Authors: Zahra Khosravi, Ali Reza Sepaskhah, Rezvan Talebnejad
Categories: Article, Saffron, Salinity, Irrigation, Planting method, Long-term, Plant sciences, Environmental sciences
Source: Scientific Reports
Authors: Zahra Khosravi, Ali Reza Sepaskhah, Rezvan Talebnejad
Saffron (Crocus sativus L.) is one of the most valuable spice crops in the world, but its productivity is highly sensitive to water availability, soil salinity, and cultivation methods. In many arid and semi-arid regions, where saffron is typically grown, water scarcity and soil salinization are growing challenges that threaten both yield and profitability. Long-term effect of irrigation water salinity in 4 levels [S1 = 0.42 dS m^− 1^ (freshwater) to 3.0 dS m^− 1^], depth of irrigation water in 3 levels (100–50% of water requirement, WR), and two planting methods (P1: basin and P2: in-furrow) on saffron growth, yield and water productivity were studied. In the basin planting, maximum saffron yield (SY) obtained only in the fourth growing season as 12.0 kg ha^− 1^ with 868 mm of rain plus irrigation water. However, in the in-furrow planting, maximum SY occurred during the second to fourth growing seasons as 13.0 kg ha^− 1^ in each year with 832 mm of rain plus irrigation water. Decreasing trend of saffron growth in all the treatments intensified from the fifth growing season onwards. In general, the mean of saffron yield over all growing seasons varied as 8.65–7.43 kg ha^− 1^ with increasing salinity from 0.42 to 3.0 dS m^− 1^. This was 8.5–7.31 kg ha^− 1^ for irrigation water varied from 100% WR to 50% WR. Results also showed that the total SY in six growing seasons in the in-furrow planting was 1.59 times that in the basin planting. The leaf dry matter (LDM), corm yield (CY), and harvest index (HI) decreased in the sixth growing season compared to that in the fifth growing season by 38.4%, 62.0%, and 82%, respectively. It is concluded that saffron is a plant highly sensitive to salinity, and water salinity of 2.0 and 3.0 dS m^− 1^ resulted in 22% reduction in total saffron yield. Mean soil salinity was doubled from 0.7 to 1.3 dS m^− 1^ at the last years that saffron cultivation should be terminated. The benefit-cost ratio (BCR) in the third to fifth growing season under different treatments was higher than 1.0 (as 1.87, 2.42, and 2.23, respectively) indicating economic feasibility of saffron crop production. However, in the sixth growing season, BCR was 0.23 that was not feasible. Therefore, according to the results of SY, CY and BCR, the farmer should terminate the saffron cultivation, and harvest the corm at the end of the fourth and fifth growing season in the in-furrow planting and the basin planting, respectively. The average economic irrigation water productivity (EWPi) in the fourth growing season was 2.2 times its value in the third growing season. However, in the fifth, compared to the fourth growing season, it decreased by 38%.
The online version contains supplementary material available at 10.1038/s41598-025-32330-1.
Due to increasing water demand and scarcity, farmers are compelled to use saline water and grow crops with less water requirement. Scarcity and salinity of water are two important limiting factors for agricultural production, especially in arid and semi-arid regions^1,2^. Due to frequent droughts in these regions, water resources are limited and of low quality, and groundwater is the primary source of irrigation water^2,3^. Therefore, farmers are forced to use saline water and cultivate crops with lower water requirements^4^.
Salinity of irrigation water and the low rainfall depth caused an increase in the soil salinity^5^. Cultivation of salinity-resistant crops, use of suitable irrigation systems with saline water, management techniques^6^, leaching of the root zone^7,8^, drainage^9^, and changes in planting methods^2,3,8^ are among the measures that were taken to deal with the challenges effectively occur in arid and semi-arid areas.
In arid and semi-arid regions, rainfall occurs mainly in autumn and winter. In these seasons, the air relative humidity is high, air temperature is low, and the field evaporation and transpiration are generally reduced. Coincidence of the growing season of the saffron crop with these seasons has caused the area under cultivation of this crop to increase, especially in Iran^10,11^.
In this regard, studies were conducted on the saffron crop, which showed that it is sensitive to deficit irrigation (drought) and highly sensitive to salinity stress^12^. Salinity and deficit irrigation stresses have affected the flowers, corms size and density, the weight of the leaves, and the quality components of saffron^2,13,14^. In Iran, since the beginning, saffron had been cultivated in the basin^15^, but the effect of water and salinity stresses could be moderated to some extent by planting saffron corms in the furrow^2,3^. This method causes a reduction in the evaporation from the surface of the soil, provides more favorable temperature conditions in the root zone, and increases the amount of soil water content and water productivity.
Since the saffron crop is considered a perennial crop, the effect of different stresses on the yield, corms, and leaves should be investigated, for several growing seasons (1–6 years). Therefore, in this research, in continuation of Dastranj and Sepaskhah’s^2^ research (data for the first and second growing seasons), the effect of irrigation water salinity [0.45 dS m^− 1^ (freshwater), 1.0 dS m^− 1^, 2.0 dS m^− 1^ and 3.0 dS m^− 1^], irrigation water levels (100% of crop water requirement, 100%WR as complete irrigation, 75% WR and 50% WR), and planting method (basin and in-furrow planting) were investigated in the last four growing seasons of six growing seasons of saffron crop. The data of the first and second growing season were obtained from Dastranj and Sepaskhah^2^.
Last part of the long-term research was conducted during four growing seasons (3rd to 6th year) on the saffron crop from 2017 to 2021 at the experimental station of School of Agriculture, Shiraz University, in Badjgah region at 29° 56′ N, 52° 02′ E and 1810 m above the mean sea level, in the southwest of Iran. The climate in the study area is semi-arid, with a long-term average air temperature, relative humidity, and precipitation of 13.4 °C, 52.2% and 370 mm, respectively.
Characteristics of the studied soil up to 0.9 m depth are shown in supplemental Table S1. Chemical analysis of the fresh and saline irrigation water is shown in Table S2. To show the variation of weather evaporative condition in different growing seasons, reference evapotranspiration (ETo) was determined by a modified Penman-Monteith equation (supplemental Fig. S1) for semi-arid environments in the study area^16^. Required meteorological data were obtained from the standard weather station at the School of Agriculture, located near the experimental field. Variation of minimum and average daily temperature and average daily relative humidity are shown in Fig. S2.
Experimental design was a split–split plot arrangement in randomized complete blocks design with four irrigation water salinity levels as the main plot, three irrigation water levels as the subplot and two planting methods as the sub-subplot with three replicates. Salinity of irrigation water levels were 0.42 dS m^− 1^ (freshwater, S1), 1.0 dS m^− 1^ (S2), 2.0 dS m^− 1^ (S3), and 3.0 dS m^− 1^ (S4). These salinity levels are chosen according the report of Yarami and Sepaskhah^3^ that saffron is highly sensitive to salinity. Therefore, in this condition, the EC of greater than 1.0 dS m^− 1^ is considered as mildly saline. The irrigation water levels were 100% of crop water requirement, WR (100%WR as complete irrigation, I1), 75% WR (I2), and 50% WR (I3). Basin (P1) and in-furrow (P2) planting were used as planting methods.
The experimental plots were prepared with dimensions of 1.5 × 2 m for the basin and in-furrow method. In 2015, the corms were planted at a depth of 15–20 cm below the soil surface at a density of 15 Mg ha^− 1^ with appropriate fertilization, as reported by Dastranj and Sepaskhah^2^. Corms were obtained from our previous research experiment in the same area. At the beginning of each growing season, chemical fertilizer of triple super-phosphate )100 kg ha^− 1^) and composted cow manure fertilizer (60 Mg ha^− 1^ in the first growing season, and 30 Mg ha^− 1^ in the next growing seasons) were applied to each plot. Chemical properties of the composted manure are presented in Table S3. Freshwater was used for crop establishment in the field at the first irrigation event in all treatments on November 4, 2015. After the first irrigation, different salinity levels and irrigation water levels were applied^2^. Saline water prepared by dissolved NaCl and CaCl2 in the freshwater in equal proportions. In the complete irrigation treatment, 15% of 100%WR was considered the leaching requirement for the saline irrigation water. The treatments continued from 2015 to 2020, and the necessary parameters were determined during the growing periods. The measured values were reported for the first and second growing seasons by Dastranj and Sepaskhah^2^.
Maximum depth of saffron root is 45 cm^17^. Crop water requirements are adjusted for this depth based on a 24-day irrigation interval^11,15^. Using the neutron scattering method, the soil water content was determined from a depth of 0–90 cm (in 30 cm intervals) before each irrigation event during the complete irrigation treatment.
The first irrigation was applied in the first decade of November every growing season (3rd to 6th ), and the next irrigation was applied after the end of flower harvesting. The water content in the root zone was used to determine the irrigation water requirement according to the following Eq^2^. :1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {\mathrm{I}}=\mathop \sum \limits_{{i=1}}^{n} \left( {{\theta _{fc}} - {\theta _i}} \right) \times {z_i}
where I is the irrigation water depth (m), \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \theta _{{fc}} $$\end{document} is the volumetric soil water content at field capacity and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \theta _{i} $$\end{document} is the volumetric soil water content in layer i before the irrigation event (m^3^ m^− 3^), ∆z~i~ is the thickness of each soil layer (m) and n is the number of soil layers. The root depth is estimated by the following equation^18^:2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {\mathrm{R}}_{{\mathrm{d}}} = {\text{ R}}_{{{\text{d min}}}} + \frac{1}{2}{\mathrm{R}}_{{{\text{d max}}}} \left( {{\mathrm{1}} + {\text{ sin}}\left( {{\mathrm{3}}.0{\text{3 }}\frac{{D_{{ag}} }}{{D_{{tm}} }} - {\mathrm{1}}.{\mathrm{47}}} \right)} \right) $$\end{document} where R~d~ is the root depth (m), R~dmin~ is the planting depth (m), R~dmax~ is the maximum root depth for saffron, which is 45 cm^17^, D~ag~ is the number of days after the first irrigation event, D~tm~ is the number of days after first irrigation event that root reaches the maximum depth for saffron that is 173 days^19^. ### Measurement of yield and yield components Yield of each growing season was harvested at the beginning of the next growing season. The appearance of flowers was usually 10 to 14 days after the first irrigation event, and the flowering period lasted about three weeks. The flowers were harvested every morning, and after weighing, the stigmas were separated from the flowers for each treatment and replication. They were then separately air-dried and weighted. The total saffron yield weight was determined for each treatment and each replication. The yield of each treatment was obtained by averaging the total yield of three replicates for each treatment. At the end of each growing season (fifth and sixth), when leaf senescence occurred, leaves were taken from one crop row in the middle of the plots to determine the leaf dry matter. To determine corms density, they were collected from a 30 cm section of one crop row. Harvest index (HI) was also calculated by dividing the saffron yield (style/stigmas) by the sum of the flowers and leaf dry matter for each treatment and growing season. ### Soil sampling for salinity determination After harvesting leaves, soil samples were collected from soil depths of 0–30 cm and 30–60 cm to determine the electrical conductivity of the soil saturated extract (EC~e~). The relationship between the relative yield and the average EC~e~ of the root zone (30–60 cm) was determined using the following equation presented by Maas and Hoffman^20^:3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \frac{{{\mathrm{Y}}_{{\mathrm{a}}} }}{{{\mathrm{Y}}_{{\mathrm{m}}} }}{\text{ = 1}} - {\mathrm{b}}({\mathrm{EC}}_{{\mathrm{e}}} - {\mathrm{EC}}_{{{\mathrm{eth}}}} {\mathrm{)}} $$\end{document} where Y~a~ is the actual yield, Y~m~ is the maximum yield, EC~e~ is the average root-zone salinity of soil saturation extract in 30–60 cm of soil layer, EC~eth~ is the threshold value of EC~e~, and b is the yield reduction coefficient of saffron yield. EC~e~ values for 50% relative yield reduction and b value were determined for each experimental treatment, i.e., irrigation water levels and planting methods. ### Water productivity and economic analysis Crop water productivity (WP~c~) is defined as the ratio of the marketable yield produced and the water used by the crop (rainfall plus irrigation water) during the same growing season. Equation for calculation of WP~c~ is given by Fernandez et al.^21^ as 4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {\mathrm{W}}{{\mathrm{P}}_{\mathrm{c}}}={\mathrm{Y}}/\left( {{\mathrm{I}}\,+\,{\mathrm{P}}} \right) $$\end{document} in which, WP~c~ is the crop water productivity (kg m^3^), Y is the marketable yield produced during the growing season (kg ha^− 1^), I is the total irrigation water applied during the same growing season (m^3^ ha^− 1^) and P is the total rainfall during the growing season (m^3^ ha^− 1^). The Irrigation water productivity (WP~i~) is obtained by dividing the marketable yield produced by the total irrigation water applied during the same growing season. This term was determined by the following equation as presented by Fernandez et al.^21^:5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {\mathrm{W}}{{\mathrm{P}}_{\mathrm{i}}}=\frac{{\mathrm{Y}}}{{\mathrm{I}}} $$\end{document} in which, WP~i~ is the irrigation water productivity (kg m^3^), Y is the marketable yield produced during the growing season (kg ha^− 1^), and I is the total irrigation water applied during the same growing season (m^3^ ha^− 1^). The economic irrigation water productivity (EWP~i~) is calculated by dividing the net benefit generated from the yield production by the irrigation water applied during the same period^21^. This ratio helps to measure the efficiency of water usage in agriculture and can be a valuable indicator for optimizing irrigation practices. To obtain this term, one must first to determine the fixed and variable cost, and then the net benefit of the marketable yield should be determined. The fixed costs are reported by Dastranj and Sepaskhah^2^. The saffron yield for each growing season is harvested at the beginning of the next growing season. Therefore, the income from the saffron yield after corm planting in the 2015 was deducted from the fixed costs of planting, and the remaining costs were taken as fixed costs^2^, which should be amortized economically in six growing seasons. The fixed cost included the initial land preparation, corm price, and corm planting. The total fixed cost for each growing season was calculated based on the initial value reported by Dastranj and Sepaskhah^2^ and Eq. (6). Variable costs for each growing season are given in Table S4. Each year’s furrow pretreatment costs were included in the variable costs for the in-furrow planting method. In the basin planting, the cost of breaking the surface crust of the soil was included in the variable costs. Regarding the harvested corms in the last growing season, the income from its sale and the harvest costs were used in the economic analysis (Table S4). The annual fixed cost is amortized for six years, calculated by the following 6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \:\mathrm{A}\hspace{0.17em}=\hspace{0.17em}\mathrm{P}\times\:\frac{i{(1+i)}^{n}}{{(1+i)}^{n}-1} $$\end{document} in which, A is the uniform annual payment, P is the present value of cost, i is the interest rate in year 2015, and n is the number of growing seasons. Total fixed costs were communicated for the next growing seasons by Eq. (6). The values of i and n in Eq. (6) for this study are 18% and 6, respectively. Furthermore, the value of initial land preparation, corm price, and corm planting were 0.28, 10.0, and 4.0 MRls ha^− 1^, respectively. Therefore, the total fixed cost for each growing season was 14.28 MRls ha^− 1^ (Table S4). The benefit of production for each treatment and growing season was considered by the following 7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {\mathrm{B}}={\mathrm{P}} \times {\mathrm{Y}} $$\end{document} in which, B is the benefit of yield produced (Rls ha^− 1^), P is the saffron price (Rls kg^− 1^), and Y is the saffron yield (kg ha^− 1^). Net benefit (NB, Rls ha^− 1^) is the difference between the benefit (B) and the sum of fixed cost and variable cost (C, Rls ha^− 1^) as 8\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {\mathrm{NB}}={\mathrm{B}} - {\mathrm{C}} $$\end{document} To determine the economic feasibility of a project, the calculation of the benefit-cost ratio (BCR) is an efficiency parameter. This term is defined as the ratio of benefit from the yield to the total cost (fix plus variable cost, Rls ha^− 1^) for each growing season, by the following 9\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \:\mathrm{B}\mathrm{C}\mathrm{R}=\frac{B}{C} $$\end{document} Finally, economic irrigation water productivity (EWP~i~) is obtained by dividing the net benefit from the yield (NB) to the applied irrigation water (W~i~) for each growing season, as 10\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {\mathrm{EW}}{{\mathrm{P}}_{\mathrm{i}}}=\frac{{NB}}{{{W_i}}} $$\end{document} in which, EWP~i~ is the economic productivity of irrigation water (Rls m^− 3^) and W~i~ is the irrigation water applied (m^3^ ha^− 1^). ### Statistical analysis The Bartlett test was performed and it showed that the data are homogenous. SAS software was used to investigate the interaction effects of salinity level, irrigation water level, and planting method each year. Analysis of variance (ANOVA) was performed using the LSD method to determine statistical significant differences between the means by *p* ≤ 0.05. The ANOVA is shown as supplemental Table S5. ## Results and discussion ### Irrigation water use Number of irrigation events was 6, 4, 4 and 5 times in the third to sixth year, respectively. The total applied irrigation water depths in complete irrigation treatment (100%WR) in the in-furrow planting compared to that in the basin planting were 14%, 5%, 5%, and 14% higher than that in the third to sixth growing seasons, respectively (Table 1). According to the data from Dastranj and Sepaskhah^2^, the growth rate in the in-furrow planting was higher than that in the basin planting; therefore, its irrigation water requirement was also higher. Table 1Saffron water consumption for different treatments in the third to sixth growing seasons (mm).Growing seasonIrrigation water levelBasin plantingIn-furrow plantingSeasonal rainfall (mm)Salinity level (dS m^− 1^)0.421230.421233rdI1 (100%WR)537.8535.1530.4525.7611.4608.3602.9597.6167.9I2 (75%WR)350.7401.3397.8394.3398.7456.2452.2448.2I3 (50%WR)233.8267.5265.2262.8265.8304.1301.5298.84thI1 (100%WR)497.6475.3470.9466.5505.3502.7498.3493.9389.5I2 (75%WR)373.2356.5353.2349.8379.0377.1373.7370.4I3 (50%WR)248.8237.6235.4233.2252.6251.4249.2247.05thI1 (100%WR)456.5436.0432.0427.9463.6461.2457.2453.1426.5I2 (75%WR)342.4327.0324.0321.0347.7345.9342.9339.8I3 (50%WR)228.2218.0216.0214.0231.8230.6228.6226.66thI1 (100%WR)475.9473.5469.4465.2541.0538.3533.6528.9187.0I2 (75%WR)357.0355.1352.0348.9405.8403.7400.2396.6I3 (50%WR)238.0236.8234.7232.6270.5269.2266.8264.4 Precipitation depth in the sixth growing season compared to the fifth growing season, decreased by 64% (Fig. S1, Table 1). Therefore, in the sixth growing season, the seasonal water application was increased due to the decrease in rainfall compared to the previous two growing seasons, and most of the crop water requirement was met by irrigation water application (Table 1). Saffron irrigation water decreased by increasing irrigation water salinity. Increasing water salinity reduced the osmotic potential of soil solutions and reduced plant water absorption^22,23^. ### Saffron yield and yield components Saffron yield (SY) measured in the third, fourth, fifth, and sixth growing seasons. Furthermore, leaf dry matter, corm yield, harvest index, and soil-saturated extract salinity (30–60 cm soil depth) measured only in the fifth and sixth growing seasons. Results of the statistical analysis (AVOVA) presented in Table S5. The growing season factor showed a significant effect on all the yield components. Therefore, the main and interaction effects of the treatments were investigated separately for each growing season. #### Saffron yield Results of the statistical analysis for saffron yield (SY) showed that in the third to sixth growing seasons, due to the difference in the depth of rainfall in different growing seasons, the interaction effect of different treatments on the SY was significant (Table S5). Average of the SY in the third to sixth growing seasons was 9.0, 12.4, 9.8, and 0.9 (kg ha^− 1^), respectively (Table 2). On average, the saffron yield in all treatments in the fourth growing season was 37% higher than that in the third growing season. However, in the fifth growing season it was 20% lower than that in the fourth growing season, and in the sixth growing season it was 91% lower than that in the fifth growing season (Table 2). It showed that SY dropped dramatically in the sixth growing season, so it is not economically feasible. Table 2Mean values of saffron yield for four growing seasons (3rd to 6th ), and leaf dry matter, corm yield, and harvest index in different salinity levels, irrigation water levels and planting methods for two growing seasons (5th and 6th ).Measured parameterSaffron yieldLeaf dry matterCorm yieldHarvest index(kg ha^− 1^)(Mg ha^− 1^)(Mg ha^− 1^)%Growing season3rd4th5th6th5th6th5th6th5th6thSalinity level (dS m^− 1^)S1 (0.42 dS m^− 1^)10.63a^*^11.93b10.13a0.89 A2.35ab1.54 A62.37b25.38 A0.32a0.05BS2 (1.0 dS m^− 1^)10.07a13.42a10.27a0.84 A2.50a1.51 A64.55a23.10 A0.29a0.05BS3 (2.0 dS m^− 1^)8.4b11.03b9.37a0.92 A2.40ab1.37 A65.75a22.73 A0.26a0.06 AS4 (3.0 dS m^− 1^)6.89c13.07a9.41a0.81 A2.27b1.44 A54.28c22.81 A0.32a0.05BIrrigation water levelI~1~ (100%WR)10.95a12.88a9.12a1.05 A2.21b1.53 A62.96a23.98 A0.27a0.07 AI~2~ (75%WR)9.18b12.78a10.03a0.84AB2.46a1.38 A60.14a24.50 A0.29a0.05 AI~3~ (50%WR)6.87c11.43b10.24a0.7lB2.47a1.49 A62.11a22.05 A0.33a0.04BPlanting methodP~1~ (Basin)5.52b11.48b8.89a0.92 A2.49a1.54 A61.41a23.13 A0.27a0.06 AP~2~ (In-furrow)12.48a13.25a10.70a0.80B2.27b1.39 A62.07a23.89 A0.33a0.05 A*Means with the same letters in each year are not significantly different (*p* < 0.05). Main effect of salinity level, irrigation water level, and the planting method (Table 2) in the third and the fourth growing seasons, the interaction effect between salinity and irrigation water level in fourth growing season (Table 3), the interaction effect between salinity level and planting method in the fourth and sixth growing seasons (Table 4), and the interaction effect between irrigation water level and the planting method in the third and sixth growing seasons showed a considerable effect on SY (Table 5). In the fifth growing season, none of the main and interaction effects of treatments were statistically significant for SY. However, the SY in the in-furrow planting was 126%, 15%, and 20% higher than that in the basin planting in the third to fifth growing season, respectively (Table 2). Table 3Interaction effect of salinity levels and irrigation water levels on saffron yield (kg ha^− 1^), leaf dry matter, and corm yield (Mg ha^− 1^)*.Growing seasonIrrigation water levelSalinity level (dS m^− 1^)0.42123Saffron yield (kg ha^− 1^)4thI1 (100%WR)11.74 cd^**^14.21a11.84 cd13.74abI2 (75%WR)13.85ab13.58ab10.80d12.88abcI3 (50%WR)10.21d12.48bc10.45d12.59abcLeaf dry matter (Mg ha^− 1^)5thI1 (100%WR)2.01DE2.56ABC2.29B-E1.97EI2 (75%WR)2.56ABC2.82 A2.17CDE2.29B-EI3 (50%WR)2.46 A-D2.13CDE2.75AB2.55ABC6thI1 (100%WR)1.42ab1.65a1.50.8ab1.53abI2 (75%WR)1.61ab1.41ab1.23b1.28abI3 (50%WR)1.60ab1.46ab1.39ab1.50abCorm yield (Mg ha^− 1^)5thI1 (100%WR)75.65 A61.38CD68.03BC46.75EI2 (75%WR)65.01BCD63.27BCD60.35D51.95EI3 (50%WR)46.44E69.02AB68.87AB64.13BCD*This interaction effect on the saffron yield in other years, corm yield in 6th and harvest index in 5th and 6th growing seasons was not significant. **Means with the same letters in each year are not significantly different (*p *< 0.05). Table 4Interaction effect of salinity levels and planting methods on saffron yield, leaf dry matter, corm yield, and harvest index*.Growing seasonPlanting methodSalinity level (dS m^− 1^)0.42123Saffron yield (kg ha^− 1^)4thBasin11.44c^**^12.69bc9.18d12.59bcIn-furrow12.42bc14.16a12.88b13.55ab6thBasin0.87B1.08 A0.92AB0.83BIn-furrow0.90AB0.60 C0.92AB0.79BCLeaf dry matter (Mg ha^− 1^)5thBasin2.30b2.42ab2.50ab2.76aIn-furrow2.39ab2.58ab2.32b1.78cCorm yield (Mg ha^− 1^)5thBasin62.80abc61.49bc70.12a51.22dIn-furrow61.93abc67.62ab61.39bc57.34 cd6thBasin24.48ABC24.46ABC24.89AB18.68DIn-furrow26.30 A21.75BCD20.58CD26.95 AHarvest index (%)6thBasin0.049b0.069a0.057ab0.051bIn-furrow0.053b0.030c0.064ab0.053b*This interaction effect on the saffron yield in 3rd and 5th, leaf dry mater in 6th and harvest index in 5th growing season was not significant. **Means with the same letters in each year are not significantly different (*p* < 0.05). Table 5Interaction effect of irrigation water levels and planting methods on saffron yield and leaf dry matter*.Growing seasonPlanting methodIrrigation levelsI1 (100%WR)I2 (75%WR)I3 (50%WR)Saffron yield (kg ha^− 1^)3rdBasin7.19 d^**^5.28 e4.09fIn-furrow14.7 a13.09 b9.64 c6thBasin1.02 A1.00 A0.74BIn-furrow1.07 A0.68B0.66BLeaf dry matter (Mg ha^− 1^)5thBasin2.28bc2.457bc2.72aIn-furrow2.13c2.45abc2.22bc6thBasin1.58 A1.57 A1.48 AIn-furrow1.47 A1.20B1.50 A*This interaction effect on the saffron yield in 4th and 5th, corm yield in 5th and 6th, and harvest index in 5th and 6th growing season was not significant. **Means with the same letters in each year are not significantly different (*p* < 0.05). High depth of rainfall in two consecutive growing season (4th and 5th in Table 5) reduced the effect of the irrigation water salinity and caused leaching of the salt in the root zone to some extent; therefore, the significant difference in the SY was shifted from low salinity levels to that in the higher salinity levels with less irrigation water levels (Table 2). ##### Interaction effects of different treatments on saffron yield *Third growing season* Results in the third growing season indicated the maximum yield was obtained in the 1.0 dS m^− 1^ salinity (20% higher), 100%WR (59% higher), and in the in-furrow planting (126% higher) that were the optimal treatments compared to the other treatments (Table 2). Results of the interaction between irrigation water levels and planting method (Table 5) showed that in this growing season a decrease in irrigation water depth caused a decrease of 34% and 43% in SY in the in-furrow plantings and basin plantings, respectively (Table 5). In the same planting, the SY decreased by increasing the salinity level at different irrigation water levels or by reducing the irrigation water level at a given irrigation water salinity (Table 6). Table 6Interaction effect of salinity levels, irrigation water levels and planting methods on saffron yield and leaf dry matter.*.Growing seasonIrrigation levelSalinity level (dS m^− 1^)0.421230.42123Basin plantingIn-farrow plantingSaffron yield (kg ha^− 1^)3rdI1 (100%WR)8.86f^**^8.57 fg6.40hij4.94jkl15.44ab16.11a14.37bc12.89 cdI2 (75%WR)7.29gh5.57ilk4.73kl3.52 lm14.69ab14.78ab12.51d10.37eI3 (50%WR)4.91kl4.58kl4.04 lm2.83 m12.61d10.81e8.38 fg6.78hiLeaf dry matter (Mg ha^− 1^)5thI1 (100%WR)2.15d-i1.94f-j2.57b- f2.48c-h1.88hij3.17ab2.01e-j1.47jI2 (75%WR)2.52c-g2.93abc2.15d-i2.27d-h2.61b-e2.70bcd2.20d-i2.31c-hI3 (50%WR)2.21d-h2.37c-h2.77bcd3.53a2.70bcd1.89 g-j2.75bcd1.56ij*This interaction effect on other yield components was not significant. **Means with the same letters in each year are not significantly different (*p* < 0.05). *Fourth growing season* Results in the fourth growing season indicated the maximum yield was obtained in the highest level of salinity (10% higher), 75%WR (same value as 100%WR), and the in-furrow planting (13% higher) that were the optimal treatments compared to other treatments (Table 2). In this growing season, results of the interaction of salinity and irrigation water levels also showed an insignificant effect of the irrigation water levels in 1.0, 2.0 and 3.0 dS m^− 1^ salinity level, but a significant effect was obtained at lower salinity (freshwater, S~1~) and 100% WR (Table 3). In salinity levels of 1.0 and 2.0 dS m^− 1^, lower SY was obtained in the basin planting compared to that in the in-furrow planting as10% and 29%, respectively, (Table 4). In this growing season, in the same salinity level, the irrigation water level did not show a significant effect on SY; however, only in the lower salinity (freshwater), the 100%WR reduced SY by 15% compared to 75%WR, and 50%WR decreased SY by 26% compared to 75%WR level (Table 3). Application of 50%WR at a salinity level of 3.0 dS m^− 1^ (i.e., I~3~S~4~) compared to 50%WR at 0.42 dS m^− 1^ (i.e., I~3~S~1~) increased SY by 23% (Table 3) due to high rainfall in this growing season that caused salt leaching from the root zone (Table 3). Maximum depth of saffron root at the end of the vegetative growth stage is 45 cm, and its planting depth is 15–20 cm; therefore, high winter rainfall is more effective in leaching salt from the root zone. *Fifth growing season* In the fifth growing season, the treatments showed non-significant effect on the SY due to high rainfall depth (Fig. S1); however, the declining trend of SY from 5th growing season onwards, despite the high density of the corms at the end of this season, indicated the lack of proper growth of the corm size to reach the flowering weight (Fig. 1). Fig. 1**a** the effect of salinity levels (S~1~ = 0.45 dS m^-1^, S~2~ = 1.0 dS m^-1^, S~3~ = 2.0 dS m^-1^, and S~4~ = 3.0 dS m^-1^), **b** irrigation levels (I~1~ = 100%WR, I2 = 75% WR, and I~3~ = 50%WR), and planting methods (P~1~ = basin planting and P~2~ = in-furrow planting) in variation of saffron yield during six growing seasons. *Sixth growing season* Results of the statistical analysis in 6th year are shown in Table S5. Salinity levels did not show significant effect on SY because rainfall reduced soil salinity during past growing season, thus high salinity level (i.e., S~4~) was optimal treatment (Table 2). The 75% WR irrigation water level showed no significant difference with the other irrigation water levels, and the basin planting compared to in-furrow planting resulted in 15% higher SY (Table 2). In this growing season, the optimal irrigation water level was 75%WR for the basin planting, and 100%WR for the in-furrow planting (Table 5); however, the density of corm in the fifth and sixth growing seasons were the same in both plantings (Table 2). The irrigation water application in the in-furrow planting was higher than that in the basin planting (Table 1), and the SY in the in-furrow planting was less than that in the basin planting (Table 2). This result indicated that corm size reduction in the in-furrow planting was higher than the basin planting in 6th growing season. High rainfall in two consecutive growing seasons (4th and 5th) gradually reduced the effect of the irrigation water salinity by leaching salt into the root zone. Therefore, it is indicated that the significant difference in the SY has shifted from low salinity levels to higher salinity levels with less irrigation water application. ##### Long-term saffron yield variation Study of long-term saffron yield during the six growing seasons (2015–2020) would greatly help to better understand the variation of changes due to the main and interaction effects of the treatments. The data provided by Dastranj and Sepaskhah^2^ included in this analysis for the 2015–2017 growing seasons. The mean SY of all treatments at different growing seasons is shown in Fig. 1. To better understand variation of the main and interaction effects of various values of yield components, it is essential to know in which growing season the maximum SY occurred for each treatment. In low salinity levels (freshwater, 0.42 dS m^− 1^), the maximum SY was produced almost during the 2nd to 5th growing seasons. However, at other salinity levels (1.0, 2.0, and 3.0 dS m^− 1^) maximum SY was produced in the fourth growing season (Fig. 1a), that may be due to heavy rainfall in the 4rd growing season by reducing the effect of salinity and deficit water on SY. In the basin planting, all the treatments have reached the maximum SY in the fourth growing season. However, the highest SY was achieved in the in-furrow plantings almost in the second growing season. SY in the in-furrow planting in the first growing season to the fifth growing season was 105%, 121%, 126%, 15% and 20% higher than that in the basin planting. However, in the sixth growing season, the SY was 13% higher in the basin planting compared to that obtained in the in-furrow planting (Fig. 1b). These results show that the in-furrow planting method is more favorable than the basin planting in the use of saline water. Also, results showed that the decreasing trend of SY in all treatments intensified from the fifth growing season onwards; however, SY in the fifth growing season decreased by 20% compared to that obtained in the fourth growing season, and this SY decrease was drastic as 91% in the sixth growing season compared to that in the fifth growing season. Highest SY of each irrigation water level occurred in the fourth growing season. Highest SY was produced with the highest irrigation water level; therefore, as the irrigation water level decreased, SY also decreased in different growing seasons. However, in the fourth and fifth growing seasons, high rainfall depth moderated some of the effects of water stress. Therefore, the 75%WR is the optimal irrigation water level (Fig. 1c). Vegetative growth of saffron is influenced by soil-plant-water relations; therefore, the very sharp decrease of SY in the sixth growing season compared to the fifth growing season may also be due to the reduction of soil nutrients and the reduction of sufficient space in the soil for corm growth. Great fluctuations in the meteorological parameters (minimum and average air temperature) during the sixth growing season showed a sharp decrease in SY (Fig. S2), which was due to decline in minimum and average air temperature in the period of 0–33 days (Fig. S2a, and b). Low temperature during flowering is also the main factor to determine leaf growth rate, flower initiation, flower appearance, and length of the flowering period^24^. Therefore, minimum air temperature during the 62 to 73 days after the first irrigation was low (Fig. S2a). This low minimum air temperature is an important factor that might have resulted in SY reduction. This occurred due to the frost being the main factor in growing season resulting in small corms size^25^. Therefore, SY of the sixth growing season reduced drastically compared to the fifth growing season. These problems in the sixth growing season caused a 91% decrease in SY, 62% of corm yield, and 38% of leaf dry matter compared to that in the fifth growing season (Table 2). *Cumulative saffron yield* Cumulative saffron yield (CSY) during the 6 growing seasons for each treatment is presented in Fig. 2. Statistical analysis of the CSY obtained during six growing seasons for each treatment revealed that the main effect of salinity level, irrigation water level and the planting method on CSY was significant. It is indicated that the interaction effect of the irrigation water level and the planting method and the interaction effect of all three treatments were simultaneously substantial. Fig. 2The cumulative saffron yield (kg ha^-1^) in different treatments, **a** the salinity levels (S~1~ = 0.45 dS m^-1^, S~2~ = 1.0 dS m^-1^, S~3~ = 2.0 dS m^-1^, and S~4~ = 3.0 dS m^-1^), **b** irrigation water levels (I~1~ = 100%WR, I2 = 75% WR, and I~3~ = 50%WR), and **c** planting methods (P~1~ = basin planting method and P~2~ = in-furrow planting method) during six growing seasons. Trend of changes in the SY for each level of treatments except for planting method was similar during six growing seasons (Fig. 2). Salinity levels of 0.42 and 1.0 dS m^− 1^ as well as 2.0 and 3.0 dS m^− 1^ did not show significant differences in SY between each other (Fig. 2a). CSY was 2%, 19% and 25% higher at lower salinity compared to different salinities during the six growing seasons (Fig. 2a). However, the increase salinity level from 0.42 to 3.0 dS m^− 1^ resulted in 16% decrease in CSY during the six growing seasons (Fig. 2a). Reducing the irrigation water level from 100%WR to 75%WR and 50%WR also resulted in 13% and 20% decrease in CSY (Fig. 2b), respectively. In the sixth growing season, the CSY obtained from in-furrow planting was 59% higher than that obtained from basin planting (Fig. 2c). However, these differences were greater in the 3rd and 4th growing seasons, as 118% and 74%, respectively. Saffron is a perennial crop with production period of 6–8 years. Our field study period was 8 years, including the results of first two years that reported by Dastranj and Sepaskhah^2^. The saffron growth and yield usually increase, and reach to a maximum at 3rd to 4th year. Then, they descend till reaching to 8th year. As it is noticed, the results of first and second year were published before reaching the maximum, therefore our results showed that at first, the yield increased, and declined afterward. It is clear that the environmental condition and agronomic management, i.e., rainfall, irrigation water and salinity variables can affect the growth and yield pattern during the long-term growing period. Consideration of the effects of these variables on the pattern of saffron growth and yield is one of the objectives of this research. To the knowledge of authors, such a study by these objectives has not been recoded and analyzed in literature. Therefore, this objective shows a novelty of present study. ##### Saffron production function Effects of growing season, irrigation water salinity, irrigation water level and planting method on the SY were significant. In addition, there were differences in precipitation depth across different growing seasons. Therefore, the relationship between SY and growing seasons, irrigation water salinity and the seasonal sum of irrigation water and precipitation during the six growing seasons were analyzed separately in the basin [Eq. (11)] and the in-furrow [Eq. (12)] planting methods. Basin:11\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {\mathrm{SY}}= - {\mathrm{9}}.{\mathrm{828}}+{\mathrm{4}}.{\mathrm{553y}} - 0.{\mathrm{7}}{{\mathrm{y}}^{\mathrm{2}}} - 0.{\mathrm{596s}}+0.0{\text{22 }}{{\mathrm{w}}_{\mathrm{i}}} - {\mathrm{4}}.{\mathrm{7}} \times {\mathrm{1}}{0^{ - \,{\mathrm{6}}}}{{\mathrm{w}}_{\mathrm{i}}}^{{\mathrm{2}}} $$\end{document} R^2^ = 75, *n* = 72, SE = 1.9, *p* value < 0.0001. In-furrow:12\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {\mathrm{SY}}= - {\mathrm{15}}.{\mathrm{41}}+{\mathrm{7}}.{\mathrm{728y}} - 0.{\mathrm{1}}.{\mathrm{29}}{{\mathrm{y}}^{\mathrm{2}}} - 0.{\mathrm{81s}}+0.0{\mathrm{51}}{{\mathrm{w}}_{\mathrm{i}}} - {\text{ 3}}.{\mathrm{3}} \times {\mathrm{1}}{0^{ - \,{\mathrm{5}}}}{{\mathrm{w}}_{\mathrm{i}}}^{{\mathrm{2}}} $$\end{document} R^2^ = 89, *n* = 72, SE = 1.6, *p* value < 0.0001. in which, SY is the saffron yield (kg ha^− 1^), y is the number growing season, s is the water salinity (dS m^− 1^), w~i~ is the sum of the irrigation water depth and rainfall (mm). All coefficients in Eqs. (11) and (12) were significant at *p *value < 5% level. Minimum of the seasonal irrigation water and precipitation sum (w~i−min~) required to start the SY production in different growing seasons (1st to 6th ) according to the irrigation water salinity levels (0.42, 1.0, 2.0, and 3.0 dS m^− 1^) is presented in Table 7. The w~i−min~ increased with increasing the salinity level, in each growing season. In high salinity levels, the water requirement for leaching salt from the root zone was increased as reported by Goet et al.^23^. ; therefore, the obtained value of w~i−min~ was high. Table 7The minimum seasonal sum of the irrigation water and precipitation (mm) required to start the flowering of saffron crop for each year.Planting methodSalinity level (dS m^− 1^)Year1st2nd3rd4th5th6thIn-furrow plantingS1 (0.42 dS m^− 1^)107.672.7259.2172.97108.0154.9S2 (1.0 dS m^− 1^)111.477.4064.3277.63111.8158.0S3 (2.0 dS m^− 1^)117.885.1472.7485.36118.2163.1S4 (3.0 dS m^− 1^)124.092.5380.7192.74124.4168.2Basin plantingS1 (0.42 dS m^− 1^)198.6133.4101.49112.3162.4239.3S2 (1.0 dS m^− 1^)206.9143.1112.17122.7171.4248.0S3 (2.0 dS m^− 1^)220.9159.4129.85139.8186.6259.8S4 (3.0 dS m^− 1^)234.5175.1146.70156.3201.3272.4 In higher water salinity, the soil osmotic potential is lower; therefore, higher w~i−min~ is required to initiate SY production. Average of w~i−min~ in the basin planting was higher than that in the in-furrow planting during 1st to 6th growing season, as 87%, 86%, 77%, 62%, 56% and 58%, respectively. These findings occurred since applied irrigation water or rainfall is distributed in a lower surface area in the in-furrow planting than that in the basin planting. Also, evaporation from the soil surface is lower in the in-furrow planting than that in the basin planting^3,26^ resulting in lower w~i−min~ for in-furrow planting. In different salinity levels, during 1st to 3rd growing season, the w~i−min~ decreased due to the corm growth and establishment in the first to third growing season; therefore, its size increased^27^. Also, according to planting depth (15–20 cm), an increase in corm size resulted in high flowering capacity^27,28^ and a decrease in w~i−min~. On the other hand, due to the increase of the corm growth during the 4th to 6th growing season, the sufficient space for growth of the corm decreased, which led to a smaller size of the corm and a decrease in soil nutrients, and increased competition between corm, which resulted in a higher w~i−min~ requirement for flower production. Furthermore, the presented saffron production function [Eqs. (11) and (12)] are empirical models that include the year of growing period, water salinity, sum of irrigation water and rainfall depth. These models and other modes are essential for economic study in insurance analysis based on yield loss due to unfavorable production parameters^29^. This shows another novelty of the presented study. #### Leaf dry matter Results of the statistical analysis of leaf dry matter (LDM) in both growing seasons (5th and 6th ) are presented in Table S5. Mean LDM in all treatments in the sixth growing season decreased by 38.4% compared to that in the fifth growing season (Table 2). Results of both growing seasons indicated that different levels of irrigation water salinity showed no significant effect on the LDM (Table 2). Heavy rainfall in the fourth and fifth growing seasons reduced the effect of irrigation water salinity on the soil salinity. Therefore, root water absorption and then vegetative growth did not show significant differences at different water salinities. Since the growing season of saffron crops coincides with autumn and winter with rather high relative humidity of the air, the negative effect of salt stress on the LDM was moderate^30^. In the fifth growing season, the optimal irrigation water level was 50%WR with 10.5% higher LDM compared to 100%WR due to the abundant rainfall in this growing season. Therefore, in the fifth growing season, the higher irrigation water level adversely affected on LDM. In the sixth growing season, irrigation water levels did not significantly affect LDM (Table 2). It should be noted that there is a reciprocal relationship between the leaf (aerial part) and the corm weight, including corm number and size^31,32^. This has explained our finding. Although the corm yield showed no significant difference in the fifth growing season (Table 2), LDM increased by 10% in basin compared to that in in-furrow planting (Table 2). The crust breaking in the basin planting could have facilitated the exit of the leaf conductor buds, and as a result, the LDM in this planting was higher than that in the in-furrow planting. However, in the sixth growing season, it reduced the positive effect of the initial crust breaking in the basin planting due to rainfall occurrence during flowering stage. In the fifth growing season, the application of high level of salinity and irrigation water (i.e., I~1~S~4~) resulted in a 20% decrease in LDM compared to the lower salinity level and irrigation water level (i.e., I~3~S~1~), (Table 3). As the high rainfall reduce the salts accumulation in this growing season, increase in the irrigation water level has caused an increase in the soil water content in the root zone, and it disrupts water uptake and crop respiration^33^. In contradiction, the higher salinity and lower irrigation levels (i.e., I~3~S~4~) increased LDM by 26% compared to lower salinity level and higher irrigation water levels (i.e., I~1~S~1~, Table 3) due to low salt accumulation in low irrigation water level. Significant effect of salinity along with the level of irrigation water (reduction of water absorption and reduction of soil water content) in different planting methods (rate of spreading and penetration of water in the soil profile) showed that according to the depth of rainfall in both (5th and 6th ) growing seasons the LDM was sensitive to full irrigation (100%WR). Increasing soil water salinity reduces the osmotic potential; therefore, the water uptake by roots is reduced, leading to the reduction of salt accumulation in the soil^34^. In the fifth growing season, the higher salinity level (3.0 dS m^− 1^, Table 4) and the lower irrigation water level (50%WR, Table 5) significantly affected the LDM in the planting method; therefore, the 3.0 dS m^− 1^ salinity level resulted in 35% increase, and the application of 50%WR, resulted in 18% increase in the LDM in the basin planting compared to that in the in-furrow planting. Due to the positive effect of rainfall in salt leaching from the root zone, other salinity levels showed no significant effect on the LDM (Table 4). In the sixth growing season, 75%WR resulted in a LDM increase in the basin planting as 31% higher than in the in-furrow planting (Table 5). According to the results of Table 6, in the in-furrow planting, 75%WR is the optimal irrigation water level for all salinity levels. However, in the basin planting, the optimal irrigation water level was 50%WR. In the basin planting, LDM increased as salinity increased or irrigation water level decreased. Therefore, in the 50%WR, by increasing salinity from 0.42 dS m^− 1^ to 3.0 dS m^− 1^, the LDM increased 60%. However, these treatments in the in-furrow planting resulted in a 42% decrease in LDM (Table 6). Again, high rainfall moderated the salinity effect on the LDM production. These results also indicated that the complete irrigation and high soil water content in the root zone resulted in adverse effect on leaf growth because of restricted root respiration^35^. #### Corm yield During the growing season, the corms are continuously propagating^36,37^. Results of analysis of variance of the effect of different treatments on the mean corm yield (CY) are given in Table S5. In the fifth growing season, mean corm yield in all treatments was 61.7 Mg ha^− 1^. In the sixth growing season, it decreased by 62% (Table 2). *Fifth growing season* Results indicated the main effect of salinity level on CY was significant only in the fifth growing season. Therefore, the 2.0 dS m^− 1^ was the optimal salinity level compared to the 0.42 dS m^− 1^ and 3.0 dS m^− 1^ salinity level, in which the CY was increased by 5% and 21%, respectively (Table 2). Main effects of irrigation water levels and planting methods, in both growing seasons, showed no significant effect on CY (Table 2) due to the highest rainfall in the fifth growing season (Table 1). Main effect of irrigation water level and the planting method did not significantly affect CY in the fifth growing season because the corm density was high and the rainfall depth was suitable. In the fifth growing season, increase in the salinity level at the same level of irrigation water or decrease in the irrigation water level at the same salinity level in medium salinity treatments (1.0 dS m^− 1^ and 2.0 dS m^− 1^) did not cause significant differences in the CY (Table 3). However, in the 0.42 dS m^− 1^ salinity level, the 100%WR compared to the 75%WR and 50%WR showed the CY increased by 16% and 63%, respectively (Table 3). In 3.0 dS m^− 1^ salinity level, the 100%WR and 75%WR did not show significant differences in the corm yield. However, the production of corm in the 50%WR increased by 23% compared to other levels of irrigation water (Table 3). The obtained CY in the higher salinity level with the higher irrigation water level (i.e., I~1~S~4~) compared to the lower salinity level with the lower irrigation water level (i.e., I~3~S~1~) was not statistically different. However, CY obtained at lower salinity level with higher irrigation water level (i.e., I~1~S~1~) was 18% higher than the higher salinity at lower irrigation level (i.e., I~3~S~4~, Table 3). Results showed that in the lower salinity level (0.42 dS m^− 1^), the irrigation water level significantly affected the CY. However, in other salinity levels, according to the depth of rainfall and corm density, the 50%WR was the optimal level of irrigation water in the fifth growing season (Table 3). In the interaction effect of irrigation water salinity and the planting method, mean corm yield reduction by increasing salinities in the basin and in-furrow planting in the fifth growing season were 18% and 8%, respectively (Table 4). In salinity levels of 0.42, 1.0, and 3.0 dS m^− 1^, CY was not significantly different. However, the 2.0 dS m^− 1^ in the basin planting compared to that in the in-furrow planting increased CY by 14% (Table 4). *Sixth growing season* In the sixth growing season, only the interaction effect of salinity levels and planting method on the CY was significant (Table S5). Results showed that the mean CY obtained in the basin planting was the same at all salinity levels except in the higher salinity level, and that level caused a 24% reduction in CY compared to the other salinity levels (Table 4). In the in-furrow planting, lower and higher salinity level resulted in the same CY values, and CY was the same at 1.0 and 2.0 dS m^− 1^ salinity levels (Table 4). Effect of salinity level in the basin planting was similar to the fifth growing season. Therefore, mean CY reduction by increasing salinities in the basin planting was 24% (Table 4). However, 0.42 and 1.0 dS m^− 1^ salinity levels did not show a significant difference in the CY in the planting methods (Table 4). In 2.0 dS m^− 1^ salinity level, the CY was 21% higher in the basin planting, and in 3.0 dS m^− 1^ salinity level in the in-furrow planting it was 44% higher (Table 4). In the sixth growing season, although the rainfall decreased, the reduction in density of the CY was not significant (Table 2) due to the low temperature of soil as a result of frost during the growing season that affected the corm growth and its propagation (Fig. S2)^38,39^. These results showed that the in-furrow planting in different salinities levels compared to the basin planting was well performed even in the sixth growing season when the rainfall depth was lower and the irrigation water level with saline water increased. Dastranj and Sepaskhah^2^ results in the first and second growing seasons showed that increasing salinity level or decreasing the irrigation water level (deficit irrigation) had a significant effect on reducing CY. They also reported that the in-furrow planting increased the CY compared to that in the basin planting. #### Harvest index Harvest index (HI) is the ratio of the saffron yield to the sum of flower and leaf dry matter. Results of statistical analysis for all treatments are given in Table 6. In the fifth growing season, effect of the treatments on the HI was insignificant (Table S5). Mean HI for all treatments in the sixth growing season decreased by 82% compared to that in the fifth growing season (Table 2). In this year, main effects of salinity level, the irrigation water level (Table 2), and the interaction effect of irrigation water salinity and the planting method on HI were significant (Table 4). In the sixth growing season, highest value of HI was obtained in 2.0 dS m^− 1^ of salinity level (Table 2). Other salinity levels with 16% decrease in HI compared to 2.0 dS m^− 1^ level did not cause a significant difference in the HI. Optimal irrigation water level was 75%WR, which its effect is statistically the same as the HI in 100% WR and led to a 25% increase in the HI compared to that in 50%WR (Table 2). In the sixth growing season, 2.0 dS m^− 1^ salinity level in the both planting methods was optimal (Table 4). This salinity level did not show a significant difference in the HI compared to other salinity levels in the basin planting; however, in the in-furrow planting, this salinity level compared to that in 1.0 dS m^− 1^ salinity level caused a 2.0 times HI (Table 4). On the other hand, Dastranj and Sepaskhah^2^ reported that irrigation water salinity level during the first and second growing seasons was not influential in the value of HI. However, the in-furrow planting increased HI compared to the basin planting as 44% and 35%, respectively in the first and second growing seasons. ### Saffron irrigation and crop water productivity Saffron irrigation water and crop water productivities in different growing seasons (WP~i~, WP~c~) are shown in Table 8. WP~i~ increased by 65% in the fourth growing season compared to that in the third growing season, while it decreased by 11% and 92% in the fifth and sixth growing seasons compared to that in the previous growing season, respectively (Table 8). Results indicated the WP~i~ in the in-furrow planting compared to that in the basin planting was 102%, 6%, and 8% higher in the third to fifth growing seasons, respectively (Table 8). However, in the sixth growing season, the WP~i~ in the basin planting was 18% higher than that in the in-furrow planting (Table 8) due to the sharp reduction of SY in this growing season in the in-furrow planting. WP~i~ value in 100%WR in the fourth growing season increased by 39% compared to the third growing season. However, in the fifth and sixth growing seasons, it decreased by 23% and 90% compared to the previous growing season. (Table 8). Low value of WP~i~ in the sixth growing season indicated that the continuation of this saffron cultivation after the fifth growing season was undesirable (Table 8). Table 8Saffron irrigation and crop water productivity for different treatments in the 3rd to 6th growing seasons.Irrigation water productivity (WP~i~) g m^− 3^Growing seasonIrrigation water levelBasin plantingIn-furrow plantingSalinity level (dS m^− 1^)0.421230.421233rdI1 (100%WR)1.65i*1.60ij2.38jkl2.16 L2.53de2.65de2.38d-g2.16 fgI2 (75%WR)2.08gh1.39ijk2.77jkl2.31 L3.68b3.24c2.77d2.31 fgI3 (50%WR)2.10gh1.7hi12.78ij2.27kl4.74a3.55bc2.78d2.27efg4thI1 (100%WR)2.13j2.89hi2.07j2.79hi2.55ij2.92ghi2.80hi2.93ghiI2 (75%WR)3.61def3.78de2.80hi3.60def3.75def3.63def3.13f-i3.55d-gI3 (50%WR)4.12 cd4.57bc3.36e-h5.21ab4.02 cd5.62a5.21a5.28a5thI1 (100%WR)2.22fgh2.02hg1.55 h1.99hg2.19hg2.36e-h2.29e-h1.61 hI2 (75%WR)2.98d-g2.85e-h2.34e-h2.73e-h3.51b-f3.26c-g3.18c-g2.95d-gI3 (50%WR)3.53b-e4.49abc4.19a-d4.57ab4.33abc5.01a5.04a5.36a6thI1 (100%WR)0.18f-i0.25c-g0.25b-g0.18ghi0.21e-h0.19f-i0.22e-h0.18ghiI2 (75%WR)0.29a-e0.36a0.21e-h0.27a-g0.23d-h0.10i0.18ghi0.17ghiI3 (50%WR)0.28a-f0.32a-d0.35ab0.30a-e0.23d-h0.15hi0.34abc0.26b-gCrop water productivity (WP~c~) g m^− 3^3rdI1 (100%WR)1.10ij1.19ijk0.89lmn0.65mno1.91ef1.87de1.95efg1.74ghI2 (75%WR)1.33i1.05klm0.69 L-o0.67o2.47b2.29bc1.92def1.72ghI3 (50%WR)1.15ijk1.05jkl0.96lmn0.71no2.53a2.09 cd1.48 fg1.65hi4thI1 (100%WR)1.09IJ1.23D-G0.93 J1.54E-H1.23 F-I1.61DEF1.19D-G1.63DEFI2 (75%WR)1.75B-E1.57B-E1.19G-J1.70 C-F1.67BCD1.47B-E1.53E-H1.86 C-FI3 (50%WR)1.35D-G1.73 C-F1.37HIJ1.96ABC1.65D-G2.23 A1.96AB1.92AB5thI1 (100%WR)1.26d-h0.95e-h0.75 h1.13fgh1.41d-h1.26c-h0.77d-h0.66ghI2 (75%WR)1.48b-g1.18b-h1.32e-h1.02d-h1.37a-d0.99a-f1.29a-f1.54b-hI3 (50%WR)1.44c-h1.65a-f0.91a-f1.91a-e1.68a-f1.76ab1.76abc1.88a6thI1 (100%WR)0.13D-H0.15 A-F0.15 A-F0.21E-H0.17B-G0.11B-H0.14B-G0.13 C-HI2 (75%WR)0.20ABC0.24 A0.13 C-H0.09 A-G0.17B-G0.06I0.12 F-I0.11GHII3 (50%WR)0.13B-G0.22 A-E0.20 A-D0.15B-G0.11 C-H0.07HI0.20AB0.14B-G*Means with the same letters in each year are not significantly different (*p* < 0.05). The mean value of WP~c~ in the third to the sixth growing season in the in-furrow planting was 36% higher than that in the basin planting. In the third to the fifth growing season, the WP~c~ in the in-furrow planting was always higher than that in the basing planting, but the planting method showed no significant difference in WP~c~ values in sixth growing season (Table 8). Irrigation level in the third and sixth growing seasons had no significant different in WP~c~. However, in the fourth and fifth growing season they were significantly different, and the 50% WR level caused the higher WP~c~ by 35% and 11% in fourth growing season and 68% and 35% in fifth growing season compared to 100% WR and 75% WR irrigation level, respectively. In the long term, the WP~c~ at 50% WR level was 1.13 and 1.3 times higher than at 100% WR and 75% WR levels, respectively. The effect of salinity level on WP~c~ was not statistically significant in the fourth and fifth growing season due to high seasonal rainfall (Table 1). However, the value of WP~c~ was different between low salinity (0.42 and 1.0 dS m^− 1^) and high salinity (2.0 and 3.0 dS m^− 1^) in 3rd growing season. In the sixth growing season WP~c~ was not significantly different due to low yield. In general, the value of WP~c~ was lower in comparison to that of WP~i~ due to addition of rainfall in denominator of Eq. 4. ### Soil saturation extract salinity Analysis of the EC~e~ at a depth of 0–60 cm showed no significant effect of the treatments on the mean EC~e~. Therefore, EC~e~ was analyzed at 30–60 cm depth according to the depth of corm planting and the maximum depth of the saffron root (45 cm). Salinity of the soil saturation extract (EC~e~) in 30–60 cm depth measured for the fifth and sixth growing seasons, and their results are shown in Table 9. Table 9Mean values of electrical conductivity of soil saturation extract (EC~e~, 30–60 cm) for different treatments in the fifth and sixth growing season (dS m^− 1^).Growing seasonIrrigation water levelBasin plantingIn-furrow plantingSalinity level (dS m^− 1^)0.421230.421235thI1 (1005WR)0.68d-g^*^0.63e-h0.98abc0.76def0.64e-h0.64e-h1.05ab1.11aI2 (75%WR)0.49ghi0.71def0.84b-e0.75def0.45hi0.63e-i1.12a1.09aI3 (50%WR)0.41i0.58f-i0.79c-f0.87bcd0.72def0.66d-h0.73def1.01ab6thI1 (1005WR)1.04GHI1.26E-I1.50 C-H1.64B-G1.05GHI1.47 C-H1.84B-E1.67B-FI2 (75%WR)1.26E-I1.94BCD2.01AB1.69B-F1.13 F-I1.17 F-I2.21ABC2.60 AI3 (50%WR)0.71I1.43D-H1.35D-H1.55 C-G0.93HI1.14 F-I1.71BF1.78B-E*Means with the same letters in each year are not significantly different (*p* < 0.05). *Fifth growing season* Mean EC~e~ in the 0–30 cm depth was 1.0 dS m^− 1^ (not shown in Table 9 for 0–30 cm depth), and in the 30–60 cm it was 0.76 dS m^− 1^ (Table 9). In this growing season, main effect of all treatments on the EC~e~ value (30–60 cm depth) was significant. An increase in the soil salt concentration caused an increase in the EC~e~ value. The difference in low salinity levels (0.42 and 1.0 dS m^− 1^) and in high salinity levels (2.0 and 3.0 dS m^− 1^) was not significant (Table 9). A decrease in irrigation water level from 100%WR to 50%WR caused an 11% decrease in EC~e~ value. In this growing season, mean EC~e~ in the 30–60 cm depth in the in-furrow planting was 16% higher than the basin planting. This result may be due to the difference in the progression of the soil water front between the two planting methods, resulting in different leaching values from the 0–30 layer to the bottom layer. Therefore, EC~e~ value in some treatments was higher in the in-furrow planting than in the basin planting (Table 9). Increasing salinity levels from 0.42 to 1.0 dS m^− 1^ did not significantly affect EC~e~ in both planting methods in the same irrigation water level (Table 9). Effect of 2.0 dS m^− 1^ salinity and 100%WR (i.e. S~3~I~1~) on EC~e~ was 29% greater in basin planting compared to 3.0 dS m^− 1^ salinity and 100%WR (i.e. S~4~I~1~). Although LDM production was not significantly different in both treatments (Table 2), value of SY and CY was higher in S~3~I~1~ treatment than in S~4~I~1~ treatment (Tables 2 and 6). An increase in water uptake from root zone in the S~3~I~1~ treatment caused the salt accumulation in this layer, which leads to higher EC~e~ value (Table 9). At 75%WR and 100%WR, increasing salinity level did not show a significant difference in EC~e~ value in the basin planting method compared to initial value in soil (Table S1). In the in-furrow planting, highest EC~e~ value obtained at higher salinity level (Table 9). In the fifth growing season and both planting method, highest EC~e~ obtained in higher salinity level and lower irrigation water level. *Sixth growing season* In sixth growing season, mean EC~e~ at the 0–30 cm depth was 2.01 dS m^− 1^ (not shown in Table 9 at 0–30 cm depth), and it was 1.5 dS m^− 1^ at 30–60 cm. Results showed that only the main effect of salinity level and level of irrigation water showed a significant effect on EC~e~ (Table S5). In this growing season, increasing the salinity level from 0.42 dS m^− 1^ to 3.0 dS m^− 1^ caused a 78% increase in EC~e~ value. In terms of irrigation water level effect on EC~e~, 100%WR and 50%WR showed similar effect, and level of 75%WR caused 22% increase compared to other irrigation water levels. *Long-term EC*~*e*~* value * Mean EC~e~ at 0–60 cm depth in growing season of 1, 2, 5 and 6 was 1.04, 1.07, 0.88, and 1.75 dS m^− 1^, respectively (not shown in Table 9 at 0–30 cm depth). In the first and second growing season, effect of the planting method on EC~e~ value was not significant; however, an increase in the salinity level increased EC~e~ value, and a decrease in irrigation water level caused a decrease in EC~e~ value^2^. Mean EC~e~ in 0–60 cm depth in the fifth growing season was decreased 15% (data not shown) compared to that in the first growing season, due to high rainfall in 1st to 5th growing season with exception in the third growing season (Table 1). However, in the sixth growing season, the decrease in rainfall depth (Table 1) and increase in irrigation water level with saline water resulted in a 21% increase in EC~e~ value compared to that in the fifth growing season. Mean EC~e~ at the 0–60 cm depth was 0.68 dS m^− 1^ at the beginning of the first growing season (Table S1); therefore, long-term effects of salinity level on increasing EC~e~ value at 0–60 cm depth was only 157%, during 1st to 6th growing season. In general, salinity of the saturated soil extract at the end of the growing season is dependent on the amount of salt entering the root zone during the growing season, the amount of water taken by the root and the amount of salt output from the root zone^23,40^. Type of crop^41^ and tillage operations^42^ in each growing season are also effective in the physical and chemical properties of the soil and the soil salinity. Furthermore, no deep tillage operations and application of several irrigation events during the growing season resulted in adverse effect on soil properties, especially its structure and soil salinity. These problems in cultivation of saffron crop after several years forced farmers to replace the saffron fields with other crops^43^. As this has been shown in our study, replacement should have occurred after five years of saffron cultivation (Fig. 2). ### Yield components relationships Saffron yield depends on the number of the flowers, and the number of the harvested flowers depends on the corms weight capable to produce flower (over 4 g of corm weight). On the other hand, propagation and corm growth depend on vegetative growth (leaf dry matter). Therefore, relationships between the total flowers fresh weight (F~f~Y), the corm yield (CY), and the leaf dry matter (LDM) were determined for 5th and 6th growing seasons. Relationships were obtained from linear regression between the mean data of each treatment in two growing seasons. Relationship between the fresh flower yield and corm yield was determined as 13\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {{\mathrm{F}}_{\mathrm{f}}}{\text{Y }}=0.0{\mathrm{6}}0{\text{1 CY}} - {\mathrm{1}}.{\mathrm{14}} $$\end{document} R^2^ = 0.81 *n* = 48 SE = 249 *p *value < 0.0001. where, F~f~Y is the fresh flower yield (Mg ha^− 1^), CY is the corm yield (Mg ha^− 1^). Relationship between the corm yield and leaf dry matter was determined as 14\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {\mathrm{CY}}={\mathrm{24}}.{\text{91 LDM}} $$\end{document} R^2^ = 0.50 *n* = 48 SE = 14.84 *p *value < 0.0001. where, LDM is the leaf dry matter (Mg ha^− 1^). Number of flowers per corm depends on its size. Corm size depends on the planting method, the corm density, the corm size of the initial planted corm, and the depth of irrigation water and rainfall during the growing season. Corm size has an inverse relationship with its number^27,28,44,45^. Therefore, according to Eq. (13), the threshold of CY for flower initialization was obtained as 19.0 Mg ha^− 1^. Rate of com growth and propagation depends on the leaf photosynthesis; therefore, increasing the leaf area or, in other words, the amount of leaf dry matter, promotes the growth and reproduction of corms during the vegetative growth period (Eq. 14). ### Yield-salinity production function To determine the yield-salinity production function [Eq. (3)] regression between the relative saffron yield (SY~a~ /SY~m~) and the average salinity of the root zone (30–60 cm) was used^20^. This function is used to estimate crop performance in response to specific levels of soil saturation extract salinity (EC~e~) at different levels of irrigation water that helps to use water of different quantity and quality^46^. At the end of the fifth and sixth growing seasons, mean EC~e~ of root zone (0–60 cm) was not very saline (0.88 dS m^− 1^ and 1.76, dS m^− 1^ respectively), which was due to the high rainfall in the fifth growing season. Therefore, the salinity of the soil in the root zone did not increase compared to the salinity of the irrigation water. Main effect of salinity on saffron yield in the fifth and sixth growing seasons was not significant (Table 2); however, EC~e~ was affected by salinity level (Table 9), and increased with higher irrigation water salinity. Relationships between the relative saffron yield and EC~e~ value (30–60 cm) were shown in Fig. 3. According to the available data to define the yield-salinity production function in the in-furrow planting method, the salinity levels divided into low (0.42 dS m^− 1^ and 1.0 dS m^− 1^, S~1~ and S~2~) and high (2.0 dS m^− 1^ and 3.0 dS m^− 1^, S3 and S~4~) levels (Table 9). The values of threshold EC~e~ (EC~eth~), yield reduction coefficient (b) for relative saffron yield and the EC~e~ for 50% yield reduction are shown in Table 10. Table 10Yield -salinity function coefficient (b), threshold ECe, and 50% yield reduction for saffron yield at different irrigation water levels, planting methods, and salinity groups.Irrigation water levelPlanting methodSalinity groupI1 (100%WR)I2 (75%WR)I3 (50%WR)Threshold EC~e~ (dS m^− 1^)In-furrowS~3~,S~4~0.400.590.58EC~e~ for 50% yield reduction (dS m^− 1^)BasinS~1~-S~4~1.370.981.26In-furrowS~1~, S~2~0.941.171.14In-furrowS~3~,S~4~1.051.291.04Yield reduction coefficient (% per dS m^− 1^)BasinS~1~-S~4~47.071.043.0In-furrowS~1~, S~2~56.045.056.0In-furrowS~3~,S~4~76.771.0109.0 Fig. 3Relationship between the relative saffron yield (Y~r~ = SY~a~/SY~m~) and mean salinity of soil saturation extract (EC~e~, dS m^-1^) in 30–60 cm depth in the basin and in-furrow planting method (P~1~ and P~2~), irrigation levels (I~1~ = 100%WR, I~2~ = 75%WR and I~3~ = 50%WR), and salinity levels (S~1~ = 0.42, S~2~ = 1.0, S~3~ = 2.0 and S~4~ = 3.0 dS m^-1^. According to Eq. (3), lower the value of the yield reduction coefficient (b) or the higher value of the EC~eth~ indicates a lower sensitivity of crop to salinity. Results of the relationships obtained for the yield-salinity production function indicated that difference in EC~eth~ at different levels of irrigation water was significant. Difference in yield reduction coefficient (b) between the basin planting at all irrigation water levels and in-furrow planting at the lower irrigation water levels was not significant (Table 10). Yield reduction coefficient at 75%WR compared to 100%WR, increased by 51% in the basin planting; instead, it decreased by 23% in the in-furrow planting at low salinity level (0.42 and 1.0 dS m^− 1^), and it was not different at high salinity level (2.0 and 3.0 dS m^− 1^). Therefore, the in-furrow planting in 75%WR decreased the adverse effect of salinity stress on the saffron yield (Table 10); however, this adverse effect was increased in high salinity levels (2.0 and 3.0 dS m^− 1^) in 50%WR. Since EC~eth~ values were much lower than the initial value of soil EC~e~ in 30–60 cm depth (as 0.67 dS m^− 1^, Table S1); therefore, these values for basin planting with S~1~ – S~4~ and in-furrow with S~1~ and S~2~ are not presented. High soil water salinity value usually decreases soil osmotic potential and reduces water absorption by plants. As a result, deficit irrigation and the increase in salinity of irrigation water reduce the osmotic potential of soil water and plant water absorption; therefore, saffron yield is reduced as reported by Goet et al.^23^. Mean EC~e~ values (Table 9) in the 5th and 6th growing seasons were higher than the EC~eth~ values at each irrigation water levels (Table 10). Similarly, these values were higher in the sixth growing season than EC~e~ for a 50% yield reduction (Table 10). Therefore, it is indicated that the saffron yield in these growing seasons was reduced compared to 4th growing season (Table 2). ### Economic analysis Values of net benefit (NB), benefit-cost ratio (BCR), and the economic irrigation water productivity (EWP~i~) in the third to the six growing seasons are shown in the Table 11. Table 11Mean net benefit, benefit-cost ratio, and economic irrigation water productivity in different salinity levels, irrigation water levels, and planting methods in different growing seasons.Measured parameterNet benefitBenefit-cost ratioEconomic irrigation water productivity(10^6^ Rls ha^− 1^)–(10^4^ Rls m^− 3^)Growing season3rd4th5th6th3rd4th5th6th3rd4th5th6th1st -6th *Salinity level (ds/m)S1 (0.42sd/m)363.1a^**^815.0B655.3a455.2 A2.1a2.35B2.16a0.21 A17.88a28.85BC20.20a12.23 A15.68abS2 (1 ds/m)333.5a938.6 A600.3a403.9 A2.0a2.43 A1.92a0.20 A16.78a38.31 A20.13a12.08 A16.06aS3 (2 ds/m)259.5b742.4B551.7a374.4 A1.8b2.29B1.94a0.22 A12.62b27.44 C19.05a10.94 A13.07cS4 (3 ds/m)191.1c908.2 A563.7a508.1 A1.7c2.41 A1.97a0.19 A10.28c33.55AB20.04a14.20 A13.90bcIrrigation water levelI1 (100%WR)364.7a882.9 A470.2a422.9 A2.06a2.37AB1.77a0.24 A13.36b26.34 C10.41c8.43B11.22cI2 (75%WR)298.5b884.1 A625.7a468.2 A1.9b2.39 A2.03a0.20AB14.02b29.99B18.55b12.62AB14.30bI3 (50%WR)195.4c781.2B691.7a383.1 A1.7c2.35B2.21a0.17B15.79a40.03 A30.82a15.44 A18.44aPlanting methodP1 (basin)147.8b802.2B520.9a469.5 A1.7b2.41 A1.95a0.23 A10.10b31.68 A17.90a14.43 A13.06bP2 (In-furrow)425.9a899.9 A664.5a401.3 A2.12a2.33B2.04a0.17B18.69a32.40 A21.81a10.30B16.30a*1st -6th : Cumulative economic irrigation water productivity during 1st to 6th growing season. **Means with the same letters in each year are not significantly different (*p* < 0.05). *Net benefit* In the third growing season, by increasing salinity level and decrease in irrigation water level, the NB reduced by 47% and 87%, respectively (Table 11). In this growing season, the NB value was increased by 2.9 times in the in-furrow planting compared to the basin planting (Table 11). In the fourth growing season, the optimal treatments for higher NB were 3.0 dS m^− 1^, 75%WR and in-furrow planting compared to other treatments with increasing NB by 11%, 13%, and 12%, respectively (Table 11). Treatments effect on NB values in the fifth and sixth growing seasons were insignificant (Table 11). Mean NB in the fourth growing season was 3.0 times that of the third growing season (Table 11). In the fifth compared to the fourth growing season, it decreased by 12%, and in the sixth compared to the fifth growing season it decreased by 15% due to the decrease in SY (Table 11). *Benefit-cost ratio *According to Eq. (8), since the benefit-cost ratio (BCR) in the third, fourth, and fifth growing seasons were highly significant and higher than 1.0, the treatments are economically feasible (Table 11). According to the mentioned reasons, reduction of the saffron yield resulted in NB decrease in the sixth growing season, and consequently a decrease in BCR (Table 11). Statistical analysis of the BCR showed that in the third growing season, optimal salinity of applied irrigation water was 1.0 dS m^− 1^, which increased the BCR by 21% compared to other salinity levels in the fourth growing season. Effect of salinity levels on BCR was not significant in the fifth growing season (Table 11). In the sixth growing season, salinity levels did not significantly affect the BCR; however, optimal irrigation water level was 75%WR. In this year, the basin planting caused a 33% increase in BCR compared to the in-furrow planting due to the higher SY in the basin planting. Therefore, according to low value of BCR in the sixth growing season, farmers should harvest the corms at the end of the fifth growing season, since in different treatments, saffron production was not economically feasible (BCR < 1). Optimal irrigation water level for BCR in 3rd and 6th growing seasons was 100%WR since in both growing seasons the rainfall was low compared to other growing seasons (Tables 1 and 11). In the fourth and fifth growing seasons, optimal irrigation water level was 75%WR since both growing seasons had high rainfall (Tables 1 and 11). In 3rd and 6th growing seasons, BCR in the in-furrow planting was also higher than that of the basin planting indicating a more feasible saffron production (Table 9). *Economic irrigation water productivity* Results obtained in the financial analysis indicated a reduction in the applied irrigation water [denominator of Eq. (5)] in the growing season that increased the EWP~i~. Therefore, in the third to sixth growing season, optimal irrigation water level is 50%WR due to increasing EWP~i~ to 18%, 52%, 196%, and 82%, respectively (Table 11). In the third growing seasons, EWP~i~ of the in-furrow planting was 85% higher than that of the basin planting. Planting method showed a significant difference in the EWP~i~ in the fourth and fifth growing seasons. In contrast, in the sixth growing seasons, EWP~i~ of the basin planting was 40% higher than that of the in-furrow planting (Table 11). These results were related to higher SY in the basin planting with higher irrigation water application due to lower rainfall during this growing season (Tables 1 and 2). Long-term statistical analysis of cumulative economic irrigation water productivity (CEWP~i~) for 1st to 6th growing season presented in Table 11. It was obtained by dividing the cumulative NB during six growing seasons by the total applied irrigation water in the same period. Optimal treatments for CEWP~i~ were 1.0 dS m^− 1^ salinity level, 50%WR, and the in-furrow planting that resulted in higher CEWP~i~ as 15%, 64%, and 25% compared to that in other treatments, respectively. In studies that different environmental and agronomic managements are considered, the preference of different variables should be indicated by economic irrigation water productivity (EWP~i~) as suggested by Fernandez et al.^21^. This is why we used this analysis along with irrigation physical water productivity (WP~i~) and compared them together. For sure, these two different water productivities are compared differently in different years of growing periods. Therefore, this finding shows another novelty of the present study. ## Conclusions In different salinity levels, irrigation water levels and the basin planting, the highest saffron yield was obtained in the 4th growing season. However, saffron yield in the in-furrow planting, and lower salinity level unlike the basin planting has been maximum in the 2nd to 4th growing season. In all treatments, saffron yield decreased in the 5th growing season onwards. Saffron yield was higher under 100%WR in comparison to 75%WR, except in the fourth and fifth growing season, as 22%, 17%, 19%, and 25%, respectively in the six growing seasons. However, in 75%WR, in six growing seasons, it was higher than 50%WR, except in the fifth growing season, as 34%, 34%, 12%, 21%, and 25%, respectively. The total saffron yield in six growing seasons and in freshwater (S~1~) was 2.4% greater than S~2~, in S~2~ 16% greater than S~3~, and in S~3~ 5.5% greater than S~4~. Saffron yield in the in-furrow planting was also 2.26, 1.15, 1.2 times those in the basin planting in the third, fourth and fifth growing seasons, respectively. However, in the sixth growing season, saffron yield was extremely low, and in the basin planting it was 15% higher than that in in-furrow planting. In general, total saffron yield during six growing seasons in the in-furrow planting was 59% higher than in the basin planting. These results show that the in-furrow planting is more appropriate than the basin planting to reduce the adverse effect of salinity and deficit irrigation on yield and its components. Results also showed that appropriate rainfall depth during the fourth and fifth growing seasons moderated the long-term effects of salinity stress and deficit irrigation in different treatments. The relationship between the flower yield and corm yield in the final growing seasons (downward trend of growth) indicated that value of corm intensity of 19.0 Mg ha^− 1^ is required to initiate flower yield. According to the corm planting depth (15–20 cm) and maximum root depth of saffron plant (45 cm), our results showed that the salinity of soil saturation extract in the layer of 30–60 cm is more effective in water uptake by the roots. Yield reduction coefficient and the threshold EC~e~ values showed that in the basin planting, the relative yield also is reduced with decrease in the depth of irrigation water (deficit irrigation); however, in the in-furrow planting, this trend was the opposite. Results showed that maximum saffron yield in all treatments except the in-furrow planting occurred only in the fourth growing season. Whereas, it occurred in second to fourth growing seasons in the in-furrow planting. However, the downward trend of saffron growth started in the 5th growing season in all treatments. Also, results of the SY, CY, NB, BCR, and EWP~i~ showed that corm should be harvested at the end of the fifth growing season because the downward trend of yield reduction of the saffron crop has started in this season. Thus, according to results of SY and NB, best salinity for all six growing seasons was lower than 2.0 dS m^− 1^, and optimal irrigation water level from the first to the third growing season was complete irrigation (100WR); however, it was 75%WR in the fourth to sixth growing season. Also, according to SY, in-furrow planting is recommended as compared to basin planting in saffron field. ## Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1