Kim Huxman Department of Biological Sciences, 4505 Maryland Parkway, Box 454004, University of Nevada, Las Vegas, Las Vegas, Nevada 89154–4004, USA. Fax: (702) 895–3956; e-mail: firstname.lastname@example.org
While investigations into shoot responses to elevated atmospheric CO2 are extensive, few studies have focused on how an elevated atmospheric CO2 environment might impact root functions such as water uptake and transport. Knowledge of functional root responses may be particularly important in ecosystems where water is limiting if predictions about global climate change are true. In this study we investigated the effect of elevated CO2 on the root hydraulic conductivity (Lp) of a C3 perennial, Larrea tridentata, and a C3 annual, Helianthus annuus. The plants were grown in a glasshouse under ambient (360 μmol mol–1) and elevated (700 μmol mol–1) CO2. The Lp through intact root systems was measured using a hydrostatic pressure-induced flow system. Leaf gas exchange was also determined for both species and leaf water potential (ψleaf) was determined in L. tridentata. The Lp of L. tridentata roots was unchanged by an elevated CO2 growth environment. Stomatal conductance (gs) and transpiration (E) decreased and photosynthetic rate (Anet) and Ψleaf increased in L. tridentata. There were no changes in biomass, leaf area, stem diameter or root : shoot (R : S) ratio for L. tridentata. In H. annuus, elevated CO2 induced a nearly two-fold decrease in root Lp. There was no effect of growth under elevated CO2 on Anet, gs, E, above- and below-ground dry mass, R : S ratio, leaf area, root length or stem diameter in this species. The results demonstrate that rising atmospheric CO2 can impact water uptake and transport in roots in a species-specific manner. Possible mechanisms for the observed decrease in root Lp in H. annuus under elevated CO2 are currently under investigation and may relate to either axial or radial components of root Lp.
Little attention has been paid to the functional aspects of root water uptake and transport under elevated atmospheric CO2. Some species respond to elevated CO2 with increased leaf water potential (ψleaf) indicating an increase in root water uptake and/or increased whole plant water use efficiency (Rogers, Runion & Krupa 1994). This is not always the case, however, as root hydraulic conductance was found to decrease for soybean plants grown in elevated CO2 (Bunce 1996), and whole plant water uptake was shown to decline in chrysanthemums (Gislerod & Nelson 1989). These findings may indicate that the capacity for water acquisition by roots may be altered when plants are grown under elevated CO2 (Norby 1994; Berntson & Bazzaz 1996). In nutrient uptake studies, BassiriRad et al. (1996, 1997) examined the impact of elevated CO2 on nutrient uptake and found that the capacity for nutrient uptake is affected by growth under elevated CO2. The potential for elevated CO2 to cause changes in the dynamics of water flow through roots therefore has important implications for plants if predictions for increased atmospheric CO2 are correct.
To increase our understanding of plant responses to global climate change, it will be important to understand below-ground responses to elevated atmospheric CO2. Several recent review articles (Stulen & den Hertog 1993; Rogers et al. 1994; Norby 1994; Gregory, Palta & Batts 1997) indicate that knowledge of root responses may be extremely important in understanding whole plant responses to elevated CO2. Growth at elevated CO2 has been shown to influence root dry mass, root : shoot (R : S) ratio, root number, root length and relative growth rate in some species (Rogers et al. 1994). Several researchers have suggested that a CO2-induced increase in root biomass would lead to increased soil exploration by roots (Gifford 1979; Curtis et al. 1990; Norby 1994; Rogers et al. 1994), potentially enhancing nutrient and water acquisition (Del Castillo et al. 1989). While these parameters are important to plant growth, they relate to the structural rather than functional aspects of roots under CO2 enrichment. Studies of root function may provide unique insight into how plants respond to increasing atmospheric CO2.
To answer questions about root water uptake, the physiological capacity of whole root system water absorption must be investigated. Estimating the efficiency by which roots take up and move water (root hydraulic conductivity, Lp) may provide helpful insight into functional root responses to elevated CO2 (Saliendra & Meinzer 1992). Species from water-limited areas could be greatly impacted by changes in root Lp. In this study we used an annual species Helianthus annuus and a perennial species Larrea tridentata from relatively water-limited systems to test the hypothesis that elevated CO2 would influence the ability of roots to take up and transport water.
METHODS AND MATERIALS
Plant growth environment
Helianthus annuus (Asteraceae) L. seedlings were planted in an environmentally controlled glasshouse under ambient (360 μmol mol–1) and elevated (700 μmol mol–1) atmospheric CO2 at the University of Nevada, Las Vegas in June 1997. The CO2 concentrations were under computer control and measured each minute with an infrared gas analyzer (Li-Cor Model 6252; Li-Cor, Lincoln, NB, USA); CO2 was injected into the air-circulation system as necessary (Huxman et al. 1998). This resulted in the maintenance of CO2 levels within ± 5% of target values at all times. The glasshouse temperature was maintained at 35 ± 5 °C with a minimum relative humidity of 10 ± 5% in the afternoon and a maximum or 60 ± 5% in the morning. Seeds of H. annuus were surface sterilized in 5% bleach, rinsed and placed in distilled water for 5 h. The seeds were germinated in a mixture of equal parts of peat moss and fine sand in 1 L conetainers. Plants were watered every other day with a 1/40 strength Hoagland's solution (Hoagland & Arnon 1950) to prevent soil drying and maintain a constant nutrient status. Plants were grown for 37 d until they had approximately six to eight leaves. At this stage of development the plants had not fully explored the soil volume.
Larrea tridentata (Zygophyllaceae) Cav. seedlings were grown from seed in 1 L conetainers for 1 month and then transplanted into 15 cm diameter × 1 m tall PVC piping in a 80 : 20 mix of sand and silt. The plants were grown for 1 year within the two CO2 treatment rooms (360 and 700 μmol CO2 mol–1) as described above. Growth conditions were the same as for the H. annuus seedlings. The plants were watered twice a week with a 1/40 strength Hoagland's solution.
At each sampling, the photosynthetic rate (Anet), stomatal conductance and transpiration were determined with a Li-Cor 6400 (Li-Cor). Leaf water potential (ψleaf) was measured on a terminal shoot with a Scholander-type pressure chamber for L. tridentata only (Soil Moisture Stress Inc., Santa Barbara, CA, USA). Stem diameter and plant heights were also measured. All measurements were made on four to six individuals from each treatment. Following Ψleaf and gas exchange measurements, the plants were harvested. Leaf area was determined with a Delta-T area meter (Delta-T Devices, Cambridge, England). Above-ground structures were immediately placed in a drying oven at 70 °C for 48 h. Roots were placed on fine mesh and carefully rinsed to remove the majority of the soil. Root length was quantified after harvest by measuring the longest roots of each intact root system. Intact root systems were transported to the lab for Lp estimation (see below). Roots were dried for dry mass determination (48 h at 70 °C) following Lp estimation.
Soil moisture measurements
Because soil moisture is known to influence root Lp (Lopez & Nobel 1991) we monitored the soil water content throughout the experimental period to maintain constant soil moisture content. Gravimetric soil water content of H. annuus was measured on a weekly basis for 6 weeks on plants that were also being harvested for biomass quantification (n = 31). Briefly, approximately 30 mg of soil was collected at a depth of 6 cm. Soil samples were dried in a 70 °C oven until the soil mass stabilized, indicating that all soil water had evaporated (approximately 2 weeks). For L. tridentata (grown in 1-m-tall pots), gravimetric soil water content was measured weekly during the experiment. For these measurements, soil samples were collected at depths of 30, 60 and 90 cm at the first two samplings and at 60 and 90 cm at the last two samplings. Soil samples (30 cm, n = 12; 60 cm and 90 cm, n = 22) were oven dried as described above. For H. annuus, there were no differences in soil gravimetric water content between the ambient CO2 treatment (5·33 ± 0·57% moisture) and the elevated CO2 treatment (4·93 ± 0·54% moisture) indicating that differences in soil moisture can be ruled out as an interacting factor influencing the root Lp measurements. For L. tridentata, there were no differences in percentage soil moisture between ambient and elevated CO2 treatments at the 30 cm (ambient, 3·25 ± 0·60; elevated, 3·92 ± 0·49), 60 cm (ambient, 7·43 ± 1·55; elevated, 6·04 ± 1·55) or 90 cm (ambient, 6·27 ± 2·01; elevated, 3·65 ± 2·96) soil depths.
Pressure chamber apparatus
A hydrostatic pressure-induced flow method was used to estimate the Lp of excised H. annuus and L. tridentata root systems. Excised roots were carefully washed free of soil and sealed into a six-port pressure chamber with the cut end of the stems protruding through the chamber lid (Smit & Stachowiak 1988). The pressure chamber contained 20 L of 1/100 Hoagland's solution (22 °C) that was continuously aerated throughout the Lp measurements. Stems were girdled with a razor blade to prevent flux of phloem exudate passing into the collection tubes. Tygon tubing was sealed around the cut end of the stem for timed collections of xylem sap into microfuge tubes.
Hydraulic conductivity measurements
Pressure gradients ranging from 0·07 MPa to 0·35 MPa in 0·07 MPa increments were applied to the excised root systems of H. annuus and L. tridentata (n = 4 to 9 per treatment) in the pressure chamber apparatus. Xylem sap was collected under pressure-induced flow for approximately 5 min at each pressure in pre-weighed microfuge tubes containing a small piece of cotton to prevent the exudate from evaporating during collection. Exudate flux was calculated from the difference in pre-collection/post-collection weight of the tubes corrected by collection time and normalized by either leaf area (m2) or root dry mass (g). This enabled us to estimate water flux (Jv) by unit leaf area (functional measure of hydraulic efficiency) and by unit root dry mass. Hydraulic conductivity was determined from the slope of the linear regression generated from a plot of applied pressure versus Jv (Lp in m3 m–2 s–1 MPa–1 for leaf area and m3 g–1 s–1 MPa–1 for root dry mass).
A first order regression analysis was used to estimate root Lp for each individual. A two-factor analysis of variance (ANOVA) was used to test the effects of CO2 environment (treatment), time (block) and the interaction of CO2 and time on all measured parameters. Significance was tested at α = 0·05.
The response of whole root system Lp to elevated atmospheric CO2 was found to differ between species (Table 1). The Lp of H. annuus roots grown in elevated CO2 was approximately one-half that of plants grown in an ambient CO2 environment on both a leaf area and root dry mass basis (Table 1, Fig. 1). Root Lp was expressed on both a root dry mass and leaf area basis. The root Lp of L. tridentata grown in ambient CO2 was not significantly different from elevated-CO2-grown plants on either a leaf area or root dry mass basis (Table 1, Fig. 2).
Table 1. . Hydraulic conductivity (Lp) of whole root systems of Helianthus annuus and Larrea tridentata on a unit leaf area and unit root dry mass basis. Data are means (± SΕ) of six to ten plants
For H. annuus there were no differences in any of the growth components measured. Above-ground (ambient, 0·32 ± 0·02 g; elevated, 0·32 ± 0·02 g) and below-ground (ambient, 0·22 ± 0·02 g; elevated, 0·23 ± 0·02 g) dry mass, R : S ratio (ambient, 0·69 ± 0·03; elevated, 0·74 ± 0·04), leaf area (ambient, 28·3 ± 0·8 cm2; elevated, 26·7 ± 0·8 cm2), root length (ambient, 27·4±1·4cm; elevated, 26·2 ± 1·5 cm), shoot height (ambient, 17·9±0·6cm; elevated, 18·2±0·7cm) and stem diameter (ambient, 2·54 ± 0·03 mm; elevated, 2·69 ± 0·03 mm) did not differ between the ambient and elevated CO2 treatments.
Similar results were found for growth parameters in L. tridentata. There were no significant differences between the two CO2 treatments in above-ground dry mass (ambient, 0·50 ± 0·10 g; elevated, 0·49 ± 0·13 g), below-ground dry mass (ambient, 0·30 ± 0·08 g; elevated, 0·27 ± 0·10 g), R : S ratio (ambient, 1·90 ± 0·46; elevated, 1·82 ± 0·57) or leaf area (ambient, 22·6 ± 4·2 cm2; elevated, 17·0 ± 5·2 cm2).
Net photosynthetic rates of H. annuus seedlings were not significantly altered by the elevated CO2 treatment (Table 2). While elevated CO2 had no significant effect on instantaneous measurements of net photosynthesis in H. annuus, there was a CO2 by time interaction (d.f. 4,30; F = 4·84; P < 0·05). No significant differences were found between the two CO2 treatments in stomatal conductance or transpiration for H. annuus (Table 2).
Table 2. . Photosynthetic rate (Anet), stomatal conductance (gs) and transpiration (E) of Helianthus annuus and Larrea tridentata at ambient (360 μmol mol–1) and elevated (700 μmol mol–1) CO2. Data are means (± SE) of four to six plants
For L. tridentata, Anet increased while gs decreased under elevated CO2 (Table 2). There was also a time by treatment interaction (d.f. 1,15; F = 13·9; P < 0·05) and an effect of measurement time (d.f. 1,15; F = 6·19; P < 0·05) on gs for L. tridentata. Transpiration decreased significantly due to the elevated CO2 treatment in L. tridentata (Table 2) and both measurement time (d.f. 1,15; F = 22·8; P < 0·05) and time by CO2 treatment (d.f. 1,15; F = 10·8; P < 0·05) had an effect on E. The ψleaf of L. tridentata was affected by CO2 treatment (d.f. 1,21; F = 7·96; P < 0·05); plants grown under elevated CO2 had slightly higher ψleaf than ambient grown plants (ambient, –2·89 ± 0·25 MPa; elevated, –2·31 ± 0·02 MPa). Measurement time had a significant effect on ψleaf where values became less negative over time (d.f. 1,21, F = 30·58, P < 0·05).
Growth under elevated atmospheric CO2 had contrasting effects on the physiological capacity for water uptake and movement in H. annuus and L. tridentata. We found a two-fold decrease in root Lp under elevated CO2 in H. annuus but no change in root Lp for L. tridentata. Similarly, in a study with soybean and alfalfa, whole plant hydraulic conductance was found to decline under elevated CO2 (Bunce 1996). In the Bunce (1996) study, it was suggested that homeostatic adjustments occur between transpiration rate and hydraulic conductance under elevated CO2. In our study, the lack of response in E and gs in H. annuus does not coincide with the observed decrease in root Lp. Therefore, the decline in root Lp in H. annuus is not a result of lower E as was found to occur in alfalfa and soybean (Bunce 1996). For L. tridentata, there was a significant decline in gs as well as reduced rates of E, but there was no decrease in root Lp. Bunce (1996) reported similar findings for alfalfa grown under elevated CO2. It is possible that the perennial habit of both alfalfa and L. tridentata may play a role in the absence of CO2-induced response of root water uptake and transport.
In this study, root length, leaf area, shoot height, above- and below-ground biomass and stem diameter were measured to determine if water flux is related to growth responses of H. annuus and L. tridentata. No changes in any growth parameter were found for either species. Similar findings have been reported for root dry mass, root length and R : S ratio for H. annuus (Rogers et al. 1994). In contrast to our study, BassiriRad et al. (1997) found that L. tridentata increased in both below- and above-ground biomass under elevated CO2 but this could be due to the shorter duration of their experiment. While a common response to doubling of CO2 is increased R : S ratio, there are examples where species are non-responsive, including one study with H. annuus where no change occurred in R : S ratio (Carlson & Bazzaz 1980). In many cases the lack of response has been attributed to temperature and/or nutrient interaction (Carter & Peterson 1983; Bazzaz 1990).
Another common response to growth under elevated CO2 is increased photosynthetic rates (Bazzaz 1990), but many species exposed to long-term elevated CO2 lose the stimulatory response (Sage, Sharkey & Seemann 1989). In this study, Anet did not increase in H. annuus under elevated CO2. In L. tridentata, there was a small but significant increase in Anet. While increased Anet is often associated with increased growth, it is possible the increase in Anet in this study may not have been sufficient enough to cause a growth response in L. tridentata. The overall lack of stimulatory effect of elevated CO2 on the growth response of H. annuus and L. tridentata could be a result of the well-watered conditions the plants were exposed to in our study. Preliminary results from our current investigation into the response of H. annuus to an elevated-CO2–drought interaction support this hypothesis (Huxman & Neuman, unpublished results). In wheat, Gifford (1979) found that elevated CO2 only enhanced root dry mass when drought was imposed. Alternatively, the lower conductivity of the H. annuus roots grown under elevated CO2 may have restricted the transport of water to the shoots for growth.
While we have not yet determined the mechanism for the decline in Lp under elevated CO2 in H. annuus, it is possible that cell wall characteristics may be important factors in the alteration of water uptake and movement in this species. In Liquidambar styraciflua, elevated CO2 induced an increase in trachied wall thickness and a decrease in the lumen diameter of stems (Conroy et al. 1990). Decreased conduit diameter could potentially impact axial conductance. A decline in conduit diameter by a magnitude of one leads to a four-fold decrease in the ability of roots to transport water (Nobel 1991). Thus, a CO2-induced decrease in root hydraulic diameter could have a large impact on root Lp. This hypothesis is currently under investigation. Alternatively, growth under elevated CO2 could cause alterations in xylem number, which could also impact axial conductivity. To date, no studies have been performed to confirm that anatomical properties play a role in changes in root water transport induced by elevated CO2. However, studies comparing hydraulic properties in contrasting species and cultivars have shown that differences in root hydraulic conductances were due to morphological and/or anatomical factors (Schulte, Gibson & Nobel 1987; Saliendra & Meinzer 1992; Gallardo et al. 1996). Temperature, water deficit and mineral nutrition are also known to alter root hydraulic properties (Newman & Davies 1988; North & Nobel 1991; Saliendra & Meinzer 1992; Harris & Bassuk 1995; BassiriRad et al. 1997).
In conclusion, the capacity for water uptake decreased in H. annuus but was unaltered in L. tridentata when these species were grown under elevated atmospheric CO2. Investigations comparing perennial and annual species may be important for predictions of functional root responses to a changing environment. While this study provides some insight into plant functioning under elevated CO2, future studies should consider developmental stages to gain a more complete picture. Future studies should also consider the influence of root architecture changes under elevated atmospheric CO2, as the location of the roots within a given soil profile may ultimately determine the potential for water uptake (Rogers et al. 1996).
The authors thank Travis Huxman for his contribution to data collection and valuable comments on drafts of this manuscript. Katrina Salsman is gratefully acknowledged for assisting with watering plants. We also thank Paul Schulte, Janet Reiber, Erik Hamerlynck and Fred Landau for their help. This research was made possible by the financial support of a NSF EPSCoR grant to the state of Nevada and by the NSF/DOE/NASA/USDA-TECO NSF grant IBN-9524036 to S.D.S. This project was also partially funded by a NSF EPSCoR WISE Research Assistantship to K.A.H.