Energy balance partitioning and evapotranspiration from irrigated Muskmelon under Semi-Arid Conditions

The Mossoró-Assu-Baraúna district, Rio Grande do Norte State (RN), is recognized by the intense production of horticulture, mainly muskmelon for export. However, this region is often devastated by intense droughts. Thus, the muskmelon production is predominantly under irrigated condition and, due to constant threat of water resources collapse on the region, a rigorous irrigation water management in the region is needed. The main objective of this article was to analyze the seasonal pattern of energy balance partitioning and evapotranspiration on irrigated muskmelon crop on the region around Mossoró-RN. The study was carried out in two areas of commercial production of muskmelons in the Mossoró-Assu-Barúna district, during two growth seasons from 2012-Jun to 2012-Nov. The components of energy balance AGROMETEOROLOGY Article Energy balance partitioning and evapotranspiration from irrigated Muskmelon under Semi-Arid Conditions Giuliana Mairana Morais de Sousa Vanomark1, José Espínola Sobrinho1, José Renato Cortez Bezerra2, Carlos Antonio Costa dos Santos3, Pedro Vieira de Azevedo3, Saulo Tasso Araújo da Silva4, Bergson Guedes Bezerra5* 1.Universidade Federal Rural do Semiárido Programa de Pós-graduação em Manejo de Solo e Água Mossoró (RN), Brazil. 2.Empresa Brasileira de Pesquisa Agropecuária Embrapa Algodão Campina Grande (PB), Brazil. 3.Universidade Federal de Campina Grande Unidade Acadêmica de Ciências Atmosféricas Campina Grande (PB), Brazil. 4.Universidade Federal Rural do Semiárido Departamento de Ciências Ambientais e Tecnológicas Mossoró (RN), Brazil. 5.Universidade Federal do Rio Grande do Norte Departamento de Ciências Atmosféricas e Climáticas Natal (RN), Brazil. *Corresponding author: bergson.bezerra@gmail.com Received: Oct. 31, 2016 – Accepted: Feb. 28, 2017 and evapotranspiration were determined by using the Bowen Ratio Energy Balance method. It was observed that more than 60% of the net radiation (Rn) was converted into latent heat flux (λE), while 21 and 11% of Rn was converted into sensible heat flux (H) and soil heat flux (G), respectively. The ratio λE/Rn varies according to the change of leaf area index (LAI) while the ratios H/Rn and G/Rn vary inversely with the LAI. The agreement λE/Rn and LAI is also evidenced by similarity between curves of crop evapotranspiration (ETc) and LAI, particularly when the melon crop reaches its maximum vegetative growth (LAI > 3). The muskmelon ETc ranged from 265 to 289 mm, values that are similar to those found by other researcher.


INTRODUCTION
Muskmelon (Cucumis melo L.) is a vegetable of great economic importance, which is cultivated in all Brazilian regions and in various parts of the world (Melo et al. 2011;Qi et al. 2015).Its cultivation has been widespread in tropical and/or semiarid regions where environmental conditions, particularly higher temperatures, low relative humidity and high insolation duration have favored its development.
Among the regions with these conditions stands out the Northeast region of Brazil, especially the Mossoró-Assu-Baraúna district, located on the northwest portion of the Rio Grande do Norte State (Medeiros et al. 2012).This district is recognized by the intense production of horticulture under irrigated conditions and it is the largest muskmelon producer in Brazil, accounting for more than half of all domestic production (Medeiros et al. 2011).
However, it is located in the Brazilian Semi-Arid region, which is often devastated by prolonged and intense droughts.The rainy season is limited to the first half of the year and about 80% of the total annual rainfall occurs in four months (from February to May).Therefore, the water deficit in the region can reach up to 800 mm•yr -1 (Bezerra et al. 2012).In this scenario, the muskmelon production in this region is predominantly under irrigated conditions, because the irregularity of rainfall does not allow commercial production on rainfed conditions.The groundwater is main source of water for irrigation in the region, which is pumped out from Jandaíra calcareous aquifer through wells of 100 m depth (Bezerra et al. 2012;2015).However, the expansion of irrigation projects has led to a substantial increase in groundwater demand in the region, leading to sharp downgrades of this aquifer to generate concerns about water security in the region and sustainability of agricultural production under irrigated conditions over long drought periods.Therefore, a careful and rigorous irrigation water management in the region is needed.
The fundamental requirement of scheduling irrigation is the determination of crop evapotranspiration (ETc).The ETc can be determined through numerous systems including lysimeters, micrometeorological methods, soil water balance, sap flow, and scintillometry (Allen et al. 2011;Bezerra et al. 2012;2015;López-Olivari et al. 2016).
The micrometeorological method based on Bowen ratio energy balance (BREB) is a relatively practical and reliable method.It has often been used to estimate the ETc of different soil-vegetation systems and in different climatic conditions, including the Brazilian Semi-Arid (Bezerra et al. 2012;2015).The energy balance explains the destination of the energy available to the system, i.e., net radiation (Rn) distributed among non-radiative soil surface flows, primarily soil cover, soil water content and the availability of solar energy control energy balance partitioning.In crops on irrigated conditions it has been reported that the portion of Rn converted into latent heat flux (or evapotranspiration) is greater than 70% (Bezerra et al. 2012;2015).
Given the concerns raised, the main objective of this study was to analyze the seasonal behavior of the energy balance partitioning, beyond evapotranspiration of muskmelon crop under irrigated conditions on the region around Mossoró-RN, Brazil.

MATERIAL AND METHODS
The study was conducted on the Mossoró-Assu-Baraúna district, located in the northwest portion of Rio Grande do Norte State (RN), Northeast region of Brazil (Figure 1).The climate of region , according to Thornthwaite (1948), is semi-arid, megathermal with water deficit during the year.The average annual rainfall is 674 mm, of which about 550 mm occur between February and May.The average annual relative humidity is 68.9%, while the average annual temperature is 27.7 °C, ranging from 27.2 °C in June to 28.4 °C in February.The soils of the area in which the experiments were carried out are predominantly red-yellow latosols (Embrapa 1971).
The experiment was conducted during two consecutives growing seasons: from June 18 to September 05 (1 st cycle) and from September 06 to November 13 (2 nd cycle) on the two studied areas.
The crop was sown in a greenhouse and 10 days after sowing (DAS) the seedlings were transplanted to the field, where the soil was protected with plastic mulch.At the time of transplanting, the rows were covered with white polypropylene agro-textile blankets (TNT), a procedure that aims to minimize the incidence of pests and diseases.Th e characteristics of the fi eld experimental plots are described in Table 1.
Th e crop was irrigated by using drip irrigation system.Th e emitters have the same plant spacing, so as to have one dripper per plant.Th e irrigation schedule adopted was the daily.
The growth season of muskmelon crop was identified based on the methodology proposed by FAO-56 (Allen et al. 1998) The components of the energy balance were determined by using BREB method.Neglecting the energy stored in the canopy and the photosynthetic energy flows, which represent less than 2% of net radiation (Rn), the energy balance expresses the energy conversion in mass flow and heat on crop (Eq. 1) (Perez et al. 1999;Bezerra et al. 2012;2015).where Rn is the surface net radiation, λE the latent heat flux, H the sensible heat flux and G the soil heat flux, all in W.m -2 .Rn was measured by using a NR-LITE net radiometer (Kipp & Zonen, Delft, The Netherland), installed about 2.5 m above the crop canopy.The values of G, in turn, were measured using two soil heat flux plate model HFP01SC Self-Calibration Soil Heat Flux Plate (Hukseflux Thermal Sensors, Delft, The Netherlands), buried in the soil to 0.02 m depth, one between plants under the mulch and the other between rows. (1) The values of λE and H were derived from the energy balance equation (Eq.1) and Bowen ratio concept (Bowen 1926;Perez et al. 1999): acquisition system CR10X (Campbell Sci, Logan, UT, USA) with energy supplied by a solar panel of 20 W. The height of the sensors remained unchanged during the crop growing since the changes on the muskmelon height was insignificant.
Th e partitioning of the available energy balance can be evaluated by analyzing the dimensionless evaporative fraction (Λ), defi ned as a ratio of latent heat fl ux to available energy fl ux, and it is usually used to characterize the energy partition over the land surface (Shen et al. 2004;Bezerra et al. 2015): where β is the Bowen ratio (Bowen 1926), calculated by following equation, only for daytime period with positive available energy (Rn -G > 0) (Perez et al. 1999;Bezerra et al. 2015).Th is equation is derived from empirical relationships between fl uxes and vertical gradients, assuming that turbulent diff usion coeffi cients for the transport of heat and water vapor are equal (Verma et al. 1978;Perez et al. 1999): where ΔT = T 2 -T 1 and Δe = e 2 -e 1 are above canopy verticals gradients of air temperature (°C) and vapor pressure (kPa), respectively, γ is psychometric constant (kPa•°C -1 ), which was calculated by using following equation: where c P is dry air specific heat at constant pressure (J•kg -1 •K -1 ), λ is latent heat of vaporization of the water (MJ•kg -1 ), which was calculated by using Eq.6, and P a is atmospheric pressure (kPa).
where T a is air temperature (°C).
The temperature and vapor pressure gradients were obtained from measurements of wet and dry temperatures measured in two levels above the canopy (0.5 and 2.0 m) by using two psychrometers constructed with thermocouple type T (copper-constantan).Electrical signals from the all sensors used in the measurement and/or computation of energy balance components were sampled every 5 s and storage averages every 20 min, through a data Daily ETc of muskmelon crop was obtained from daily values of λE (MJ•m -2 •day -1 ) (Eq. 2) converted to millimeters (mm) from value of latent heat of vaporization of the water (Eq.6).Reference evapotranspiration (ET 0 ) was calculated by the FAO-56 method (Allen et al. 1998), based on meteorological data collected on the weather station of National Institute of Meteorology (INMET), located about 400 m to experimental areas.
Th e consistency analysis of BREB data was performed according to the criteria established by Perez et al. (1999) and Payero et al. (2003).Extremely inaccurate H and fl uxes obtained by the BREB method occur when β ≈ -1, which correspond to the night-time period and to precipitation or irrigation events.Th us, data collected during these events or when -1.25 < β < -0.75, were eliminated as proposed by Perez et al. (1999), andPayero et al. (2003).Given these criteria, data deemed physically inconsistent were discarded, including cases that are outside the limits of instrumental resolutions.
Leaf Area Index (LAI) was measured weekly from the fi rst Day Aft er Transplanting (DAT) until the end of growing season totalizing eleven measurements during fi rst cycle in the Area 1 and nine measurements during second cycle of Area 1 and fi rst and second cycles of the Area 2. Leaf Area was measured using LI-3100C Area Meter (LI-COR, Lincoln, NE, USA).Th e LAI was derived from leaf area measurements and crop spacing.

RESULTS AND DISCUSSION
Total irrigation amount applied on the Area 1 was 583.9 and 468.0 mm during the two consecutive cycles, (2) respectively.In the Area 2, in turn, total irrigation was 293.7 and 352.9 mm for the first and second muskmelon cycle, respectively.
The mean monthly values of meteorological variables beyond monthly total precipitation observed during growing season of muskmelon on both cycles were showed in Table 2.Note that during 1 st cycle T air ranged from 26.7 °C to 26.9 °C while during 2 nd cycle T air ranged from 26.9 °C to 27.2 °C.This increase of the T air during 2 nd cycle in relation to 1st cycle was due to radiative effect because mean monthly solar radiation during 2 nd cycle was almost 1 MJ•m -2 •day -1 higher than 1 st cycle (Table 3).This increase occurred because during transition between two consecutives cycles occurred spring equinox in the Southern Hemisphere, when consequently the Southern hemisphere gets warmer.
The direct effect of T air increase is in establishing the length of the growing season (Bezerra et al. 2012;2015).Note that during 2nd cycle T air was almost 0.5 °C warmer than 1 st cycle.Consequently, 2 nd cycle was shorter than 1 st in both studied areas, according to Table 3 which shows the dates of the events that marked the changes of physiological stages of the crop and the length thereof in each assessed portion.On Area 2, 2 nd cycle was 4 days shorter than 1 st cycle, while on Area 1 the 2 nd cycle was 15 days shorter than 1 st cycle (Table 3).However, excessive extension of the 1 st cycle in Area 1 should not only be attributed to air temperature.During the ripening the muskmelon crop was infested by the whitefly pest.So, to minimize the crop yield losses, the producer extended the irrigation and consequently the crop growing season was lengthened.Note that late-season of the 1 st cycle of the Area 1 was almost twice late-seasons of others cycles.
Also according to Table 2 the demand atmospheric water during the 2 nd cycle was greater than during the 1st cycle, since ET 0 was almost 1 mm higher.It is also noted that the wind speed during the 2 nd cycle is greater than during the first cycle.The VPD, in turn, during the 1 st cycle (Figure 2a) ranged from 1.01 kPa (Jul-27) to 2.42 kPa in August17.During 2 nd cycle (Figure 2b) VPD ranged from 1.40 kPa (September 21) to 2.41 kPa recorded in October 5.

Area
Growing Figure 2 shows the behavior of the daily solar radiation (Rg) and VPD values during growing season of muskmelon crop.During first cycle (Figure 2a) the highest Rg values occurred on July 4 (32 MJ•m -2 •day -1 ), while on July 27 was registered the lowest Rg value, 13.2 MJ•m -2 •day -1 .During second cycle (Figure 2b) the highest Rg value occurred on November 2 (29.8 MJ•m -2 • day -1 ), while on October 15 was recorded the lowest value, 16.32 MJ•m -2 •day -1.
The accumulated ET 0 was 527.7 mm and 514.34 mm during 1 st (Figure 3a) and 2 nd (Figure 3b) cycles, respectively, showing the high atmospheric water demand of the local.The behavior of the daily ET 0 was similar to daily global solar radiation as well as VPD.The months of maximum water demand were August (1 st cycle) and October (2 nd cycle), whose daily mean values of ET 0 were of 7.2 mm and 7.5 mm, respectively.The minimum and maximum values of ET 0 recorded during 1 st muskmelon cycle were 3.9 mm•day -1 on July 27 and 8.6 mm•day -1 on September 03.During 2 nd cycle daily values of the ET 0 ranged from 6.2 mm on September 11 to 8.8 mm on October 6.
The applied irrigation on muskmelon crop in the two studied areas during two consecutive cycles in 2012 is shown in Table 4. Still in the Table 4 we can see the changes of the energy balance partitioning and of the leaf area index (LAI) during growing season of muskmelon crop.
Analyzing the data related to Area 1 in the Table 4, one can see that from Initial to Middle season the LAI increased from 0.01 cm 2 •cm -2 to 4.23 cm 2 •cm -2 during first cycle and from 0.05 cm 2 •cm -2 to 4.22 cm2•cm -2 during 2 nd cycle.In turn, the percentage of converted to lE increased from 58% to 73% and from 67% to 72% during 1 st and 2 nd cycles, respectively.On the other hand, the LAI decreased from 4.23 to 2.49 cm 2 •cm -2 during first cycle and from 4.22 cm 2 •cm -2 to 2.34 cm 2 •cm -2 during second cycle, from Middle to Late season, while lE decreased from 73% to 71% during first cycle, and from 72% to 62% in the second cycle.
In Area 2, also according to Table 4, the same behavior is observed on this account LAI increased from 0.01 to 3.54 cm 2 •cm -2 from Initial to Middle season, during 1 st cycle and from 0.18 cm 2 •cm -2 to 4.57 cm 2 •cm -2 during the 2 nd cycle.There was also an increase of the values lE from   58% to 66% (1 st cycle) and from 72% to 79% (second cycle).In contrast, there was a decrease in the LAI values from Middle to Late season (from 3.54 cm 2 •cm -2 to 2.51 cm 2 •cm -2 ) during 1 st cycle and from 4.57 cm 2 •cm -2 to 3.41 cm 2 •cm -2 during 2 nd cycle, while the lE/Rn decreased from 66% to 64% (1 st cycle) and from 79% to 66% (second cycle).
The increase lE/Rn from initial to middle season, in accordance with the increase of the LAI, is expected since in this period the muskmelon crop has high growth and development changing from seedlings (very low LAI) to middle season when it reaches the apex of its development during only about 50 days (Table 3).During middle season (from flowering to start of fruit maturity) the crop reaches the prime of their photosynthetic, physiological and metabolic activities and requires maximum water use (Bezerra et al. 2015).
Note that during Mid-season λE/Rn values were greater than 70%, except during the first cycle of the Area 2. This value is slightly higher than the values found in other studies in which was used drip irrigation system such as López-Olivari et al. (2016) 2010) found a ratio of 0.86 between λE and Rn for the full growing season of the soybean and corn crops near Ames, Iowa, Midwestern US.Teixeira et al. (2008) reported an average of 89% of the Rn was converted into λE and 11% was converted into H during flowering season of the mango crop in the semi-arid region of the São Francisco River basin, Northeast Brazil.In cultivation of crotalaria, near to Tottori City, Southwest part of Japan, Takagi et al. (2009) recorded partition 80% of the available energy (Rn -G) for ETc and only 20% for H.However, in this study the muskmelon crop was mulched and irrigated by using drip irrigation system whose dripper is located below mulch.Thus, the water loss is reduced and considerable portion of Rn is converted into H, which is almost double that the G/Rn (Table 4).Moreover, the reduction of LAI and the LE/Rn values during Late season occurs due to decreased water irrigation supply and muskmelon crop senescence.
In contrast, percentages of converted into G (G/Rn) and H (H/Rn) varied inversely with the LAI, i.e., decreased from Initial to Mid-season and increased from Middle to Late seasons.This behavior is physically expected since the values of G and H are controlled by the soil water availability and soil cover (Shen et al. 2004;Bezerra et al. 2015).Note that when the LAI values increased (from Initial to Mid-Season) G and H values decreases.During Late season the muskmelon crop is senescence, the leaves age and fall, and strongly decreases the ground cover.In this period H and G values increases.The lower G values occurred during Middle season, when LAI > 3.
The mean values of G/Rn values observed during two cycles in each area were 11%, ranging from 8% (Midseason) to 17% (Initial season).Borges et al. (2015), studying the muskmelon crop under irrigated conditions, reported that the portion of Rn converted in to G ranged from 20% (Initial season) to less than 10% (Mid-season), whose average was 14%.
Values of Λ reflect the condition of soil moisture in the root-zone, so that there is a direct relationship between them (Bezerra et al. 2013).According to Table 4 higher values of Λ was observed during Mid-season, period in which the crop requires maximum water supply, consequently maximum soil water content.
Figures 4 and 5 show the daily behavior of the energy components on the muskmelon crop during growth season.The maximum value of Rn and LE for Area 1 (Figure 4) takes place between 11:20 am and 12:20 pm local time on the 1 st cycle, and between 11:00 a.m. and 12:00 p.m., local time, during 2 nd cycle.In the area 2 (Figure 5) the maximum value of Rn and LE takes place between 11:40 am and 12:40 pm in the first cycle and between 11h00 am local times and 12:20 pm local time during second cycle.Similar behavior was found by Bezerra et al. (2015) on the cotton crop under irrigated conditions on the western region of Rio Grande do Norte State.
The maximum values of G occurred between 10:00 a.m. and 13:00 p.m. local time (Figures 4 and 5).The peak of the G values ranged in comparison with the other components, which may be explained by irrigation because despite being performed always in the morning period there was no standardization of times: sometimes occurred at the beginning, middle, or late in the morning.The influence of irrigation on G has been reported in the literature (for example, Abu-Hamdeh and Reeder 2000; Bezerra et al. 2012;2015).As the irrigation events occurred in the morning, soil water content in this period has always been higher than in the afternoon.According Abu-Hamdeh and Reeder (2000) increasing soil water content increases thermal conductivity and thus increases G.At about 12:30 pm local time, the soil water content decreased because of soil evaporation resulting in a decline in the G.
ETc of muskmelon crop was 362.2 mm and 289 mm on the Area 1 during first and second cycles, respectively, and 265 mm and 279 mm on the Area 2, during 1 st and 2 nd cycles, respectively.However, there was difference more than 70 mm between the first and second cycles of the Area 1.The reason for this high difference was the late-season lengthening due to extension of the irrigation and consequently extension of late-season of the Area 1 first cycle.The effect of irrigation extension is evidenced by LAI extension from eighth to ninth measurement (Figure 6a).
Except the ETc value of 1 st cycle of the Area 1, the other results are similar to values reported by Borges et al. (2015) (273.3 mm) and Sousa et al. (2000) (281 mm).
However, a comparison of these values with other studies is difficult because ETc values are influenced by many local factors such as climate, soil characteristics, cultural practices, water management, growing season length , and responds strongly to the magnitude of solar radiation received (Alberto et al. 2011;Bezerra et al. 2012).According to Figure 6, which shows the daily ETc, the maximum daily values were 6.7 and 6.2 mm•d -1 during two consecutive cycles on the Area 1 (Figures 6a and 6c), respectively, and 6.2 mm•d -1 and 7.1 mm•d -1 on the Area 2 during fi rst and second cycle (Figures 6b,d), respectively.Such daily amounts of ETc corroborate to results obtained by Borges et al. (2015) in the same region of Mossoro-RN.Melo et al. (2011), in order to determine the evapotranspiration and production of irrigated muskmelon Gália also in Mossoró-RN, found values of 6.5 mm•d -1 , while Oliveira et al. (2010) reported values of 5.2 mm•d -1 in the Juazeiro, north of Bahia State.Both studies were conducted in the Brazilian Semi-Arid region.Th e diff erences between the ETc daily maximum values in the current study and the values found in other areas of Semi-Arid region may be associated with several factors such as the high spatial variability and climate parameters (relative humidity, wind speed, and the air temperature) in the region.
Note that daily ETc values varied considerably.The maximum daily ETc values occurred due to increased soil evaporation, especially during the Initial season and Crop-development -when crop does not provide complete ground cover (LAI < 3, Table 4 and Figure 6).

CONCLUSION
Th e energy balance partitioning and evapotranspiration over the muskmelon crop under irrigated conditions on the Brazilian Semi-Arid region was observed for during two consecutive growing seasons and in two distinct areas.Th e seasonal variations and its relations throughout the

AKNOWLEDGEMENTS
We thanks to Coordination for the Improvement of Higher Education Personnel (CAPES) for granting a Ph.D. scholarship to the first author, to Embrapa (National Center for Cotton Research), to CNPq for providing Research Productivity Grant for the 4th and 5th authors and Federal Rural University of Semi-Arid to provide the Bowen ratio systems used.Still thanks to Dinamarca and Norfruit farms, which gently provided the respective production areas for carry out experimental campaigns and to INMET for providing the meteorological data used during the experimental campaigns.
, which are: Initial season = from transplanting to approximately 10% ground cover; Crop development = from 10% ground cover to effective full cover or start of flowering; Middle season = from start of flowering to start of maturity; Late-season = from the start of maturity to harvest or end of water use.

Figure 2 .
Figure 2. Daily values of the Vapor Pressure Deficit (VPD) and global daily solar radiation (Rg) recorded during the first (a); and second (b) Muskmelon cycle in Mossoró (RN), 2012.

Figure 3 .
Figure 3. Daily values of the reference evapotranspiration (ET 0 ) and rainfall recorded during first (a); and second (b) cycle of muskmelon on Mossoró(RN).

Figure 4 .
Figure 4. Mean daily of the energy balance fl uxes determined by BREB method in the fi rst (left panel) and second cycles (right panel), on Area 1, during the growing season of the muskmelon crop: (a) Initial season; (b) Crop development; (c) Mid-season and (d) Late-season in Mossoró (RN).

Figure 5 .
Figure 5. Mean daily of the energy balance components determined by BREB method in the first (left panel) and second (right panel) cycles, on Area 2, during the growing season of the muskmelon crop: (a) Initial season; (b) Crop development; (c) Mid-season and (d) Late-season in Mossoró (RN).

Figure 6 .
Figure 6.Seasonal variations of the ETc and LAI observed in the fi rst cycle of Area 1 (a) and Area 2 (b) and the second cycle of Area 1 (c) and Area 2 (d) in Mossoró (RN).
discussed for reaching the following conclusions: approximately 60% of the net radiation was used as latent heat (λE), 21% as sensible heat (H) and 11% as soil heat flux (G).The values vary according to the variations of the leaf area index (LAI), while and varied inversely to LAI.These results reveal important role of the muskmelon crop vegetative growth in the partitioning of the energy balance.The muskmelon ETc obtained from BREB on the climatic conditions of Mossoró-RN ranged from 265 to 289 mm, excluding the value of first cycle of Area 1, due to reason previously discussed.The values range is similar to values found by other researchers on the region.

Table 1 .
Description of the assessed plots in the two areas studied and the two muskmelon cycles: plot area (ha), cultivated variety, spacing (m), planting dates, placing and removal of blanket and early harvest in Mossoró-RN, 2012.

Table 2 .
Mean monthly solar radiation, air temperature, relative humidity, wind speed, vapor pressure deficit, reference evapotranspiration, and total monthly rainfall observed during muskmelon growing season on Mossoró, 2012.

Table 3 .
The growing season length of muskmelon in Mossoró-RN, 2012.

Table 4 .
Kool et al. (2016)ocated in the Pencahue Valley, Región del Maule, Chile, andKool et al. (2016)in a desert vineyard located central Negev highlands, Israel.However, it should be considered that Rainfall, irrigation, average values of energy balance partition, evaporative fraction and leaf area index for each growing season in each muskmelon crop cycle inMossoró-RN, 2012.