GAS EXCHANGE IN DIFFERENT VARIETIES OF BANANA PRATA IN SEMI-ARID ENVIRONMENT

This study aimed to evaluate gas exchange of banana prata in two production cycles in semiarid environment. Six cultivars were used as treatments arranged into a completely randomized design with five replications and four plants per plot. For physiological characteristics, it was considered a factorial arrangement of 6x14x2, six cultivars, 14 periods (months), two readings, 8h and 14h in each period. The rates of gas exchange, the carboxylation efficiency and the instantaneous efficiency of water use were higher at 8h and lower at 14h, with rare exceptions. The ‘BRS platina’ had a higher leaf temperature, higher transpiration and lower water use efficiency. ‘Prata-Anã’, ‘FHIA-18’ and ‘Maravilha’ expressed lower leaf temperature and lower transpiration. The ‘Maravilha’ is the most efficient in water use. Transpiration increases linearly with the leaf temperature, while the instantaneous efficiency of water use decreases linearly.


INTRODUCTION
The cultivation of banana tree is significant in agricultural systems in the agro-ecological zones of the tropics (AzEvEDO et al., 2010) with great economic and social importance.in Brazil, the 'Prata', 'Pacovan' and 'Prata-Anã' cultivars are the most widespread. in the Southwest of Bahia (DONATO et al., 2009) and in the North of Minas Gerais, the most planted cultivar is the 'prata-Anã' that despite the high commercial value, it is susceptible to the yellow and black sigatoka, and panama disease.
in the search of solutions, the Brazilian program for Banana plant improvement, coordinated by Embrapa Mandioca e Fruticultura, developed banana 'prata' hybrids with different degrees of commercial acceptance.
The physiological characters evaluation may be important for recommending cultivars because it allows establishing a genotypic variation of physiological responses of banana to the environment (TuRNER et al., 2007).The existence of this variation would infer on changes in transpiration rates, stomatal conductance and photosynthesis as physiological indicators of the presence of stress (LuCENA, 2013); in addition it contributes to the identification and the selection of superior individuals.The extrapolation of these results may subsidize specific production systems for various cultivars.
Additionally, physiological characteristics studies are quite common in cultivars of Cavendish type (ROBiNSON; GÁLAN SAÚCO, 2012), however they are scarce in cultivars of 'prata' type prevalent in Brazil.Given the above, the aim of this study was to evaluate the physiological characteristics of six 'prata' type bananas in two production cycles in semi-arid environment.

MATERIAL AND METHODS
The experiment was established in the area of instituto Federal Baiano, Campus Guanambi, in the State of Bahia, Brazil.The original soil is classified as typical dystrophic Red-Yellow Latosol, weak A, medium texture, hypoxerophytic caatinga phase, flat to mildly hilly relief, with annual average of precipitation and temperature, 680 mm and 26 °C respectively.
in planting, on 05/11/2010, we used seedlings that were acclimatized in plastic bags, 30 cm tall, and planted in 3.0 x 2.5 m spacing.The introduction and cultivation followed the recommendations for the crop (RODRiGuES et al., 2008).The plants were irrigated by micro sprinkling with Netafim® self-compensating emitters, flow 120 L h -1 , wet diameter of 7.4 m, with red nozzle of 1.57 mm, spacing of 6 m between side lines and 5 m between emitters.
The irrigations were based on the evapotranspiration reference (ETo) determined daily by the penman-Monteith method, and on the data from an automatic weather station vantage pro integrated Sensor (Davis instruments, Wayward, CA, EuA) located 100 m from the area.The crop coefficients to determine the ETc were defined according to the phenological stages of the crop.
We evaluated the gas exchanges, the leaf temperature and incident radiation on the third or fourth leaf (leaf three or four) counting from the apex to the base, with the help of Lcpro+® Portable Photosynthesis System (ADC BioScientific Limited, uK) infrared gas analyzer (iRGA), always with the radiation shield facing the sun, with ambient temperature and irradiance, and airflow of 200 ml min -1 .
There were 14 monthly evaluations in two reading times, 8h and 14h, covering the period from October 2010 to November 2011, corresponding to the early flowering of the first cycle to the beginning of the harvest of the second production cycle.
We measured the incident radiation in the leaf For statistical analysis of the data of the characteristics evaluated, we adopted the following: a) For the vegetative and yield characteristics, we used six treatments, the cultivars arranged in a randomized design.The data were submitted to variance analysis and the averages compared by Tukey test at 5% error probability (p <0.05) in SAEG software (SAEG, 2009).b) For the physiological characteristics, we adopted the arrangement in factorial 6x14x2, six cultivars, 14 evaluation periods (months) and two reading times each period, arranged in a completely randomized design.The data were submitted to analysis of variance and proceeded to the split of interactions according to their significance.The F and Turkey test compared the averages from those variables (p<0.05) for reading time factor and cultivars factor, respectively; and the Skott-Knott criterion grouped them for the evaluation period factor (months).We also realized correlation studies between the different variables and the adjusted linear models for those significant and of greater magnitudes.

RESULTS AND DISCUSSION
The differences in physiological characteristics evaluated in several months did not allow a grouping of the averages by the Scott-Knott criterion (p<0.05) according to the periods of the year.The grouping was random; probably because of the assessments made with devices are pointwise values, influenced by weather conditions of the moment (SANTOS et al., 2013).These results contradict the expectation of obtaining physiological differences grouped according to the periods, to the two times of reading, which would make it possible to compare the physiological responses of different cultivars in months with similar climatic characteristics.
The physiological variables, leaf temperature (T leaf ), transpiration rate (E) and instantaneous efficiency of water use (A/E) varied with the cultivar regardless of the month and reading time (Table 1).
The 'BRS platina' showed higher Tleaf (37.39 o C) and 'Prata-Anã' showed the lower (35.90 o C), with a small percentage variation of 4.15%.The transpiration rate (E) varies similarly to Tleaf, with the higher value (7.04 mmol H 2 O m -2 s -1 ) measured on the 'BRS platina' and the lower 6.19; 6.40 and 6.41 mmol H 2 O m -2 s -1 on the 'prata-Anã', 'Maravilha' and 'FHIA-18', respectively.The instantaneous efficiency of water use (A/E) in leaf was higher (3.45 µmol CO 2 m -2 s -1 /mmol H 2 O m -2 s -1 ) in the 'Maravilha' and lower (2.96 µmol CO 2 m -2 s -1 / mmol H 2 O m -2 s -1 ) in the 'BRS platina'.There is a tendency of direct relation between the T leaf and E and a contrary relation between the T leaf and the A/E, proven by the correlation study (Figure 1).The higher A/E observed in 'Maravilha' indicates the highest cultivar efficiency regarding the use of water resource, which is the main obstacle in banana production.Productive efficiency and better use of water do not contradict the recommendation for a new cultivar that requires other desirable characteristics, such as market acceptance despite of the observation meets the research needs to understanding the mechanisms of tolerance to drought (VANHOVE et al., 2012;MUTHUSAMY et al., 2014;KiSSEL et al., 2015).
The photosynthetically active radiation incident on the leaf surface (Qleaf), leaf temperature (Tleaf), transpiration rate (E), the instantaneous efficiency of water use (A/E), and the quantum efficiency of photosynthesis (A/Q leaf ) also varied with periods and reading times regardless of the cultivar (Table 2).
The Q leaf varied between times in most of the months (64.28%) and the higher values were recorded in the morning.Significant changes were observed between the months in the two times of evaluation.The higher value (1.650,94 µmol photons m -2 s -1 ) was recorded in February 2011 and the lower (485.65 µmol photons m -2 s -1 ) was recorded in October 2010 coincident with a cloudy day and presence of rain because it is a typical month of high radiation.The higher value recorded is between the radiation recommended, 1.500 and 2.000 µmol photons m -2 s -1 , and the lower value is in the range where photosynthesis is severely reduced, below 1.000 µmol photons m -2 s -1 (TuRNER et al., 2007).
The Tleaf recorded in 'prata' type banana varied between times, in all the evaluated months, regardless of the cultivars, except in January and May 2011 (Table 2).inall cases, the lower Tleaf occurred at 8h and the higher occurred in the afternoon, as observed by Donato et al. (2013).Significant differences were also observed between the months in the two times, regardless of the cultivar.The Tleaf had a percentage variation of 43.23%, the lowest value was 30.60 ° C and the highest value was 43.83 o C.
The transpiration rates evaluated in 'prata' type banana differ between times in every evaluation month, regardless of the cultivar (Table 2), with the occurrence of lower values in the morning and of higher values in the afternoon at 78.57% of the cases.variations were also observed between the months in both times, regardless of the cultivar.in October 2010, at 14h, the lowest transpiration rate was recorded (3.58 mmol H 2 O m -2 s-1), coincident with cloudy and rainy day illustrated by the lower radiation (485.65 µmol photons m -2 s -1 ) ( The efficiency of water use (A/E) in the leaf of 'prata' type banana ranged between times in every month except for January 2011 that always presented the higher values in the morning (Table 2) because of the higher photosynthetic rates registered at 8h (Table 4) and the lower transpiration (Table 2).There was also significant variation between the months in both times.
The photochemical efficiency of photosynthesis represented by the relation between photosynthesis and photosynthetically active radiation (A/Q leaf ) incident on the third leaf of 'prata' type banana, showed little variation between the months evaluated at 8h (Table 2). in the second time the variation was higher.No significant differences were recorded between times in most months.in all cases, the highest photochemical efficiency was observed at 8h, except in October 2011. in C3 plants, the quantum photosynthesis productivity is raised close to 30 °C and decreases a lot, particularly in banana trees above 34 oC (ROBiNSON; GALÁN SAuCO, 2012), which explains the lower quantum efficiency at 14:00 attested by higher leaf temperature values that were measured at this time.
All the cultivars showed significant variation in gs between the months and in the two times evaluated (Table 4).The stomatal conductance differed between times in 47.61% of the cases.These differences were observed for all cultivars in October and November 2010, and in January and May 2011 in most cases, probably because these were the hottest months of the evaluated period (Table 2).
in most of the cases, the highest values of gs occurred in the morning and the lowest in the afternoon, which can be justified by the occurrence of winds in the morning (DONATO et al., 2012) that contributes to the removal of the boundary layer and consequently decreases resistance.The stomatal conductance is the inverse of stomatal resistance to the steam.its value refers to the potential amount of water that could flow over the leaf surface and it is different from steam flow (or transpiration).The reduction of gs in the afternoon can also be related to the stomatal sensitivity to the air aridity, strongly influenced by temperature.Its values decrease with the increasing of steam pressure deficit and temperature.Factors that influence gs interfere in the acquisition of carbon in plants.
The cultivars did not express differences in gs values in both times every month evaluated, with the exception of November 2010, September and November 2011, at 8h, and November 2011, at 14h (Table 5).Ekanayake et al. (1994) argue the cultivars that restrict the stomatal conductance in drought conditions are considered "water economic."Thus obtaining cultivars with increased tolerance to abiotic stresses passes the identification of characteristics that grants drought tolerance (RAvi et al., 2013) as transpiration efficiency (KISSEL et al., 2015).
it is unlikely that the soil water content may change significantly between measurements, but the leaf steam pressure difference to the air will increase as the temperature increases.Therefore, the variation in stomatal conductance between cultivars is a strategy for the detection of stomatal sensitivity to the steam pressure deficit, which may or may not be related to drought tolerance.Turner et al. (2007), in their review, state that the gas exchange of the leaf is a more sensitive method for determining the response of banana tree to water deficit compared to traditional volumetric or thermodynamic measurements of leaf water status, such as the relative water content.
The change in gas exchange in banana leaves due to the stomatal response to water deficit, steam pressure deficit, leaf temperature and the intensity and quality of solar radiation have been reported in several studies.Different studies show that the stomatal of banana trees can respond to low relative humidity, as well as the reduction of soil moisture, and there is a genetic variation among cultivars, in relation to this feature (EKANAYAKE et al., 1994;THOMAS et al., 1998;KiSSEL et al., 2015).
Recorded changes in stomatal conductance (gs) (Table 3), photosynthesis rate (A) (Table 4), carboxylation efficiency (A/Ci) (Table 5), transpiration rate (E) and efficient use of water (A/E), due to the change of the radiation (Q Leaf ) and leaf temperature (T leaf ) (Table 2) were also reported by Donato et al. (2013).
The photosynthetic rate of all cultivars was different between the months in both reading times (Table 4).The values were higher at 8:00 and lower at 14h in most months (66%), except for the 'prata-Anã' in November 2011 (Table 5).
The cultivars showed the same rate of photosynthesis in each reading time in all evaluation periods, except in October 2010 and in November Rev. Bras.Frutic., v. 38, n. 2: (e-600) DOI 10.1590/0100-29452016600 March/April 2016 Jaboticabal -SP 2011, at 8h, and in June and in November 2011 at 14h (Table 5).
The photosynthetic rate, stomatal conductance (gs) and the efficiency of water use (A / E), in November 2010, July, August, September and November 2011 were lower at 14:00 than at 8h, despite subjected to the same radiation (Q leaf ).Even with the difficulty of separating the effect of the radiation changing (Q leaf ) and leaf temperature (T leaf ) in gas exchange in field experiments (LUCENA, 2013), the present study data allow us to state that the decrease in rates, observed at 14h, is the reflection of increasing temperature, as observed by Donato et al. (2013).The leaf temperature varied a percentage of 43.23%, the highest value, 43.83 °C recorded in November 2010 at 14h, to the lowest, 30.60 °C, in September 2011 at 8h (Table 2).
The rise in temperature increases evapotranspiration demand and directly influences all metabolic and physiological processes of the plant.probably, the rise in temperature affects the functioning of the enzyme system at a higher intensity than the stomatal closure because the transpiration rate (E) increased at 14h for all cultivars in most of the months (Table 2).However, the ecophysiological behavior results from the balance of the various environmental factors (DONATO et al., 2013).This is also proven by the decrease of A/E and the increase of the transpiration on linear basis with the increase of temperature (Figure 1).if transpiration increased, logically there would not be stomatal restriction.
The ratio A/Ci is a measure of the rubisco carboxylation efficiency and its decrease expresses a shift toward oxygenized activity.This characteristic varies between months, in both times for all cultivars (Table 5).The relation A/Ci of the cultivars was similar between reading times, in most of the evaluation months (60.72%). in 39.28% of the cases, the values were always higher at 8:00 and lower at 14h, except for the 'BRS platina' in December 2010.proof of the Rubisco carboxylase activity change to oxygenize, with the temperature increase observed from 8:00 to 14:00, is in the reduction of ration between photosynthesis and internal CO 2 concentration (A/Ci) observed in 39.28 % of the cases.Considering the percentage difference, regardless of differences detected by the Tukey test (p <0.05), they increase this variation to 67.85% for all cultivars and months.
The cultivars evaluated showed similar A/ Ci relation.The averages did not differ between the evaluated times and month, except in October 2010 and June 2011 (Table 5).The highest value (0.13 µmol CO 2 m-2 s -1 /µmol CO 2 mol -1 ) was recorded in 'BRS platinum' in January 2011 at 8h and the lowest (0.04 µmol CO 2 m -2 s -1 /µmol CO2 mol -1 ) in Jv42-135 evaluated in February and May 2011 at 14h.This relation may clarify the factors that limit the photosynthesis, analyzing the adjusted curve of A/Ci.
The great temperature for carboxylation of the prevailing CO 2 in plants with C 3 photosynthetic mechanism, such as the banana, is around 22 °C, while the great temperature for growth and development is approximately 27 o C (ROBiNSON; GALÁN SAÚCO, 2012).The balance between carboxylase and oxygenize activities of Rubisco is ruled by its kinetics, temperature and concentration of CO 2 and O 2 substrates.under environmental CO 2 concentration, the increase in temperature modifies the kinetic constants of rubisco, and increases oxygenation rate preferably to carboxylation, consequently increases photorespiration and decreases net photosynthesis.
The plant can respond differently to environmental conditions, proved by the maintenance of the photosynthesis rate (A) (Table 4) and stomatal conductance (gs) (Table 3), in the evening, most of the cultivars, in December 2010, March, April and June 2011.Despite the increase in leaf temperature (T leaf ), the transpiration rate (E) and the reduction in water use efficiency (A/E) (Table 2), we did not observe decrease in carboxylation efficiency (A/Ci) at 14h for most cultivars, which shows the greatest effect on the intensity of the radiation (Qleaf) in gas exchange.
Seneviratne et al. ( 2008) also found a reduction in photosynthesis rate when the radiation level decreased while studying the different levels of shading and photoinhibition under conditions of high intensity of light.probably the same happened in this study, in February 2011, when all cultivars were subjected to a higher radiation qleaf (1650.94µmol photons m -2 s -1 ) at 8h they had minor photosynthetic rates for the period, except the 'Prata-Anã' (Table 4).
The photosynthesis rate is increased by the presence of growth organs in the plant and reduced with the increase of shadowing or senescence (age) of the leaf.The lowest value of A (8.28 µmol CO 2 m -2 s -1 ) was recorded in May 2011 at 14h in Jv42-135 and the highest (27.10 µmol CO 2 m -2 s -1 ) recorded in January 2011 at 8h in 'Maravilha'.The photosynthesis showed a variation of 227.29%, proving that the photosynthetic rates of banana trees can reach values of 25 to 30 µmol CO 2 m -2 s -1 (TuRNER et al., 2007;ROBiNSON;GÁLAN SAÚCO, 2012).
The studied cultivars expressed the same photosynthetic rate at 8h and 14h on all evaluated months, with the exception of October 2010 at 8h, June 2011 at 14h and November 2011 at 8h and 14h (Table 4).
Several authors established associations between gas exchange of plants and the weather.Field studies reveal the integrated effects of environmental conditions on the physiology of banana trees, so correlations between these answers and climatic factors indicate trends, since there is influence of unmeasured factors.Higher precision in the associations between gas exchange and climatic factors is obtained in environments with controlled conditions (ROBiNSON; GÁLAN SAÚCO, 2012).Nevertheless, vanhove et al. (2012) state that experiments conducted in vitro and in greenhouse increase the experimental control, but it has lower physiological relevance compared to field studies.
The banana trees type 'prata' presented, in November 2010, higher photosynthetic rates, stomatal conductance and transpiration compared to other months, regardless of the cultivar.The higher transpiration values (E), 10.11 and 11.96 mmol H 2 O m-2 s -1 , in the two times, 8h and 14h, respectively, were due probably to the higher volume of rainfall recorded in the month (276.50 mm) that consequently increased soil moisture.The maximum Tleaf, 39.23 o C and 43.83 ° C, observed at 8h and 14h, respectively, are a result of the ambient temperature above 34 ºC, stressful for the banana tree (ROBiNSON; GÁLAN SAÚCO, 2012), as well as the second highest radiation level (Q leaf ) 1298.11 µmol photons m -2 s -1 and 1212.36 µmol photons m -2 s -1 recorded in the morning and afternoon respectively.The high transpiration rates illustrate the plant cooling mechanism to relieve the heat stress caused by higher soil moisture.
The correlation study between the variables showed direct association, significant and of high magnitude, only between the rate of transpiration and leaf temperature of banana trees type 'prata' cultivated in semi-arid environment (Figure 1).The adjusted linear models estimate an increase of 0.70; 0.47; 0.48; 0.42; 0.50 and 0.44 units in transpiration for each additional unit on the leaf temperature of cultivars 'Maravilha', BRS FHiA-18, FHIA-18, BRS Platina, 'Prata-Anã' and JV42-135 respectively.
The correlation study also established an inverse association, significant and high magnitude, only between the relation A/E and the Tleaf (Figure 1).The adjusted linear regression models predict a decrease of 0.26; 0.26; 0.37; 0.32; 0.29 and 0.28 units in the relation (A/E) for each increase of one unit in Tleaf in the cultivars 'Maravilha ', ' BRS platinum ', The increase in leaf temperature, due to the air temperature rise, increases transpiration and reduces water use efficiency for all crops, as reported by Donato et al. (2013).Changes in transpiration rates show stomatal opening as a cooling mechanism and showed that the reduction in photosynthetic rates in warmer times was more influenced by enzymatic involvement caused by temperature increase than by stomatal closure.Lucena (2013) found a high, positive and significant correlation (p <0.05) between transpiration and the air temperature to 'BRS Platina' and 'Prata-Anã'.The author also found a high correlation, inverse and significant (p <0.05) between the A/E and the air temperature.in his study, without water limitation, 'BRS platinum' and 'Prata-Anã' had the same productivity (14,375 kg.ha -1 ) in the first cycle and maximum values for the next 20.6 and 22.2 µmol CO 2 m -2 s -1 , respectively.in this study under the same conditions (local, level, population, spacing and original soil class), but with higher soil fertility and better management of plants, the maximum values of A for 'BRS platinum' and 'Prata-Anã 'were higher, 26.44 and 25.52 µmol CO 2 m -2 s -1 , respectively.The yield in the first cycle was higher, 25.4 and 21.3 t.ha -1 , respectively.
For DaMatta (2007), the reduction of production is associated with a decline in photosynthetic rates induced by low water availability in the soil, either by a direct effect of dehydration in the photosynthetic apparatus or by an indirect effect through stomatal closure, which restricts the absorption of CO 2 .Santos et al. (2013) evaluated 'Tommy Atkins' mangos under different irrigation regimes in the same region of this study and concluded that the gas exchange influenced the growth, development and production and they were related to water conditions of the plant, depending on the soil water status and weather conditions.
The showed data highlight the significant influence of temperature on gas exchange, either in a direct way with protein denaturation or indirectly by the sensitivity of stomatal to the effect of steam pressure deficit.However, correlations between these characteristics are not always found.

TABLE 2 -
Radiation incident on the leaf surface (Qleaf), leaf temperature (Tleaf), transpiration rate (E), water use efficiency (A/E), and photochemical efficiency of photosynthesis (A/Qleaf), evaluated on the third leaf of banana trees type 'Prata' in the first and second production cycle, at 8h and 14h inGuanambi, BA, 2010-2012.

TABLE 3 -
Stomatal conductance (gs), mol H evaluated on the third leaf of banana tree type 'Prata' in the first and second production cycle, at 8h and 14h in Guanambi,BA, 2010BA,  -2012.  .

TABLE 4 -
photosynthesis rate (A), µmol CO , evaluated on the third leaf of 'Prata' type banana in the first and second production cycle, at 8h and 14h in Guanambi,BA, 2010BA,  -2012.  .

TABLE 5 -
Efficiency of carboxylation (A/Ci), µmol CO , evaluated on the third leaf of 'Prata' type banana in the first and second production cycle, at 8h and 14h, in Guanambi,BA, 2010BA,  -2012.  .