Abstract: A field experiment was conducted during 2001-2003 to evaluate the effect of water stress on the yield and yield components of four rice cultivars commonly grown in Mazandaran province, Iran. In northern Iran irrigated lowland rice usually experiences water deficit during the growing season include of land preparation time, planting, tillering stage, flowering and grain filing period. Recently drought affected 20 of 28 provinces in Iran; with the southeastern, central and eastern parts of the country being most severely affected. The local and improved cultivars used were Tarom, Khazar, Fajr and Nemat. The different water stress conditions were water stress during vegetative, flowering and grain filling stages and well watered was the control. Water stress at vegetative stage significantly reduced plant height of all cultivars. Water stress at flowering stage had a greater grain yield reduction than water stress at other times. The reduction of grain yield largely resulted from the reduction in fertile panicle and filled grain percentage. Water deficit during vegetative, flowering and grain filling stages reduced mean grain yield by 21, 50 and 21% on average in comparison to control respectively. The yield advantage of two semidwarf varieties, Fajr and Nemat, were not maintained under drought stress. Total biomass, harvest index, plant height, filled grain, unfilled grain and 1000 grain weight were reduced under water stress in all cultivars. Water stress at vegetative stage effectively reduced total biomass due to decrease of photosynthesis rate and dry matter accumulation.
INTRODUCTION
Rice (Oryza sativa L.) is the staple food for more than two-third of the world`s population (Dowling et al., 1998). Stable and high yields of rainfed lowland rice are important for food security in many of the subsistence farming systems in Asia (Cooper, 1999). About 7.5% of total rice production comes from irrigated lowland production (Bouman and Toung, 2001). It has been estimated that more than 200 million tons of rice are lost every year due to environmental stresses, diseases and insect pests (Herdt, 1991; Chen and Murata, 2002). Drought stress is a major constraint for about 50% of the world production area of rice. Yield losses from drought in lowland rice can occur when soil water contents drop below saturation (Bouman and Toung, 2001). Rice crops are susceptible to drought, which causes large yield losses in many Asian countries (Bouman and Toung, 2001; Pantuwan et al., 2002a), however, some genotypes are more drought resistance than others, out-yielding those exposed to the same degree of water stress. The development of drought resistant cultivars may be assisted if mechanisms of drought resistance are known. Iran is a semi-arid country with 240 mm annual precipitation and 600,000 ha paddy field area; all of them are irrigated lowland rice with continuously submerge irrigation regime by keeping 3-5 cm standing water all of the growing period. Rice is the second main consumed in Iran, with wheat being the main staple for most of the population and Capital rice consumption per year is 36-38 kg. Two northern province near the Caspian Sea bank with annual precipitation of 700-1000 mm covering 70-80% of paddy filed cultivation area in Iran. Unfortunately the main parts of rainfall are out of rice cultivation season. Irrigation dominates water use in Iran and surface water storage has been increased by construction of numerous multi-purpose dams and reservoirs along rivers flowing from the Zagros and Elburz mountains. In northern Iran irrigated lowland rice usually experiences water deficit during the growing season include of land preparation time, planting, tillering stage, flowering and grain filing period. Recently drought affected 20 of 28 provinces in Iran, with the southeastern, central and eastern parts of the country being most severely affected (Reynolds, 2001). Iranian rice cultivars consist of local and improved variety so that, local variety are low yield (less than 4 t ha-1 based on paddy), tall plant, high quality, generally short-term maturity (less than 90 days from transplanting to harvesting) and marketable with high price; meanwhile, improved variety are high yield (more than 6 t ha-1 based on paddy), short plant height, mid and long-term maturity (110-130 days), medium quality and cheaper than local variety. There is three methods for variety improvement in Iran include of: Selection; Introduction from IRRI and breeding techniques. Water stress may occur at different growth stages and be of varying duration and intensities, thereby affecting growth and yield. Different reports showed that rice grain yield affected by water stress. If water stress occurs at tillering stage, caused to reduce number of reproductive tiller and panicle per hill (Wopereis et al., 1996). However some experiment showed that water stress between panicle initiation and flowering stage to cause reduce number of grain per panicle and when water stress event at flowering and early grain filling period, grains sterility and panicles fertility will be reduce (Boonjung and Fukai, 1993; Garrity and O` Toole, 1994). Water stress after the flowering stage caused to reduce grain weight (Bouman and Tuong, 2001). However, it`s showed that different varieties performance vary response to water stress, some of them are susceptible at vegetative stage and other at flowering and grain filling period (Pantuwan et al., 2002a). There is very little information on drought resistance of rice genotypes in Iran used in lowland production. This study was carried out to describe the differences in yield and yield component of commonly grown rice cultivars when crops were stressed at different growth stages.
MATERIALS AND METHODS
A field experiment was carried out in Rice Research Institute of Iran-Deputy of Mazandaran Province (Amol city) located in the north of Iran (52° 22` N, 36° 28` E and 28 masl) during 2001-2003 at the rice cultivation season with mean temperature 16.4, 22.9, 25.9, 27.3, 25.4°C from April to September, respectively. The research field had a loamy-silty soil and moderate climate (Table 1).
This experiment was laid out to evaluate varietal performance of four rice cultivars in terms of yield and yield components as affected by water stress. The experiment was designed as a split-plot, factorial in a randomized complete block design with three replications. Main-plots were four water stress regimes (water stress in vegetative stage, water stress in flowering stage, water stress in grain filling stage and control or no water stress). Subplots were 3x4 m with 0.5 m and main plots with 1 m apart. Aimed at water stress is shortage irrigation up to 20 days for reducing saturate water so that narrow crack will appear in the soil surface. The control was irrigated as required to ensure and keep a 2-5 mm level of standing water throughout crop growth. Plants in water stress treatments were grown under favorable water conditions with supplementary surface irrigation throughout the crop cycle while irrigation was interrupted to induce drought stress at around vegetative, flowering and grain filling stages. Sub-plots were four contrasting cultivars, Tarom, Khazar, Fajr and Nemat (Table 2). A mixed commercial fertilizer was applied at the rate of 92 kg N ha-1,44 kg P ha-1 and 83 kg K ha-1; all of phosphorous, potassium and half of nitrogen fertilizer applied at basal and other 50% nitrogen fertilizer has applied as a top dressing at panicle initiation. Making nursery and management for raising seedling have done with lowland traditional wet nursery. Seedlings 30-35 days old (4-5 leaves stage) were used for transplanting and three seedlings were transplanted to each hill, spaced at 25x25 cm. Since, growth duration of cultivars were different together, for synchronously of cultivars flowering, Nemat transplanting date (as a long-term variety) was 10 days earlier than others (April-25) and Tarom transplanting date was later than others as a short term variety (May-5). Plant height was measured on 10 randomly selected hills by measuring the distance from
| Table 1: | Soil characteristic of experimental field and weather condition |
| Table 2: | Description of cultivars under experiment |
the soil surface to the tip of the highest panicle within each hill. Grain yield was determined from a harvest area of 4 m2 (64 hills) adjusting to 14% moisture content and yields refer to rough grain yield. All plants from the harvested area were dried at 70 °C for total dry matter determination and harvest index was calculated as grain yield per total dry matter. Panicle number per m2 was determined at dough stage. To estimate the fertility of panicles, five randomly sampled panicles per plot were counted for filled and unfilled grains and the percentage of filled grains was calculated. Data were analyzed by Analysis of Variance (Proc Anova). The 2-year data underwent a repeated-measures data analysis by using a combined analysis of variance across years. All statistical tests were carried out using the Statistical Analysis System (SAS, 1997).
RESULTS AND DISCUSSION
Water deficit during vegetative, flowering and grain filling stages reduced mean grain yield by 21, 50 and 21% respectively in comparison to control. Grain yield differed among the four cultivars. In two years, Nemat had the largest grain yield in the control treatment. When drought stress was imposed at the vegetative stage, the grain yield of Nemat, Khazar and Fajr were significantly reduced. Nemat had the highest reduction (25%), while Tarom had only a slight reduction (14.5%). Bouman and Toung (2001) showed that different cultivars might have different responses to the same drought stress timing and intensity. Compared to the well-watered conditions, panicle number was only slightly reduced in Tarom (Table 3). Total biomass, harvest index, plant height, filled
| Table 3: | Grain yield and plant parameters of four rice cultivars grown under four water stress treatments in north of Iran |
| A: Common letter(s) within the column do not differ significantly at 5% level of significance analyzed by DMRT, B: W0 Control; W1: Water stress at vegetative stage; W2: Water stress at flowering stage; W3: Water stress at grain filling stage, C: Common letter(s) within the row do not differ significantly at 5% level of significance analyzed by DMRT | |
grain, unfilled grain and 1000 grain weight were reduced under water stress in all cultivars. Water stress at vegetative stage effectively reduced total biomass due to decrease of photosynthesis rate and dry matter accumulation (Table 3).
The combine analysis of variance showed that the effect of year on yield is significant and yield at second year is more than first year due to increase number of tillers and harvest index at second year (Table 4). There is significant interaction between water stress and kind of varieties (1% level-Duncan test) so that cultivar of Nemat had a highest yield and total biomass at control treatment and cultivars of Tarom and Khazar had a lowest yield at treatment of water stress at flowering stage (Table 5).
Rahman et al. (2002) reported that plant height, tiller number, panicle number, panicle length, number of filled grains per panicle, 1000-grain weight, harvest index (HI), total dry matter (TDM) and yield were decreased with stress. Grain yield was reduced dramatically in all cultivars with drought starting at panicle initiation or at flowering. Water stress at flowering reduced grain yield more than other water stress treatments. The reduction in yield largely resulted from the reduction in fertile panicle number and filled grain percentage. Fukai et al. (1999) reported that maintenance of leaf water potential just prior to flowering is associated with higher panicle water potential, reduced delay in flowering time and reduced spikelet sterility and hence contributes to higher yield. On the other hands, yield losses from the normal level due to water stress are useful in assessing drought resistance. The different variety has differently drought resistance mechanism, for a variety, which is also different, at the different stages. However, the complex drought resistance occurs at any growth stage in crop, with the different response and the different mechanism to drought tolerance (Na et al., 2007). Some researcher reported that grain yield could be drastically reduced (about 60%) if drought occurs during flowering time (Boonjung and Fukai, 1996). Pantuwan et al. (2002b) reported that drought stress that developed prior to flowering generally delayed the time of flowering of genotypes and the delay in flowering was negatively associated with grain yield, fertile panicle percentage and filled grain percentage. Genotypes with a longer delay in flowering time had extracted more water during the early drought period and as a consequence, had higher water deficits. They were consistently associated with a larger yield reduction under drought. Castillo et al. (1992) reported that draining rice fields at either vegetative or reproductive phases caused significant yield loss. Evaluating the effect of different durations of water stress at various growth stages showed that water stress at any stage would reduce yield (Salam et al., 2001; IRRI, 2002). However, the duration of these stresses was more closely related to yield reduction than to stage at which the stress occurred. Islam et al. (1994b) observed that yield losses resulting from water deficit are particularly severe when drought strikes at booting stage.
Plant height was significantly affected by water stress at booting, flowering and grain filling stage (Table 3) over the control. This result agrees with Islam et al. (1994b), who found that moisture stress reduced plant height under 20% soil saturation at booting and flowering stages. Similar result has also been reported by Islam (1999). The decrease in height might be either due to inhibition of length of cells or cell division by water deficits.
Water stress during vegetative stage reduced tiller number, while stress at the reproductive and grain-filling stage reduced grain number and weight. Rahman et al. (2002) also reported that the number of tillers per hill was decreased significantly under moisture stress at different growth stages except that at flowering stage. This agreed with Islam et al. (1994a). Bouman and Toung (2001) found that drought before or during tillering reduces the number of tillers and panicle per hill. When late season drought was the main cause of low yield, late-maturity cultivars (such as Nemat) were not suitable as panicle development was severely impaired.
| Table 4: | The combine analysis of variance on grain yield and plant parameters of different rice cultivars under water stress treatments |
| *Significant different at 0.05 level (Duncan test); **Significant different at 0.01 level (Duncan test); ns: Not significant | |
| Table 5: | Interaction effects between water stress treatments and rice cultivars on the grain yield and plant parameters |
| Common letter(s) within the column do not differ significantly at 5% level of significance analyzed by DMRT, W0: Control; W1: Water stress at vegetative stage; W2: Water stress at flowering stage; W3: Water stress at grain filling period | |
Total grain number per panicle was drastically reduced when drought stress occurred at flowering. This reflected the reduced crop growth due to drought during flowering. Rahman et al. (2002) and Islam et al. (1994a) also showed that the number of filled grains per panicle decreased significantly with the moisture stress at booting, flowering and grains filling stages compared with control. The proportion of unfilled grain in the drought stress at flowering stage was 46% compared with 22% in well-watered (control) conditions.
The 1000-grain weight in the drought stress at grain filling stage was 17% smaller than control. Thus, the yield reduction in drought stress at flowering stage mostly resulted from reduction in total grain number per panicle (increase in unfilled grain and a greatly decreased proportion of filled grain) and 1000-grain weight respectively. Similar results on 1000-grain weight under water stress at booting and flowering stages had been showed by Islam (1999) and Islam et al. (1994b). Stress during different growth stages might decrease translocation of assimilates to the grains, which lowered grain weight and increased the empty grains. HI values indicate the efficient translocation of assimilates towards sink. Lower HI values under stress at booting and flowering stages indicate that it was more harmful in translocation of assimilates towards the grains over grains filling stage (Rahman et al., 2002).
It has been argued that under severe drought stress, when yields are reduced to below 50% of those under favorable conditions the relationship between yield under favorable and stress conditions break down (Ceccarelli and Grando, 1991). Results of this experiment suggest that genotypes had no capability in expressing their genetic yield potential under these conditions. It appears that the yield advantage observed under favorable conditions of semi-dwarf cultivars (Fajr and Nemat) which required less assimilate for vegetative organs was not maintained under water-limiting conditions.
The results also suggest that Tarom is drought-tolerant and is able to retain green leaves longer than other cultivars under drought conditions. Retention of green leaves in seedlings under drought conditions has been used as a selection criterion for drought resistance (De Datta et al., 1988). Alternatively, cultivars with green leaf retention may process dehydration-tolerance mechanism, which allow the plants to maintain metabolic activity, despite low leaf water potential, for example, as a result of high osmotic adjustment (Fukai and Cooper, 1995). Other experiment also confirmed a positive relationship between green-leaf retention and grain yield among 35 lines (Henderson et al., 1995). From the above results it can be concluded that cultivars required for Iranian conditions, where frequent drought develops, are those with appropriate phenological development to escape late drought and an ability to maintain growth during drought that may develop late in the season. Consideration of these characters in plant-breeding programs should increase the efficiency of plant improvement in the region.
ACKNOWLEDGMENTS
This study was carried out within the Ph.D project funded by the Tarbiat Modares University and Rice Research Institute of Iran-Deputy of Mazandaran (Amol) that supervised by Dr. Z. Tahmasebi Sarvestani. We also gratefully thank the assistance of Dr. H. Pirdashti and M. Rahimi on data collection.