Irrigation of ‘Prata-Anã’ Banana with Partial Root-Zone Drying in a Semi-Arid Environment

Abstract

Considering the uncertainty of rainfall and prolonged droughts in semiarid regions, optimizing water management through techniques like partial root-zone drying (PRD) is crucial for sustainable banana production. This study aimed to evaluate the ‘Prata-Anã Gorutuba’ banana under irrigation by PRD. The experimental design was randomized blocks with five irrigation strategies (PRD7 50%–50% ETc and 7-day frequency of alternation of the irrigated side—FA, PRD14 50%–50% ETc and 14-day FA, PRD21 50%–50% ETc and 21-day FA, FX 50%–50% ETc and fixed irrigation, and irrigation with 100% ETc on both sides of the plant—FULL) with five replicates. Soil water content, physiological, vegetative, yield characteristics, and water productivity were assessed over two production cycles. PRD on the dry side lowered soil water content below optimal levels for banana cultivation, increased transpiration, and decreased photosynthesis and instantaneous water use efficiency with rising temperatures, while photosynthesis increased with stomatal conductance. PRD reduced plant vigor and delayed flowering in the first cycle. Compared to full and fixed irrigation, PRD conserves water while maintaining crop yields. Water productivity was higher under PRD, with PRD14 (50% ETc and 14-day alternation) offering the best water use efficiency while maintaining yield, making it suitable for ‘Prata-Anã Gorutuba’ banana cultivation. The study recommends PRD for sustainable banana farming in regions with limited water resources, contributing to sustainable agricultural practices and better water management.

1. Introduction

Brazil is among the ten countries with the largest area equipped for irrigation in the world, with irrigated land increasing significantly from 462,000 hectares in 1960 to 6.95 million hectares in 2015, and is projected to expand to 10 million hectares by 2030 [1,2]. However, water consumption without technical criteria has undermined the sustainable use of water resources in both public and private irrigation projects [3]. Irrigation accounts for about 70% of all freshwater consumed worldwide [4], highlighting the urgent need for strategies to use water rationally in irrigated agriculture.

Climate change has increased the vulnerability of Brazil’s semi-arid regions, such as the Northeast, to more prolonged and severe droughts [5]. The 2012–2017 drought, the most severe in recent decades, significantly affected water availability, causing crises in about 1300 municipalities and impacting millions of people [6]. In this context, optimizing water management is essential for productive and environmental sustainability. One promising technique is partial root-zone drying (PRD), which has been shown to increase water use efficiency in various crops [3,7,8,9,10].

The PRD technique involves alternating the side of the root system that is irrigated, where one part of the roots is irrigated while the other part is left to dry, inducing the production of chemical signals such as abscisic acid, which reduces stomatal opening and decreases transpiration [11]. This technique can maintain satisfactory photosynthetic rates while saving water, thereby increasing water use efficiency [12]. Studies on melon, papaya, orange trees, and olive trees, for example, have shown that PRD can maintain or even increase productivity and fruit quality while reducing water consumption [13,14,15].

Figure 1 illustrates the principle of the PRD irrigation technique. In this technique, one side of the root system is irrigated while the other side is left to dry. This management induces the production of chemical signals, such as abscisic acid, which reduces stomatal opening and decreases transpiration. Thus, PRD can maintain satisfactory photosynthetic rates while saving water, increasing water use efficiency.

Banana cultivation, especially the ‘Prata-Anã Gorutuba’ variety, requires a significant amount of water (600–1500 mm) for optimal growth and production [3]. In the semi-arid region of Brazil, the availability of water resources has limited banana production, highlighting the importance of irrigation strategies such as PRD. Additionally, bananas are one of the main fruits cultivated in Brazil, with great economic and social importance, especially in semi-arid regions where agriculture is a crucial source of income [16].

Although PRD has shown benefits in other crops, there are still gaps in knowledge regarding its application in banana plants, especially under specific soil and climate conditions [17,18]. Recent studies indicate that the PRD technique can not only save water but also improve nutrient use efficiency and plant resistance to water stress, which is essential for the sustainability of agricultural production in semi-arid regions [13,18].

The PRD technique can enhance water use efficiency without significantly compromising the productivity and quality of the ‘Prata-Anã Gorutuba’ banana fruits in regions with semi-arid conditions. The hypothesis is that the application of PRD will adequately maintain plant hydration and physiological functioning, resulting in a reduction in total water requirements and contributing to improved sustainability in banana production. Given these considerations, the objective of this study is to evaluate the ‘Prata-Anã Gorutuba’ banana under the PRD technique in northern Minas Gerais, analyzing soil water behavior, growth, physiology, productivity, and fruit quality. This research aims to provide technical information that can assist in maximizing the rational use of water in banana production in the semi-arid region of Brazil. Studying the application of PRD in banana plants is crucial not only for optimizing water use efficiency but also for ensuring the economic viability and environmental sustainability of the crop in areas vulnerable to climate change.

2. Materials and Methods

2.1. Experiment Characterization

The experiment was conducted in the Mocambinho Experimental Field, belonging to the Agricultural Research Company of Minas Gerais (EPAMIG Norte), Irrigated Perimeter of Jaíba, in Jaíba, MG, Brazil. The geographical location is 15°07′47″ S latitude, 43°57′02″ W longitude and 625 m altitude. According to Köppen’s classification, its climate is BSh (hot climate of Caatinga), with summer rains and well-defined dry periods in the winter [19].

Maximum and minimum temperatures, reference evapotranspiration, crop coefficient, accumulated irrigation and rainfall during the experimental period, between 12 August 2016 and 20 January 2018, were collected by the INMET automatic weather station located near the experimental field (Figure 2). The maximum and minimum temperatures were 30.4 a 19.3 °C, respectively, during the period.

The soil was red-yellow oxisol, with a sandy clay loam texture, whose physical–hydraulic attributes are described in

Table 1. Physical and physical–hydraulic characterization of the soil of the experimental area.

Depth (m)	Sand	Silt	Clay	BD (kg dm−3)	Moisture (m3 m−3) at Tension (kPa)
	(g kg−1)				10	33	100	500	1500
0.00–0.20	673	51	276	1.44	0.1817	0.1533	0.1381	0.1316	0.1280
0.20–0.40	712	35	253	1.40	0.1864	0.1559	0.1376	0.1279	0.1268

BD—bulk density.

and Figure 3. These soil analyses were conducted following Embrapa’s [20] recommendations. The textural analysis was performed using the dispersion method followed by sieving, while soil density was determined using undisturbed samples collected with a specific auger. Soil moisture as a function of water tension was measured with a Richards pressure extractor. The soil water contents equivalent to the upper and lower limits of water availability in the soil are 0.1841 cm³ cm⁻³ (10 kPa) and 0.1274 cm³ cm⁻³ (1500 kPa) (Table 1). The minimum soil water content for banana cultivation without water stress in this soil is equivalent to 0.1554 cm³ cm⁻³ (35 kPa), which corresponds to 49.5% of available water [21].

The orchard was planted on 12 August 2016 at 2.5 × 2.0 m spacing with a density of 2000 plants ha⁻¹, and the families were grown with one shoot, following the alignment of the cultivation row. To control weeds, manual weeding was performed in the plant rows and mechanical weeding was carried out between the rows with the help of a brush cutter. Disease and pest control were carried out through chemical sprays. During the experiment, periodic defoliation was conducted to reduce the inoculum of fungal diseases, improve the aeration of the banana plantation, increase light penetration, and facilitate phytosanitary management. After flowering, following the opening of the inflorescence, the heart was removed according to recommended practices for the crop in the semi-arid climate [23]. Fertilization was performed through fertigation every 15 days during the months of the experiment, based on recommendations from soil analyses.

The irrigation system used was drip, with two lateral lines by each plant row and three pressure-compensating emitters on each side of the plant, 0.25 m apart from it with a 4 L h⁻¹ flow rate. One fixed line per plant row with three emitters per plant was also used for one irrigation strategy.

Four irrigation strategies were used in the study based on the 50% reduction in the gross irrigation depth (GID) and on the 7-, 14- and 21-day frequency of alternation (FA) of the irrigated side of the plant. Another strategy was to keep irrigation on a fixed side of the row of plants, in addition to treatment with irrigation on both sides of the plant, without alternation and 100% GID. The experimental design followed a randomized block with five replicates and six usable plants per plot. They were as follows: (i) PRD7 50%–50% GID and 7-day FA; (ii) PRD14 50%–50% GID and 14-day FA; (iii) PRD21 50%–50% GID and 21-day FA; (iv) FX 50%–50% GID and only one side irrigated; and (v) FULL–100% GID and the two sides of the plant irrigated without alternation. Figure 4 provides a detailed diagram of the planting density and the irrigation system layout utilized in the study.

Regarding the water management of banana cultivation, the full irrigation depth was estimated daily using Equation (1).

$$\mathrm{E}\mathrm{T}\mathrm{c}=\mathrm{E}\mathrm{T}\mathrm{o} \mathrm{K}\mathrm{c} {\mathrm{K}}_{\mathrm{L}}-\mathrm{P}$$

(1)

where ETc—crop evapotranspiration, mm day⁻¹; ETo—reference evapotranspiration, mm day⁻¹; Kc—crop coefficient, dimensionless; K_L—location factor, dimensionless; and P—effective precipitation, mm day⁻¹.

Reference evapotranspiration (ETo) was calculated by a modified FAO–Penman–Monteith equation [24] with data obtained from an automatic meteorological station located near the experimental area. The automatic weather station used was a Campbell Scientific model CR3000, and for this study, the following variables were utilized from the provided data: air temperature (°C), relative humidity (%), atmospheric pressure (hPa), solar radiation (W m⁻²), wind speed (m s⁻¹), and rainfall (mm). The crop coefficients (Kc) used were those recommended for bananas [25] and are presented in Figure 2b. The location factor (K_L) was calculated according to Equation (2).

$${\mathrm{K}}_{\mathrm{L}}=0.10{\left(\mathrm{S}\mathrm{A}\right)}^{0.5}$$

(2)

where K_L—location factor, dimensionless; and SA—shaded area, %.

Soil water content was measured on the irrigated side and on the side exposed to partial root-zone drying. Soil water content was monitored using time-domain reflectometry (TDR) probes 0.10-m-long and installed 0.20 m away from the plants at 0.30 m depth. Soil water tension data, collected with Watermark sensors, were converted into soil water content using the Van Genuchten model [22], according to Equation (3).

$$\mathsf{\theta }=\mathsf{\theta }\mathrm{r}+{\displaystyle \frac{(\mathsf{\theta }\mathrm{s}-\mathsf{\theta }\mathrm{r})}{[1+(\mathsf{\alpha }{\left|\left.\mathrm{\Psi }\right|\right.}^{\mathrm{n}}{]}^{\mathrm{m}}}}$$

(3)

where: θ—soil volumetric water content, cm³ cm⁻³; θr—soil residual water content, cm³ cm⁻³; θs—soil saturation water content, cm³ cm⁻³; Ψ—soil matric potential, kPa; and α, n, m—fitted parameters referring to the soil.

Available water in the soil was calculated by Equation (4), using the upper limit (0.18405 cm³ cm⁻³–10 kPa) and lower limit (0.12740 cm³ cm⁻³–1500 kPa) of soil water availability.

$$\mathrm{A}\mathrm{W}={\displaystyle \frac{{\mathsf{\theta }}_{\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{u}\mathrm{a}\mathrm{l}}-{\mathsf{\theta }}_{\mathrm{P}\mathrm{W}\mathrm{P}}}{{\mathsf{\theta }}_{\mathrm{F}\mathrm{C}}-{\mathsf{\theta }}_{\mathrm{P}\mathrm{W}\mathrm{P}}}}100$$

(4)

where AW—the available water in the soil, %; θ_actual—soil water content at 0.30 m depth at a given time, cm³ cm⁻³; θ_FC—soil water content equivalent to field capacity, cm³ cm⁻³; and θ_PWP—soil water content equivalent to the permanent wilting point, cm³ cm⁻³.

2.2. Gas Exchange and Phytotechnical Characteristics Determinations

Stomatal conductance—gs (mol m⁻² s⁻¹)—and T_leaf (°C) evaluations were performed together with the readings of soil water content using the Leaf Porometer SC-1 (Decagon Devices, Inc., Pullman, WA, USA) at four times after planting (319, 320, 361 and 362 days after planting—DAP). At 361 days after planting (DAP), the infrared gas analyzer (IRGA) LCpro+ Portable Photosynthesis System (ADC BioScientific Limited, Hoddesdon, UK) with radiation and ambient temperature, airflow of 200 mL min⁻¹, was used to evaluate the physiological variables solar radiation incident on the leaf—Q_leaf (μmol m⁻² s⁻¹ photons), leaf temperature—T_leaf (°C), internal CO₂ concentration—Ci (μmol mol⁻¹), stomatal conductance—gs (mol m⁻² s⁻¹), transpiration—E (mmol m⁻² s⁻¹), photosynthesis—A (μmol m⁻² s⁻¹ of CO₂), intrinsic carboxylation efficiency—A/Ci (μmol m⁻² s⁻¹/μmol mol⁻¹), quantum efficiency of photosynthesis—A/Q_leaf (μmol m⁻² s⁻¹ of CO₂/μmol m⁻² s⁻¹ of photons) and instantaneous water use efficiency—A/E (μmol m⁻² s⁻¹/mmol m⁻² s⁻¹), according to Arantes et al. [26] and Ramos et al. [27]. All the measurements were obtained on the third or fourth leaf of the plants in two periods, in the morning and in the afternoon.

Phytotechnical characteristics were evaluated in two crop cycles. Pseudostem perimeter (PP) was measured at a 0.20 m height from the soil, and plant height and number of living leaves were evaluated at flowering. The mass of hands, number of fruits per bunch, mass, length and diameter of the fruit of the second hand, and number of living leaves were evaluated and yield of hands and water use efficiency were estimated at harvest.

The data were subjected to the normality test and analysis of variance, and the means were compared by the Tukey test at a 5% significance level. The data of variables that did not show normal distribution were subjected to non-parametric analysis, and the Kruskal–Wallis test was applied at a 5% significance level. The statistical analyses were performed using the Experimental Designs package of R software version 4.0.5 [28].

3. Results

3.1. Soil Water Content

The water consumption patterns in the treatments that received 100% and 50% ETc water replacement, as well as the accumulated precipitation during the two banana cultivation cycles, are presented in Figure 5a. It is observed in this figure that the accumulated precipitation during the first banana cultivation cycle (August 2016 to May 2017) was 676 mm, and during the second cycle (May 2017 to January 2018) it was 271 mm, totaling 947 mm over the entire experimental period. It is also observed in Figure 5a that for the treatment where the banana received 100% ETc water replacement, the accumulated irrigation depth in the first and second cycles was 908 mm and 1519 mm, respectively, resulting in a total of 2427 mm over the entire experimental period. In the treatment with 50% ETc, the accumulated irrigation depth in the first and second cycles was 521 mm and 760 mm, respectively, resulting in a total of 1281 mm. The soil water content on the irrigated side was greater than 0.1554 m³ m⁻³ (35 kPa) for all times and irrigation strategies (Figure 5b). Under fixed irrigation (FX), on the dry side of the plant, the soil water content remained lower than 0.1274 m³ m⁻³ (θ_PWP) at all times evaluated, indicating the absence of available water. In the full irrigation (FULL) strategy, the soil water content remained close to the value of field capacity (θ_FC) at all times evaluated, indicating that the crop evapotranspiration demand was met. In the treatments with PRD, the soil water content on the dry side was lower than on the irrigated side, reaching values below 0.1554 m³ m⁻³ (35 kPa).

Figure 6 presents the available water (AW) in the soil on both sides of the plant for the different treatments. It can be observed that in the treatments using the PRD technique (Figure 6a–c), as the AW increased on one side, it decreased on the other due to the alternating irrigation. When this did not occur, it was due to rainfall. It is also noted that the shorter the alternation period, the fewer days the AW on the dry side remained below the minimum availability for banana cultivation without water stress (35 kPa). In the irrigation strategy PRD7 50% (Figure 6a), the AW on the dry side of the plant required five days to reach the value of 50% ETc (35 kPa) and, due to the shorter interval of alternation of the irrigated side, it did not remain in this situation for more than four days.

When the banana crop was irrigated on only one side continuously (Figure 6d), the AW on the irrigated side never dropped below the minimum water availability limit. This consistent availability of water was also observed in the full irrigation treatment (Figure 6e), where the AW remained above the critical threshold at all times. These observations highlight the effectiveness of the PRD technique in managing water stress by alternating the irrigated side and maintaining AW within acceptable limits.

3.2. Gas Exchange

Although no statistical difference was observed between the irrigation strategies and the times of stomatal conductance (gs) measurement, at 319 DAP, all plants subjected to PRD with alternation of the irrigated side showed a tendency for reduced gs in the afternoon, unlike the plants under FX and FULL (

Table 2. Stomatal conductance—gs, measured at four evaluation times (319, 320, 361 and 362 days after planting—DAP *) and in two periods (morning and afternoon) in ‘Prata-Anã Gorutuba’ banana subjected to different irrigation strategies. Jaíba, MG, Brazil, 2017–2018.

Irrigation Strategies	319 DAP		320 DAP		361 DAP		362 DAP
	Morning	Afternoon	Morning	Afternoon	Morning	Afternoon	Morning	Afternoon
PRD7 50%	0.353	0.277	0.474	0.275	0.474	0.414	0.275	0.563
PRD14 50%	0.460	0.377	0.456	0.442	0.456	0.430	0.442	0.559
PRD21 50%	0.409	0.331	0.430	0.319	0.430	0.473	0.307	0.572
FX 50%	0.434	0.475	0.441	0.282	0.441	0.507	0.295	0.369
FULL	0.385	0.447	0.457	0.329	0.457	0.526	0.329	0.676
CV (%)	43.35	32.81	20.63	32.41	20.63	36.98	33.16	36.69

* The means between the irrigation strategies and the times of measurement did not differ according to Tukey’s test at a 5% significance level. The measurements were performed during the change in alternation of the irrigated side.

). In the case of plants under FULL, the better water condition in the soil and milder atmosphere compared to the other times, with a vapor pressure deficit (VPD) of 2.19 kPa, contribute to explaining the increase in gs in the afternoon.

The physiological variables internal CO₂ concentration (Ci), transpiration (E), stomatal conductance (gs), photosynthesis (A), leaf temperature (T_leaf) (

Table 3. Physiological variables evaluated at 361 days after planting (DAP) in two reading periods (morning and afternoon) in ‘Prata-Anã Gorutuba’ banana subjected to different irrigation strategies—Ci—internal CO₂ concentration, E—transpiration, A—photosynthesis, gs—stomatal conductance and leaf temperature—T_leaf. Jaíba, MG, Brazil, 2017–2018.

Physiological Variable	Period	PRD7 50%	∆ (%)	PRD14 50%	∆ (%)	PRD21 50%	∆ (%)	FX 50%	∆ (%)	FULL	CV (%)
Ci (µmol mol−1)	Morning	211.0	+1.9	220.40	+6.5	228.40	+10.3	209.40	+1.2	207.00	5.80
	Afternoon	205.8	–8.4	206.40	–8.1	209.40	–6.8	227.80	+1.4	224.60	7.36
E (mmol m−2 s−1)	Morning	6.93	+15.5	7.03	+17.2	7.16	+19.3	6.57	+9.5	6.00	11.23
	Afternoon	8.94	+7.5	8.51	+2.3	9.04	+8.7	10.21	+22.7	8.32	14.53
A (µmol m−2 s−1)	Morning	23.03	+13.1	23.12	+13.5	20.94	+2.8	23.94	+17.5	20.37	11.66
	Afternoon	20.01	+19.8	19.11	+14.4	18.22	+9.1	20.83	+24.7	16.70	12.21
gs (mol m−2 s−1)	Morning	0.36	+72.9	0.40	+89.6	0.39	+84.0	0.37	+77.9	0.21	18.08
	Afternoon	0.32	+12.9	0.30	+5.8	0.30	+5.8	0.41	+44.7	0.28	25.06
Tleaf (°C)	Morning	34.36	+2.7	34.02	+1.7	34.57	+3.4	33.48	+0.1	33.45	1.92
	Afternoon	38.94	+1.1	38.88	+0.9	39.60	+2.8	39.19	+1.7	38.53	3.02

), carboxylation efficiency (A/Ci), quantum efficiency of photosynthesis (A/Q_leaf), and water use efficiency (A/E) (

Table 4. Physiological variables evaluated at 361 days after planting (DAP) in two periods (morning and afternoon) in ‘Prata-Anã Gorutuba’ banana subjected to different irrigation strategies—instantaneous water use efficiency (A/E), carboxylation efficiency (A/Ci) and quantum efficiency of photosynthesis (A/Qleaf). Jaíba, MG, Brazil, 2017–2018.

Irrigation Strategies	A/E (µmol m−2 s−1/mmol m−2 s−1)		A/Ci (µmol m−2 s−1/µmol mol−1)		A/Qleaf (µmol m−2 s−1/µmol m−2 s−1)
	Morning	Afternoon	Morning	Afternoon	Morning	Afternoon
PRD7 50%	3.440	2.267	0.110	0.099	0.013	0.010
PRD14 50%	3.368	2.341	0.105	0.092	0.013	0.010
PRD21 50%	3.026	2.033	0.092	0.085	0.011	0.010
FX 50%	3.675	2.048	0.115	0.092	0.013	0.010
FULL	3.428	2.150	0.099	0.074	0.011	0.009
CV (%)	8.53	14.81	14.06	14.29	14.53	12.90

) did not differ significantly between irrigation strategies. Thus, the variations in the variables as a function of the reading times and the increase or decrease in the percentage of the values of the variables evaluated under the deficit irrigation and full irrigation strategies are presented. Ci and gs decreased between the morning and afternoon periods for irrigation strategies with PRD and side alternation and increased for FX and FULL. The highest values of T_leaf were recorded in the afternoon period. The lowest values of T_leaf were recorded in plants under FULL in the two periods evaluated, and the highest ones were found under PRD with 21-day FA. Transpiration (E) was higher in the afternoon under all irrigation strategies (Table 3).

The values of photosynthetically active radiation—Q_leaf—varied little between periods. Carboxylation efficiency (A/Ci), water use efficiency (A/E) and quantum efficiency of photosynthesis (A/Q_leaf) decreased between reading periods (Table 4). The largest decreases for A/Ci and A/E were recorded under FX and FULL. The values of E increase with T_leaf (Figure 7a), while WUE and A decrease (Figure 7b,d) and A increases with gs (Figure 7c).

3.3. Vegetative Characteristics

In the first production cycle, plant height, pseudostem perimeter, and the interval of days to flowering were influenced by irrigation strategies, with significant differences. However, in the second cycle, the means did not differ statistically, while the number of living leaves was similar between the irrigation strategies in the two crop cycles (

Table 5. Vegetative characteristics of ‘Prata-Anã’ banana, clone ‘Gorutuba’, subjected to different irrigation strategies in two production cycles. Jaíba, MG, Brazil, 2017–2018.

Characteristics	Cycle	Irrigation Strategies
		PRD7 50%		PRD14 50%		PRD21 50%		FX 50%		FULL		CV (%)
Plant height (m)	1st	2.12	ab	2.24	a	2.10	b	2.19	ab	2.19	ab	3.35
	2nd	2.92	a	3.22	a	3.00	a	2.92	a	3.20	a	5.37
Pseudostem (m)	1st	0.65	ab	0.67	a	0.66	ab	0.63	b	0.67	a	2.84
	2nd	0.73	a	0.80	a	0.76	a	0.74	a	0.81	a	6.05
Days to flowering (days)	1st	168.3	a	132.2	bc	148.1	ab	133.4	ab	127.4	c	16.06
	2nd	307.9	a	329.2	a	343.9	a	327.4	a	311.4	a	10.22
N° living leaves at flowering (un)	1st	18.5	a	18.4	a	18.5	a	18.5	a	18.6	a	5.04
	2nd	14.8	a	13.8	a	14.1	a	14.4	a	14.7	a	4.92

Means followed by equal letters for irrigation strategies do not differ significantly by the Kruskal–Wallis test at a 5% significance level.

). Banana plants under full irrigation on both sides showed greater height than plants under PRD21 50%, greater pseudostem perimeter than those under FX 50% on only one side of the row, and flowered earlier than those under FX, PRD7 50%, and PRD21 50% in the first crop cycle (Table 5). In the second production cycle, the vegetative characteristics did not differ between the irrigation strategies.

3.4. Yield Characteristics and Water Use Efficiency

The irrigation strategies caused no significant differences in the number of fruits per bunch in the second production cycle, fruit length, fruit mass, and yield of hands in the first production cycle, fruit diameter, number of living leaves, and total leaf area evaluated at harvest in both production cycles (

Table 6. Yield characteristics of the ‘Prata-Anã Gorutuba’ banana subjected to different irrigation strategies in two production cycles. Jaíba, MG, Brazil, 2017–2018.

Characteristics	Cycle	Irrigation Strategies
		PRD7 50%		PRD14 50%		PRD21 50%		FX 50%		FULL		CV (%)
Number of fruits per bunch (un)	1st	101	b	106	ab	108	ab	105	ab	116	a	6.20
	2nd	106	a	120	a	113	a	108	a	126	a	9.63
Length of central fruit of the 2nd hand (cm)	1st	14.3	a	14.3	a	14.8	a	14.0	a	14.5	a	5.36
	2nd	16.3	b	17.0	ab	17.4	ab	17.6	ab	18.3	a	4.77
Diameter of central fruit of the 2nd hand (mm)	1st	31	a	33	a	33	a	32	a	33	a	3.46
	2nd	36	a	36	a	35	a	35	a	37	a	2.69
Mass of central fruit of the 2nd hand (g)	1st	92.8	a	108.1	a	110.3	a	109.9	a	112.8	a	16.09
	2nd	126.7	b	138.2	b	140.4	ab	143.9	ab	157.5	a	7.04
N° living leaves at harvest (un)	1st	6.5	a	7.4	a	6.6	a	7.4	a	7.4	a	25.54
	2nd	10.4	a	10.0	a	10.3	a	10.0	a	11.1	a	7.20
Total leaf area (m2)	1st	3.61	a	4.44	a	4.13	a	4.62	a	4.52	a	28.21
	2nd	7.10	a	7.68	a	7.87	a	7.58	a	9.30	a	16.34
Water productivity—WP (kg ha−1 mm−1)	1st	40.51	ab	45.18	a	48.07	a	46.16	a	29.83	b	14.04
	2nd	47.46	a	52.70	a	51.13	a	50.56	a	30.49	b	10.29
Yield of hands (Mg ha−1)	1st	7.10	a	7.68	a	7.87	a	7.58	a	9.30	a	16.34
	2nd	29.85	b	33.14	ab	32.16	ab	31.80	ab	38.35	a	11.35

Means followed by equal letters, in the row, do not differ from each other by Tukey test at 5% significance level.

). Full irrigation promoted a higher number of fruits per bunch in cycle I, longer fruit length, and higher yield of hands in cycle II compared to the PRD7 50% strategy, and greater fruit mass in cycle II compared to the PRD7 50% and PRD14 50% strategies.

The water use efficiency of the ‘Prata-Anã Gorutuba’ banana, here referred to as water productivity, differed between the irrigation strategies adopted, both in the first and in the second cycle (Table 6). For both cycles, ‘Prata-Anã Gorutuba’ banana plants under PRD with 50% ETc with alternation of the side to be irrigated every 7, 14, and 21 days, and without alternation, had higher water use efficiency compared to those under full irrigation, except for PRD with alternation of 7 days (PRD7 50%) in the first cycle, which led to the same efficiency as full irrigation.

4. Discussion

4.1. Soil Water Content

Evaluations on the dry side at 319 DAP and 361 DAP, when the interval between the alternation of the irrigated side was at its maximum under the strategies with alternation of the irrigated side (7, 14, and 21 days), showed lower soil water content in treatments with PRD than in treatments with full irrigation. The water deficit caused by irrigation with PRD 50% can influence root distribution, root deepening, and the increase in root length density. This is consistent with findings by Coelho et al. [3], who observed that the water deficit with PRD 50% did not limit the distribution of the root system of the ‘BRS Princesa’ banana. The statement claiming that, under water deficit, the plant increases the investment in root production, increasing the explored volume of soil to absorb a greater amount of water necessary to maintain its basic functions, is classic [29].

Coelho et al. [3] found a higher rate of water absorption in ‘BRS Princesa’ banana plants under PRD 50%, subjected to 21-day fertigation application on the irrigated side, compared to those under the shortest alternation intervals, which indicates an adaptation of their active roots, compensating for the lower availability of water on the non-irrigated side. In this situation, the plants had 85% of their roots distributed up to 0.20 m depth, indicating that the availability of water below this depth would be insufficient to meet their water needs. On the other hand, the shortest interval of alternation of sides, 7 days, allowed water absorption in deeper layers and lower reduction in water availability in the soil. Possibly due to the predominance of superficial active roots due to the PRD technique [30]. During the experimental period, the low rainfall and the well-defined dry period favored the irrigation strategies, since the water absorbed by the crop was supplied most of the time by irrigation. This evidenced the effect of partial root-zone drying on the ‘Prata-Anã Gorutuba’ banana.

For ‘BRS Princesa’, Coelho et al. [3] observed that under PRD7 50%, it took six days for the soil water content to reach 0% of the available water. In plots with plants under PRD14 50%, the side exposed to partial root-zone drying required six days for AW to reach 50% (35 kPa) and remained below this threshold for up to 11 days. Under the irrigation strategy PRD21 50%, it also took six days for AW to reach 50% (35 kPa), and AW values remained below 50% for up to 27 days. In contrast, plants under FX 50% maintained AW values close to field capacity throughout the entire experimental period, similar to those under FULL irrigation.

The longest interval of alternation of the irrigated side of the plants caused a more pronounced decrease in available water and prolonged water stress conditions. These findings are consistent with those reported by Coelho et al. [3] in their study with the ‘BRS Princesa’ banana under PRD. However, in all evaluated irrigation strategies, the duration of the water stress period was shorter compared to the present study, suggesting a higher tolerance of ‘BRS Princesa’ to water stress, possibly due to enhanced regulation of water absorption rates by the roots.

In plants under the PRD21 50% irrigation strategy, it took six days for the available water to reach 50% (35 kPa), and AW values remained below 50% for up to 27 days. In plants under FX 50%, the AW on the irrigated side fluctuated with values close to soil water content at field capacity throughout the entire experimental period. A similar situation was observed in plants under FULL irrigation.

To enrich the understanding of the adaptability and effectiveness of PRD in different contexts, it is useful to compare these results with findings from similar studies on other species and varieties and under different climatic conditions. For instance, Kumar et al. [30] found that PRD improved water use efficiency without significantly reducing yield in various drought-resistant fruit species and varieties. Additionally, Iqbal et al. [31] demonstrated the adaptability of PRD to semi-arid climates, emphasizing the need for local adjustments to optimize irrigation strategies. These comparisons highlight the importance of considering varietal differences and climatic conditions when implementing PRD strategies, as they can substantially impact water absorption patterns and overall plant performance.

4.2. Gas Exchange

During the night, when temperatures are cooler and transpiration rates decrease, plants restore their water status, creating more favorable conditions for gas exchange in the morning. As solar radiation and air temperatures rise throughout the day, atmospheric conditions become more challenging. To maintain internal water balance, plants must reduce gs to minimize water loss. However, under fluctuating light conditions, gs represents only one aspect of plant physiology affecting overall water use efficiency and carbon assimilation. These processes are also influenced by genotype-specific traits, including various transpiration phenotypes associated with drought tolerance [32].

The gs is regulated by a series of processes, including chemical signaling in response to root water deficit conditions. One accepted mechanism involves increased synthesis and concentration of abscisic acid (ABA) in the roots, reduced ethylene levels, and maintained relative growth of the roots. Hydraulic signals include decreased water potential and osmotic pressure at the root tip, increased turgor, reduced water potential gradient between soil and root, and decreased water uptake by the roots [33]. ABA acts as a chemical messenger transported to the shoots, inducing stomatal closure to regulate plant water status and prevent excessive transpiration. Even under drought conditions, banana plants maintain internal water balance through osmotic adjustment [34]. Stomatal closure is a key mechanism for conserving water under water stress, triggered by root signaling and exacerbated by increases in vapor pressure deficit (VPD) with rising temperatures and decreasing relative humidity. According to Eyland et al. [35], gas exchange assessment is highly sensitive for detecting water stress in bananas.

At 361 DAP, the afternoon atmospheric conditions intensified, with a vapor pressure deficit (VPD) of 3.20 kPa. The gs increased in plants under FULL, FX 50%, and PRD21 50%, but decreased in those under PRD7 50% and PRD14 50%. This suggests that partial root-zone drying heightened sensitivity to adverse environmental conditions.

During the evaluation at 362 DAP, gs increased despite more severe afternoon atmospheric conditions, characterized by a VPD of 4.45 kPa across all irrigation strategies. Typically, gs would be expected to decrease under such VPD conditions, as observed in ‘BRS Princesa’ with VPD exceeding 1.60 kPa [3]. The increase in transpiration could serve as a mechanism for latent heat dissipation, enhancing thermal comfort for banana plants in response to higher afternoon temperatures, despite the observed decrease in gs [26,27]. However, these responses may vary among cultivars, influenced by genetic, biochemical, molecular [36], and morphological factors [37].

Coelho et al. [3] found that higher values of gs in ‘BRS Princesa’ banana under PRD were associated with greater availability of water on both sides of the plant so that the flow of water from the roots to the shoots remained positive even under adverse climatic conditions. The higher values of Ci are associated with higher values of gs. In turn, the values of T_leaf and E increased under all irrigation strategies between morning and afternoon, while the values of A decreased (Table 3). These results corroborate those reported by Coelho et al. [3]. The greater increase in Ci between periods in plants under FULL can be attributed to the better balance in gas exchange due to the favorable water condition in the soil. There was an increase of at least 4 °C in T_leaf in the afternoon (Table 3); however, leaf temperature above 34 °C indicates that the plants are in a situation of stress most of the day and compromises the photosynthetic reactions in banana [35,38]. As for transpiration, the increase may have been due to the higher temperature with increased VPD, even with a decrease in gs under the strategies with PRD and side alternation. This is justifiable as a defense strategy for latent heat loss in response to higher temperatures [26].

The values of photosynthetically active radiation—Q_leaf—varied little between periods; however, photosynthesis decreased in the afternoon under all irrigation strategies, due to the limitation imposed by temperature, attested by the higher T_leaf. In the afternoon, there were changes in gs but decreases under PRD strategies with side alternation and increases under FX and FULL. In any case, there is evidence of a stomatal effect with the decrease in gs [27], suggesting impairment in the enzymatic reactions of photosynthesis with a probable decrease in the carboxylase activity of ribulose 1,5-bisphosphate carboxylase/oxygenase (RuBisCO) [26,39].

The largest decreases in A/Ci and A/E were recorded under FX and FULL. This helps to explain that the biggest reason for the decrease in photosynthesis rates was the alteration in the carboxylase activity of RuBisCO and/or in the permeability of the membranes due to the high temperatures because gs increased under the irrigation strategies FX and FULL. This suggests a lower contribution of the stomatal component to the restriction to CO₂ entry, also corroborated by the increase in E under all irrigation strategies. Although the actual proof of a change in RuBisCO activity requires the determination of enzymatic kinetics, such as maximum carboxylation rate of RuBisCO, photosynthetic electron transport rate, use of triose phosphate, daytime respiration and mesophyll conductance [40,41], the results strongly suggest that [26,39]. Additionally, in general, in the morning there are better conditions for plants to perform photosynthesis, milder temperatures, better quality radiation for the light reactions of photosynthesis with predominant wavelength in the red and far-red range, favorable water conditions, higher relative humidity and lower values of VPD.

The correlations between E and T_leaf, A/E and T_leaf, A and gs, and A and T_leaf ratify the discussion concerning Table 3 and Table 4. Higher air temperature favors the increase in T_leaf, with a consequent increase in transpiration. This occurred as a compensatory mechanism of energy exchange by the loss of latent heat, respiration and reduction in net photosynthesis mainly due to enzymatic impairment in the reactions that depend on RuBisCO, even with the increase in gs in some cases. This makes it possible to infer the contribution of temperature and/or stomatal closure to the reduction in photosynthesis in banana plants [26,27].

4.3. Vegetative Characteristics

Despite some variations, the differences in vegetative and cycle characteristics recorded in the first crop cycle when the root system develops suggest that full irrigation (100%) on both sides of the plant favors the vegetative growth of the banana plant compared to some strategies with PRD. The irrigation strategy PRD14 50% led to similarity in phytotechnical characteristics when compared to the strategy with full irrigation. In the second production cycle, the vegetative characteristics did not differ between the irrigation strategies. This, combined with the few differences detected in the first cycle, suggests that the 50% reduction in the gross irrigation depth applied can be used with the inherent advantages of water and energy saving.

The height and pseudostem perimeter of the banana reflect its vigor and express the cumulative growth and development throughout the production cycle. On the other hand, the measurements of gas exchange are point-specific and reflect the moment of evaluation, which may not represent the stress conditions caused by the irrigation strategies adopted during the period of cultivation [39]. This justifies the influence of irrigation strategies on those characteristics (Table 5) and the absence of influence on these (Table 3 and Table 4).

Additionally, long-term studies are essential to understand the impact of irrigation strategies on fruit productivity and quality, especially in light of changing climate conditions and variations in plant water requirements. For example, Cháves et al. [42] indicate that water use efficiency can be significantly affected by changes in climate conditions, requiring adjustments in irrigation practices to maintain productivity. Farias et al. [43] and Blanchy et al. [44] discuss how different water management strategies can influence fruit quality, highlighting the importance of considering both quantitative and qualitative aspects in irrigation management. These studies suggest that, although PRD may offer advantages in terms of water and energy savings, it is crucial to continuously evaluate its long-term implications to ensure the sustainability and economic viability of banana production.

4.4. Yield Characteristics and Water Use Efficiency

The length and diameter of ‘Prata-Anã Gorutuba’ fruits under all irrigation strategies meet the classification standard of Abanorte [45] for top-quality bananas, except for fruit diameter under the PRD7 50% strategy in cycle I. According to the rules for commercial classification of the ‘Prata-Anã’ banana currently in force in the north of Minas Gerais and southwest and west of Bahia, released by the Programa Banana na Medida Certa [45], top-quality bananas must have an external fruit length greater than 14 cm and diameter greater than 31.75 mm. Thus, the average values of these variables, regardless of the irrigation strategy adopted, meet the proposed commercial classification as top quality [45] and corroborate Magalhães et al. [46] regarding the maintenance of commercial quality with reduction in irrigation depth. According to the data presented, fruit production under PRD is one of the alternatives available for banana production with water saving, which is quite relevant in areas with water scarcity.

The higher values of water use efficiency are justified by the real reduction in the irrigation depth applied between strategies with deficit and full irrigation, whereas for PRD7 50%, in the first production cycle, the lower water use efficiency is justified by the reduction in yield (Table 6). When associating the results of yield and water productivity (Table 6), it is possible to observe the importance of refining studies using PRD with a reduction in the applied depth, since it can maintain yield and increase water use efficiency in irrigation. The water productivity values found in the present study are higher than those reported by Coelho et al. [3], who found values between 40 and 43 kg ha⁻¹ mm⁻¹ in the ‘BRS Princesa’ banana under PRD.

To deepen the understanding of PRD effects, future research should investigate the application of PRD in different climatic and soil contexts, as well as explore how PRD can interact with sustainable agricultural practices, such as soil cover and nutrient management. Studies on the genetic adaptation of different banana varieties to PRD and the evaluation of emerging technologies for real-time soil moisture and plant health monitoring would also be valuable. Additionally, long-term cost–benefit analyses and the socio-economic impact of PRD on farming communities could provide further insights into its feasibility and widespread acceptance.

5. Conclusions

Soil water content for plants under PRD on the dry side decreases to values below the optimal level for banana cultivation without water stress.

Transpiration increases, while photosynthesis and instantaneous water use efficiency decrease with temperature, and photosynthesis increases with stomatal conductance.

Partial root-zone drying reduces vigor and increases the interval of days to flowering in the first production cycle.

Water productivity is higher in the ‘Prata-Anã Gorutuba’ banana under a partial root-zone drying strategy.

PRD with 50% of the gross irrigation depth and alternation of the irrigated side every 14 days results in greater water use efficiency with the maintenance of yield, being suitable for cultivation of the ‘Prata-Anã Gorutuba’ banana.

Future research should explore long-term PRD impacts on various banana cultivars and climatic conditions, optimize alternation intervals, and manage soil health. Large-scale PRD applications can significantly conserve water, mitigate climate change effects, and enhance crop sustainability, emphasizing the importance of water footprint management in sustainable agriculture.