agriculture though, salinity is a widespread and prevalent problem. Soil salinity in the Coachella Valley of California ranges from 1 to 150 dS·m (S. Aslan, personal communication) and salinity problem is of increasing concern there where the Colorado River water has been a major source of irrigation water (salinity = 1.1 to 1.2 dS·m) for many years (Bower et al., 1969). Additionally, high-quality water for agricultural purposes is becoming increasingly scare due to changing environmental standards and rising demands from urban areas. The future of irrigated agriculture, in the desert and elsewhere, will need to include the use of waters containing higher levels of soluble salts. Unfortunately, little information is available on the response of cowpea growth to increasing salinity. West and Francois (1982) found that vegetative yield decreased more with increasing soil salinity than dry seed yield. Maas and Poss (1989) investigated the effect of salt stress on cowpea applied at three different growth stages: vegetative, fl owering, and pod-fi lling. They found that pod and seed yields were most sensitive to salinity during the vegetative stage. Vegetative growth was reduced by salinity during all three growth stages studied but the effect was much less when the stress was imposed during the last two stages. However, these prior investigations were concerned exclusively with ‘CB5’, ‘California Blackeye No. 5’. Other cultivars such as ‘Iron Clay’ may be better suited for the Coachella Valley (Abdul-Baki, 1998). Therefore, any information regarding salt Table 1. Cowpea cultivars used in the experiment. Cultivars 1 to 7 (counted from the top), are popular Coachella Valley cover crops with different origins and growth traits. Cultivar 9, ‘CB27’, is a new variety which demonstrates high heat tolerance and broad-based resistance to both Fusarium wilt and root-knot nematodes. Cultivar 10, ‘CB46’, is a new, widely planted cultivar and now accounts for 80% to 90% of the area planted in the U.S. Cultivar 8, ‘CB5’, is an old standard cultivar which now accounts for only 5% to 10% of the area planted. The fl owering term normal refers to cultivars that fl ower in normal growth period. Late or photosensitive refers to cultivars that do not fl ower until September or October. DLS refers to delayed leaf senescence cultivars. Cultivar Flowering time Origin Iron Clay (standard variety) Late Southeastern U.S. Speckled Purple Hall Late Southeastern U.S. IT89KD-288 Photosensitive Nigeria IT93K-503-1 Photosensitive Nigeria UCR 134 Photosensitive Botswana UCR 671 Photosensitive Botswana UCR 730 Photosensitive Kenya CB5 (California Blackeye) Normal California CB27 (new variety) Normal California CB46 (standard variety) Normal California 7964 (DLS) Normal Southeastern U.S. 8517 (DLS) Normal California FebruaryBook 1 225 12/14/05 10:59:10 AM HORTSCIENCE VOL. 41(1) FEBRUARY 2006 226 tolerance among cowpea cultivars could be useful in addressing the current issues involved in desert agriculture. There is mixed evidence in the literature as to whether salt tolerance can vary among cultivars within a given crop. For example, Francois (1996) investigated four sunfl ower hybrids and found similar threshold values for seed yield (4.8 dS·m). Conversely, among bermudagrass cultivars, Francois (1988) reported widely ranging threshold values of 2.7, 8.4, and 10.3 dS·m for the cultivars ‘Tifton 10’, ‘Tifway II’, and ‘Tifton 86’, respectively. Just as important, it seems that salt tolerance among cultivars may vary with one component of yield but not another. Francois et al. (1988) showed that two cultivars of triticale, ‘Cananea’ and ‘Beaguelita’, had similar threshold values for relative grain yield but differed with respect to relative straw yield. Finally, Leng et al. (2001) identifi ed wide genotypic differences in relative salt tolerance within rice and found that their relative salt tolerance ranking varied depending on the parameter measured. Thus, we examined the hypothesis that there exists wide salt tolerance in the vegetative growth among cowpea cultivars. We investigated the growth response of 12 major cowpea cultivars, which may be useful as cover crops in desert agriculture, to increasing salinity levels using sand cultures: ‘UCR 730’, ‘IT89KD288’, ‘Iron Clay’, ‘IT93K-503-1’, ‘Speckled Purple Hall’, ‘UCR 134’, ‘UCR 671’, ‘CB5’, ‘CB27’, ‘8517’, ‘7964’, and ‘CB46’. We also used the salt tolerance model of van Genuchten and Hoffman (1984) to ascertain the salinity level resulting in a 50% decrease (C 50 value) in the growth parameter studied. Materials and Methods Experimental design. The cowpea experiment was set up as a standard split plot design, with salt as the main plot variable, and cowpea cultivar as the subplot variable. The experiment consisted of 12 cultivars of cowpea (Table 1) subjected to seven different levels of salinity (Table 2). The entire experimental design was replicated across three plots (1 plot = 1 sand tank, 21 sand tanks used in total). In all, there were 252 (12 × 7 × 3) observations for each yield response variable. Seeds were obtained from the Cowpea Research and Breeding Group, Department of Botany and Plant Sciences, University of California, Riverside. Planting. The 12 cultivars were planted in greenhouses in two rows in each sand tank based on a random cultivar map generated by SAS PLAN procedure. The greenhouses were located at Riverside, California (lat. 33°58'24"N, long. 117°19'12"W). The tanks measured 1.2 × 0.6 × 0.5 m deep and contained washed sand having an average bulk density of 1.2 Mg·m. At saturation, the sand has an average volumetric water content of 0.34 m·m. The plant were spaced 17 × 20 cm apart. Two seeds were sown a half-inch deep at each space on 18 July 2000. After gemination, the plants were thinned to one at each space. Growth conditions. The temperature, radiation, and humidity were automatically recorded hourly at a point slightly above the plant canopy. Over the course of the investigation the air temperature ranged 32 to 35 °C day and15 to 18 °C night. Relative humidity ranged from 43% to 52%. Plants were irrigated three times daily with a base nutrient solution made up with City of Riverside municipal water. The composition of base nutrient solution (BNS) for irrigation was a modifi ed (about 80% strength) Hoagland solution consisting of (in mol·m): 2.5 Ca(NO 3 ) 2 , 4.0 KNO 3 , 2.0 KCl, 3.0 NH 4 NO 3 , 0.36 KH 2 PO 4 , 1.5 MgSO 4 , 0.10 Fe as sodium ferric diethylenetriamine pentaacetate, 0.023 H 3 BO 3 , 0.015 MnSO 4 , 0.0012 ZnSO 4 , 0.0003 CuSO 4 , and 0.0001 H 3 MoO 4 . The BNS served as the control treatment. Each irrigation was of 5 min duration (three times daily) which allowed the sand to become completely saturated, after which the solution drained into 765-L reservoirs for reuse in the next irrigation cycle. Water lost by evapotranspiration was replenished automatically each day with deionized water to maintain constant electrical conductivities in the solutions. The pH was adjusted weekly using concentrated H 2 SO 4 and maintained between 6.5 and 7.5. Salt treatment. Salinization commenced 7 d after planting and continued for up to 6 consecutive days until the highest salt level was achieved. Equivalent amounts of salts were added incrementally each day to avoid osmotic shock to the seedlings. These irrigation waters were prepared to simulate the increasingly saline waters derived from Colorado River water using NaCl, CaCl 2 , and MgSO 4 . Final ion compositions are shown in Table 2. The fi nal electrical conductivities of the irrigation waters (EC i ) were 2.6 (control, the lowest salinity we obtained using tap water for the preparation of the irrigating solution), 3.7, 5.4, 8.3, 12.1, 17.3, and 20.1 dS·m. Biomass determination. Plants were harvested on 9 Sept. 2000, 53 d after planting when the cultivars with a normal fl owering period passed fl ower initiation stage. The harvested plants were immediately brought to the laboratory for processing. Root zone up to stem base was washed with tap and then deionized water to remove surface ions. The root surface was dried using paper. The washed plants were separated into leaf blades, stems including petioles, and roots, then placed into paper bags and oven-dried at 45 °C for 7 d. Before drying, the fresh leaf-blade area was measured using LI-COR 3100 area meter. The dried biomass was weighed using Sartorius Universal balance (accuracy = 0.01 g) for larger samples and Mettler AC 100 (accuracy = 0.0001 g) balances for smaller samples. Data analysis Salt-tolerance model. The data for each cultivar was analyzed using a salt-tolerance model proposed by van Genuchten and Hoffman (1984), y = δexp(αx – βx) [1] where y is biomass yield, x is salinity, and δ, α, and β are empirical parameters. Equation [1] describes the relationship between yield and salinity as a smooth, continuous equation. It can be converted into an ordinary regression equation using a log transformation, which facilitates stabilizing biomass variance. The standard split-plot analysis of variance (ANOVA) model Under a log transformation, Eq. [1] becomes log(y) = log(δ) + αx – βx + ξ = β 0 + β 1 x + β 2 x + ξ [2] which is a simple quadratic regression equation. In Eq. [2], ξ represents the residual error component, which is assumed to be independent and follow the standard normality. Additionally, β 2 is assumed to be ≤0. When multiple cultivars are being analyzed, Eq. [2] could be expanded into one of the two more general linear models shown below: log( y j ) = β 0j + β 1j x + β 2j x + ξ [3] log( y j ) = β 0j + β 1 x + β 2 x + ξ [4] where y j is the yield for the j cultivar. These three models can be categorized as follows: Model Eq. type Interpretation [3] Full interaction Unique salt-tolerance curves [4] No interaction Constant salt-tolerance curves having unique intercepts [2] No interaction Identical salt-tolerance curves (no cultivar differences) Table 2. Salt levels and composition for irrigation solution added to simulate increasing Colorado River water salt compositions. Solutions were prepared using City of Riverside municipal water with an EC value around 0.55 dS·m and having about (in meq·L): 3.3 Ca, 1.6 Na, 0.1 K, 0.83 Cl, 1.36 NO 3 , and 1.44 SO 4 , which were included in the salt composition. Salinity Osmotic level potential Salt composition (meq·L) (dS·m) (MPa) Ca Mg Na K SO 4 Cl 2.6 –0.08 8.3 3.0 1.6 6.1 4.4 2.8 3.7 –0.15 8.3 6.0 12.1 6.1 7.4 13.3 5.4 –0.19 11.1 12.1 22.5 6.1 13.5 26.5 8.3 –0.31 18.4 25.3 45.2 6.1 26.7 54.5 12.1 –0.47 26.0 38.8 68.5 6.1 40.2 87.4 17.3 –0.69 38.5 62.3 108.6 6.1 63.7 140.0 20.1 –0.82 44.0 77.8 135.6 6.1 79.2 172.7 FebruaryBook 1 226 12/14/05 10:59:13 AM 227 HORTSCIENCE VOL. 41(1) FEBRUARY 2006 The above various salt-tolerance models can be directly analyzed using the split-plot ANOVA model in SAS GLM procedure (SAS Institute, 1994).