Impact of deficit irrigation on citrus production under a sub-humid climate: a case study

Limited water availability is one of the major factors affecting the productivity of agricultural crops in the tropics. The whole agricultural sector is going to be shrunk due to less water availability caused by more demand for water for drinking and industrial purposes in the near future. In this scenario, the development of optimal water supply strategies through efficient irrigation methods is one of the options to sustain crop production (Behera & Panda 2009).

Deficit irrigation (DI) is a strategy in which a reduced quantity of water is supplied over the full water requirement of the crops. The precise application of DI requires a thorough understanding of the yield response to water and of the economic impact of the irrigation strategy on crop production (Zhang & Oweis 1999). In the regions where water resources are limited, DI can be more profitable for the farmers in the way it maximizes crop water productivity instead of maximizing the yield per unit land area (English et al. 2002). The saved water can be utilized for other purposes or to irrigate extra units of land, resulting in higher crop production. In other words, DI aims at stabilizing productivity and at obtaining maximum crop production by bringing more area under irrigation, resulting in higher crop water productivity.

Citrus is a widely grown fruit crop, covering around 146 countries in the world (CCRI 2015). As an evergreen perennial crop, citrus requires 1,200–2,000 mm of water for its annual life cycle (Huchche et al. 1999). The major citrus-growing regions are mainly confined to tropical and sub-tropical climates, where water scarcity is one of the prime factors affecting citrus production (Ghosh 2007). Once the plant attains a desirable canopy, it is not advisable to supply water for its vegetative growth (González-Altozano & Castel 2000). Moreover, the higher vegetative growth in mature plants reduces the productivity and quality of citrus fruits (Davies & Albrigo 1994). Once the crop develops a widespread root zone, it draws some quantity of water from the soil beyond irrigation water. If water is supplied in full to fill the total evapotranspiration of the plants, the water uptake from the non-irrigated rhizosphere reduces. Thus, applying irrigation water in full as per the crop water requirement may cause low water productivity in mature citrus orchards. However, the plants undergo severe stress when soil water is very low and the water uptake by the roots fails to compensate for the optimal crop water requirement. Hence, the accuracy in water application, creating desirable stress which is important for citrus production in water-scarce areas.

Earlier, some studies documented the response of citrus to DI. Castel & Buj (1990) revealed that irrigation at 60% crop evapotranspiration (ET_c) during flowering and fruit setting stage did not affect the fruit yield and quality in ‘Salustiana’ orange (Citrus sinensis Osbeck) in Spain. Chartzoulakis et al. (1999) observed that a 60% reduction in irrigation water quantity resulted in a 36% reduction in fruit yield without affecting fruit quality in ‘Bonanza’ orange (C. sinensis) in a sandy clay loam soil in Chania Crete, Greece. González-Altozano & Castel (1999) reported that DI at 25% or 50% ET_c reduced the fruit yield up to 62% with higher total soluble solids (TSS) in juice compared with full irrigation (FI, 100% ET_c) in ‘Clementina de Nules’ (Citrus clementina Hort.ex Tan.) in a sandy loam soil in eastern Spain. Withholding water supply during the initial and final fruit growth period reduced the fruit yield up to 24% with higher TSS and acidity in juice compared with FI in ‘Lane late’ orange (C. sinensis Osbeck) in a loamy soil in a semi-arid region of Spain (Pérez-Pérez et al. 2008). Gasque et al. (2010) observed that DI with 40% and 60% reduction in the water supply at the initial fruit growth phase did not affect the fruit yield and quality of ‘Navalina’ sweet orange (C. sinensis Osbeck) in a sandy loam soil in Spain. García-Tejero et al. (2010) demonstrated that DI at 55% ET_c during flowering and fruit growth combined with irrigation at 70% ET_c during maturity reduced the fruit yield by 15–25% with higher TSS and acidity in juice without affecting the juice content in ‘Navelina’ sweet orange in a sandy loam soil of south-west Spain. Ballester et al. (2013) concluded that DI at 50% ET_c during the initial fruit growth phase saved 19% water and improved TSS and acidity in juice, without reducing fruit yield in Navel Lane Late (C. sinensis (L) Osbeck) citrus plants in Valencia, Spain, in a stony clay loam soil. Panigrahi et al. (2014) reported that DI at 50% ET_c during the early fruit growth phase resulted in higher TSS in juice without affecting the fruit yield of Kinnow mandarin (Citrus reticulate Blanco) in a sandy loam soil in a semi-arid environment of India. Morianou et al. (2021) concluded that the fruit yield under DI at 60% ET_c was 11–56% lower than FI in different varieties of grapefruit (Citrus paradisi Mac.) grown in a loamy soil in sub-tropical climate of Chania, Crete, Greece. However, DI improved the TSS, citric acid, ascorbic acid, phenolic contents and maturation index of fruits. Overall, the studies indicate that the level and timing of imposing water stress along with its duration are the main factors responsible for the success of DI in citrus. Moreover, orchards and crop characteristics such as soil, climate and cultivar also play an important role in the success of DI (Treeby et al. 2007; Panigrahi & Srivastava 2011).

Nagpur mandarin (C. reticulata Blanco), a loose-skin world-famous citrus cultivar, is commercially grown in around 2.0 lakh hectares of central India as an irrigated crop in vertisol (Kuchanwar et al. 2017). Surface irrigation (basin, furrow) using groundwater is the dominant method of water application in the region. The substantial water loss through preferential pathways/deep cracks developed at sub-optimum soil water content (SWC) in the vertisol is a common problem, causing low water use efficiency in agriculture (Singh & Srivastava 2004). However, the area under the citrus crop is exponentially increasing due to climate suitability and better financial return compared to other crops in the region. For the last few years, the decline of groundwater due to overexploitation has become a major concern for farmers. The water shortage affects the productivity and quality of citrus produced in this region (Panigrahi et al. 2010). The strategies like drip irrigation, continuous trenching and mulching have been found as water-saving techniques in Nagpur mandarin (Panigrahi et al. 2012, 2017). Furthermore, DI with a drip system may result in substantial water-saving by improving fruit quality and water productivity compared with FI in the crop under the water scarcity situation of central India. Financial analysis of citrus production under any irrigation system is also important from the farmers’ perspective (Panigrahi et al. 2013). However, the information on optimal DI regime under a drip system in relation to water use, fruit yield, fruit quality and production economics of a mature mandarin cultivar of citrus in a high clay content soil under a sub-humid tropical climate is limited worldwide. Keeping this in mind, a field experiment was undertaken to study the effects of different DI regimes on water use, fruit yield, fruit quality and production economics of drip-irrigated ‘Nagpur’ mandarin orchard grown in a vertisol (high clay content soil) under a hot sub-humid tropical climate of central India.

The SWC was monitored twice a week at 0–0.2 m, 0.2–0.4 m, 0.4–0.6 m and 0.6–0.8 m depths in each treatment plot using access tubes and neutron moisture meter (Troxler model-4300, USA). The mean monthly SWC for each depth was estimated under different treatments. The soil water suction in different soil layers was measured using ceramic cup tensiometers.

The leaf physiological parameters (net photosynthesis, P_n; stomatal conductance, g_s; and transpiration rate, T_r) of the experimental plants were measured once a month, from 8 am to 3 pm, using CO₂ gas analyzer (model-301PS, CID Bio-Science, USA). Four mature leaves (third or fourth leaf from the tip of the shoot) per plant in different directions (north, south, east and west) were selected for measurement. The leaf water use efficiency (LWUE) was calculated as the ratio of P_n to T_r (P_n/T_r), as suggested by García-Sánchez et al. (2007). The relative leaf water content (RLWC) and leaf water concentration (LWC) for the experimental plants were determined in different treatments following the standard procedures suggested by Bowman (1989) and Peñuelas et al. (1997).

The vegetative growth parameters such as plant height (PH), stem height, stock girth (SG) diameter and scion girth (SCG) diameter were measured for all experimental plants using measuring tap and vernier calliper, and their pooled annual incremental magnitudes were compared. The canopy (hemispheroid shape) volume was calculated based on the formula 0.5233 H W², where H is the difference between tree height and stem height, and W is the canopy width (Obreza 1991). The PH is the distance between the top of the crown of the plant from the ground surface and the stem height is the distance between the ground surface and the base of the nearest branch on the stem.

The number of fruits, fruit weight and weight of total fruits from each experimental plant under various treatments were recorded. The total fruit yield was estimated by multiplying the mean fruit yield of the experimental plants with a total number of plants per hectare (278) in different treatments. The water productivity (yield per unit quantity of water used) was calculated as the ratio of total fruit yield (kg ha⁻¹) to total water used per hectare (m³ ha⁻¹) in different treatments. Five fruits per plant were taken randomly for the determination of fruit quality parameters (juice, acidity and TSS). Juice was extracted manually by a juice extractor and its percentage was calculated on a weight basis with respect to fruit weight. The acidity was estimated by volumetric titration with standardized sodium hydroxide, using phenolphthalein as an internal indicator (Ranganna 2001), and the TSS was determined by digital refractometer (Atago model-PAL 1, Japan).

The relative economics under various irrigation treatments was determined based on the seasonal cost of production (cost of fertilizers, pesticides, energy for pumping of irrigation water), labour cost for basin cleaning, irrigation, fertilizer application, spraying and fruit harvesting, and cost of drip irrigation system. The annual cost of the drip irrigation system was calculated considering its initial cost of installation (INR 82,000 ha⁻¹) with a bank interest rate of 6% per annum. The useful life of the drip system was considered as 5 years. The gross income from the produce (fruits) was estimated using the prevailing wholesale market price of INR 10,000 t⁻¹. The net seasonal income was determined by subtracting the seasonal cost of production from the gross income of the produce. The net economic water productivity was worked out as the ratio of net economic return and total water used under different treatments. The benefit–cost ratio was calculated as the ratio of net income to total seasonal production cost.

The data were subjected to analysis of variance, and separation of means was obtained using the Duncan multiple range test which was performed by using Least Significant Difference values at a 5% probability level obtained following the method described by Gomez & Gomez (2010).

The monthly irrigation water applied under different irrigation regimes under the drip system was lowest in December (5.8–19.5 L day⁻¹ plant⁻¹) and highest in June (18.8–62.6 L day⁻¹ plant⁻¹) (Table 1). The increase in water application was due to the increasing rate of daily pan evaporation rate from December (2.5–3.0 mm) to June (8.0–10.0 mm) during the study years. Earlier studies by Autkar et al. (1989) and Panigrahi et al. (2012) recorded a similar trend of water use by citrus plants from December to June under central India conditions. Overall, the quantity of water applied under 30% FI, 50% FI, 70% FI and 100% FI regimes were 806 m³ ha⁻¹ yr⁻¹; 1,343.4 m³ ha⁻¹ yr⁻¹; 1,880.7 m³ ha⁻¹ yr⁻¹; and 2,686.8 m³ ha⁻¹ yr⁻¹, respectively. The past studies indicated that the annual crop water use was 1,942–2,189 m³ ha⁻¹ for Salustian orange in Spain (Castel & Buj 1990), 2,371–2,570 m³ ha⁻¹ for Bonanza orange in Greece (Chartzoulakis et al. 1999), 2,822–2,953 m³ ha⁻¹ for lemon in central Iran (Abu-Awwad 2001) and 3,277–4,147 m³ ha⁻¹ for Kinnow mandarin in northern India (Panigrahi et al. 2014) against 2,686.8 m³ ha⁻¹ for Nagpur mandarin in the present study. The variations in water use by citrus plants are attributed to differences in cultivar and age of the plants used in varied agro-climates of the study regions.

The irrigation treatments had significant effects on leaf P_n, g_s and T_r (Table 2). The P_n was higher under a higher level of irrigation with the highest magnitude in FI, and then decreased with the irrigation regime from 70% FI to 30% FI, indicating the negative effect of soil water deficit on the photosynthetic capacity of citrus plants. Earlier, Chartzoulakis et al. (1999) and Bhatnagar et al. (2011) also observed the reduction of P_n with a soil water deficit in Bonanza orange in Greece and Kinnow mandarin in a semi-arid environment of north India, respectively. However, the reduction in P_n was higher between DI at 50% FI and DI at 30% FI (31.7%) than that between DI at 70% FI and DI at 50% FI (11.7%) and between FI and DI at 70% FI (4.7%). The higher reduction in  P_n indicated the occurrence of severe water stress conditions for mature mandarin plants under DI at 30% FI in the region. The g_s and T_r values reduced with a reduction in irrigation regime from FI to DI at 30% FI. However, in contrast to the trend observed in relation to P_n under DI, the magnitudes of reduction in g_s and T_r between FI and 70% FI (g_s, 17.9%; T_r, 15.4%) was higher than that between 70% FI and 50% FI (g_s, 15.7%; T_r, 12.4%) and between 50% FI and 30% FI (g_s, 12.2%; T_r, 8.2%). It indicated that the threshold value of SWC for obtaining optimum P_n of the Nagpur mandarin plant exists between FI and 70% FI; whereas the threshold value of SWC for g_s and T_r exists in between the irrigation regime at 50% FI and 30% FI. Overall, the reduction in g_s and T_r was higher than that in P_n under DI (except at 30% FI), reflecting more sensitivity of g_s and T_r towards soil water deficit than P_n of citrus plants. The reduction in P_n, g_s and T_r of leaves under DI was earlier observed by Vu & Yelenosky (1988) in ‘Valencia’ Orange and Ribeiro et al. (2009) in Satsuma mandarin. However, the mean LWUE (μmol CO₂ fixed per mmol H₂O transpired) in DI at 50% FI was higher than that in other treatments. Higher LWUE is due to a marginal decrease in P_n associated with the higher decrease in T_r of the plants under DI at 50% FI over other treatments. The RLWC and LWC under different treatments followed a similar trend of P_n in the mandarin plants.

The annual incremental growth (PH, SG, SCG and canopy volume (CV)) of the plants showed that only PH and CV were significantly influenced by irrigation treatments (Table 3). The highest increase in PH (0.38 m) and CV (9.48 m³) was observed in FI followed by DI at 70% FI (0.35 m and 9.41 m³, respectively). This may be due to better photosynthesis and partitioning of a higher amount of photosynthates towards vegetative growth under FI compared with other treatments. The minimum vegetative growth was observed under DI at 30% FI. The lower vegetative growth under a lower level of irrigation corroborates the findings of García-Tejero et al. (2010) in ‘Salustiano’ orange under DI in Spain. However, the results of the present study differ from Chartzoulakis et al. (1999) in Bonanza orange in Greece which indicated the significant effect of DI on the trunk diameter of the plants. Furthermore, Gasque et al. (2010) did not observe any significant effect of DI on CV in ‘Navalina’ sweet orange in Spain. These differences in the effects of DI on vegetative growth parameters of citrus plants from the present study are ascribed to crop cultivar, soil, climate and irrigation scheduling methods used in the studies.

The fruit yield was higher under a higher level of irrigation (Table 4). However, the yield under FI was statistically at par with that under DI at 70% FI and DI at 50% FI. The fruit yield at 30% FI was significantly lower compared with other treatments. The lower photosynthesis rate and partitioning of the lower portion of photosynthates towards yield parameters might be the reason for the low yield under irrigation at 30% FI compared with other irrigation treatments. More fruits (480 per plant) with lower fruit weight (98.5 g) was observed in FI compared with DI at 70% FI (number of fruit, 435; fruit weight, 102.6 g) and DI at 50% FI (number of fruit, 389; fruit weight, 102.6 g). The higher number of fruits probably caused lower fruit weight under FI and 70% FI compared with 50% FI treatment. The fruit yield was significantly lower in DI at 30% FI due to a lower number of fruits and lower fruit weight than in other treatments. Similar results of lower fruit yield due to reduced fruit number and fruit weight under DI were also reported by Pérez-Pérez et al. (2008) in ‘Lane late’ orange and García-Tejero et al. (2010) in ‘Salustiano’ orange. However, González-Altozano & Castel (1999) observed that the lower fruit yield in ‘Clementina de Nules’ were due to smaller fruits, not fewer fruits under DI in eastern Spain. In contrast, Gasque et al. (2010) indicated that the fruit yield under DI was affected due to a lower number of fruits, rather than fruit weight. These differences in relation to the effects of DI on fruit number and fruit weight are due to the methods of irrigation scheduling adopted, cultivar used and growing environment of the regions taken for the studies. The higher water productivity was observed under DI at 50% FI (8.54 kg m⁻³) compared to FI (4.89 kg m⁻³) and DI at 70% FI (6.60 kg m⁻³). The higher water productivity under DI at 50% FI and 70% FI was attributed to a higher increase in fruit yield with comparatively less water supply under these treatments compared with FI.

Fruit quality (juice percentage, acidity and TSS) under various irrigation treatments showed that the FI produced fruits with higher juice contents (40.3%) compared with other treatments (36.2–40.1%). However, the higher TSS and lower acidity of fruits were observed in DI at 50% FI (TSS, 10.3 °Brix; acidity, 0.83%) than that in DI at 70% FI (TSS, 10.1 °Brix; acidity, 0.84%) and FI (TSS, 9.8 °Brix; acidity, 0.85%). The fruits with the lowest TSS (9.6 °Brix) and acidity (0.87%) were harvested in DI at 30% FI. The reduction in juice percentage is one of the reasons for the enhancement of soluble solids concentrations in fruits under DI except at 30% FI. Secondly, the higher TSS and lower acidity of fruits under optimum water stress (DI at 50% FI, 70% FI) were probably due to the enhanced transformation of acids to sugars in dehydrated juice sacs which are required to maintain the osmotic pressure of fruit cells (Huang et al. 2000; Navarro et al. 2010). Earlier studies also demonstrated comparatively better fruit quality (higher TSS and lower acidity) of citrus fruits under optimal DI over FI (Panigrahi et al. 2014). In contrast, Pérez-Pérez et al. (2008) and García-Tejero et al. (2010) reported higher TSS and acidity in juice under DI in ‘Lane late’ orange and ‘Navelina’ sweet orange, respectively, in Spain. Furthermore, Gasque et al. (2010) did not observe any significant effect of DI on the fruit quality of ‘‘Navalina’ sweet orange in Spain. The differences in the effects of DI on fruit quality are ascribed to crop type, timing and duration of water stress and agro-climate of the study regions.

Economic analysis shows that FI produced the maximum gross income, followed by DI at 70% FI and DI at 50% FI due to higher fruit yield under a higher level of irrigation (Table 5). However, the net income generated under DI at 50% FI was the highest, followed by DI at 75% FI. This was due to the higher cost of production caused by higher investment in drip irrigation (INR 82,000 ha⁻¹), labour charges for irrigation and electrical energy used in pumping water under FI. These results corroborate the findings of Panigrahi et al. (2013) under DI in Kinnow mandarin grown in a semi-arid environment in north India. The DI at 50% FI produced the highest net economic water productivity (INR 70.19 m⁻³) and benefit:cost ratio (4.60) among the treatments. On the other hand, the high water-stressed plants that were under DI at 30% FI resulted in the lowest economic net return among the treatments.

DI was found as a potential water-saving strategy in a matured citrus orchard. Irrigation with a 50% reduction in water supply could increase the water productivity by 75% without affecting the fruit yield significantly compared with FI in Nagpur mandarin in vertisol. Moreover, it was demonstrated that a matured mandarin plant could be grown with the application of 10 − 31 L day⁻¹ water under a drip system in a hot-dry climate of central India. The significant variation in SWC in the top 0.40 m soil, and irrigating the plant at 30% depletion of SWC under 50% FI suggest taking this depth and level of SWC for drip irrigation scheduling in the crop. Overall, substantial water-saving and increased water productivity with better quality fruits under 50% deficit water supply suggest its adoption in matured Nagpur mandarin orchards in the study region, and elsewhere with a similar agro-climate to the study region. This will help in increasing the areas under irrigation, resulting in higher production of quality citrus fruit.

The author acknowledges the support and facilities rendered by Director, National Research Centre for Citrus, Nagpur, India during this study.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.