LEAF GAS EXCHANGE IN TWO DWARF COCONUT GENOTYPES IN THE SOUTHEAST OF BAHIA STATE, BRAZIL.

Net photosynthetic (A) and leaf transpiration (E) rates and stomatal conductance to water vapour (gs) of Malayan Yellow Dwarf (MYD) and Brazilian Green Dwarf (BGD) coconut accessions (Cocos nucifera var. ‘nana’ L.) were studied and discussed in terms of the technical aspects related to light-response curves in field conditions. Measurements of gas exchange were performed during four days, in April and may 2000, at the Cocoa Research Center Experimental Station (Una - BA, Brazil). The A, gs and E parameters were significantly (P < 0.05) different between the two genotypes. The mean maximum values of A, gs and E were 10.4 and 12.0 µmol CO 2 m -2 s -1 , 0.21 and 0.35 mol H 2 O m -2 s -1 and 3.07 and 3.69 mmol m -2 s -1 for MYD and BGD, respectively. For both genotypes a good fitting of the light-2 response curve models were obtained, indicating that A and gs were dependent of the photosynthetically active radiation incident on leaf surface (Qi), in spite of high genotipic variation. Interesting results were achieved when an empirical multiplicative model was used. The model relating A or gs with Qi and with leaf-to-air water vapour pressure deficit inside the chamber (VPD L ) was tested for both genotypes and showed a negative influence of the latter on the stomatal behavior and consequently on A. Such effect was more pronounced in BGD than in MYD. These and others relationships involving leaf gas exchange and microclimatic variables in coconut palm trees are discussed


INTRODUCTION
The southeast area of Bahia State (Brazil), traditionally known as "cacaueira", is currently facing an unprecedented economical and social crisis, mostly due to the low international prices of the cocoa beans associated with the high incidence of the fungal "witches broom" disease in their crops. The necessity to develop other agricultural products in this region has been demanding a lot of information, mainly with respect to the yield potential of these alternative crops under a changing environment. In this sense, the cultivation of dwarf coconut genotypes for the production of tender coconut water has been turned into a profitable and widespread activity for the farmers of this and other Brazilian States.
The understanding of the physiological responses of the coconut to the environmental conditions is relevant when the objective is to reach high yields in different field conditions. Then, it is essential to identify the responses of the gas exchange parameters to environmental variables. Some information about responses of physiological parameters of coconut palm trees to environmental factors in Brazilian conditions are available for tall (Passos andSilva, 1990, Prado et al., 2001) and dwarf varieties (Cintra et al., 1992;Passos et al., 1999). In other countries, studies concerning photosynthesis and respiration Ramadasan, 1974, 1975;Shivashankar et al., 1982;Jayasekara and Jayasekara, 1995), water relations (Kasturi Bai et al., 1988;Shivashankar et al., 1991;Rajagopal and Kasturi Bai, 1999) and leaf gas exchange (Kasturi Bai et al., 1997;Braconnier, 1998) are also of great relevance for the management of tall varieties and hybrids.
The use of mathematical models to estimate stomatal and nonstomatal limitations of photosynthesis have been developed (Mansfield and Davies, 1985;Ball et al., 1987;Leuning, 1990;Thornley and Johnson, 1990). These models provide a simplified description of a real system, thereby generating information that would otherwise remain hidden and would help predictions of the results to a reasonable extent (Iqbal, et al., 1997). The use of empirical mathematical models, adjusted from simultaneous  measurements of gas exchange and environmental variables, allows one to approach the effects of those variable on gas exchange (Jarvis, 1976;Yang, et al., 1998;Day, 2000;Mielke et al., 2000).
The aim of this work was to study the leaf gas exchange parameters in two adult dwarf coconut genotypes growing in the southeast of Bahia State, Brazil, and to test the effectiveness of an empirical mathematical model that explains the responses of photosynthesis and stomatal conductance to water vapour to the light and to the leaf-to-air water vapour pressure deficit.

MATERIALS AND METHODS
The measurements were performed during the following four days 04/28, 05/12 and 25, and 06/02 of the year 2000, always from 8:00 to 15:00 h, in the Cocoa Research Center Experimental Station, in Una-BA (15º15'45"S, 39º05'49"W, 105 masl). The site of study employed included plants of cocoa (Theobroma cacao L.) and 'cupuaçu' (T. grandiflorum L.) as accessory species, and plants of the dwarf coconut (Cocos nucifera var. nana L., Palmae) genotypes Malayan Yellow Dwarf (MYD) or Brazilian Green Dwarf (BGD) as the main culture, in an 'intercrop' design. The times of measurements chosen aimed to verify the effects of leaf-to-air water vapour pressure deficit on the gas exchange parameters.
Light-response curves were obtained using a portable photosynthesis system LI-6400 (Li-Cor Inc., Nebraska, USA) equipped with an artificial light source 6400-02B RedBlue #SI-925. The "Light curve" routine of the software OPEN 3.3, in seven levels of photosyntheticaly active radiation (Q i , µmol photons m -2 s -1 ), 50, 100, 200, 400, 800, 1300 and 1600, in a decreasing order, was set. Each reading was recorded in an interval of time from 120 (minimum) to 150 s (maximum). A flow rate of 400 µmol s -1 was set, with a maximum coefficient of variation (C.V.) allowed of 1%, although most of readings has been saved with C.V. < 0.8%.
The rates of net photosynthesis (A, µmol CO 2 m -2 s -1 ) and leaf transpiration (E, mmol H 2 O m -2 s -1 ), and the stomatal conductance to water vapour (gs, mol H 2 O m -2 s -1 ) were estimated from values of CO 2 assimilation and of variation in the air humidity inside de chamber, both determined by the infrared gas analizer (IRGA) of the portable photosynthesis system. The instantaneous water use efficiency (WUEi, µmol CO 2 mmol H 2 O -1 ) was calculated by the quotient between A and E, for each reading. The rate of dark respiration (Rd, µmol CO 2 m -2 s -1 ) corresponded to values of A when Q i = 0 µmol photons m -2 s -1 . Values of internal (C i , µmol mol -1 ) and atmospheric CO 2 molar fractions (Ca, µmol mol -1 ), air and leaf temperatures (Ta and T L , o C) and leaf-to-air water vapour pressure deficit inside the chamber (VPD L , kPa) were supplied or calculated by the software of the equipment.
Eleven plants of MYD and eight plants of BGD, with 5 years old and 4 to 5 m of height were used for the estimation of the light response curves. In each plant, the measurements were made in one leaflet of the middle portion of the rank 9 leaf, counted from top down. In order to access the leaf let with the equipment, were used fixed and mobile towers.
For the maximum values of gas exchange, were considered only the readings obtained when Q i ? 800 µmol photons m -2 s -1 . Comparisons were accomplished between the means of measured microclimatic variables and between the mean maximum values of the leaf gas exchange variables, using the Turkey's procedure (P < 0.05).
The parameters derived from the light-response curves were obtained by applying empirical multiplicative models adjusted by the non-linear estimates routine of Statistica for windows, version 5.0 (StatSoft, Inc., 1995). The model y = f(Q i )f(VPD L ) was used, where y correspond to A or gs, following the equations: where A max is the light saturated rate of net photosynthesis (µmol CO 2 m -2 s -1 ), ? A is the apparent quantum efficiency of photosynthesis (µmol CO 2 µmol photons -1 ), gs max is the light saturated stomatal conductance to water vapour (mol H 2 O m -2 s -1 ), ? gs is the apparent quantum efficiency of stomatal conductance (mol H 2 O µmol photons -1 ) and VPD o is the values of VPD L when A (or gs) = 0.

RESULTS AND DISCUSSION
The microclimatic parameters were highly variable for the two genotypes (Table 1) as a consequence of the measurements being taken in different days and times of the day. VPD L , was significantly higher at the days of MYD measurements. Ta and Tf did not differ significantly. Such microclimatic situation was providential once one of the goals of this work was to verify the responses of A and gs to VPD L .
Significant differences between the two genotypes were observed for the maximum (Qi ? 800 µmol photons m -2 s -1 ) values of A and gs ( Table 2). The BGD genotype showed higher values of A (12.0 µmol CO 2 m -2 s -1 ), gs (0.35 mol H 2 O m -2 s -1 ), E (3.69 mmol H 2 O m -2 s -1 ) and Ci/Ca (0.75), whereas the MYD genotype showed high values of WUEi (3,98) and Rd (1.84 µmol CO 2 m -2 s -1 ). The maximum values of A obtained in this study were higher than others obtained for tall and hybrid coconut trees (Kasturi Bai et al., 1997;Braconnier, 1998). For example , Kasturi Bai et al. (1997) found values of A ranging from 6.48 to 8.26 µmol CO 2 m -2 s -1 for nonirrigated and irrigated plants, respectively. On the other hand, values of gs are lower than that obtained by Passos et al. (1999). These authors observed values of 0.48, 0.67 and 0.32 mol H 2 O m -2 s -1 for the gs of Jiqui Green Dwarf, Malayan Yellow Dwarf and Malayan Red Dwarf, respectively. However, they used 10 years old plants and a steady state parameter (LI-COR 1600) to assess the parameters, which may help to explain the differences observed in our data.
The values of Ci/Ca are good indicators for the stomatal limitation of photosynthesis at field conditions (Ehleringer and Cerling, 1995;Ishida et al., 1996;Mielke et al., 2000). The higher the values of Ci/Ca, the lower will be the stomatal limitation to the influx of CO 2 and WUE i in the species. It was verified that the values of Ci/Ca were relatively constant for Qi levels ? 800 µmol photons m -2 s -1 (Figure 1). The maximum values of Ci/Ca and WUEi (Table 2) indicate the two genotypes seem relatively conservatives with respect to water use when compared with other tropical tree species (Hogan et al., 1995;Ishida et al., 1996). The significant differences observed for the values of Ci/Ca between the two genotypes also suggest that the BGD is less conservative regarding water use than the MYD, although no significant differences in WUEi between the two genotypes were observed (Table 2). Nevertheless, both genotypes could be considered as being efficient with respect to the water use if compared with tall and hybrids coconut genotypes. For example, Kasturi Bai et al. (1997) found values of WUE ranging from 1.5 to 2.0 µmol CO 2 mmol H 2 O -1 for one tall variety and for two hybrids. The WUE data obtained in this work are contrasting to those obtained by Kasturi Bai et al. (1996), in which the hybrids are more efficient than the varieties and also that the MYD genotype is less efficient than tall and hybrid genotypes.
The light response curves of A and gs indicate different tendencies between the two genotypes ( Figure 2). For estimates of A, the tested models were corrected for the rate of dark respiration (Rd, µmol CO 2 m -2 s -1 ), estimated when Qi = 0 µmol photons m -2 s -1 (Table 2). Using the adjusted models, the values of compensation irradiance (?) were calculated as 48.4 and 32.5 µmol photons m -2 s -1 for MYD and BGD, respectively. The parameters derived from the light response curves are shown in Table 3. For both genotypes, these parameters were higher than the limits found for other tree species. For example, Barker et al. (1997), working with ten tropical tree species previously acclimated to the gap environment, observed values of quantum efficiency of A (? A ), compensation irradiance (??? and light saturated A (A max ) varying from 0.02 to 0.04 µmol CO 2 µmol photons -1 , 15.8 to 39.4 µmol photons m -2 s -1 and 2.2 to 10.1 µmol CO 2 m -2 s -1 , respectively. According to Luttige (1997), the limit values for ? and A max in sun plants of tropical forests are between 20 and 30 µmol photons m -2 s -1 , and 10 and 20 µmol CO 2 m -2 s -1 , respectively.
The incorporation of a linear function (y = 1 -VPD L /VPD o ) to the light response equations has been used successfully (Dolman, 1988;Dolman and Van Den Burg, 1988;McCaughey and Iacobelli, 1994;and Mielke et al., 1999) to estimate gs in different tree species, being an appropriated indicator of the stomatal sensibility to VPD L . For the adjustment of the models, 51 and 48 simultaneous readings of A, gs, Qi and VPD L were used for MYD and BGD, respectively, with the omission of the values when Qi = 0 µmol photons m -2 s -1 . The equations obtained were the following: For both genotypes, A and gs showed a different and negative response to the increases in VPD L , with gs ? 0 mol H 2 O m -2 s -1 when VPD L ? 11.4 and 3.6 kPa for MYD and BGD, respectively. Those data indicate that MYD seems to be more capable than BGD in maintaining the stomata open under conditions of higher values of VPD L , thereby maintaining a positive rate of carbon assimilation at the hottest hours of the day (Figure 3). The significantly higher A/gs value for MYD genotype (Table 2) reinforces the idea of a higher photosynthetic efficiency to each value of stomatal conductance in that genotype.
In summary, both dwarf coconut genotypes showed high photosynthetic rates and high water use efficiency when compared with talls and hybrids genotypes. The light-response curves obtained at field conditions indicated that A and gs were highly dependents of Qi. The application of a multiplicative empirical model relating A or gs with Qi and VPD L , was efficient in identifying useful integrative relationships among the data, such as the negative response of the stomata to increasing VPD L . Besides, the inclusion of a linear function relating the effects of VPD L on A or gs allowed a better use of the data collected in the field. The Malayan Yellow Dwarf genotype showed a higher photosynthetic efficiency than the Brazilian Green Dwarf for the same value of stomatal conductance, being capable of maintaining a positive rate of carbon assimilation at the hottest and more dry hours of the day. collaboration in the collection of data in the field, and Dr. Leandro L. Loguercio for valuable comments in the manuscript. The financial support was provided by UESC.      . Relationships between net photosynthesis (A) and phothosyntheticaly active radiation (Qi) and between stomatal conductance to water vapour ( gs) and Qi for Brazilian Green Dwarf (?) and Malayan Yellow Dwarf (o) at field conditions. The light response curves of A and gs are derived of t he models adjusted from the plotted values (for the equations, see text). Qi (µmol m -2 s -1 ) Figure 3. Relationship between net photosynthesis ( A) and leaf-to-air water vapour pressure deficit inside the chamber ( VPD L ) and between stomatal conductance to water vapour (gs) and VPD L for two genotypes of coconut palm at field conditions. Brazilian Green Dwarf (?); Malayan Yellow Dwarf (o).