COMPARISON OF SPRING AND SUMMER SOWING OF SUGAR BEET GENOTYPES AT DIFFERENT HARVEST DATES TO SHIFT FROM TRADITIONAL CROP TO CASH CROP IN CENTRAL IRAN

This study was carried out in Karaj, Iran in 2017 and 2018 to assess the efficacy of summer sowing (June 22) versus spring sowing (April 20) of six sugar beet genotypes at three harvest times (October 13, November 2, and November 23) via the measurement of catalase (CAT), malondialdehyde (MDA), and agronomic traits. Results showed that in both sowing dates, higher growth and temperature were related to higher CAT activity and MDA content, and the maximum MDA and CAT activity were observed in 1700−1900 growth degree days (GDD). Genotypes responded to the shortening of the growth period differently. The best genotypes for summer sowing were found to be ‘Paya’, ‘IR7’, and ‘Pars’ when a combination of the least response to delayed sowing and the highest root yield in the summer sowing conditions was considered. Compared to the spring sowing, the summer sowing decreased white sugar yield (WSY) of all cultivars by 28.3−50.5% in the first year and 5.3−32.4% in the second year. ‘Paya’ and ‘IR7’ were the most capable cultivars in preserving WSY so that they maintained 70% of their yields. In addition, the genotypes exhibited their highest WSY at the November 23 harvest date so that root, raw sugar and white sugar yields were 41.21, 6.35 and 5.02 t ha −1 higher at the November 23 harvest date than at the October 13 harvest date, respectively. Based on the results, if summer-sown sugar beets are considered as a cash crop in rotation with grains and there is no limitation on water supply, it can then be recommended to farmers as it can make good profits for them.


INTRODUCTION
Temperature rise has influenced crop production all around the world by increasing drought frequency, but some parts of Africa and the Middle East are especially at risk (38,43). Vegetation cover in countries like Iran has been influenced by climate change and the resulting summer heat waves and droughts, without the possibility to shift crop cultivation to other parts of the year effectively (7,11,31). Most regions utilize a wide range of adaptive measures in agronomy (e.g. irrigation, nutrition, tillage methods, and so on) to alleviate the adverse impacts of climate change on crop production. Sowing date adjustment is a strategy in focus that can play a significant role in accommodating crops and mitigating the negative effects of climate change (3,21,49). It has been hypothesized that yield potential is influenced by the length of radiation absorption period, photosynthesis rate, and total evapotranspiration so that increasing this period will contribute to increasing crop yield (5,16). For any changes in sowing dates, crop water use efficiency should be first measured (8; 34). Like many other crops, sugar beet is well fitted in rotations with other crops, and various methods have been proposed to reduce water use in sugar beet cultivation. Examples include developing drought-tolerant cultivars (49), selection for yield and secondary traits (8,29), deficit irrigation (25,28), and plant selection and breeding approaches to achieving genotypes with shorter growth period (8,17). However, these cultivars inhibit crop yield loss by late-season stresses (34). These cultivars cannot be recommended for regions with a long growth period. So, breeding programs had better focus on cultivars with a flexible growth period than early-maturing cultivars by considering the environmental conditions (2,41,44). Also, many researchers have proposed to change the sowing date to escape from drought at sensitive growth stages, which is root swelling in sugar beets. A negative impact of stress in all plants is the increased synthesis of reactive oxygen species (ROS), which degrades DNA, lipids, proteins, and other biomolecules (48). The products of the biodegradation caused by the peroxidation of lipids and amino acids are malondialdehyde (MDA) and dityrosine, respectively. So, these two compounds are commonly measured as biomarkers of degradation by heat and drought stress. In normal conditions, cells have an efficient and adequate antioxidant defensive system against ROS. Antioxidant enzymes play a critical role in the adaptability and survival of plants during stress (49). In stressful conditions, ROS synthesis, however, outweighs the plants' capability to handle them and this causes oxidative damages and has negative impacts on plant growth and yield (39). In addition, heat and drought in sugar beets reduce root yield, raw sugar, and sugar yield and increases α-amino significantly, but they have a very slight impact on sodium and calcium (although root sodium content increases and sugar percentage decreases by extensive irrigation). On the other hand, farmers are highly willing to maximally use land and radiation by sowing an earlymaturing plant that is adaptive to deficit irrigation and produces adequate yield (as intermediate summer sowing) between two consecutive autumn crops (11). The intermediate summer sowing between two autumn crops (e.g. wheat, barley, and canola) as a cash crop is prevailing in many parts of Iran. In summer sowing, crop yields decline as growth period shortens. But, this yield loss has been partially controlled in recent years by the introduction of early-maturing cultivars. However, the summer sowing of many economically valuable plants, like sugar beets, still needs to be studied (44). So, although it is assumed that the summer sowing of sugar beets reduces water use owing to the shortening of the period in which irrigation is required, it is predicted that the crop yield may decline because early harvest of sugar beets in cold temperate regions is a major factor responsible for the loss of their yield and sugar content (33). The issue of crop yield and quality decline is a critical point in deciding on the use of this approach to decreasing water use because the final crop is harvested at the end of the growing season and the distinction between different sugar beet cultivars during the growing season is not precisely revealed for the improvement of sound crop management. Thus, the present study aims to compare the intensity of heat-induced 234 oxidative damage and root biomass accumulation trend in spring and summer sowing and to assess the efficacy of summer sowing, water use efficiency, and yield and quality of sugar beet cultivars in different sowing dates in a cold temperate region of Iran. MATERIALS AND METHODS Experimental site and plant growth conditions: The field experiment was conducted on clay-loam soil at the Motahari Sugar Beet Research Station, Karaj, Iran as a factorial arrangement based on randomized complete block design with four replications during 2017−2018. The area is located at latitude 35°59' N and longitude 51°6' E with an altitude of 1312 m above the sea level. To simplify the comparison of the growing season weather, the monthly total precipitation and temperature were considered from March to December at the Karaj Agricultural Research Farm ( Figure 1). To determine soil characteristics, soil sampling was performed before the experiment. To do this, field soil sampling was done from the depth of 0−30 cm at eight spots. Then, the collected samples were sent to the laboratory to determine soil texture and chemical composition. Properties of experimental soil samples are given in Table 1. Experimental treatments and soil preparation: A factorial experiment was applied based on randomized complete block design (RCBD) with four replication, it involved sugar beet cultivars and sowing and harvesting dates with four replications. The six sugar beet cultivars from Betaseed included 'IR7' (G1), 'Pars' (G2), 'Paya' (G3), 'Jolgeh' (G4), '261*276.P.77.SP.19' (G5), and '(7112*261)*5RR-87-HF.33' (G6). These cultivars were selected based on their disease tolerances, resistance to bolting, high yield potential, and suitability for energy beet production (personal communication, Steve Libsack). There were two sowing dates: spring (April 20, 2017 and April 23, 2018) and summer (Jun 22, 2017 and Jun 25, 2018). Harvesting dates included October 13 (H1), November 2 (H2) and November 23 (H3). To prepare the seedbed, the soil was disked followed by rotary tillage. Each plot included 6 sowing rows of 8 m length and 50 cm spacing. At the harvesting time, after removing the first and 6 th rows of each plot and 0.5 m from both ends of each row, an area of 14 m 2 was harvested. Fertilization included a preplant application of 82 kg ha −1 N (as urea), 37 kg ha −1 P (as triple superphosphate), and 56 kg ha −1 K (as Potassium sulfate) and again at 4−6 weeks after sowing, which varied due to the weather conditions and early season growth patterns of the sugar beets among site-years.   After sowing, irrigations were performed normally until the 4−leaf stage. Then, a Class−A evaporation pan was mounted and the irrigations were performed with the schedule provided in Tables 2 and 3 based on 90 mm of evaporation. A WSC−flume type IV was mounted at the entrance of water into the farm and a WSC−flume type III was mounted at its exit and Eq. (1) and (2)  (2) in which Q denotes the quantity of water flowing into the farm and H denotes water height in the WSC−flume (cm). Water use efficiency (WUE) was calculated by Eq. (3).
in which SY denotes sugar yield (g) and WU denotes the quantity of water use (m 3 ) (14).

Root growth trend
To take care of the marginal effect, plants were sampled from the middle of the plots. One month after emergence, to check the trend of the root growth, the plants were harvested and after removing their roots and leaves, their fresh and dry weights were recorded. The measurement times, presented in Table 3, were based on the growing degree day (GDD), which is calculated by minimum daily temperature T min , maximum daily temperature T max , and base temperature T b of the plant growth. The numerical value of GDD expresses the thermal efficiency of the growth days in plant evolution (6). We calculated GDD using Eq. (4) based on the daily accumulation from the first irrigation (i) until the harvest time (n) assuming a base temperature of 3°C (27): in which if T max was greater than 30°C, it would be assumed to be 30°C and if T min was lower than 2°C, it would be set at 3°C.

Catalase (CAT) and malondialdehyde (MDA) calculations
CAT and MDA were measured in five steps according to Table 3 (Aebi, 1984). Malondialdehyde (MDA) measurement. For measurement of MDA, 0.2 g fresh leaf tissues were ground in 5 ml of 0.1% TCA. After centrifuging at 4,000 rpm for 20 min, 2.5 ml 0.5% thiobarbituric acid (TBA) in 20% TCA were added to the supernatants. Extracts were incubated in a water bath at 95° C for 30 min and immediately cooled in ice. Afterwards, all samples were centrifuged at 4,000 rpm for 30 min and the absorbance was measured at 532 nm and 600 nm. The MDA concentration was calculated by subtracting the non-specific absorption at 600 nm from the absorption at 532 nm using an absorbance extinction coefficient of 155 mM −1 cm 1 (36). At harvest time, plants from an area of 1 m 2 were sampled after removing 0.5 m from both ends of the plots and two borderline rows to account for the marginal effect. The samples were, then, placed in plastic bags immediately and were sent to the plant physiology laboratory of Sugar Beet Seed Institute to record their traits.
In the laboratory, different parts of the plants (including leaves, crowns, and roots) were separated and their fresh and dry weights (after oven-drying at 105 °C for 24 h) were determined (28). After the roots were rinsed in the laboratory, an automatic device was used to prepare root pulp. Then, 26 g of the pulp was mixed with 177 ml of dilute lead acetate solution to get a transparent extract. All these operations were performed in a Venema G 2 . The infiltrated extract was poured into specific glasses and was sucked into the sugar beet quality analysis device (Betalyzer) by a sucker and its sugar percentage was determined. Sugar percentage was determined by polarimetry, sodium (Na) and Potassium (K) content by flame-photometry, and aminonitrogen (amino-N) by spectrophotometry.
After the values were determined, white sugar content, raw sugar yield, and white sugar yield  in which WSC represents white sugar content (%), SC represents sugar content of the roots (%), RSY denotes raw sugar yield or white sugar yield (t ha −1 ), RY shows root yield (t ha −1 ), and WSY denotes white sugar yield (t ha −1 ).

Statistical analyses
An analysis of variance (ANOVA) was carried out in the SAS9.3 software package (SAS Institute Inc., Cary, NC) using the ANOVA procedure. The table of ANOVA includes the simple and interactive effects of sowing date, harvest date, and beet cultivars. Wherever the interaction of the factors was significant, the results were presented for the interactive effects, and the simple effects of the factors were ignored. Means were compared by the Least Significant Differences (LSD) test at the p ≤ 0.05 level.

Sugar beet root growth trend
The results showed that the maximum root growth in the first year was achieved for all cultivars from 134 DAP (GDD = 2584) until 195 DAP (GDD = 3521) in the spring sowing and from 92 DAP (GDD = 1823) until 132 DAP (GDD = 2370) in the summer sowing. But, in the second year, an increase was observed in the root growth of G3, G4, and G6 and a decrease in the root growth of G1, G2, and G5 after 166 DAP (GDD = 2130). In most root growth steps, G4 had the highest root weight in both sowing dates. The second highest root weight was related to G3 in the spring sowing and G5 in the summer sowing ( Figure 2). In summary, the maximum root growth was within 130−190 DAP in the spring sowing and 90−135 DAP in the summer sowing, so any stresses within these time frames would reduce root and sugar yields severely. Plants are threatened in these periods with various stresses, e.g. water deficiency, high temperatures, and hot winds, which farmers are unable to control (5). In most studies on comparing spring and summer sowing of sugar beets under climatic conditions that have been similar to our study, the highest dry matter and leaf areas have been obtained from the early sowing, which has resulted in significantly higher root and sugar yields (40).

Trend of catalase (CAT) activity and malondialdehyde (MDA) variations
The results for MDA measurement of sugar beet cultivars during their growth period revealed that it was increased in all cultivars in the spring sowing from 572 GDD to 1277 GDD, but from 1277 GDD until 3303 GDD, the rate of MDA synthesis almost reached a plateau in all cultivars except for G6 ( Figure  3). In the summer sowing, the rate of MDA variations was ascending until 1823 GDD for 237 all cultivars, but it started to decline in G2 and G4 and increase in G3, G5, and G6 from 1823 GDD to 2151 GDD ( Figure 4). The results of the measurement of CAT activity indicated that in the spring sowing, the maximum CAT activity was observed in all cultivars from 1277 GDD to 1836 GDD except for G4 whose activity was peaked at 1836 GDD and started to decline after that. The highest CAT activity at different growth steps was observed in G1 and the lowest in G4 until 1836 GDD and G2 after 1836 GDD ( Figure 5). But, in the summer sowing, CAT activity was increased in all cultivars except for G5 from 572 GDD until 1725 GDD at different rates, and then they went down. In most steps, the highest CAT activity was observed in G5 whose maximum activity was obtained at 1823 GDD ( Figure 6). The results indicated that although cultivars exhibited various CAT activity when sown in spring or in summer, the maximum CAT activity in the spring sowing (152 μmol min −1 mg −1 protein for G1) did not differ from that of the summer sowing (163 CAT activity (μmol min −1 mg −1 protein for G5) significantly. CAT is the most important antioxidant enzyme that is activated when a plant is exposed to stress (18,37). The occurrence of stress during plant growth periods is unavoidable in Iran's climate, especially in hot months, so it will be useful to change the sowing date to avoid the coincidence of sensitive growth steps with stressful conditions (16). It was proven by the measurement of CAT and MDA so that the trend of CAT activity was consistent in the spring-sown plants, but MDA was increased until 1860 GDD and then started to decrease. In the summer sowing, CAT activity increased until 1720 GDD and then started to decrease whereas MDA had an ascending trend in most cultivars. So, it can be inferred that the plants were exposed to stress and that the summer sowing encountered stress to a greater extent than the spring sowing.  Table 4 shows that LDW was significantly influenced by the interaction of G × H in both years, S × H in the first year, and S × G in the second year. Based on the comparison of means for the interaction of S × G, the summer sowing increased LDW by 2.6−14.1% in G2 and G4 and by 6.2−18.5% in the other cultivars. Also, the comparison of means for the interaction of S × H indicated that in both spring and summer sowing, LDW was decreased as harvest time was extended. Also, it was found about the interaction of G × H that in both years, LDW was increased from the first to third harvest in G2 and G4 in both years, but it was decreased in the remaining cultivars (Table 4). Leaf production in sugar beets is a function of environmental conditions, nutrition, genetics, and growth period length, and in addition to leaf number and area, leaf duration is also important (47).  20) found that the leaf and root growth of sugar beets were increased with temperature rise during the first 65 days of sugar beet growth, but the relationship between the environmental temperature and growth rate diminished from 121 to 145 days after sowing. However, it should be noted that the rise of environmental temperature impairs dry matter accumulation in C 3 plants (24).

Crown fresh weight (CFW)
The results of ANOVA revealed that the simple effect of S, G, and H in both years and the interactive effect of S × H were significant on CFW (Table 4). In both years, the summer sowing was related to lower CFW than the spring sowing and the later harvests increased CFW significantly versus the subsequent harvest dates. The interaction of S × H, also, showed that the third harvest had higher CFW in both spring and summer sowing and that the harvest dates in spring sowing had higher CFW than that in the summer sowing. Among the sugar beet cultivars in both years, G6 had the lowest CFW of 421 g m −2 in 2017 and 314 g m −2 in 2018 (Table 2). Similarly, Nagib et al. (30) concluded that delayed harvest increased CFW, especially in late-maturing cultivars. According to Loel and Hoffmann (24), earlier sowing dates produced higher crown dry weight and root yield was significantly correlated to crown weight, total dry matter, and leaf weight. Reinsdorf et al. (35) argue that genotypes with higher resistance and more sugar concentration have smaller crowns.

Root dry weight (RDW) and root yield (RY)
The results showed that the interactions of G × H and S × G in both years and S × H in the first year were significant for RDW (  (Table 5). Although RY of cultivars showed various responses to sowing and harvest dates, RY of all cultivars was decreased when sowing was delayed, whereas the RY of some cultivars was higher in the second harvest than in the third harvest and this difference was more remarkable in the spring sowing date (Figure 7). The main effect of harvest date also implies that in the first year, RY was higher in the second harvest date than in the first and third harvest dates, but in the second year, the second and third harvest dates did not differ significantly. Delayed sowing reduced yield at all harvest times, and the decrease was 32.6% in the first harvest (H1), 42.0% in the second harvest (H2), and 44.7% in the third harvest (H3) compared to similar harvest dates under spring sowing (Table 5). On the other hand, different genotypes responded to the shortening of the growth period differently. The best genotypes for summer sowing, determine by mixing two components of the weakest response to delayed sowing and the highest yield in summer sowing conditions, were found to be 'Paya', 'IR7', and 'Pars'. It has been documented by preceding studies that yield escalation by earlier sowing is more significant in years with appropriate climatic conditions (5,13,17). Assuming that the response of RY to growth period duration is linear (19; 40; 46), the comparison of spring sowing (April 9−19) and summer sowing (June 21 −30) showed that each day delay in sowing in the

240
Karaj region, Iran would cost 400 kg ha −1 on RY, which is much higher than 200 kg ha −1 RY per each day delay in the Netherlands (42) and 134−162 kg ha −1 in Turkey (4,21). This is, however, in full agreement with previous studies in the region (13). The amount of RY decline in this study is consistent with the reported values, which is about 6−9% (4). Irrespective of the type of mathematical relation, the length of growth period has always been regarded as one of the main factors describing RY variations (10,40).  (4) also asserted that SC was increased by 0.07 units per each day delay in harvest. It seems that since 75% of root dry matter includes sucrose (40), the reason for the higher sugar content at delayed harvest is the increase in the root dry matter percent. Also, in another study, postponing the harvest time improved sugar beet quality but SC loss and technical quality of sugar beet after harvest and during storage in silo were the minimum in early sowing and delayed harvest (26).

Raw sugar yield (RSY) and white sugar yield (WSY): Based on the results, the interaction of S × H and S × G in both years
and G × H in the first year were significant for RSY (  (Table 5). It has already been reported that the range of differences among genotypes is decreased when sowing is retarded (23). It is, thus, reasonable to recommend the use of late-maturing genotypes for early sowing dates and early-maturing genotypes for late sowing dates or early harvest dates. Research on the effect of environmental factors on the growth and yield of six sugar beet cultivars in 62 sites showed that the sowing date was generally the most influential factor on the interaction of cultivar × environment (15). Means comparison for the 242 interaction of G × H indicated that all cultivars had higher WSY in the second and third harvest than in the first harvest in both years, but the highest WSY was in the third harvest for some cultivars and in the second harvest for the other cultivars. In both years, the highest WSY was obtained from G6 whose WSY was 9.46 t ha −1 in the third harvest in 2017 and 7.66 t ha −1 in the second harvest in 2018 (not differing from that of the third harvest, 7.35 t ha −1 , significantly). However, it should be noted that although the second and third harvests did not differ significantly, the second harvest had 6.3% higher WSY than the third harvest in the spring sowing and the third harvest had 11.8% higher WSY than the second harvest in the summer sowing ( Table  5). The results show that the harvest at 161 DAP was the best for sugar beets in the Karaj region. On the other hand, RSY in the second year was higher than that in the first year by 33.04% in the spring sowing and 10.33% in the summer sowing. Given the significance of total intercepted radiation and environmental temperature at different growth stages for sugar yield (12,26,31), this difference may partially be attributed to the appropriate climatic conditions of 2017. In 2017, the average monthly temperature was 1.9°C, 0.8°C, and 2.4°C (1.69 °C on average) higher in March-April, April-May, and May-June than their counterparts in 2018, respectively ( Figure 1). Since the rate of sugar yield decline with the decrease in temperature is 10% per °C over this period (41), it can be inferred that the lower RSY and WSY in 2018 versus 2017 was related to the difference in the climatic conditions of these two years. The interaction of S × H showed that the second harvest in the spring sowing and the third harvest in the summer sowing had higher yields than the other harvests (Table 5). Overall, RSY and WSY were 35.5% and 33.8% lower in the summer sowing than in the spring sowing, respectively. As per one day delay in sowing, WSY was reduced by 2.24 g m −2 in 2017 and by 1.08 g m −2 in 2018, whilst Hoberg et al. (15) reported over 3 g m −2 decline of yield as per one day delay in sowing in the Netherlands. In Ireland, a month delay in the sowing of sugar beet (from early-March to early-April) resulted in about 1.1 t ha −1 decline of sugar yield (9). Similar findings in Turkey (4) revealed that when sowing was retarded by one week, sugar yield was decreased by about 8.3−9.7%. The different rates of sugar yield decline with the delay in sowing in different studies may be associated with the part of the growing season that the plants have missed. For example, the rate of sugar yield loss in April and May was estimated to be 4 and 15 g sugar per m 2 per day delay in sowing (42). In total, the decline of sugar yield by late sowing of sugar beets has been proven by Kirchhoff et al. (21). Quality of sugar beet extract: Na, K and amino-N concentrations of the extracted juice of sugar beet cultivars, which represent sugar quality, were influenced by sowing and harvest dates and their interaction ( Table 6).
The results revealed that the most important factor underpinning the components of juice quality was the harvest date as only the main effect of harvest date was significant on all qualitative traits in both years whereas the sowing date influenced Na content in the second year and K content in both years. The interaction of S × H showed that both in spring and summer sowing conditions, the first harvest had higher Na, K and amino-N content than the other harvests, but their concentrations were significantly higher in summer sowing than in spring sowing ( Table  6). The increase in K and N content of sugar beet roots in summer sowing has been reported by Lauer (23), Larney et al. (22), and Webster et al. (47), too. This has been attributed to the increase in K uptake compared to the internal consumption of the element in delayed sowing (5). On the hand, the interaction of S × G for the qualitative traits revealed that although some cultivars did not exhibit any statistically significant differences between spring and summer sowing, Na content in summer sowing was lower than that of the normal sowing for all cultivars except for G2 in both years whereas G1, G4, and G5 in the summer sowing and G2, G3, and G6 in the normal sowing produced higher K and amino-N. In addition, the interaction of G × H for the qualitative traits indicated that all sugar beet cultivars in the first harvest showed higher Na content in 2017, higher K content in 2018 and higher amino-N in both years than the second 243 and third harvests. Furthermore, the comparison of the main impact of the harvest time indicated that in all qualitative traits, the first harvest had the highest mean value and the third harvest had the lowest mean value.
In fact, as harvest time was postponed, K and amino-N content of the roots were decreased (Table 6). Sugar beet root quality is influenced by many factors such as genotype, organic matter, soil nutrients, and agronomic operations like N application (25). The highest concentrations of such elements as N, Na, and K in roots reduce sugar content.  Among the cultivars, 'IR7', which had the highest sugar content, had the lowest N content. This inverse relationship between sugar content and root K content has been reported by other researchers too (10,15,22,32). That the extension of growth period reduces these elements is caused by the full assimilation of nutrients and their mobilization to the shoots and crowns, but when growth period is shortened, there is not adequate time for assimilation and mobilization of nutrients absorbed from the soil (35).

Water use efficiency (WUE)
The amount of water use was almost equal in the two experimental years (11265 and 11017 m 3 ha −1 in 2017 and 2018, respectively), but since RSY was higher in the first year (6.74 t ha −1 ) than in the second year (6.08 t ha −1 ), the first year had higher WUE than the second year (805 g m −3 versus 607 g m −3 ; Table 6). Also, the interaction of S × G × H was significant for WUE in both years (Table 6). Figures 9 and 10 show that the sugar beet cultivars (except for G2 and G5 in the second year) had significantly higher WUE in the spring sowing than in the summer sowing. In addition, the second harvest was related to the highest WUE in G2, G3, and G4 in the spring sowing and in G1, G3, and G5 in the summer sowing, whereas no significant difference was observed between the second and third harvests in G5 and G6 in the spring sowing and G2 and G4 in the summer sowing ( Figure  11) whereas, in the second year, the second and third harvests had higher WUE than the first harvest for all cultivars although there was not a statistically significant difference between the second and third harvests for some cultivars. 2018 S1 S1 S1 S2 S2 S2 245 the summer-sown G6 plants at the third harvest (806 g m −3 in the first year and 991 g m −3 in the second year). In addition, among the studied cultivars, G2 was the most stable (in terms of WUE difference between 2017 and 2018) and G6 was the least stable ( Figure  8). *, ** and ns are significant at levels of 5%, 1% and non-significant, respectively. In each column, mean values followed by the same letter(s) are not statistically different based on the LSD test at P < 0.05.
The most important environmental factor is the water supply, which is affected by soil type, climatic conditions, and the actual rate of precipitation. Limited water supply influences leaf and root growth dramatically and leaf senescence is accelerated and its regeneration is limited with the increase in high temperatures and the failure to supply adequate water (3,7,45). Since Iran is located in a semi-arid region of the world, WUE is an important factor for sugar beet cultivar selection. On the other hand, a goal of the attempts to find the best sowing and harvest dates is to achieve the maximum WUE. CONCLUSIONS 1. The results indicated that in the spring and summer sowing, most cultivars exhibited maximum CAT and MDA in 1700−1900 GDD, so sound management of plants in this period can reduce yield loss by heat stress.
2. Sugar accumulation in the summer sowing was 23% lower than that in the spring sowing. However, various cultivars responded to the changes in sowing date differently, and the cultivars 'Paya' and 'IR7' could preserve 85% of their yields.
3. If it is assumed that the condition for the summer sowing is the enhancement of WUE, the results showed that the summer sowing had no advantage over the spring sowing when considering the further decline of crop yield in the summer sowing vis-à-vis the amount of water use. 4. Nonetheless, if summer-sown sugar beets in rotation with grains are considered a cash crop and there is no limitation on supplying their water requirement, the summer sowing can then be recommended to farmers to earn good profits.