Effect of High Temperature Stress and Supplemental Irrigation on Flower and Pod Formation in Two Canola (Brassica napus L.) Cultivars at Mediterranean Climate

MATERIALS AND METHODS

Study was conducted at the Agricultural Research Station of Gonbad (45 m a.s.l., 37 N, 55 E), Golestan province, Iran during 2005-2006 and 2006-2007, under irrigated and rainfed conditions. P and K were applied as basal to supply 50 P₂O₅ kg ha^-1 and 50 K₂O kg ha^-1 using triple super phosphate and potassium sulfate, respectively. N was applied at 75 kg ha^-1 (as urea) with a third applied as basal, a third used to side-dressed at the beginning of stem elongation and the rest applied at the beginning of flowering. The experiment was designed as a randomized complete block design arranged in split plot with 3 replications. Two cultivars of spring canola (namely, Hyola401 and RGS003) were allocated as subplots and five different sowing dates were used as main plots. The sowing dates were 9 Nov., 6 Dec., 5 Jan., 4 Feb. and 6 Mar. in 2005-2006 and 6 Nov., 6 Dec., 5 Jan.,

Fig. 1:	Maximum and minimum daily temperatures (a and b) and daily rainfall (c and d) during the experiment

4 Feb. and 6 Mar in 2006-2007. The best sowing date for canola in the area is 6-21 Nov., but we planted canola in the late sowing dates to synchronize the reproductive stages of the crop with heat and drought stress periods.

Plots were over planted and after seedling establishment, the plants were thinned to the desired spacing between plants of 5 cm (1000000 plant ha^-1). Each subplot consisted of eight rows. Main plots and subplots were 2 and 0.4 m apart, respectively. A 3 m pathway separated replicates. The plots were hand weeded during the season. Two days after beginning of flowering, all of the flowers and buds (each main and branch racemes) were counted and recorded from 10 plants selected at random from within each plot. The plants were tagged and numbered and at the physiological maturity the tagged plants were removed from the plot and the number of pods (complete pods) and incomplete pods (the pods without seed or with abnormal figures) were counted (each main and branch racemes, separately). In each plot, samples during growth season were taken from 10 plants of rows of 2 and 3. At harvested time (2 days after physiological maturity) rows 5, 6 and 7 were cut by hand and air dried. Then the plants were threshed with a stationary combine to determine seed yield. Phonological observations were made on a regular basis with the Harper and Berkenkamp (1975) growth stage key. Maximum and minimum temperatures and rainfall measured at a near-by weather station (Fig. 1). Mean temperatures was calculated for the same growth phases as the sum of the daily temperatures divided by the number of days during the growth phase. Under supplemental irrigation conditions, plots were irrigated at the beginning of stem elongation, flowering and seed filling stages. Two days before irrigation times, soil samples were dried for 24 h at 105C and weighed. Then soil water content was measured and plots were reached to field capacity with irrigation (Zhang et al., 1999). All obtained data were tested by the analysis of variance using SAS (1989) and mean grouping was done according to the LSD (p = 0.05).

RESULTS AND DISCUSSION

Seed yield of canola is determined by the number of flowers per unit land area which produced pods (Morrison and Stewart, 2002). In this study, the combined analysis revealed statistical significant effect of year on flower and pod parameters (data not shown); therefore we analyzed each year data, separately (Table 1). In each year, the number of flowers, pods and pods/flowers ratio were significantly affected by irrigation, sowing dates and cultivars (Table 1). Seed yield was positively related to pods per plant in the two cultivars with 38 and 64% of the variation for Hyola401 and 71 and 84% of that for RGS003 under irrigated and rainfed conditions, respectively (Fig. 2). With increase in number of flowers per plant, the number of pods increased. Pod number produced per plant is mainly set at flowering period, when the number of flowers and pods are determined. The relationship between the number of flowers and pods in the main

Table 1:	Means of some traits of canola during the experiment
Means followed by the same letter(s) within each column are not significantly different according to the LSD (p = 0.05). MR: Main Raceme, BR: Branch Racemes

Fig. 2:	Relationship between seed yield and the number of pods per plant, (a) Hyola401 at irrigated conditions, (b) Hyola401 at rainfed conditions, (c) RGS003 at irrigated conditions and (d) RGS003 at rainfed condition

sraceme was positively strong with 88 and 72% of the variation for Hyola401 and 88 and 78% of that for RGS003 under irrigated and rainfed conditions, respectively (Fig. 3). This suggests that increase in flower number, particularly in the main raceme under supplemental irrigation, could be an important factor that contributes to increase in pod number in canola.

In the main raceme, the ratio of pods/flowers for Hyola401 and RGS003 was 57 and 52% in 2005-2006 and 57 and 49% in 2006-2007, respectively (Table 1). In branch racemes, the ratio of pods/flowers for Hyola401 and RGS003 was 52 and 39% in 2005-2006 and 41 and 32% in 2006-2007, respectively (Table 1). In each year, the number of flowers that were converted into pods in the main raceme was more than that of branch racemes. McGregor (1981) reported that less than 45% of the flowers formed on B. napus cultivars produced pods. Canola usually produces nearly twice as many flowers as pods. Tayo and Morgan (1975) and McGregor (1981) reported that under typical conditions, canola differentiates more floral primordial than its photosynthetic capacity can support.

The amount of vegetative growth and the weather conditions (temperature, rainfall and radiation) during flowering period determines the formation of buds and flowers and hence the final pod number. In this study, other factors such as days to flowering and temperature during flowering period influenced flower and pod production of canola. There was a positive relationship between the number of flowers per plant with the number of days to flowering (Fig. 4) with 38 and 57% of the variation for Hyola401 and RGS003, respectively. The relationship between the numbers of pods per plant with days to flowering was also very strong with 81 and 76% of the variation for Hyola401 and RGS003, respectively (Fig. 4). This was in agreement with the result of Thurling (1974). There was a strong negative relationship between pod number and the ratio of pods/flowers with mean temperature during the period (Fig. 5) with 84 and 73% of the variation for Hyola401 and 71 and 61% of that for RGS003, respectively. Since optimal temperatures during flowering for B. napus is 20C (Angadi et al., 1999), therefore high temperatures particularly in late sowing dates (Table 2) accelerated the rate of plant development, reduced the length of the flowering period and the number of flowers and pods and thereby seed yield potential as shown by Entz and Flower (1991).

In each year, sowing date significantly affected the number of flowers and pods per plant and the ratio of pods/flowers (Table 1). With delay in sowing, flowers and pods per main raceme decreased, ranging from 53 to 12 flowers and 34 to 2 pods over the two years. In 2006-2007, the number of flowers per branch racemes increased in

Fig. 3:	Relationship between the number of flowers and pods in main raceme (MR). (a) Hyola401 at irrigated conditions, (b) Hyola401 at rainfed conditions, (c) RGS003 at irrigated conditions and (d) RGS003 at rainfed conditions

Fig. 4:	Relationships between days to flowering with the number of flowers per plant (a) Hyola401 and (b) RGS003 and the number of pods per plant (c) Hyola401 and (d) RGS003 for the cultivars

Fig. 5:	Relationships between pod number per plant (a) Hyola401 and (b) RGS003 and the ratio of pods/flowers per plant (c) Hyola401 and (d) RGS003 with mean temperature during flowering period for the cultivars

Table 2:	The mean max. temperature (C) during flowering and seed filling periods

sowing date 5 compared to others, due to good weather conditions and high rainfall in March. (Table 1, Fig. 1). However, the number of flowers in branch racemes that could produce pod decreased significantly due to further severe water and heat stress. Therefore the number of pods per branch racemes in sowing date 5 was the same as sowing date 4 due to severe low ratio of pods/flowers in sowing date 5. Seed yield of canola was found to be closely related to temperature during flowering and early seed development but rainfall from early seed development through seed filling is essential to achieving full yield potential (Johnston et al., 2002). In this experiment, late sowing synchronized critical flowering period with heat stress (Table 2) and thus reduced the duration, flower, pod number and seed yield (Table 1). In North Dakota, the reduction in canola yield with delay in sowing was attributed to fewer pods per plant, leading to a lower harvest index (Johnson et al., 1995). In Gonbad area, with a typical terminal heat and drought stress particularly in late sowing dates (Table 2), heat stress during flowering of canola can prematurely end flowering, resulting in limited flower and pod formation per unit area. However the reproductive period (flowering and seed filling periods) of the early-sown canola is longer than that of the late-sown crop, therefore, contributing to a higher flower and pod number. The chance that rainfed early-sown canola will face heat and drought stress during flowering and seed filling periods was not as high as they would be if the crop was sown late.

In each year, with delay in sowing, the ratio of pods/flowers decreased (Table 1) and that was due to increase in terminal heat stress, though the rate of decrease was not the same. This was confirmed by the strong positive relationship between incomplete pods and flowers in the main raceme in sowing date 5, which accounted for 84 and 89% of the variation for Hyola401 and RGS003, respectively (Fig. 6). There was not any significant relationship between incomplete pods and

Fig. 6:	Relationship between flower number per plant and incomplete pods in Main Raceme (MR) at sowing date 5 (SD5) for the cultivars

flowers in the main raceme in early sowing dates. The rate of decrease in the ratio of pods/flowers in 2006-2007 was more than that of 2005-2006 ranging from 65-34% in main raceme, 47-34% in branch racemes and 52-34% in total plant in 2005-2006 and 69-13% in main raceme, 49-18% in branch racemes and 55-17% in total plant in 2006-2007 (Table 1). With delay in sowing, the rate of decrease in the ratio of pods/flowers in the main raceme was more than that of branch racemes due to decrease or/even increase in flowers per plant in branch racemes compared to the main raceme.

In 2005-2006, supplemental irrigation was significantly increased the number of flowers and the ratio of pods/flowers (both main and branch racemes), compared to rainfed conditions, but there was no significant difference between flower number and pods/flowers ratio in 2006-2007 (Table 1). Under supplemental irrigation conditions, the ratio of pods/flowers in the main raceme, branch racemes and total plant was 58, 42 and 48% in 2005-2006 and 53, 32 and 37% in 2006-2007, respectively (Table 1). Under rainfed conditions, the ratio of pods/flowers in the main raceme, branch racemes and total plant was 51, 39 and 43% in 2005-2006 and 49, 32 and 37% in 2006-2007 (Table 1). It appeared that the benefit of supplemental irrigation in 2005-2006 was more than that of 2006-2007. The cumulative rainfall from October to June in 2005-2006 and 2006-2007 was 393 and 448 mm, respectively (Fig. 1) and as such, supplemental irrigation amount was higher in the first year of the experiment (data not shown). When the amount of rainfall from March to May between the two years were compared and, flowering and seed filling periods of canola in the area with normal sowing dates, the difference was higher. The cumulative rainfall from March to May in 2005-2006 and 2006-2007 was 98 and 229 mm, respectively (Fig. 1). Therefore water stress was more at the time of flowering and seed filling periods in 2005-2006 than 2006-2007, providing a better opportunity for supplemental irrigation to influence canola flower number and pods/flower ratio in 2005-2006 (Table 1). This was confirmed with the result of seed yield in the experiment. However seed yield of canola under irrigation conditions, sowing dates and cultivars in 2005-2006 (2331 kg ha^-1) was more than that of in 2006-2007 (2110 kg ha^-1), due to better weather conditions and distribution of rainfall during crop growth season in 2005-2006 (Table 1, Fig. 1).

The mean annual rainfall of Gonbad area is about 450 mm. Because approximately there is no enough rainfall in April till May, coinciding with the seed filling period of canola in the area, occurrence of drought and crop water stress is common. The heat (Table 2) and drought stress that normally occurs in the area significantly lowers the seed yield of canola. Increased production, as a result, depends to a greater extent on further improvement of soil water availability. Therefore drought effects can be minimized by management practices such as supplemental irrigation (Oweis et al., 2004; Fan et al., 2005). Brandt and McGregor (1997) reported that for each millimeter increase in rainfall, there was a corresponding seed yield increase. The number of flowers (each main and branch racemes) in the two cultivars over years, sowing dates and irrigation conditions were approximately the same, but in both years, the ratio of pods/flowers of Hyola401 was more than that of RGS003 (Table 1). The mean 2 year of the ratio of pods/flowers for Hyola401 and RGS003 was 57 and 50% in main raceme, 42 and 35% in branch racemes and 47 and 40% in total plant, respectively (Table 1).

CONCLUSION

Pod formation and seed yield of canola in semiarid environments can be increased by minimizing the crop exposure to high temperatures and water stress that often occurs during the growing season. In this study, earlier sowing more closely matched incoming rainfall and reduced terminal temperature stress that increased the number of flowers and pods. Delay in sowing led to more rapid development of the crop, decreased days to flowering, flowering period and flower and pod number per plant. Sowing date and supplemental irrigation were very important management tools in minimizing the negative impact of high temperature and water stress during the reproductive stage in canola. In this study, a great proportion of the variation in flower, pod formation and seed yield in canola were related to environmental conditions during the critical period of flowering. Supplemental irrigation was a viable option that could be used to increase and stabilize canola pod number and seed yield.