Leaf water relations of cotton in a free-air CO2-enriched environment

As part of an intensive study of crop response to CO 2 enrichment in a free-air CO2 enrichment (FACE) experiment in the field, we determined aspects of the water relations of a cotton crop on selected dates in 1991. The atmosphere was enriched from 370 #mol CO2 mo1-1 (control) to about 550 #mol mo1-1 in free air during daylight hours. Under full irrigation, CO2 enrichment decreased stomatal conductance and single-leaf transpiration only toward the end of the season, and these changes led to increased leaf water potentials only at that time of year. Under water-stressed (deficit irrigation) conditions, CO2 enrichment decreased conductance throughout the season but there was no corresponding consistent effect on leaf water potentials, As with the fully irrigated controls, CO2 enrichment increased leaf water potentials only at the end of the season. CO2 enrichment increased season-long biomass accumulation 39% under full irrigation and 34% under deficit irrigation. These results are consistent with previous studies of cotton in open-top chambers that found only small effects of CO2 enrichment on internal water relations of cotton, and no water stress-induced increase in crop responsiveness to elevated CO2. 1. I n t r o d u c t i o n The C O 2 concent ra t ion o f the a tmosphere has been increasing since the late * Corresponding author at: USDA-ARS, National Program Staff, Bldg. 005, BARC-West, Beitsville, MD 20705, USA. 0168-1923/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0168-1923(93)05033-O 172 N.C. Bhattacharya et al. / Agricultural and Forest Meteorology 70 (1994) 171-182 nineteenth century, and is expected to continue increasing for the forseeable future. This change is of interest to agriculture because CO2 is the raw material for the photosynthetic production of crop biomass, and availability of CO2 can be an important limitation to productivity. In most crops, atmospheric CO2 enrichment stimulates photosynthetic rate, growth, and yield, although the degree of stimulation varies depending on such factors as plant type, developmental stage, environment, and nutritional status (Kimball, 1983, 1985; Jones et al., 1984; Strain and Cure, 1986; Hileman et al., 1992; for recent reviews, see Bhattacharya, 1993; Bowes, 1993). In addition, CO2 enrichment often slows transpiration rate by partially closing stomata, thereby restricting the diffusion of water vapor from the leaf to the atmosphere (Bhattacharya, 1993; Bowes, 1993). Most studies of CO2 responses have been carried out in controlled environments (glasshouses or growth chambers). There are several reasons to question whether results from typical controlled environments might be applicable to the field; for example, growing plants in pots causes artificial restriction of rooting (Thomas and Strain, 1991), and the thermal environment is modified compared with that in the field (Radin et al., 1987). To overcome such deficiencies, some investigators have grown crops in CO2-enriched open-top chambers in the field (Rogers et al., 1984; Kimball et al., 1986; Bhattacharya et al., 1990; Kimball and Mauney, 1993). Results with cotton (Gossypium hirsutum L.) differed from those obtained in controlled environments. Over several years, the degree to which CO2 enrichment stimulated growth and yield in the field substantially exceeded the typical stimulation seen in controlled environments (Kimball and Mauney, 1993). Interpretation was difficult, however, because in these experiments there was a significant effect of the open-top chambers themselves that might have altered the results. One of the greatest world-wide limitations to crop yield is the occurrence of water stress (Boyer, 1982). Thus an additional question of great interest is the influence of water stress on crop responses to CO2. Most research on this topic has also been carried out in controlled environments (Sionit et al., 1980; Wulff and Strain, 1982; Morison and Gifford, 1984; Tolley and Strain, 1984). In general, high CO2 slowed transpiration and reduced the severity of the water stress; as a result, CO2 enrichment exerted a larger stimulatory effect on water-stressed plants than on unstressed plants (Dahlman et al., 1984; Dahlman, 1992). This subject is especially complex because of the dual effect of CO2 enrichment on photosynthetic assimilation and on transpiration; again, applicability of these findings to the field needs to be demonstrated. Kimball and Mauney (1993) studied water stress-CO2 interactions in open-top field chambers. As with the main effects of CO2 enrichment, the results again differed from those in controlled environments, i.e. water stress did not increase the relative stimulation by CO2. The free-air CO2 enrichment (FACE) system offers a means of studying direct CO2 effects, and CO2-environment interactions, on crop growth under natural conditions in an open field. Here we report studies of the water relations of CO2-enriched cotton at two levels of water supply. N. C. Bhattacharya et al. / Agricultural and Forest Meteorology 70 (1994) 171-182 2. Materials and methods 173 All experiments were carried out at the University of Arizona Maricopa Agricultural Center (33.07°N, 111.98°W, elevation 358 m). Seeds of cotton (Gossypium hirsutum L., cv. 'Deltapine 77') were planted on 16 April 1991 (day of year (DOY) 106) in rows 1.02 m apart in a 4 ha field with a drip irrigation system 0.18-0.24 m below the surface. The soil was a reclaimed Trix clay loam (fine-loamy, mixed (calcareous), hyperthermic Typic Torrifluvents). The field was divided into four blocks, each of which was split into two sub-blocks. One of the two sub-blocks was supplied with ample irrigation water throughout the cropping period ('wet' treatment; 1048 mm of water). The amount of water to be supplied was estimated as potential evapotranspiration from a grass reference crop (ETo) multiplied by LAI (leaf area index)/3 for LAI < 3; otherwise ETo. Deficit irrigations were initiated in the other sub-block on DOY 140, with water applied thereafter at 0.67 of the amounts for the wet treatment ('dry' treatment; 792 mm of water). Total precipitation from planting to harvest was 41 mm. Details of crop growth have been given by Mauney et al. (1994). Within each block, two rings of 25 m diameter each spanned the border between sub-blocks. One ring was a FACE array of vent pipes to maintain a CO2-enriched atmosphere. The other was an unenriched control ring. The distance between tings in each direction was 100-150 m. CO2 was injected into the FACE plots from 26 April (DOY 116) to 16 September (DOY 259) between 05:00 and 19:00 h at a set point of 550 #mol mo1-1 . The system for maintaining CO2 at the set point has been described elsewhere (Lewin et al., 1994). The ambient concentration of CO2 in the control rings was also monitored, and averaged 370 #mol mo1-1 (Nagy et al., 1994). Leaf water potential was estimated as the xylem pressure potential with a Scholander pressure chamber (Soil Moisture Equipment Corporation, Santa Barbara, CA), following procedures of Bhattacharya et al. (1991). Briefly, the most recently expanded mainstem leaf (fourth or fifth leaf) of a plant was shaded, enclosed in a plastic bag, and then excised with a sharp razor blade. The excised leaves in closed bags were placed in a humidified cooler chest and transported to a nearby laboratory for immediate measurement (Bhattacharya et al., 1990). For each sub-block, measurements were replicated on leaves excised from eight plants selected randomly. Stomatal conductance and leaf transpiration rate were determined with a Li-Cor LI-1600 steady-state porometer (Li-Cor Instruments, Lincoln, NE). For each leaf, abaxial and adaxial measurements were made sequentially and added. For each sub-block, measurements were replicated with 8-10 leaves. Apparent hydraulic conductance of the plants was estimated by a procedure dependent upon leaf water potential and transpiration rate. At steady state, the relationship between leaf water potential and transpirational flux can be expressed as J = LAk~ (1) in which Jis water flux, L is the hydraulic conductance between soil and leaf, and A~ is the water potential difference between the two sites (Radin et al., 1991). L was 174 N.C. Bhattacharya et al. / Agricultural and Forest Meteorology 70 (1994) 171-182 estimated as the slope of a plot of J on A g , assuming that qffsoil --0. Variation in J and A ~ was generated by time of day, with measurements taken from dawn to midday. Single-leaf transpiration rate was assumed to represent whole-plant flux (Radin et al., 1991). Plants were harvested for estimation of crop dry weight from a specially designated 'destructive harvest ' area of the plots (Mauney et al., 1994). They made up one-third of the plants from 3 m of row on each harvest date. To avoid sampling bias, sections of row and individual plants were selected for harvest on DOY 130, while the plants were still seedlings. Tagged individual plants were harvested, separated into components, and dried for 10 days at 60°C to obtain dry weights of roots, stems, and bolls. All data reported are averages of the four blocks (replications), together with their corresponding standard errors. We also estimated the probabilities of differences between treatments (t-test). 3. Results and discussion 3.1. L e a f water relations Leaf water potentials at 06:00 h Mountain Standard Time (MST) (predawn)