Simulating Irrigation Requirements for Vegetable Crops in a Humid Region Considering a Changing Climate

Meaghan Kilmartin¹, Chandra A. Madramootoo^1,*

Published in Journal of Natural Resources and Agricultural Ecosystems 3(1): 1-16 (doi: 10.13031/jnrae.16048). Copyright 2025 American Society of Agricultural and Biological Engineers.

¹ Department of Bioresource Engineering, McGill University, Sainte-Anne-de-Bellevue, Quebec, Canada.

^* Correspondence: chandra.madramootoo@mcgill.ca

The authors have paid for open access for this article. This work is licensed under a Creative Commons Attribution NonCommercial NoDerivatives 4.0 International License https://creative?commons.org/licenses/by-nc-nd/4.0/

Submitted for review on 26 April 2024 as manuscript number NRES 16048; approved for publication as a Research Article and as part of the Climate Change Impacts on Natural Resource and Agricultural Systems Collection by Associate Editor Dr. Younggu Her and Community Editor Dr. Kati Migliaccio of the Natural Resources & Environmental Systems Community of ASABE on 17 November 2024.

Citation: Kilmartin, M., & Madramootoo, C. A. (2025). Simulating irrigation requirements for vegetable crops in a humid region considering a changing climate. J. Nat. Resour. Agric. Ecosyst., 3(1), 1-16. https://doi.org/10.13031/jnrae.16048

Highlights

• Water deficiency of 50% did not reduce yields.
• Potato needs a minimum of 247 mm of water by the 2080s.
• Irrigation requirements of squash are 124 mm for 2080s.

Abstract. This study assessed the supplemental irrigation needs in a humid vegetable producing region of Quebec using the AquaCrop model. Three irrigation treatments were investigated, comprising a maximum allowable depletion (MAD) of 20%, 35%, and 50% of plant-available water (AW) for potato and squash cultivated on sandy soils. The AquaCrop models were calibrated to field measurements of soil moisture in 2022. Potato and squash models simulated soil moisture with a strong agreement to field sensor measurements, with a Willmott index of agreement of 0.91 and 0.84, respectively. Statistical analysis revealed that both the historical weather and the irrigation treatment (20%, 35%, or 50% depletion of AW) had a significant effect on net irrigation requirement. The MAD of 50% AW demonstrated significant water-saving potential, particularly in dry years where crop-water demand was significantly higher. Under the 50% MAD, the mean net irrigation requirement for potato and squash in historical dry years were 265 mm and 151 mm, respectively. The models were then used to predict the impact of climate change on irrigation requirements for the periods 2050 and 2080. The irrigation requirement for potato is predicted to significantly increase by the 2080s compared to the historical period (1997–2021), under the CMIP-6 SSP5-8.5 high emissions scenario. Projected irrigation requirements for squash remained stable. Irrigation treatment significantly impacted the net irrigation requirement, with an MAD set at 50% AW significantly reducing irrigation requirements across all climate periods.

Keywords.Climate change, Crop modeling, Humid regions, Irrigation water requirement.

In many regions, water is a limiting factor in agricultural production. The changing weather and climate, increasing population, and groundwater depletion are all factors that exert pressure on limited water supplies (Kisekka et al., 2017). Since agricultural production is a major consumer of water, it is important to use water more efficiently to ensure sustainable crop production (De Souza and Koo-Oshima, 2021). Interest in the issue of water limiting agricultural production has been very high in areas with very limited rainfall in the arid and semi-arid zones, where without irrigation, production is almost negligible (Salman et al., 2021). However, in today’s changing climate, optimizing crop water productivity is increasingly important in semi-humid and humid zones. There exists a growing uncertainty and variability of rainfall amount and distribution in all regions. This is augmenting concerns for farmers, water managers, and government agencies regarding water allocation and the timing and quantities of supplemental irrigation (Paz et al., 2007; Anapalli et al., 2016; Danielescu et al., 2022). Meanwhile, maximizing the economic yield through optimizing irrigation practices is paramount for feeding a growing global population (Licker et al., 2010; van Ittersum and Cassman, 2013). Predicting crop-water requirements to improve water use efficiency is thus essential in all climatic regions.

This study is focused on an agricultural region located in Lanoraie, Quebec. The growing season in Lanoraie is increasingly drier and more variable due to climate change, with 2021 being the driest and second-warmest growing season in the last 25 years.

Recent advances have led to a plethora of methods and devices to estimate crop water requirements and water stress, often integrating combined remote sensing and ground data into crop and soil water models (Cahn and Johnson, 2017; Corbari et al., 2019; Tolomio and Casa, 2020; Pereira et al., 2021). Crop models have diverse applications in agricultural water management, ranging from on-farm irrigation scheduling to regional assessments and planning. Enhanced detection of crop water stress to improve irrigation scheduling helps optimize water use at the field scale, especially for high-value crops, such as vegetables (Ihuoma and Madramootoo, 2017). Crop models can be embedded in decision support systems for irrigation scheduling to provide real-time, site-specific information for farmers and technicians (Gallardo et al., 2020; Zhai et al., 2020) and used to assess the effects of deficit irrigation (Paredes et al., 2014; Montoya et al., 2017) or climate change (Khordadi et al., 2019) on crop water requirements and yields. AquaCrop is a water-driven model developed by the FAO, based on crop yield response to water that is particularly suited for water-limited conditions. It requires a relatively low number of input parameters and is well known for its balance of simplicity, robustness, and accuracy (Steduto et al., 2009; Vanuytrecht et al., 2014). AquaCrop can also be applied to a variety of crops (Geerts et al., 2009; Stricevic et al., 2011; Bello and Walker, 2017), which has led to its widespread use.

The specific objectives of this study are to: (1) evaluate the ability of AquaCrop to simulate soil moisture content for the major irrigated crops in Lanoraie for water resource management and planning; (2) estimate the net irrigation requirements (I_net) for the historical dry, average, and wet years of the study area, for the major irrigated crops, soil types, and irrigation methods used; (3) predict the impact of climate change on the future irrigation requirements of the two major irrigated crops.

Materials and Methods

Simulating Irrigation Requirements

The AquaCrop model offers an updated approach for estimating crop productivity with respect to water supply and agronomic management based on current concepts of plant physiology and soil-water budgeting (Raes et al., 2009). The model's robustness and ability to depict the impacts of water stress at specific growth stages make it valuable for assessing irrigation strategies and studying factors like climate, soil type, field management techniques, and sowing dates on crop production and water use efficiency. In this study, AquaCrop was employed under different irrigation strategies for each of the four calibrated crop models to estimate the historical irrigation requirements for the last 25 years (1997–2021). Furthermore, the models were used to predict the impact of climate change on the near (2050s) and far (2080s) future irrigation requirements.

Estimating Historical Irrigation Requirements

Each of the calibrated AquaCrop models was used to estimate the net irrigation requirements for the last 25 years (1997–2021) for three different irrigation treatments. The climate data obtained were input as a climate file. Agronomic management practices, such as planting date and weed control, were kept consistent to grower specifications. The crop file was converted from calendar day format to growing degree days to account for the variability in the duration of the crop cycle due to temperature over the growing season, which was also required for predicting climate change scenarios (Tsakmakis et al., 2019).

The irrigation file was set to determine the net irrigation requirements. Three irrigation treatments were investigated based on the maximum allowable depletion (MAD) of the available water content (AW) of the soil. Considering potato is a shallow-rooted crop, commonly grown on medium to coarse soils, it is sensitive to water stress (King et al., 2020; Stark et al., 2020). The recommended soil water content (SWC) for potatoes is typically optimal when kept between 60 and 85% of AW at different stages of the growth cycle (Allen et al., 1998; King et al., 2020). In Quebec, Dubé and Rochette (1985) suggest a minimum of 50% AW for all growth stages, a level also used by Boisvert et al. (1992) in Ottawa for irrigated potato production experiments. Squash is a shallow-rooted crop sensitive to excess soil water from seeding to maturity (Amer, 2011; OMAFRA, 2022). Given the relatively shallow depth of roots, in the top 40–50 cm of soil, it is crucial to maintain SWC above 50% of AW to prevent detrimental water deficits (Allen et al., 1998; Hess et al., 1997). Therefore, the AquaCrop model irrigation was set to trigger when 20%, 35%, or 50% of the total available water (TAW) in the soil had been depleted. In all cases, water was applied to bring the root zone soil moisture back up to field capacity (FC).

Five dry, five average, and five wet years were identified by comparing the total rainfall over the crop growth period from 1997–2021. The mean simulation results of five years were calculated to represent the historical dry, average, and wet growing season weather conditions.

The simulated output was then compared statistically to evaluate the effects of irrigation treatment and weather on net irrigation requirements, dry yield, and water productivity. The statistical analysis was used to recommend the optimum irrigation strategy in dry, average, and wet growing season conditions.

Assessing the Impact of Climate Change on Irrigation Requirements

Additionally, the models were used to predict the impact of climate change on future irrigation requirements to the 2050s and 2080s. Each AquaCrop model was run for an ensemble of five GCMs under the SSP5-8.5 scenario. The CO₂ concentrations for SSP5-8.5 were selected from the AquaCrop database, which contains CO₂ files corresponding to the IPCC’s five SSP climate scenarios (Raes et al., 2022). All other parameters remained consistent with the historical simulations. The three irrigation treatments were investigated, corresponding to a MAD of 20%, 35%, and 50% AW.

The simulated net irrigation demands were compared to the baseline period (1997–2021), similarly to Busschaert et al. (2022), to assess changes in the mean irrigation requirements. The simulated results for each model in the ensemble were averaged over 30 years, typical of climate change crop simulation studies (Bird et al., 2016; Govere et al., 2022). The 2050s represents a 30-year average over 2036–2065, and the 2080s represents a 30-year average over 2066–2095.

The objective was to estimate the mean I_net for near and far future climate periods and relate this to the historical baseline. The effect of irrigation treatment and climate on I_net was investigated with a statistical analysis.

Statistical Analysis of Significance

The statistical significance of the irrigation treatment and weather or climate were tested using a two-way ANOVA with factorial interaction, executed with the GLM procedure in SAS Studio version 3.81 (SAS Institute Inc., Cary, NC). The significant differences of the F-test were determined at p = 0.05. Additionally, the least squared means (LS-means) were computed for each main effect as well as their interaction, using the LS-means statement with a confidence level kept at p = 0.05.

Field Sites

In order to apply the AquaCrop model, two field sites were selected. The study area is located between 45.82° and 46.09°N and 73.44° and 73.17°W, enveloping the Lanoraie peatland complex and surrounding agriculture. The study area covers a total of 32,400 ha, including 12,000 ha of cropland and 7,600 ha of wetland. The nearest weather station, l’Assomption (Climate ID 7014160), is located at 45.81°N 73.43°W. The average temperature and rainfall over the growing season from May to September are 17.7°C and 480 mm. Potatoes, other vegetables, and small fruits are the main irrigated crops. Water for irrigation is mainly sourced from excavated reservoirs and ponds adjacent to streams, as well as by 12 small dams operated by the municipalities.

The major irrigated crop is potato, accounting for nearly two-thirds of the net irrigated acreage. Other vegetables and gourds collectively represent 25% of the irrigated cropland. Within this group, winter squash was selected as a common crop. The cultivated soil type is largely homogeneous, with sandy-textured soils covering 76% of irrigated cropland. The highest cultivated acreage is on the soil type Lanoraie fine sand, followed by Saint-Thomas fine sand. Field sites were selected to represent the primary irrigated crop groups and soil types in cooperation with local producers. The sites were: L1) potato grown on Lanoraie fine sand (fig. 1); L2) squash grown on Saint-Thomas fine sand (fig. 2). Field data were collected over the 2022 growing season to refine the AquaCrop models. Weather data were collected from the nearest weather station, in L’Assomption, managed by Environment and Climate Change Canada (ECCC). For the 2022 growing season (May to September), the average temperature was 18°C and the total rainfall was 425 mm, above the historical 25-year average. Weather, groundwater level, and water quality were assumed to be consistent over the study area. Each farmer provided crop, field, and irrigation management information. Cultivation techniques followed traditional farming practices for the area to maximize crop yield and quality.

Figure 1. L1 field site of potato grown on Lanoraie fine sand, showing geographic location of the field, the soil moisture sensors for data collection, and the water intake locations of nearby producers.

Crop Management

The crop growing season, field management, and irrigation at both field sites are summarized in table 1. At field site L1, potato tubers (cultivar Solanum tuberosum FL-2137) were sown on 11 May, and seedlings emerged on 5 June. Flowering began on 2 July and lasted 28 days. Plant senescence occurred on 15 August, and the harvest was conducted on 22 September, when tubers reached maturity. Potatoes were grown on a hill furrow design with a row spacing of 0.90 m and plant spacing of 0.30 m. The canopy cover reached a maximum of 90%. Weed management was very good, meaning weeds accounted for only 5% of the canopy cover over the season. Irrigation was applied via a center pivot system with an application pressure of 103 kPa. There were four irrigation events in 2022: 8 July (18 mm), 14 July (18 mm), 21 July (22 mm), and 1 August (25 mm).

Figure 2. L2 field site of squash grown on Saint-Thomas fine sand, showing geographic location of the field, the soil moisture sensors for data collection, and the water intake locations of nearby producers.

Winter squash seedlings (Cucurbita maxima) were transplanted at field site L2 on 13 June, with a row spacing of 1.2 m and plant spacing of 0.45 m. Flowering began 13 July and lasted approximately 35 days. The field was nearly entirely covered at maximum canopy cover, and senescence started 20 August. Squashes were clipped from their vines 25 August and left to ripen in the field. Weed management was very good, i.e., 5% relative weed cover. A depth of 25 mm of water was applied to the field by sprinkler irrigation on three occasions during the growing season.

Soil Physical Characteristics

The soil from both field sites was classified as sandy soil. Bulk density was determined in the laboratory from soil core samples collected in the field. Soil texture was reported by IRDA Quebec pedological data. Soil hydraulic properties, such as FC, wilting point (WP), saturation point (SAT), and saturated hydraulic conductivity (K_sat), were estimated using the Soil Water Characteristic Curve (SWCC) pedo-transfer function model developed by Saxton and Rawls (2006). FC and WP were subsequently calibrated to the field conditions, within a reasonable range for the soil class.

Soil Moisture Data

At each field site, Decagon capacitance sensors (Decagon Devices Inc., Pullman, WA) were installed in duplicate at two depths within the crop’s root zone to monitor soil moisture. Two sensors measured soil moisture closer to the surface, at a 5 cm depth, while two others measured soil moisture deeper in the root zone at depths varying between 20 cm and 30 cm, depending on the crop. They were connected to EM50 data loggers (METER Group, Pullman, WA, USA) and monitored throughout the growing season. Sensor data were retrieved via the EC2H20 Utility 1.8 software (METER Group, Pullman, WA, USA).

Volumetric water content (VWC) (cm³ cm^-3) was logged at five-minute intervals throughout the growing season. Field-specific calibration of soil moisture sensors is an important step in ensuring the accuracy of the sensor readings (Robinson et al., 2008). Therefore, soil core samples of known volume were collected during eight field visits to the sites throughout the 2022 growing season. The gravimetric samples were collected in duplicate at each sensor depth and converted to VWC using the measured bulk density. Similar to Wang et al. (2023), measured soil bulk density and gravimetric water content were used to correct the VWC recorded by the sensors.

Consistent with the methods of Danielescu et al. (2022), the corrected VWC was averaged for the replicates at each depth to obtain a representative daily value. Next, daily VWC measurements (cm³ cm^-3) were multiplied by the thickness of the root zone to achieve an equivalent depth of water column (mm) and allow comparison with simulated soil water content (SWC) in the root zone.

Meteorological Data

Historical Climate Data

For the historical analysis, the meteorological data, including precipitation (mm), air temperature (°C), relative humidity (%), and wind speed (m s^-1), were extracted from the L’Assomption weather station of ECCC on an hourly basis for the years 1997–2022 (ECCC, 2022) and processed to calculate daily maximum and minimum values for temperature and relative humidity, as well as total daily precipitation and average wind speed. Net solar radiation was obtained from the NASA Langley Research Center Prediction of Worldwide Energy Resource Project (NASA, 2022). Relative humidity, wind speed, and solar radiation were used to calculate ET_o using the FAO Penman-Monteith Equation (Allen et al., 1998), and thereafter determining the ET_c of the main crops for the last 25 years. Mean annual atmospheric CO₂ concentrations are provided in AquaCrop from the Mauna Loa Observatory.

Future Climate Data

To predict the impact of a changing climate, daily downscaled and bias-adjusted variable data were obtained from ClimateData.ca (2022). ClimateData.ca was created through a collaboration between the Pacific Climate Impacts Consortium, Ouranos Inc., the Prairie Climate Centre, ECCC, Centre de Recherche Informatique de Montre´al, and Habitat7. The latest phase of the Coupled Model Intercomparison Project (CMIP-6) ensemble was selected as it is the most current global climate model data available. The ensemble consists of projections from five climate change models: IPSL-CM6A-LR, CNRM-CM6-1, MIROC6, M-ROC-ES2L, and CNRM-ESM2L, under the SSP5-8.5 emission scenario. The ensemble was selected based on the data format and availability from ClimateData.ca (2022) to ensure compatibility with the AquaCrop model’s calendar. The SSP5-8.5 (high-emissions scenario) was chosen to maintain consistency between the work by the IRDA team for the SCELANEAU project.

Each GCM was statistically downscaled to the study area using BCCAQv2 methodology. BCCAQv2 is a hybrid method combining results from Bias Corrected Constructed Analogs (BCCA) (Maurer et al., 2010) and Quantile Delta Mapping (QDM) (Cannon et al., 2015). CMIP-6 datasets have been downscaled to a predefined resolution of approximately 6 x 10 m (0.0833° latitude x 0.0833° longitude). Mean annual CO₂ concentrations corresponding to the SSP5-8.5 scenario were selected in AquaCrop.

Two future periods were defined for the projected simulation results to compare near and far effects on irrigation requirements to the baseline period (1997–2021). The 2050s represents a 30-year average over 2036–2065, and the 2080s represents a 30-year average over 2066–2095. The climate normal for the baseline period was compared with the normal for the 2050s and 2080s under the SSP5-8.5 scenario to investigate the projected changes in temperature and precipitation. For estimating future irrigation requirements, each year was run in the model using daily climate data. Simulated net irrigation was subsequently averaged over the 30-year periods to evaluate the effect of climate change.

AquaCrop Model Calibration

The AquaCrop models for potato and squash were calibrated with the 2022 field data from each site. We acknowledge the need for additional years of data to increase the reliability of model calibration and validation. However, due to funding constraints and the conclusion of the project phase, only one year of data was available for analysis. Each model was initialized with parameters corresponding to the field conditions, encompassing climate and crop, soil, and management variables. ET_o was calculated by the FAO-56 Penman-Monteith Equation and the geographic coordinates and elevation of each field were specified.

Model Parameters

The crop parameters for the potato and squash models are presented in table 2. The conservative crop parameters, those that do not change significantly with time, location, and management, are provided by AquaCrop in the default potato crop file. These values remain the same for any potato cultivar grown in any region of the world (Hsiao et al., 2009). Since squash is not a defined crop in AquaCrop, values were extracted from literature and calibrated to local conditions if they simulated a response in SWC. These conservative parameters include the crop coefficient for transpiration at full canopy, the canopy decline coefficient, biomass water productivity, and the reference harvest index. Non-conservative parameters, which are management and cultivar specific, are also required. Planting method, plant density, dates of planting, emergence, start of senescence, and maturity were specified by farmer observations. Planting dates mirrored the actual agricultural practices in the study area, ensuring that the model was aligned with the local agricultural calendar. The crop response to soil salinity or soil fertility stress was not considered.

The soil physical attributes derived from the SWCC are reported in table 3. Soil water attributes, such as FC and WP, derived from the SWCC pedo-transfer function, were evaluated as sensitive parameters. The TAW content for the investigated soil depth is estimated in the last column of table 3. TAW is used to simulate different irrigation regimes based on MAD. The model automatically adjusted the curve number (CN) to 46 and the readily evaporable water (REW) to 2 mm, according to the sandy soil type of L1 and L2. The CN was subsequently calibrated while remaining within the range for soil hydrological group A. The initial condition for soil moisture was assumed at FC on 1 May. Capillary rise was negligible since the groundwater table was below 2 m. Management parameters associated with irrigation and field practices, such as irrigation events and weed management, were defined as specified by the farmer questionnaire to mimic the irrigation practices employed in the region. This site-specific approach was pivotal in accounting for the conditions and practices of the study area.

Sensitivity Analysis and Calibration

An iterative parameter adjustment process was initiated to fine-tune the model. This calibration primarily focused on parameters associated with soil water infiltration and crop growth. Additionally, a sensitivity analysis (SA) was carried out to identify the parameters that had the most significant influence on model predictions (Cao and Petzold, 2006), providing insights into the critical factors affecting soil moisture dynamics. The SA followed the methodology outlined by Geerts et al. (2009). The parameter inputs were changed one at a time while the others remained constant. The interval of variation of the inputs was chosen as -25% to +25% of the standard value. Model outputs under the changed inputs were compared to the basic output. Variables were assigned a sensitivity level of high, medium, or low if the model’s response to the changes was greater than 15%, between 2% and 15%, or less than 2%, respectively (Geerts et al., 2009). The results of the SA helped narrow the calibration efforts. The aim was to minimize the disparity between the model's simulated soil moisture levels and the observed field measurements through a trial-and-error approach.

Statistical Metrics for Model Evaluation

To evaluate the calibrated AquaCrop model's performance, a comprehensive comparison was conducted between its simulated soil moisture values and the observed field data. Due to the individual strengths and weaknesses of statistical indicators, model performance evaluation requires the use of multiple metrics for a comprehensive assessment (Willmott, 1982; Feng et al., 2022). Consistent with recent AquaCrop model performance studies (Razzaghi et al., 2017; Paredes et al., 2018; Terán-Chaves et al., 2022; Wang et al., 2023), the goodness-of-fit of each crop model was evaluated using the following statistics: the Pearson correlation coefficient (r), the root mean square error (RMSE), the normalized root mean square error (NRMSE), the Nash-Sutcliffe model efficiency coefficient (EF), and the Willmott index of agreement (d).The r, RMSE, NRMSE, EF, and d are described in equations 1 to 5 (Nash and Sutcliffe, 1970; Willmott, 1982; Moriasi et al., 2007):

(1)

(2)

(3)

(4)

(5)

where

O_i = observed values

S_i = simulated values

= mean observed values

= mean simulated values

n = number of observations.

According to FAO (2015), r values > 0.90 indicate a very good agreement, while values of 0.70 to 0.90 and 0.50 to 0.70 are considered good and acceptable, respectively. Values below 0.50 indicate poor agreement. A simulation can be considered excellent if the NRMSE is less than 10%, good if between 10% and 20%, fair if between 20% and 30%, and poor if greater than 30% (Jamieson et al., 1991). An EF of 1 indicates a perfect fit, while an EF of 0 implies the model is only as accurate as the observed mean, and EF <?0 suggests the observed mean is a better estimator than the model (Legates and McCabe, 1999). An EF greater than 0.40 is considered acceptable for SWC in crop modeling studies (Geerts et al., 2009; Paredes et al., 2014; Razzaghi et al., 2017). The index of agreement (d) ranges from 0, meaning complete disagreement, to 1, meaning predicted and observed data are identical (Willmott, 1982; Legates and McCabe, 1999). The simulation results were considered good when the majority of the statistical indicators of the model evaluation were classified as good to very good (Terán-Chaves et al., 2022).

It is common practice in hydrological modeling to reserve a separate dataset for model validation, distinct from the data used in the calibration process (Arsenault et al., 2018). However, due to the inaugural phase of the project and limited timeframe, only data from one growing season were collected. Therefore, it was not possible to validate the models. These AquaCrop models, calibrated against soil moisture field data, served as the foundation for subsequent irrigation simulations.

Results and Discussion

Historical Weather and Trends

Analyzing historical weather data (1997-2021) for the growing season (1 May to 30 September) revealed a 1-in-25 dry season in 2021, an average season in 1998, and a wet season in 2015, corresponding to 280 mm, 478 mm, and 602 mm of rainfall, respectively. There are substantial year-to-year fluctuations in total growing season rainfall, with a decreasing trend from 2015 to 2021. Concurrently, there was a general upward trend in mean temperature. The 2021 growing season was the driest and second warmest, with a total of 280 mm of rainfall and an average temperature of 18.9°C. This aligns with producers' concerns about increased pressures on irrigation resources and the difficulties they faced during the dry spells of 2021. Interestingly, the growing season ET_o shows less variation year-to-year, ranging between 525 mm and 612 mm, and appears uncorrelated to rainfall. Therefore, rainfall, as well as its distribution throughout the growing season, is expected to have an important influence on the soil water balance. The reference year in which field data were collected was 2022.

Table 4 ranks the historical growing season weather data from least to greatest rainfall. Considering the preceding 25 years, the 2022 growing season exhibited an above-average total rainfall (524 mm), a typical mean temperature (17.9°C), and standard ET_o (553 mm). Despite a 10-day dry spell in early May, rainfall events were distributed across the season. Air temperature reached a maximum of 33.1°C on 7 August and was accompanied by a rainfall event, which subsequently cooled temperatures.

Projected Climate and Trends

Climate variables predicted by the CMIP-6 model ensemble were imperative to examine the potential impact of climate change on irrigation demands in the study region. Under SSP5-8.5, the projected mean temperature over the growing season exhibits a discernible upward trajectory, with a simultaneous amplification in the variability among model predictions as the temporal scale extends further into the future. The precipitation during the growing season displays a rising variability from one year to another, alongside fluctuations within the ensemble. These short-term fluctuations and anomalies highlight the importance of averaging data over a 30-year period to provide a statistically stable representation of underlying climate patterns.

The future changes in mean temperature and precipitation over the growing season compared to the historical baseline (1997–2021) are presented in table 5, considering 30-year means. For all models of the ensemble, there is a systematic increase in mean temperature. From the ensemble median, an increase of 2.2°C and 6.7°C is predicted by 2050 and 2080, respectively. The higher increase is logically associated with the farther period given the pessimistic emission scenario. The growing season rainfall is projected to decrease by 6.5 mm by 2050 and by 15.4 mm by 2080 compared to the baseline, considering the ensemble median. However, there are within-ensemble disagreements regarding the direction of change in precipitation. Additionally, this assessment omits the seasonal distribution of rainfall. This stresses the importance of simulating irrigation needs on an annual basis and thereafter evaluating the 30-year mean of net irrigation requirements.

Soil Moisture Dynamics

Soil moisture was monitored at the two field sites. Continuous capacitance sensor readings of VWC were calibrated with reference gravimetric measurements. Daily mean soil moisture data for each field site are presented below.

Figure 3 presents the average daily VWC for the L1 field site, alongside rainfall and irrigation depths. Overall, VWC at 5 cm depth (depicted in green) responded more prominently to rainfall events compared to 30 cm depth (depicted in orange), attributed to the superficial root system of the potato crop and the plant water uptake. The maximum root density is concentrated in the upper 30 cm of soil (Opena and Porter, 1999; Paredes et al., 2018; King et al., 2020; Djaman et al., 2022), which is where rapid water absorption occurs (Fulton, 1970). For irrigated potatoes on sandy soils, Alva (2008) found that water uptake by the plants was primarily between 0 to 30 cm of soil, with the majority of water uptake at 10 cm. As a result, only the effects of large rainfall events are observed in VWC readings at 30 cm, during which the water infiltration rate surpassed the plant’s water absorption rate.

Over the data collection period (23 June to 24 August 2022), the VWC at 5 cm and 30 cm was approximately 0.087 (mean) ±0.034 (standard deviation) cm³ cm^-3 and 0.067 ±0.015 cm³ cm^-3, respectively. The minimum root zone VWC occurred in July and August, coinciding with peak absorption during tuber initiation and bulking. During these critical stages, tubers grow rapidly and consist of 72% to 86% water (Stevenson et al., 1994; Pavlista, 1995). VWC exceeded FC after notable rainfall and irrigation events, but these peaks were short-lived (one to two days), reflecting the low water retention of the sandy soil. In agreement with field observations, soil moisture was well below the SAT throughout the season and did not reach below the permanent WP. However, the low VWC observed at the start of July and the middle of August, correlated to an absence of rainfall or irrigation, suggests that the crop experienced periods of soil water deficit. In general, the temporal and depth-related soil moisture dynamics at the L1 site aligned with the observed rainfall patterns, irrigation practices, and the developmental stages of the potato crop.

Figure 4 displays the average daily VWC at L2, including rainfall and irrigation applications. VWC at 5 cm and 20 cm depths follows a similar pattern throughout the growing season. Root zone VWC typically exceeded surface levels, especially after dry periods, likely due to surface evaporation and higher root water absorption near the surface. Soil moisture peaks due to rainfall events are followed by a gradual decrease in VWC, pointing to higher water retention compared to the L1 site.

Throughout the growing season, from 17 June to 24 August 2022, the mean VWC at 5 cm was 0.136 ±0.049 cm³ cm^-3, while at 20 cm it was 0.164 ±0.026 cm³ cm^-3. Peaks in VWC correlated with large rainfall events, excluding the 28 mm precipitation on 12 July. The relatively stable soil moisture on that day may indicate different weather conditions between the weather station location and the field site. This could also suggest increased water absorption by the developing roots. Additionally, the impact of the 8 July irrigation event on VWC was likely attenuated by increased water uptake during flowering and fruit formation.

Calibration of AquaCrop Models

To estimate irrigation requirements, the AquaCrop model simulates the soil water balance in daily time steps. In the soil water balance, the root zone can be regarded as a reservoir. By tracking the inflow and outflow of water at the boundary, the fluctuations of SWC in the root zone can be monitored. The observed SWC in the root zone was compared to the SWC simulated by AquaCrop for each crop during the 2022 growing season.

Following the methodology outlined by Geerts et al. (2009), FC and WP were highly and moderately sensitive parameters, respectively. Conversely, the maximum canopy cover, saturated hydraulic conductivity, and CN had low sensitivity.

AquaCrop's simulated SWC for the effective root zone of the potato crop aligned well with observed trends, responding effectively to precipitation and irrigation inputs. Overall, there was a strong agreement between observed and simulated SWC, indicated by a Pearson correlation (r) of 0.88 and Willmott index of agreement (d) of 0.91 as shown in table 6. The model estimated SWC with satisfactory accuracy for the 2022 field conditions, with an RMSE of 5.3 mm, an NRMSE of 21.8%, and an EF of 0.54. These results are similar to the findings of Montoya et al. (2016), Razzaghi et al. (2017), and Wang et al. (2023), who used AquaCrop to simulate SWC in potato fields under various irrigation treatments. They found r values between simulated and observed SWC for different irrigation levels varying between 0.67 and 0.86, 0.58 and 0.96, and 0.63 and 0.95, respectively. Furthermore, for eight calibration fields of quinoa, Geerts et al. (2009) simulated SWC in the 30 cm root zone with an average EF of 0.59. Although this study’s RMSE was lower than the range reported by Wang et al. (2023) (6.1 to 11.5 mm), it is important to note that they simulated SWC across a greater soil depth (0 to 50 cm) and therefore had a lower error when normalized. Minor discrepancies, with slight overestimations and underestimations, were observed at the extreme ranges of SWC. Similar trends were noted by Montoya et al. (2016) when SWC approached or exceeded FC or neared WP. These variations in soil water dynamics could stem from the parametrization of soil water characteristics or from disparities in rainfall amounts at the field site compared to the weather data. Considering the results from these other studies, the calibrated potato model performed well at simulating SWC and could be enhanced with field data from additional sites and growing seasons.

Figure 3. Daily volumetric soil water content (VWC) at 5 cm and 30 cm depths during the 2022 season at the L1 field site.

Figure 4. Daily volumetric soil water content (VWC) at 5 cm and 20 cm depths during the 2022 season at the L2 field site.

For squash grown on Saint-Thomas fine sand, FC was also identified as a highly sensitive parameter in simulating SWC, while the crop transpiration coefficient (K_cTr) showed moderate sensitivity. Due to the absence of a default crop file for squash in AquaCrop, a K_cTr calibrated to 0.90 was selected. This value is within the reference range of winter squash values of 0.90 to 0.95 (Allen et al., 1998; Pereira et al., 2021), as well as calibrated values of 0.85 to 0.98 for zucchini squash (Darouich et al., 2020). Parameters such as initial and maximum canopy cover, canopy decline coefficient, shape factor for root zone expansion, WP, saturated hydraulic conductivity, and CN exhibited low sensitivity. Reference values of crop coefficients can serve as an essential starting point for calibrating conservative crop parameters that are not in the model’s database.

In general, the model estimated SWC in the root zone of the squash crop with acceptable accuracy, with an r of 0.76, RMSE of 8.7 mm, NRMSE of 19.2%, EF of 0.35, and d of 0.84 as shown in table 6. AquaCrop’s simulated SWC matches the response of the observed SWC to rainfall and irrigation events, except in the case of the precipitation on 12 July. It is hypothesized that there were differences between the data from the L’Assomption weather station and the actual conditions in the field. Furthermore, there is a slight tendency to underestimate SWC, particularly at the start of the crop period. Similar trends were observed by Mebane et al. (2013), who validated the AquaCrop model for maize production in Pennsylvania. They also used field bulk density and sand, silt, and clay proportions from soil survey data to derive soil water characteristics from the pedo-transfer function model. Therefore, the model’s estimated soil hydraulic characteristics may differ from the actual values in the field, which could be responsible for the disagreement of simulated and observed SWC. There could also be error associated with the estimated irrigation depths and schedule provided by the producer. However, for the planning and management of regional irrigation requirements, a bias toward underestimation of SWC is preferred to overestimation to ensure that sufficient water needs are met.

Simulated Historical Irrigation Requirements

The simulated irrigation requirements for each crop model are presented below for historical weather conditions and the future climate change scenario. Crop production and water productivity outputs are included. While the model’s crop parameters were not calibrated to field data for yield and water productivity outputs, the default crop parameters from AquaCrop’s library provide ways to approximate incomplete information and have shown to accurately simulate production output (Raes et al., 2009). For example, Tsakmakis et al. (2019) found that cotton seed yield was only accurately simulated when the model’s default crop file was used.

Historical Irrigation Requirements of Potato

The simulated historical net irrigation requirements (I_net), crop yield (Y), and ET water productivity (WP_ET) for potato are presented in table 7 for the historical dry, average, and wet years and three irrigation treatments. The growing season rainfall, ET_o, and mean temperature are provided for reference. Each value represents the mean result of five years, representing the dry, average, and wet historical growing season weather conditions. The irrigation requirement decreases with increasing MAD from 20% to 50% of the AW content, since the lower depletion threshold requires smaller but more frequent irrigation applications. This is due to the shorter depletion time for 20% AW from the FC of the soil, compared to the other treatments. As the irrigation requirement decreases, the total dry yield remains constant and results in a greater ET water productivity. These results are consistent with recommended irrigation strategies for potatoes and the general agreement that to maximize yields, AW should not drop below 50% to 65% in the zone of maximum root activity (Singh, 1969; Wright and Stark, 1990; Allen et al., 1998; Shock et al., 2007). Beyond this threshold, crop water stress is predicted to impact yield production.

The statistical significance of the effects of weather and irrigation treatment obtained from the ANOVA are summarized in table 8. In this summary, each value represents the mean of all simulation outputs for the respective treatment in the first column. The F-test revealed that both weather and irrigation treatment had a significant impact on irrigation requirement (p < 0.0001), while the interaction effect did not. In terms of historical weather, as expected, only the wet growing season conditions had significantly lower net irrigation and higher yield than the dry and average years.

Regarding irrigation treatment, all values of I_net were significantly different from each other, showing a significant reduction in irrigation requirements with a MAD of 50% of AW. Danielescu et al. (2022) found that the total amount of supplemental irrigation for potato production in Prince Edward Island also differed significantly between minimal, moderate, and extensive irrigation scenarios, following the same decreasing trend with increased depletion. On the other hand, irrigation treatment had no significant effect on yield, except in the case of rainfed conditions, confirming the importance of supplemental irrigation for potato production. As a result, the 50% depletion had significantly higher water productivity than each of the other irrigation treatments. Although tuber quality parameters were not measured in this study due to the project’s constraints, the irrigation treatment could impact the quality of the potato, with respect to the requirements of the processing sector and end-use market. Yari et al. (2021) found that in the dry year (2015) of their two-year field study in southern Alberta, there were significant differences in tuber quality parameters (such as specific gravity and glucose content) and marketable yield between the irrigation treatments, with normal irrigation producing better quality tubers compared to high and low irrigation.

In the dry season scenario, the LS-Means test of the interaction effect revealed that the simulated I_net was statistically different under each irrigation treatment, whereas in the average and wet seasons, the MAD35% and MAD50% treatments resulted in similar net irrigation requirements. These results highlight the reduced irrigation water required during historical wet growing seasons, as well as the meaningful impact of a lower water depletion threshold in terms of increasing water productivity. Notably, in dry seasons of highest concern for water availability, the irrigation scheduling of potato producers has the greatest impact on their net irrigation usage.

Historical Irrigation Requirements for Squash

The historical simulation results for the squash model are presented in table 9, where each value is the five-year mean corresponding to the historical weather condition. The simulated I_net for squash follows the same trends as the potato simulations, decreasing with augmenting depletion thresholds, as well as from dry to wet growing season conditions. Similarly, as the irrigation requirement decreases, the yield remains constant, and the ET water productivity increases. Due to the low water holding capacity of the soil, allowing the soil moisture to deplete to the recommended 50% of AW (Maughan et al., 2015) reduces the frequency of irrigation events without permitting the crop to experience drought stress.

The statistical significance of the effects of weather and irrigation treatment obtained from the ANOVA are summarized in table 10. The evaluated effects followed similar trends for the historical squash simulations as for potato. The growing season weather had a significant effect on I_net, yield, and WP_ET, but only resulted in significantly different results in the wet years (lower I_net, higher yield, and higher WP_ET). The significant effect of irrigation treatment resulted in statistically different irrigation requirements with respect to each treatment. Meanwhile, squash yield and WP_ET decreased significantly in the case of rainfed conditions only. Overall, the MAD50% irrigation treatment significantly reduces the net irrigation requirement and increases the water productivity.

In summary, the statistical analysis of historical simulations revealed that both historical weather (wet, average, or dry growing season) and irrigation treatment (20%, 35%, or 50% depletion of AW) had a significant effect on the net irrigation requirement. The results highlighted the water-saving potential of irrigating at a MAD of 50%AW, especially in dry years where crop-water demand was significantly higher. For both major crops in the study area, the historical AquaCrop simulations found that a MAD of 50%AW significantly reduced net irrigation requirements and increased water productivity without significantly reducing crop production.

Impact of Climate Change on Irrigation Requirements

The models were used to predict the impact of climate change on irrigation requirements up to the 2050s and 2080s, compared to the historical baseline (1997–2021). The 30-year means were calculated for each timeline (baseline, 2050s, and 2080s), model (in the ensemble), and crop (potato and squash). Irrigation treatments of 20%, 30%, and 50%MAD of AW were simulated.

Projected Irrigation Requirements for Potato

The median and range of the resulting net irrigation requirements for potato are presented in figure 5. The 30-year average simulation results, represented by the median model result and black bars indicating the range, highlight a significant increase in potato irrigation requirements by the projected 2080s period, compared to the historical baseline, under the selected high emissions scenario. These results support previous findings simulated with the AquaCrop model, evaluated for future irrigation requirements across different climate scenarios. For example, Busschaert et al. (2022) indicated a potential increase in net irrigation requirement of 30% in far future periods (2071–2100) under a high-emission scenario (SSP3–7.0), particularly impacting agriculture in central and southern Europe.

In addition, the results showed an increase in tuber yield production. This aligns with findings in global studies on potato crop modeling in the context of climate change, such as those by Jennings et al. (2020), which suggest increased yields with adaptations. Simulating with AquaCrop, Govere et al. (2022) found a beneficial effect of climate change on wheat yield and water use in Zimbabwe under RCP4.5 and RCP8.5 for projected 2040s and 2080s. They attributed the simulated increase in yields and decrease in crop water use to the CO₂ fertilization effects, which had a dominant effect over the projected higher temperatures. However, a specific modeling study on potato production in Prince Edward Island (Adekanmbi et al., 2023) demonstrated a potential decrease in yield due to climate change without adaptations from current agricultural practices. A recent study on the impact of climate change on fully irrigated, well-fertilized potato crop across the USA under the RCP 8.5 scenario of high emissions indicated that while increasing temperatures will likely reduce potato yields, this will be mostly compensated by elevated atmospheric CO₂ (Zhao et al., 2022). Furthermore, yields may improve in most regions by planting potatoes earlier as an adaptation to avoid hot summers. Despite higher temperatures, water use by potatoes is predicted to decrease as a result of a shorter growing season and greater water use efficiency under elevated atmospheric CO₂ (Zhao et al., 2022).

Notably, the statistical analysis demonstrates that the irrigation treatment plays a significant role in influencing the net irrigation requirement, with a MAD set at 50% of AW showing a considerable reduction. The irrigation treatment did not have a significant effect on yield. Overall, ANOVA results underscore the substantial impact of both irrigation treatment and climate on net irrigation requirements, indicating an expected increase under the selected climate model scenario, mitigated by the MAD 50%AW treatment. In summary, the simulation results are in line with the anticipated greater water demand due to weather effects, counterbalanced by the potential for increased crop yield attributed to higher CO₂ concentration.

Projected Irrigation Requirements for Squash

The median and range of the resulting net irrigation requirements for squash are presented in figure 6. For the squash model, the future simulations resulted in stable net irrigation requirements, as well as lower yields in the 2050s and 2080s under the high emissions scenario. These results may be attributed to the observed reduction in the crop's growth cycle, associated with changes in climatic conditions. The mean simulated crop growth cycle was 98 days in the historical baseline period, but it decreased to 77 days by the 2050s and 68 days by the 2080s. While the default values of the conservative AquaCrop parameters (Hsiao et al., 2009; Raes et al., 2009) for the potato crop file are presumed applicable to a wide range of conditions, the conservative parameters for the squash model, such as upper and lower temperature thresholds for leaf growth, were obtained from the default general vegetable file. Future research and model parametrization could further investigate the effect of climate change on the crop cycle of winter squash.

Figure 5. Median and range of projected irrigation requirements under CMIP-6 SSP5-8.5 for potato.

Figure 6. Median and range of projected irrigation requirements under CMIP-6 SSP5-8.5 for squash.

Consistent with the potato simulations, the F-test indicated a significant effect of irrigation treatment on the net irrigation requirements for squash, without significantly changing the yield. The MAD of 50% AW significantly reduces the net irrigation requirements in all climate periods, emphasizing the role of the producer’s selected irrigation regime on irrigation water use.

Conclusions

Overall, the potato and squash AquaCrop models exhibited a good agreement between simulated and observed soil moisture. The prevailing weather conditions of the growing season also had a significant impact on the net irrigation. The climate effect of the CMIP-6 high emissions scenario (SSP5-8.5) was investigated for the two crops in the study area with AquaCrop calibrated to local field conditions. Future irrigation needs were estimated by incorporating climatic data from the CMIP-6 model ensemble, downscaled statistically to the study area. Under this scenario, the future irrigation requirements of each crop had distinct responses to climate change. For potato, the future climate and irrigation strategy both played a significant role in influencing net irrigation requirements, projecting a significant irrigation increase by the 2080s. Projections for squash suggested stable seasonal water needs, potentially attributed to shortened crop growth cycles simulated for the 2050s and 2080s. Across all climate periods, maintaining a MAD of 50%AW emerged as the recommended irrigation management strategy, as it consistently reduced net irrigation requirements for both crops without significantly impacting the yield.

Further model calibration is suggested to investigate the effect of climate change on shortened crop growth cycles. Future research could explore the optimization of crop varieties and planting dates to adapt the agricultural region to climate change, as highlighted by other studies (Khordadi et al., 2019; Sharma et al., 2021). Conducting future model simulations with climate change data modeled under additional shared socio-economic pathways would provide a more expansive and comprehensive understanding of the potential effects of climate change on net irrigation requirements, considering the variability among climate projections. Studying the temperature thresholds and heat stress responses across multiple crop varieties could further explore the impact of climate change on crop cycle length and aid in the selection of cultivars better suited for future climates.

Acknowledgements

This study was conducted with funding from the Quebec Ministry of Agriculture (MAPAQ), through its PALCCA program to the Fédération de l’UPA de Lanaudière in support of its SCELENEAU project. We thank Messrs. Charles Bergeron, Dominic Brochu, and Sylvestre Delmotte of the UPA for their technical and logistical support. We also thank the summer research assistants, Farhan Ahmad and Miranda Xiao, who assisted with the fieldwork and data collection, and the growers on whose fields the study was conducted. We thank Emily Lafond for reviewing and editing the paper. We are also grateful to NSERC and McGill University, through a James McGill Professorship grant to CAM, for their financial support.

References

Adekanmbi, T., Wang, X., Basheer, S., Nawaz, R. A., Pang, T., Hu, Y., & Liu, S. (2023). Assessing future climate change impacts on potato yields — a case study for Prince Edward Island, Canada. Foods, 12(6), 1176. https://doi.org/10.3390/foods12061176

Allen, R. G., Pereira, L. S., Raes, D., & Smith, M. (1998). Crop evapotranspiration — Guidelines for computing crop water requirements — FAO Irrigation and drainage paper 56. Rome, Italy: United Nations FAO. Retrieved from http://www.fao.org/docrep/x0490e/x0490e00.htm

Alva, A. K. (2008). Setpoints for potato irrigation in sandy soils using real-time, continuous monitoring of soil-water content in soil profile. J. Crop Improv., 21(2), 117-137. https://doi.org/10.1080/15427520701885311

Amer, K. H. (2011). Effect of irrigation method and quantity on squash yield and quality. Agric. Water Manag., 98(8), 1197-1206. https://doi.org/10.1016/j.agwat.2011.03.003

Anapalli, S. S., Fisher, D. K., Reddy, K. N., Pettigrew, W. T., Sui, R., & Ahuja, L. R. (2016). Vulnerabilities and adapting irrigated and rainfed cotton to climate change in the lower Mississippi Delta region. Climate, 4(4), 55. https://doi.org/10.3390/cli4040055

Arsenault, R., Brissette, F., & Martel, J.-L. (2018). The hazards of split-sample validation in hydrological model calibration. J. Hydrol., 566, 346-362. https://doi.org/10.1016/j.jhydrol.2018.09.027

Bello, Z. A., & Walker, S. (2017). Evaluating AquaCrop model for simulating production of amaranthus (Amaranthus cruentus) a leafy vegetable, under irrigation and rainfed conditions. Agric. For. Meteorol., 247, 300-310. https://doi.org/10.1016/j.agrformet.2017.08.003

Bird, D. N., Benabdallah, S., Gouda, N., Hummel, F., Koeberl, J., La Jeunesse, I.,... Woess-Gallasch, S. (2016). Modelling climate change impacts on and adaptation strategies for agriculture in Sardinia and Tunisia using AquaCrop and value-at-risk. Sci. Total Environ., 543, 1019-1027. https://doi.org/10.1016/j.scitotenv.2015.07.035

Boisvert, J., Dwyer, L., & Lemay, M. (1992). Estimation of water use by four potato (Solarium tuberosurn L.) cultivars for irrigation scheduling. Can. Agric. Eng., 34(4), 319-325.

Busschaert, L., de Roos, S., Thiery, W., Raes, D., & De Lannoy, G. J. (2022). Net irrigation requirement under different climate scenarios using AquaCrop over Europe. Hydrol. Earth Syst. Sci., 26(14), 3731-3752. https://doi.org/10.5194/hess-26-3731-2022

Cahn, M. D., & Johnson, L. F. (2017). New approaches to irrigation scheduling of vegetables. Horticulturae, 3(2), 28. https://doi.org/10.3390/horticulturae3020028

Cannon, A. J., Sobie, S. R., & Murdock, T. Q. (2015). Bias correction of GCM precipitation by quantile mapping: How well do methods preserve changes in quantiles and extremes? J. Climate, 28(17), 6938-6959. https://doi.org/10.1175/JCLI-D-14-00754.1

Cao, Y., & Petzold, L. (2006). Accuracy limitations and the measurement of errors in the stochastic simulation of chemically reacting systems. J. Comput. Phys., 212(1), 6-24. https://doi.org/10.1016/j.jcp.2005.06.012

ClimateData.ca. (2022). [dataset].

Corbari, C., Salerno, R., Ceppi, A., Telesca, V., & Mancini, M. (2019). Smart irrigation forecast using satellite LANDSAT data and meteo-hydrological modeling. Agric. Water Manag., 212, 283-294. https://doi.org/10.1016/j.agwat.2018.09.005

Danielescu, S., MacQuarrie, K. T., Zebarth, B., Nyiraneza, J., Grimmett, M., & Levesque, M. (2022). Crop water deficit and supplemental irrigation requirements for potato production in a temperate humid region (Prince Edward Island, Canada). Water, 14(17), 2748. https://doi.org/10.3390/w14172748

Darouich, H., Karfoul, R., Eid, H., Ramos, T. B., Baddour, N., Moustafa, A., & Assaad, M. I. (2020). Modeling Zucchini squash irrigation requirements in the Syrian Akkar region using the FAO56 dual-Kc approach. Agric. Water Manag., 229, 105927. https://doi.org/10.1016/j.agwat.2019.105927

de Souza, M., & Koo-Oshima, S. (2021). Food and agriculture. In The United Nations World Water Development Report 2021: Valuing water (pp. 67-78). UNESCO. https://doi.org/10.18356/9789214030140c011

Djaman, K., Koudahe, K., Saibou, A., Darapuneni, M., Higgins, C., & Irmak, S. (2022). Soil water dynamics, effective rooting zone, and evapotranspiration of sprinkler irrigated potato in a sandy loam soil. Agronomy, 12(4), 864. https://doi.org/10.3390/agronomy12040864

Dube´, P., & Rochette, P. (1985). Les de´ficits hydriques. 2, 27-114.

ECCC. (2022). Historical data [dataset]. Environment and Climate Change Canada. Retrieved from https://climate.weather.gc.ca/historical_data/search_historic_data_e.html

FAO. (2015). Statistical indicators. In AquaCrop new features and updates version 5.0 (pp. 82-85). Rome, Italy: United Nations FAO. Retrieved from https://www.pmf.unizg.hr/_download/repository/AQUA_CROP_upute_model.pdf

Feng, D., Li, G., Wang, D., Wulazibieke, M., Cai, M., Kang, J.,... Xu, H. (2022). Evaluation of AquaCrop model performance under mulched drip irrigation for maize in Northeast China. Agric. Water Manag., 261, 107372. https://doi.org/10.1016/j.agwat.2021.107372

Fulton, J. M. (1970). Relationship of root extension to the soil moisture level required for maximum yield of potatoes, tomatoes and corn. Can. J. Soil Sci., 50(1), 92-94. https://doi.org/10.4141/cjss70-015

Gallardo, M., Elia, A., & Thompson, R. B. (2020). Decision support systems and models for aiding irrigation and nutrient management of vegetable crops. Agric. Water Manag., 240, 106209. https://doi.org/10.1016/j.agwat.2020.106209

Geerts, S., Raes, D., Garcia, M., Miranda, R., Cusicanqui, J. A., Taboada, C.,... Steduto, P. (2009). Simulating yield response of quinoa to water availability with AquaCrop. Agron. J., 101(3), 499-508. https://doi.org/10.2134/agronj2008.0137s

Govere, S., Nyamangara, J., & Nyakatawa, E. Z. (2022). Beneficial effect of climate change on wheat yield and water footprints in the Middle-Manyame sub-catchment, Zimbabwe. J. Water Clim. Change, 13(8), 2895-2910. https://doi.org/10.2166/wcc.2022.038

Hess, M., Bill, M., Jason, S., & John, S. (1997). Oregon State University Western Oregon squash irrigation guide. 514, 737-6304. Corvallis, OR: Oregon State University, Department of Bioresource Engineering.

Hsiao, T. C., Heng, L., Steduto, P., Rojas-Lara, B., Raes, D., & Fereres, E. (2009). AquaCrop — The FAO Crop model to simulate yield response to water: III. Parameterization and testing for maize. Agron. J., 101(3), 448-459. https://doi.org/10.2134/agronj2008.0218s

Ihuoma, S. O., & Madramootoo, C. A. (2017). Recent advances in crop water stress detection. Comput. Electron. Agric., 141, 267-275. https://doi.org/10.1016/j.compag.2017.07.026

Jamieson, P. D., Porter, J. R., & Wilson, D. R. (1991). A test of the computer simulation model ARCWHEAT1 on wheat crops grown in New Zealand. Field Crops Res., 27(4), 337-350. https://doi.org/10.1016/0378-4290(91)90040-3

Jennings, S. A., Koehler, A.-K., Nicklin, K. J., Deva, C., Sait, S. M., & Challinor, A. J. (2020). Global potato yields increase under climate change with adaptation and CO2 fertilisation. Front. Sustain. Food Syst., 4. https://doi.org/10.3389/fsufs.2020.519324

Khordadi, M. J., Olesen, J. E., Alizadeh, A., Nassiri Mahallati, M., Ansari, H., & Sanaeinejad, H. (2019). Climate change impacts and adaptation for crop management of winter wheat and maize in the semi-arid region of Iran. Irrig. Drain., 68(5), 841-856. https://doi.org/10.1002/ird.2373

King, B. A., Stark, J. C., & Neibling, H. (2020). Potato irrigation management. In J. C. Stark, M. Thornton, & P. Nolte (Eds.), Potato production systems (pp. 417-446). Cham, Switzerland: Springer. https://doi.org/10.1007/978-3-030-39157-7_13

Kisekka, I., DeJonge, K. C., Ma, L., Paz, J., & Douglas-Mankin, K. (2017). Crop modeling applications in agricultural water management. Trans. ASABE, 60(6), 1959-1964. https://doi.org/10.13031/trans.12693

Legates, D. R., & McCabe Jr., G. J. (1999). Evaluating the use of “goodness-of-fit” measures in hydrologic and hydroclimatic model validation. Water Resour. Res., 35(1), 233-241. https://doi.org/10.1029/1998WR900018

Licker, R., Johnston, M., Foley, J. A., Barford, C., Kucharik, C. J., Monfreda, C., & Ramankutty, N. (2010). Mind the gap: How do climate and agricultural management explain the ‘yield gap’ of croplands around the world? Global Ecol. Biogeogr., 19(6), 769-782. https://doi.org/10.1111/j.1466-8238.2010.00563.x

Maughan, T., Drost, D., & Allen, L. N. (2015). Vegetable irrigation: Squash and pumpkin. Utah State University Extension. Retrieved from https://digitalcommons.usu.edu/extension_curall/745/

Maurer, E. P., Hidalgo, H. G., Das, T., Dettinger, M. D., & Cayan, D. R. (2010). The utility of daily large-scale climate data in the assessment of climate change impacts on daily streamflow in California. Hydrol. Earth Syst. Sci., 14(6), 1125-1138. https://doi.org/10.5194/hess-14-1125-2010

Mebane, V. J., Day, R. L., Hamlett, J. M., Watson, J. E., & Roth, G. W. (2013). Validating the FAO AquaCrop model for rainfed maize in Pennsylvania. Agron. J., 105(2), 419-427. https://doi.org/10.2134/agronj2012.0337

Montoya, F., Camargo, D., Ortega, J. F., Córcoles, J. I., & Domínguez, A. (2016). Evaluation of Aquacrop model for a potato crop under different irrigation conditions. Agric. Water Manag., 164, 267-280. https://doi.org/10.1016/j.agwat.2015.10.019

Montoya, F., García, C., Pintos, F., & Otero, A. (2017). Effects of irrigation regime on the growth and yield of irrigated soybean in temperate humid climatic conditions. Agric. Water Manag., 193, 30-45. https://doi.org/10.1016/j.agwat.2017.08.001

Moriasi, D. N., Arnold, J. G., Van Liew, M. W., Bingner, R. L., Harmel, R. D., & Veith, T. L. (2007). Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Trans. ASABE, 50(3), 885-900. https://doi.org/10.13031/2013.23153

NASA. (2022). Prediction of worldwide energy resource (POWER) project funded through the NASA earth science program (2.0.10) [dataset]. NASA.

Nash, J. E., & Sutcliffe, J. V. (1970). River flow forecasting through conceptual models part I — A discussion of principles. J. Hydrol., 10(3), 282-290. https://doi.org/10.1016/0022-1694(70)90255-6

OMAFRA. (2022). Pumpkin and squash production | ontario.ca. Ontario Ministry of Agriculture, Food and Rural Affairs. Retrieved from http://www.ontario.ca/page/pumpkin-and-squash-production

Opena, G. B., & Porter, G. A. (1999). Soil management and supplemental irrigation effects on potato: II. Root growth. Agron. J., 91(3), 426-431. https://doi.org/10.2134/agronj1999.00021962009100030011x

Paredes, P., D’Agostino, D., Assif, M., Todorovic, M., & Pereira, L. S. (2018). Assessing potato transpiration, yield and water productivity under various water regimes and planting dates using the FAO dual Kc approach. Agric. Water Manag., 195, 11-24. https://doi.org/10.1016/j.agwat.2017.09.011

Paredes, P., de Melo-Abreu, J. P., Alves, I., & Pereira, L. S. (2014). Assessing the performance of the FAO AquaCrop model to estimate maize yields and water use under full and deficit irrigation with focus on model parameterization. Agric. Water Manag., 144, 81-97. https://doi.org/10.1016/j.agwat.2014.06.002

Pavlista, A. D. (1995). EC95-1249 potato production stages: Scheduling key practices. University of Nebraska-Lincoln Extension. Retrieved from https://digitalcommons.unl.edu/extensionhist/1584

Paz, J. O., Fraisse, C. W., Hatch, L. U., Garcia y Garcia, A., Guerra, L. C., Uryasev, O.,... Hoogenboom, G. (2007). Development of an ENSO-based irrigation decision support tool for peanut production in the southeastern US. Comput. Electron. Agric., 55(1), 28-35. https://doi.org/10.1016/j.compag.2006.11.003

Pereira, L. S., Paredes, P., Hunsaker, D. J., López-Urrea, R., & Mohammadi Shad, Z. (2021). Standard single and basal crop coefficients for field crops. Updates and advances to the FAO56 crop water requirements method. Agric. Water Manag., 243, 106466. https://doi.org/10.1016/j.agwat.2020.106466

Raes, D., Steduto, P., Hsiao, T. C., & Fereres, E. (2009). AquaCrop — The FAO crop model to simulate yield response to Water: II. Main algorithms and software description. Agron. J., 101(3), 438-447. https://doi.org/10.2134/agronj2008.0140s

Raes, D., Steduto, P., Hsiao, T. C., & Fereres, E. (2022). AquaCrop on-line reference manual. Retrieved from https://www.fao.org/aquacrop/resources/referencemanuals/en/

Razzaghi, F., Zhou, Z., Andersen, M. N., & Plauborg, F. (2017). Simulation of potato yield in temperate condition by the AquaCrop model. Agric. Water Manag., 191, 113-123. https://doi.org/10.1016/j.agwat.2017.06.008

Robinson, D. A., Campbell, C. S., Hopmans, J. W., Hornbuckle, B. K., Jones, S. B., Knight, R.,... Wendroth, O. (2008). Soil moisture measurement for ecological and hydrological watershed-scale observatories: A review. Vadose Zone J., 7(1), 358-389. https://doi.org/10.2136/vzj2007.0143

Salman, M., García-Villa, M., Fereres, E., Raes, D., & Steduto, P. (2021). The AquaCrop model - Enhancing crop water productivity: Ten years of development, dissemination and implementation 2009-2019. FAO Water Report No. 47. Rome, Italy: United Nations FAO. https://doi.org/10.4060/cb7392en

Saxton, K. E., & Rawls, W. J. (2006). Soil water characteristic estimates by texture and organic matter for hydrologic solutions. Soil Sci. Soc. Am. J., 70(5), 1569-1578. https://doi.org/10.2136/sssaj2005.0117

Sharma, A., Deepa, R., Sankar, S., Pryor, M., Stewart, B., Johnson, E., & Anandhi, A. (2021). Use of growing degree indicator for developing adaptive responses: A case study of cotton in Florida. Ecol. Indic., 124, 107383. https://doi.org/10.1016/j.ecolind.2021.107383

Shock, C. C., Pereira, A. B., & Eldredge, E. P. (2007). Irrigation best management practices for potato. Am. J. Potato Res., 84(1), 29-37. https://doi.org/10.1007/BF02986296

Singh, G. (1969). A review of the soil-moisture relationship in potatoes. Am. Potato J., 46(10), 398-403. https://doi.org/10.1007/BF02869560

Stark, J. C., Thornton, M., & Nolte, P. (2020). Potato production systems. Springer Nature. https://doi.org/10.1007/978-3-030-39157-7

Steduto, P., Hsiao, T. C., Raes, D., & Fereres, E. (2009). AquaCrop — The FAO crop model to simulate yield response to water: I. Concepts and underlying principles. Agron. J., 101(3), 426-437. https://doi.org/10.2134/agronj2008.0139s

Stevenson, W. R., Curwen, D., Kelling, K. A., Wyman, J. A., Binning, L. K., & Connell, T. R. (1994). Wisconsin’s IPM program for potato: The developmental process. HortTechnology, 4(2), 90-95. https://doi.org/10.21273/horttech.4.2.90

Stricevic, R., Cosic, M., Djurovic, N., Pejic, B., & Maksimovic, L. (2011). Assessment of the FAO AquaCrop model in the simulation of rainfed and supplementally irrigated maize, sugar beet and sunflower. Agric. Water Manag., 98(10), 1615-1621. https://doi.org/10.1016/j.agwat.2011.05.011

Terán-Chaves, C. A., García-Prats, A., & Polo-Murcia, S. M. (2022). Calibration and validation of the FAO AquaCrop water productivity model for perennial ryegrass (Lolium perenne L.). Water, 14(23), 3933. https://doi.org/10.3390/w14233933

Tolomio, M., & Casa, R. (2020). Dynamic crop models and remote sensing irrigation decision support systems: A review of water stress concepts for improved estimation of water requirements. Remote Sens., 12(23), 3945. https://doi.org/10.3390/rs12233945

Tsakmakis, I. D., Kokkos, N. P., Gikas, G. D., Pisinaras, V., Hatzigiannakis, E., Arampatzis, G., & Sylaios, G. K. (2019). Evaluation of AquaCrop model simulations of cotton growth under deficit irrigation with an emphasis on root growth and water extraction patterns. Agric. Water Manag., 213, 419-432. https://doi.org/10.1016/j.agwat.2018.10.029

van Ittersum, M. K., & Cassman, K. G. (2013). Yield gap analysis — Rationale, methods and applications — Introduction to the Special Issue. Field Crops Res., 143, 1-3. https://doi.org/10.1016/j.fcr.2012.12.012

Vanuytrecht, E., Raes, D., Steduto, P., Hsiao, T. C., Fereres, E., Heng, L. K.,... Mejias Moreno, P. (2014). AquaCrop: FAO’s crop water productivity and yield response model. Environ. Model. Softw., 62, 351-360. https://doi.org/10.1016/j.envsoft.2014.08.005

Wang, H., Cheng, M., Liao, Z., Guo, J., Zhang, F., Fan, J.,... Wang, X. (2023). Performance evaluation of AquaCrop and DSSAT-SUBSTOR-Potato models in simulating potato growth, yield and water productivity under various drip fertigation regimes. Agric. Water Manag., 276, 108076. https://doi.org/10.1016/j.agwat.2022.108076

Willmott, C. J. (1982). Some comments on the evaluation of model performance. Bull. Amer. Meteor. Soc., 63(11), 1309-1313. https://doi.org/10.1175/1520-0477(1982)063<1309:SCOTEO>2.0.CO;2

Wright, J. L., & Stark, J. (1990). Potato. In Irrigation of agricultural crops - Agronomy (pp. 859-888). Madison, WI: ASA-CSSA-SSSA.

Yari, A., Gilbert, L., Madramootoo, C. A., Woods, S. A., & Adamchuk, V. I. (2021). Optimum irrigation strategy to maximize yield and quality of potato: A case study in southern Alberta, Canada. Irrig. Drain., 70(4), 609-621. https://doi.org/10.1002/ird.2556

Zhai, Z., Martínez, J. F., Beltran, V., & Martínez, N. L. (2020). Decision support systems for agriculture 4.0: Survey and challenges. Comput. Electron. Agric., 170, 105256. https://doi.org/10.1016/j.compag.2020.105256

Zhao, C., Stockle, C. O., Karimi, T., Nelson, R. L., van Evert, F. K., Pronk, A. A.,... Asseng, S. (2022). Potential benefits of climate change for potatoes in the United States. Environ. Res. Lett., 17(10), 104034. https://doi.org/10.1088/1748-9326/ac9242