Water stress and forage production in Tetrachne dregei Nees , Panicum coloratum L . and Eragrostis curvula ( Schrad ) Nees

1 INTA EEA Anguil “Ing. Agr. Guillermo Covas”, Ruta Nac. 5, km 580, CC11 (6326) Anguil, La Pampa, Argentina. Facultad de Agronomía, UNLPam, Ruta 35, km 335, CC300 (6300) Santa Rosa, La Pampa, Argentina. Address Correspondence to: Lic. María de los Ángeles Ruiz, e-mail: mruiz@anguil.inta.gov.ar, fax (54) 2954-495057. Recibido/Received 17.X.2007. Aceptado/Accepted 22.X.2007. Abstract. Tetrachne dregei, Panicum coloratum and Eragrostis curvula are perennial forage C4 grasses, introduced from South Africa to the pampean semiarid region. This work was carried out to compare water stress tolerance; forage production and quality of T. dregei, P. coloratum and E. curvula. Studies were conducted under greenhouse and field conditions. In the greenhouse, watering was stopped after eighty one days of plant emergency in the water stress treatment. Water potential (ψ), stomatal resistance (SR) and shoot and root weights were evaluated. Under water stress, ψ diminished earlier and SR increases were higher in P. coloratum than in T. dregei and E. curvula. Plant survival in T. dregei and E. curvula was higher than that of P. coloratum. Under field conditions (INTA,Agricultural Experimental Station Anguil, La Pampa, Argentina), biomass production of T. dregei was lower than that of the other species (p<0.05) during the first year, but forage production was higher (p<0.05) in E. curvula, followed by T. dregei, than in P. coloratum in the following years. In spring, P. coloratum showed a greater forage digestibility (p<0.05) than E. curvula and T. dregei; P. coloratum and T. dregei had more Revista Internacional de BOTANICA EXPERIMENTAL


INTRODUCTION
Rainfall in the Pampean semiarid region decreases from NE to SW with a high inter-annual variation.Since soils have a low water retention capacity (Casagrande & Vergara, 1996), it is important to include water stress resistant grasses as a rancher management strategy.Tetrachne dregei Nees (Green grass, Td), Panicum coloratum L. (Kleingrass, Pc) and Eragrostis curvula (Schrad) Nees (Weeping lovegrass, Ec) are C 4 perennial forage grasses, introduced from South Africa and neighbouring countries.It is important to compare T. dregei performance under water stress conditions with that of E. curvula and P. coloratum.This is because of the well-known perfomance of the last two perennial grasses in the Pampean semiarid region (Cairnie, 1984;Covas, 1991;Petruzzi et al., 2003;Ruiz et al., 2004).
Eragrostis curvula shows a high performance under semiarid conditions (Voigt, 1991;Colom & Vazzana, 2001, 2003), but it has good forage quality characteristics for cattle grazing during a short time period (Stritzler et al. 1996).Panicum coloratum was introduced with the purpose of obtaining a better forage quality (Stritzler et al., 1996;Petruzzi et al., 2003).
Rainfall limitation and soil with low water retention capacity cause water deficit in plants; decreases of water potential can affect seriously leaf growth, CO 2 exchange rate, tillering and root/shoot ratio in several plant species (Echenique & Curveto, 1986;Boyer, 1995;Ben Haj & Tardieu, 1997;Brevedan et al., 2004;Blum, 2005).Under water stress, plants can increase stomatal resistance and reduce transpiration in seconds (Pugnaire et al., 1994;Colom & Vazzana, 2003).Plants also present other mechanisms to control transpiration such as leaf abscission or increases of leaf specific density (Pugnaire et al., 1994).
Our working hypothesis was that T. dregei shows a water stress tolerance similar to that in E. curvula, which is well known because of its good performance in semiarid environments.Panicum coloratum was included in this study because of its good forage quality.The objectives of this work were to compare the water stress tolerance, forage production and quality of T. dregei, E. curvula and P. coloratum in a semiarid environment.

MATERIALS AND METHODS
G Gr re ee en nh ho ou us se e t tr ri ia al l. .A greenhouse study was conducted at the Facultad de Agronomía, Universidad Nacional de La Pampa, Santa Rosa, La Pampa Province (36°34´ S, 64°16´ W; 210 m.a.s.l) from March 2002 to October 2002.Greenhouse environmental conditions were: average temperature 17 °C, humidity 69%, average photosynthetic photon flux density 525 mmol/m 2 /s.Seeds were sown in PVC pots, filled with 5 kg of regional soil (entic haplustol).After 30 days, only one plant was kept in each pot.
A Complete Randomized Design with a factorial treatment (3 species X 2 water conditions; 6 replicates) was used.Treatments were water stress and control.In the water stress treatment, plants were first watered near field capacity and thereafter watering was withheld during 81 days after plant emergence.At the same time, control plants were maintained near field capacity throughout the experimental period.At the time water was withheld, number of tillers was 1-3 in T. dregei, 1-4 in E. curvula and 2-4 in P. coloratum, and number of expanded leaves in the main stem was 5 in T. dregei and E. curvula, and 8 in P. coloratum.
Stomatal resistance (SR) was determined weekly after withholding water by using a Delta T Porometer, AP4-UM-2, 2.28 Version, 1991.At the same time, leaf water potential (ψ) was evaluated with a Scholander pressure chamber.
Leaf area was measured at the end of the experiment using a LICOR, LI 3000 leaf area meter.Shoot and root dry weights were determined.This allowed calculation of specific leaf area (leaf area/leaf weight) and root/shoot ratios.

RESULTS AND DISCUSSION
G Gr re ee en nh ho ou us se e t tr ri ia al l. .The interaction water level x species was significant (p<0.05) for ψ and SR in all five dates.This was because E. curvula showed the least differences in the control and water stress treatments.In P. coloratum, ψ and SR of watered plants showed significant differences (p<0.05)compared to plants of the water stress treatment at 24 days after withholding water (DAWW).At 61 DAWW, ψ of water-stressed plants showed values lower than -1 MPa, and SR increased to 80 s/cm (Fig. 2 and 3).Leaf wilting was observed between 57 and 66 DAWW.
Water potential and SR in control plants of T. dregei showed significant differences (p<0.05) in comparison to stressed plants after 24 and 35 DAWW, respectively.From 45 to 60 DAWW, however, stomatal resistance was not different between treatments.Plants under water stress started to wilt gradually by 104-157 DAWW; these plants showed a 40 day-delay in exhibiting this symptom in comparison to plants of P. coloratum.Water potential and SR of T. dregei showed the highest differences among individual plants at the wilting stage.
Water potential and SR showed significant differences (p<0.05) between treatments after 65 days of water withholding in E. curvula.This occurred 41 days later than in the other two species.Plants of E. curvula wilted by 93-106 DAWW, almost simultaneously with T. dregei.However, differences among individual plants were lower than those in T. dregei.Similar to results in our study, Colom & Vazzana (2001, 2003) found high stability of water potential, stomatal conductance, photosynthetic rates and leaf pigments content in cultivars of E. curvula exposed to water stress.
Leaf area of T. dregei and P. coloratum was significantly higher (p<0.05) in control than water-stressed plants.Eragrostis curvula, however, did not show significant differences between both treatments.While mortality of some mature leaves was observed, young leaves were kept alive in water-stressed plants of T. dregei.This is similar to results of Blum (2005) in Sorghum.Because of this, green leaf area of T. dregei plants was lower under water stress than in the control treatment.These plants, however, showed a higher level of stomatal conductance and ψ for a longer time period than P. coloratum.In 2005, a dry year (543 mm), E. curvula showed the highest yield (p<0.05) among all the study species.This is similar to results in 2003 (Fig. 5).It has been reported that cultivars of E. curvula can differ in their response capacity to water stress (Echenique & Curvetto, 1986).
During spring (September to November), forage digestibility was higher in P. coloratum than in E. curvula and T. dregei, which had similar  values for this parameter (Table 1).Panicum coloratum and T. dregei showed higher protein concentrations than E. curvula (Table 1).
During summer (December to April), although significant differences in DMD, ADF and metabolizable energy were not detected (p>0.05), E. curvula reached higher fiber concentrations and lower dry matter digestibility than the other two species.Tetrachne dregei showed the highest (p<0.05) protein concentrations.Stritzler et al. (1996) found that T. dregei generally produces better forage quality than other warm-season perennial grasses in semiarid Argentina.In previous experiments at different locations of the pampean semiarid region, Ruiz et al. (2004) reported higher protein concentrations in green and differed forage in P. coloratum than in E. curvula.
Water stress and forage production in C 4 grasses C Co on nc cl lu us si io on ns s. .Under water stress, E. curvula exhibited a better performance than P. coloratum and T. dregei.However, T. dregei showed more forage quality than E. curvula and it presented good response to water stress.Panicum coloratum exhibited the lowest drought resistance, but its forage quality was superior to that of E. curvula and T. dregei.Also, it showed a higher forage production during the first growing period.Since T. dregei showed a good water stress resistance and forage quality, it should be considered an important forage resource in semiarid regions.However, it is necessary to point out that T. dregei has a slow initial growth, while at the same time the initial growth of P. coloratum is good.Ruiz MA et al., ΦYTON 77 (2008) 14).The experimental design was a complete randomized block with eight replicates.Sowing was performed in November 2001.Plots (2.1 x 3.0 m each) had row interspaces of 0.70 m.Rainfall during the study is shown in Fig. 1.For 2003 and 2005, annual rainfall was 449 or 543 mm, respecti-vely; these values were lower than or similar to the historical long-term mean, respectively (Meteorological Service, EEA Anguil).Clipping of forage was made in April and November of 2002, 2003 and 2004, and in November of 2005.It was performed in the central row (two linear meters) in each plot, and fresh weight was determined.Afterwards, 200 grams were oven-dried at 60 ºC to constant weight.In November 2003 and April 2004, the following forage quality determinations were made: acid detergent fiber (ADF), dry matter digestibility (DMD), metabolised energy (Mcal/kgDM) and protein concentration (PC).Data were statistically analysed by ANOVA and LSD test (p<0.05).

Fig. 3 .
Fig. 3. Effect of water stress on leaf stomatal resistance (s/cm) (SR) of T. dregei, P. coloratum and E. curvula during the water stress period.DAWW: days after withholding water.

Table 1 .
Forage quality of T. dregei, P. coloratum and E. curvula in spring and summer 2003, Anguil, La Pampa.Different letters in the same column, within each season, indicate significant differences, LSD (p<0.05).