Effects of Different Water Supply and Corm Planting Density on Crocin , Picrocrocin and Safranal , Nitrogen Uptake and Water Use Efficiency of Saffron Grown in Semi-Arid Region

Saffron’s color, taste and odor result from the chemicals crocin, picrocrocin and safranal, respectively. Hence, in addition to quantitative yield, secondary metabolites content are known as crucial factors for a successful saffron production. Moreover, enhancing resources efficiency, especially water and nitrogen, is becoming increasingly important for agricultural improvement in arid and semi-arid regions. Thus, the effects of irrigation levels and corm planting on crocin, picrocrocin and safranal content, water use efficiency (WUE) as well as nitrogen use efficiency (NUE) of saffron were investigated as a two-year field experiment based on a randomized complete block design arranged in split-plot with three replicates. The irrigation levels (100, 75 and 50% of saffron water requirement) and corm planting pattern (50, 100, 200 and 300 corms m) were allocated to main and sub-plots, respectively. Based on the results, crocin and picrocrocin content increased with decreasing irrigation levels. The highest WUES (WUE based on dry stigma yield) was obtained when 50% of saffron water requirement was supplied. However, the lowest WUEC (WUE based on daughter corms yield) and NUEC (NUE based on daughter corms yield) were obtained when 50% of saffron water requirement was applied. Irrespective of irrigation levels, WUES, WUEC and NUEC increased with increasing the planting density. The results demonstrated that although relatively severe water stress increases WUES and secondary metabolites in saffron stigmas, it could decrease WUEC and NUEC through affecting daughter corm growth.

Water use efficiency (WUE) is defined as yield of marketable crop produced per unit of water used in evapotranspiration (Dong et al., 2011).It is generally believed that in the future, water availability will become increasingly scarce, particularly in arid and semi-arid regions, due to rapid urbanization, higher population growth and expanding areas of irrigation (Abbaspour and Sabetraftar, 2005;Chiew et al., 2011;de Souza and Costa da Silva, 2014).Hence, enhancing WUE is becoming increasingly important for agricultural improvement in these regions (El-Hendawy et al., 2008;Barati et al., 2015).
According to saffron's irrigation schedule, an optimal irrigation schedule consists of five to six irrigation rounds (Koocheki et al., 2014;2016).These irrigation rounds are usually performed in mid-summer (for flowering induction), in early October (for flowering acceleration), in November (after flower picking and leaves appearance), in December (after winter weeding), in March and finally in April (supplementary irrigation for optimum daughter corm growth).However, in some arid and semi-arid regions, saffron fields are irrigated only once (in October), mainly due to water shortage, causing a significant reduction in flower and corm yields (Kafi et al., 2002;Koocheki et al., 2014).Hence, in spite of being a crop compatible with arid and semi-arid regions, with low water requirements (Alizadeh et al., 2009;Sepaskhah and Kamgar-Haghighi, 2009;Yarami et al., 2011), water shortage is the most important challenge in sustainable saffron production (Yarami and Sepaskhah, 2015).Therefore, it is critical to determine the amount of water by which saffron can produce maximum yield.
In addition to WUE, more attention should be paid to nitrogen use efficiency (NUE) as an important index in saffron sustainable production (Koocheki and Seyyedi, 2015).
Nitrogen use efficiency, which is defined as the ratio of the crop yield to the total input of N applied, is split into acquisition and physiological efficiency (Lea and Azevedo, 2006;Salvagiotti et al., 2009).Due to being a perennial species, at least in field conditions (Kumar et al., 2009;Babaei et al., 2014), as well as having dynamic N allocation among leaves or underground organs (Ourry et al., 1988;Dordas, 2009), it appears that NUE in saffron is more complicated than within other annual plants (Koocheki and Seyyedi, 2015).
Optimum planting pattern based on mother corms density is one of the most factors affecting daughter corms behavior, resulting in more flower yield and of better quality (Kumar et al., 2009;Koocheki et al., 2011;2014).Dense corm planting pattern can increase saffron production, especially during early years (Koocheki et al., 2011(Koocheki et al., , 2012(Koocheki et al., and 2014)).Accordingly, it was hypothesized that dense corm planting pattern would increase WUE and NUE.Therefore, this experiment was aimed to study the crocin, picrocrocin and safranal content, WUE and NUE in response to different levels of saffron water requirement (SWR) and corm planting patterns.).The soil (0-30 cm) has bulk density 1.29 g cm -3 , EC 1.13 dS m -1 ; organic carbon 0.54%; available N 18 mg kg -1 ; available P mg kg -1 ; available K 165.19 mg kg -1 ; clay, sand and silt, 49.80, 18.23 and 31.79%,respectively.

Experimental design and field management
The experimental design was a randomized complete block arranged in split-plot with three replicates.The irrigation levels (100, 75 and 50% of SWR equal to no water stress, mild water stress and relatively severe water stress) and corm planting pattern (50, 100, 200 and 300 corms m -2 ) were allocated to the main and sub-plots, respectively.
SWR was calculated according to total potential evapotranspiration values in the first (523 mm) and second (640 mm) year of the experiment, respectively (Yarami et al., 2011).More information about determination of SWR and irrigation schedule is given in Table 1.
Before mother corms (4-6 g) planting, composted cattle manure (25 ton ha -1 ) was mixed into the soil and then plots were established.The plots were 2.5 × 1.5 m in size and 0.5 m apart.Composted cattle manure (N 1.65%; P 0.41%; K 0.87%; organic carbon 28.36%) was applied just in the first year of the experiment as the level of soil organic carbon was low (0.54%).
Mother corms planting was done on 17 th of June 2012.Interrow distance for each density was 20 cm.Irrigation (Table 1) was performed using polyethylene irrigation network equipped with counter.During both years of the experiment, weeds were controlled manually when required.

Flower and corm measurements
In the first year, flowers were manually picked up from mid to late of November 2012 and dried stigma yield was measured.Stigmas were dried (Fig. 2) in an oven at 30 °C for 48 h.At the end of first growing season, daughter corm yield was determined using a 0.5 × 0.5 m quadrate per plot.The rest of the daughter corms were devoted for the second year.In the second year, flowers were manually picked up from early to late of November in 2013.Other measurements were performed same as in the first year of the experiment.

Statistical analysis
Analysis of variance (ANOVA) and Duncan's multiple range tests were performed using SAS 9.3 software (SAS, 2011).Saffron traits were analyzed as split-split plot arrangement in time.The irrigation levels, corm planting density and harvesting (in the first and second year) were considered as first, second and third factors, respectively.

Flower quality
Although the effect of irrigation levels on crocin and picrocrocin content was significant, safranal content was not affected by irrigation levels (Table 3).In addition, the effect of planting density and harvesting year were not significant on above mentioned compounds (Table 3).
Interestingly, crocin and picrocrocin content increased with decreasing SWR levels.In other words, the highest crocin and picrocrocin content was observed under relatively severe water stress (50% of SWR) condition (Table 4).
The mechanism of saffron organic compound synthesis in response to water stress is not fully understood.As mentioned earlier, crocin and picrocrocin are the most important compounds found in saffron stigmas (Tarantilis et al., 1995;Escribano et al., 1996), so that higher amounts of these compounds leads to an increase in the quality of saffron (D'Auria et al., 2004;Srivastava et al., 2010;Koocheki et al., 2016).On the other hand, it has been reported that the quantity of essential oils and secondary metabolites in plants would increase in response to drought stress (Singh-Sangwan et al., 1994;Reddy et al., 2004).In addition, an increase in 336 Determination of crocin, picrocrocin and safranal In the first and second year, crocin, picrocrocin and safranal were measured based on ISO 3632 trade standard (ISO/TS 3632, 2003), using UV-vis spectrometric method.Crocin (440 nm), picrocrocin (257 nm) and safranal content (330 nm) were expressed as direct readings of the absorbance of 1% aqueous solution of dried stigma saffron (Lage and Cantrell, 2009).

Determination of WUE WUE (water use efficiency) was calculated as follows:
WUES= Dry stigma yield (g ha -1 ) / Total water use (mm) WUEC= Daughter corm yield (kg ha -1 ) / Total water use (mm) The total water used (TWU) was measured using the following equation (Dong et al., 2011): TWU = P + I + ∆W Where: P is the precipitation (mm), I is the irrigation (mm), ∆W is the soil moisture change (mm).
Due to plots design, there was no surface water runoff under the conditions of this experiment.The soil water drainage below the crop root zone (mm) and capillary water rise to the root zone (mm) were considered to be negligible.

Determination of NAE and NUE
Nitrogen concentration (g kg -1 ) in daughter corms (plus corm tunics) and aerial part was measured based on Kjeldal method (AOAC, 2000).On the basis of dry stigma and daughter corms yields, nitrogen acquisition efficiency (NAE) and nitrogen use efficiency (NUE) were calculated using the following equations (Brennan et al., 2014;Koocheki and Seyyedi, 2015): NAE (%) = (Nt / Na) × 100 NUES (g g -1 ) = SY / Na NUEC (g g -1 ) = CY / Na Where: Nt is g N in the total plant m -2 , Na is g N applied m -2 , SY is g dry stigma yield m -2 , and CY is g daughter corms yield m -2 .
Applied N was determined by the sum of following resources: 1-initial N content into soil before establishment of the trial (based on soil bulk density at the depth of 30 cm), 2-N added by composted cattle manure, and 3-N content in mother corms (Table 2).protein content, peroxidase and superoxide dismutase activity, in response to drought stress, has been documented by Maleki et al. (2011).Consequently, it seems that increase in crocin and picrocrocin synthesis in response to drought stress is a compatibility mechanism in saffron.

Interaction between irrigation and planting pattern
Irrespective of irrigation levels, WUES, WUEC, NAE, NUES and NUEC increased with increasing planting density (Table 5).For instance, in full irrigated plants (100% of SWR), an increase in planting density from 50 to 300 corms m -2 increased WUES and NUEC by 4 and 2 times, respectively.
As mentioned above, dense corm planting pattern not only improved flower yield during early years, but also promoted sustainable production of saffron (Koocheki et al., 2011;2016).According to the literature, an increase in planting density can be a good approach to deal with water loss in arid regions (Stroosnijder et al., 2012).Hence, it seemed that dense planting pattern can increase NUE in saffron through more N uptake and help to reduce water loss.
In each level of planting density, the highest WUES was obtained when 50% of SWR was applied (Table 5).This might be due to flowering adaptation mechanisms to drought stress (Sepaskhah, 2009).In other words, drought stress is an incentive factor for flowering which in turns resulted in maximum WUES.
Regardless of the planting density, the lowest and highest WUEC were obtained when saffron plants were irrigated with 50 and 75% of SWR, respectively.Under same conditions, in terms of planting density, the lowest NAE, NUES and NUEC were recorded when 50% of SWR was applied (Table 5).For example, when 50 corms m -2 were cultivated, a reduction in irrigation water from 100 to 50% of SWR decreased NAE by 41.98%.
It appeared that mild water stress (supplying 75% of SWR), stimulated daughter corms growth through increasing root growth and better nutrients uptake, especially N. In other words, a slight reduction in water availability would increase corm yield per unit of available water.Nevertheless, considering the sensitivity of saffron to water shortage (Sepaskhah and Yarami, 2009;Renau-Morata et al., 2012;Yarami and Sepaskhah, 2015), sever stress would negatively affect daughter corm growth, NAE, NUES and NUEC.In this regard, Renau-Morata et al. (2012) observed a decrease in the photosynthetic rate of saffron under water stress.Motalebifard et al. (2013) showed that water deficit stress caused a significant reduction in tuber numbers, yield and WUE of potato (Solanum tuberosum L. cv.'Agria').

Interaction between irrigation and harvesting (year)
The results revealed that when 75 and 100% of SWR was applied, more WUES, WUEC, NAE, NUES and NUEC were registered in the second year rather than the first year of the experiment (Table 6).For instance, in full irrigated plots, NAE in the second year increased by 32.6% compared with the first year (Table 6).
Saffron is a plant that propagates by mean of corms, which are renewing each year.Above and underground organs grow more from year to year (Kumar et al., 2009;Koocheki et al., 2014), thus more nutrients are absorbed from the soil over the time (Koocheki et al., 2016).Moreover, higher growth of aerial part and roots cause lower water loss.According to the previous figures, developed above ground organs stimulate root growth and in turns improve plant capability to uptake water and nutrients.Therefore increasing plant density would increase WUE in crops (Sani et al., 2008;Stroosnijder et al., 2012).

Interaction between planting density and harvesting (year)
From the results obtained, when 50, 100 and 200 corms m -2 were cultivated, more WUES, WUEC, NAE, NUES and NUES were recorded in the second year compared with the first year (Table 7).However, when 300 corms m -2 were cultivated, more WUES and WUEC were recorded in the first year compared with the second year (Table 7).This might be due to more small corms, formed at the end of the first year, in 300 corms m -2 treatment.In this regard, Koocheki et al. (2012) found that an increase in planting density up to 400 corm m -2 increased the ratio of small daughter corms to total daughter corm.It has been reported that there is a positive relationship between corm size and flowering ability (Gresta et al., 2008;Douglas et al., 2014).Therefore, small corms formation in the first year and less flower production in the second year may be considered as the main reasons for reduction of WUES and WUEC in saffron.

Correlation between water consumption and N uptake
There was a positive and significant correlation between WUEC and NAE (Fig. 3A), WUEC and NUES (Fig. 3B) and WUEC and NUEC (Fig. 3C).These results suggested that effective approaches for increasing WUEC can be practiced through stimulating daughter corms growth and more N uptake from the soil, which improve flower yield per each unit of absorbed N. From the other point of view, considering the key role of N in stimulating vegetative growth, daughter corms formation and flower production (Chaji et al., 2013), it seems that higher NUE causes less soil evaporation and more saffron yield per unit of consumed water.Generally, the results demonstrated that dense planting pattern can be an effective approach for increasing WUE and NUE in saffron production.Furthermore, the results indicated that although relatively severe water stress increases WUES and secondary metabolites in saffron stigmas, it could decrease WUEC, NAE, NUES and NUEC through affecting daughter corm growth.Considering the positive relationship found between WUEC and NAE, WUEC and NUES, as well as between WUEC and NUEC, it can be stated that the crucial factors affecting daughter corms growth can lead to more efficient use of water and nitrogen in saffron crop.
Two-year field experiment was carried out during 2012-2013 and 2013-2014 growing seasons, at experimental station of Faculty of Agriculture, Ferdowsi University of Mashhad (latitude: 36 • 15′ N; longitude: 59 • 28′E; elevation: 985 m).The study site was classified in semi-arid region located in Northeast of Iran.Monthly rainfall and average temperature during both growing seasons are given in Fig. 1.The soil was clay (US system) and alkaline in reaction (pH 8.16

Fig. 1 .
Fig. 1.Monthly rainfall and average temperature during both growing seasons (from June 2012 to May 2014)

Table 1 .
Determination of saffron water requirement (SWR) and total amount of applied water (TAAW) in the first and second years of the experiment SWR was calculated according to Kc coefficients at initial, middle and final growth stages in the first and second years of the experiment, respectively.

Table 3 .
Analysis of variance for the studied traits of saffron

Table 4 .
Effects of irrigation, planting density and harvesting (year) on studied traits of saffron Values followed by the same letter were not significantly different at p ≤ 0.05 (DMRT).WUES: water use efficiency (WUE) based on dry stigma yield; WUEC: WUE based on daughter corms yield; NAE: nitrogen acquisition efficiency; NUES: nitrogen use efficiency (NUE) based on dry stigma yield; NUEC: NUE based on daughter corms yield.

Table 5 .
Interaction effects of irrigation and planting density on studied traits of saffron Values followed by the same letter were not significantly different at p ≤ 0.05 (DMRT).Values in the parenthesis indicate standard deviation (±) of means.WUES: water use efficiency (WUE) based on dry stigma yield; WUEC: WUE based on daughter corms yield; NAE: nitrogen acquisition efficiency; NUES: nitrogen use efficiency (NUE) based on dry stigma yield; NUEC: NUE based on daughter corms yield.

Table 6 .
Interaction effects of irrigation and harvesting (year) on studied traits of saffron Values followed by the same letter were not significantly different at p ≤ 0.05 (DMRT).Values in the parenthesis indicate standard deviation (±) of means.WUES: water use efficiency (WUE) based on dry stigma yield; WUEC: WUE based on daughter corms yield; NAE: nitrogen acquisition efficiency; NUES: nitrogen use efficiency (NUE) based on dry stigma yield; NUEC: NUE based on daughter corms yield.

Table 7 .
Interaction effects of planting density and harvesting (year) on studied traits of saffron