Drip fertigation could improve source-sink relationship of aerobic rice ( Oryza sativa L . )

The present study investigated variations among different water and fertilizer levels of aerobic rice. The data collected to assess the source-sink relationship of aerobic rice were leaf number, individual leaf size, soluble protein content and leaf area duration expressed as source strength and moreover sink strength expressed using spikelet number, thousand grain weight, grain filling duration and grain filling rate. The higher sink strength and optimum source strength was recorded in 125% pan evaporation (PE) with 100% recommended doses of fertilizers (RDF) treatment. Besides, higher sink-source ratio and lower source limitation was a reason for the balanced source and sink strength observed and also recorded higher grain yield. The treatments receiving 100% PE and 75% RDF level showed considerable reduction in the grain yield through source sink related characters.


INTRODUCTION
Rice is an important food crop for a large proportion of the world's population.It is staple food in the diet of the population of Asia.Rice provides 35 to 60% of the dietary calories consumed by nearly more than 3 billion people (Fageria et al., 2003).Globally, it is also the second most cultivated cereal after wheat.Unlike wheat, 95% of the world's rice is grown in less developed nations, primarily in Asia, Africa and Latin America.China and India are the largest rice producing and consuming countries in the world (Fageria, 2007).By the year 2025, it is estimated that it will be necessary to produce about 60% more rice than what is currently produced to meet the food needs of a growing world population.In addition, the land available for crop production is decreasing steadily due to urban growth and land degradation.Hence, increases in rice production will have to come from the same or an even less amount of land.This means appropriate rice production practices should be adopted to improve rice yield per unit area (Fageria, 2007).
In a previous study, we suggested that aerobic culture could save water without any yield penalty compared with flooded culture, but with some risks of poor performance (Kato et al., 2009).High crop yields are determined by ability of plants to produce high levels of photoassimilate and/or to partition large proportions of carbohydrate efficiently into harvested organs (Faville et al., 1999).Assimilate producing plant parts such as leaves are known as the source, and plant parts to which assimilate is translocated to grains are known as the sink (Fageria et al., 2007).In crop plants, physiological basis of dry matter production is dependent on source-sink concept, where source is a potential capacity for photosynthesis and sink is a potential capacity to utilize photosynthetic products.If sink is small and large, the yield cannot be high if source capacity is limited.A developing leaf is a *Corresponding author.E-mail: vanithacrp@gmail.comsink, but when fully grown it becomes a potential source.Sink size has a potential capacity for maximum production of a crop (Venkateswarlu and Visperas, 1987).Source is the first organ to respond to management practices.Therefore, source may suffer if stress affects its balance with sink (Venkateswarlu and Visperas, 1987).Source is more sensitive to cultural, nutritional and climatic factors.Nutritional and cultural practices often stimulate source thus making it more responsive.Since several parameters concerning the physical frame of the plants are almost exhaustively studied, it is likely that the manipulation of functional traits might enhance the yield plateau especially under low-yielding aerobic environment.Realizing this aspect, attempts are made to elucidate the 'source' (leaf) and 'sink' (panicle) interrelationships to improve the functional efficiency of the rice plants intended for aerobic environment.

MATERIALS AND METHODS
Field experiment was conducted in Wetland, Tamil Nadu Agricultural University, Coimbatore, India during dry season (2007) (11°N, 77°E).The experimental plots were dry-ploughed and harrowed.Raised flat beds were formed and laid out with double channels around all the plots.Before sowing, the wet seeds were treated with the Azophosmet biofertilizer at the rate of 200 g 10 kg -1 of seeds and sprouted, biofertilizer treated seeds were dry-sown by hand dibbling at 3 cm depth in rows 20 cm apart and covered with soil, in the field for all the treatments except the conventional practice (T1) at seeding rates of 30 kg ha -1 .A pre-emergence herbicide of pendimethalin at 1.25 kg a.i.ha -1 was applied 3 days after the first irrigation and later hand weeding was taken at 35 days after sowing for maintaining weed free environment.

Fertigation
The fertigation schedule indicating that the nutrient requirement at different pheonological stages and quantity of nutrients to be applied for 75, 100 and 125% recommended dose of NPK (150:50:50 kg ha -1 ) respectively (Figure 1a to c).All the three fertilizers viz., nitrogen, phosphorus and potassium were supplied through fertigation in the form of water soluble fertilizers as per the drip fertigation treatments once in a week.In the case of conventional method (T1), entire dose of P was applied basally before sowing.In the case of N, the recommended dose was given in four equal splits at basal, tillering, panicle initiation and first flowering; while, K was given in two equal splits at basal and panicle initiation stages.Recommended doses of FeSO4 (50 kg ha - 1 ) and ZnSO4 (25 kg ha -1 ) were applied as the basal dressings before sowing in all the ten treatments.
Depth or volume of irrigation water was measured using the discharge of the delivery hose connected in the pump, time of irrigation, and surface area of the plot.Calibration of water discharge from the delivery hose was done by measuring the discharged water using graduated cylinder at a certain time.It was done in a series of trials at different sections or length of delivery hose.With the given depth of irrigation, size of plot and average discharge of the delivery hose, the time of irrigation for every plot was computed.Drip fertigation treatments comprised of three water and fertilizer levels.
The following treatments were employed in the present study: T1: Surface irrigation given at 3 cm depth with IW/CPE ratio of 1.0 +

Source characteristics
Four hills from each replication and treatment were removed and the leaf was counted as a single unit.The number of leaves per hill was presented by calculating the mean of four hills.Leaf area duration (LAD) was determined by using the formula of Power et al. (1967) and the values expressed in days.The individual leaf size (ILS) was calculated by using the formula of ILS = Total leaf area (cm 2 hill -1 ) / Total leaf number hill -1 and expressed in cm 2 leaf -1 .Soluble protein was estimated from the rice leaves by the method of Lowry et al. (1951) and expressed as mg g -1 fresh weight.

Sink characteristics
For panicle growth modeling, at the time of flowering, about one hundred panicles in the each treatment in three replications were tagged (Jaiswal et al., 1982).The panicles were sampled periodically and their dry weights collected at the chosen time interval were plotted as a function of time from anthesis until the final harvest.Panicle data from the treatments were fitted with different panicle growth models as proposed by Thornley (1976).Models are exponential (negative), Gompertz, logistic, cubic, polynomial, quadratic and Richard's and a hybrid of Richard's-Quadratic model.Among them, the hybrid model of Richard's -Quadratic was found to be best fit.The Richard's -Quadratic (hybrid) equation used by Mohandass et al. (1988) was Wt = Wf [1 + e θ(t) ] -1/n .Where, Wt and Wf were panicle dry weight at time 't' and at maturity respectively, and θ(t) is the quadratic polynomial in 't'.The value of 'n' in this model was found to be 0.1.From these models and functions (Figure 2 and Tables 1  and 2), observations were made of the dates on which 90% of the final panicle weights were attained and the relative duration of the grain filling duration (GFD) of each treatment was calculated as the length of the time interval from zero flowering (anthesis) to the predicted date of 90% of final grain yield (Jones et al., 1979).The dry matter accumulated by the panicle at 90% of the final panicle weight was divided by the GFD to arrive at the grain filling rate (GFR).Sink capacity was calculated as the product of spikelets/ m 2 and test weight (Yoshida, 1981).Sink-limitation was worked out as spikelet number x 1000 grain weight / Leaf Area at flowering (Rao and Murty, 1976).Approximate source-limitation (Sa) was calculated by taking into account the changes in dry weight of panicle (ΔY) and the biomass (ΔB) due to variation in treatments as: Sa = (ΔY) / (ΔB) (Gifford et al., 1973).Grain yield adjusted to 14% moisture content was obtained from the whole area of the plot.

Statistical analysis
The data collected were subjected to statistical analyses in the randomized block design using ANOVA (AGRES version 7.01) following the method of Gomez and Gomez (1984).

Source characteristics
Leaves are the dominant primary source for producing assimilates in crop plants.Photosynthetic activities of plant parts other than leaf blades are very small for rice.The source strength is the major factor influencing the sink strength (Yoshida, 1981).Source strength has got two components, viz., source size and its activity.The source activity is further split into two sub-components that is, photosynthetic efficiency (in terms of soluble protein content, is an indirect measurement of photosynthetic capacity) and longevity of source (LAD).It is noticed that the source size (in terms of both leaf number and ILS at flowering stage) increased with increase in water availability (Table 3).Interestingly, the reduction in soluble protein content with lower water availability (100% PE) was found to be 12.7 and 19.5% in comparison with 125 and 150% PE level of drip irrigation respectively.Similar trend was evident with LAD also with greater magnitude of reduction in the case of lower water availability situation.These trends implied that under water stressed environment at 100% PE level, fair crop productivity could be achieved when optimum leaf area was maintained with greater and sustained photosynthetic efficiency.This association appeared valid only when proper efficient varieties and good agronomic management practices (such as the fertigation at optimum RDF level) were ensured for the stressed situations.

Sink characteristics
Sink strength consists of two major components, viz., sink size and sink activity (Wareing and Patrick, 1975).The sink activity is again sub-divided into two components and they are the GFD and GFR (Pearson and Hall, 1984).The duration of the reproductive growth period is very important because number of grains for plants are determining during this period.It has been reported that numbers of spikelets per ear in cereal crops can be increased by increasing length of the reproductive growth period (Yoshida, 1972).The data are furnished in Table 4, which indicated that the sink capacity decreased (-19.1%)under 100% PE compared to 125% PE level of drip irrigation.But, the values showed increasing trend with increase in fertilizer levels (75% RDF: 682.84 to 125% RDF: 749.0 gm -2 ).Nevertheless, drastic decline in the sink capacity could be compensated when supplemented with the fertigation.The sink size, that is, the spikelet number per unit area (Yoshida, 1981) was also reduced with deficit and excess water availability situations (-13.2 and -5.7% in 100 and 150% PE levels respectively as against 125% PE) but compensated fairly well with fertigation practice.Grain filling, a crucial determinant of grain yield in rice crop, is characterized by duration and rate of filling (Yang et al., 2008).Grainfilling period is the duration from anthesis to physiological maturity and beyond.The GFD showed variations with the values ranging from 25.4 (100% PE) to 27.1 (150% PE) for water treatments.In the case of fertilizer levels, the GFD increased with increase in RDF levels.
However, the reduction in GFR was found to be phenomenal with the lesser water availability treatment (100% PE: -34.4%) in comparison with the moderate water supply (125% PE).Influence of fertigation was phenomenal in safely narrowing down such reduction in GFR values.This signified that between the two components of sink activity, the GFR played a vital role for panicle weight and grain yield increase especially in less water applied + fertigation practice.This was in corroboration with the findings of Mohandass et al. (1988) showing that the GFR possessed significant positive association (r = 0.953**) but the duration exhibited poor correlation (r = 0.278) with the grain yield under lowlight stress environment.Studies on the contribution of GFD and GFR to grain yield in rice on a panicle basis further showed that the grain yield of different genotypes was mainly determined by GFR (Cho et al., 1987;Jones et al., 1979).

Source-sink limitation
An attempt has also been made to explain yield variations in the chosen drip fertigation treatments in terms of source-and / or sink-limitation.The results of the Table 5 indicated that though the sink-source ratio was higher with moderate level of water (125% PE) and fertilizer (100% RDF) supply, the differences due to drip fertigation treatments were less appreciable.
On the contrary, the Sa values increased steadily from 100% (0.602) to 150% (0.985) PE level of drip irrigation.In the case of fertilizer levels, higher Sa (0.910) was observed at 125% RDF level.Thus, the values of Sa was centering around 1 especially for the medium and higher water supplying (125 and 150% PE levels) treatments, which in the scale of 0 -1 of Gifford et al. (1973) explained that all additional assimilates produced was fully used up by the developing grains, and thus the grain growth was entirely source-limited in these water treatments.However, no solid conclusion could be arrived at in respect of the fertilizer levels tried, meaning that detailed studies are required in future to unravel the mystery.
Contribution to grain yield in cereal crops had conventionally been assessed using yield per plant and various yield attributes, consequently ignoring the function of other organs such as ear awns and flag leaf.These plant parts considerably affect grain yield and its attributes during grain filling stage (Ahmed et al., 2004; Khaliq et al., 2008).The grain yield of rice is often influenced by sink capacity rather than source strength under stress-free environment (Fukai et al., 1991).Higher grain yield of 5643 kg ha -1 was registered with the drip fertigation practice at 125% PE + 100% RDF level.Similar observation of yield reduction was noticed in both the low water as well as fertilizer supplying treatments.The reduction in grain yield due to deficit water supply was more (27.2%) in 100% PE as compared to 125% PE level respectively (Figure 3).

Conclusion
Summarizing the influence of drip fertigation system on sink characteristics in aerobic rice, it could be inferred that the yield variations observed under varied water as well as fertigation levels could be mainly due to the alterations in sink capacity and its activity.Again, major component of sink activity, viz., grain filling rate was greatly influenced by water and fertilizer levels.This indicated that further crop improvement and management strategies intended for stabilizing the yield under aerobic environment should be aimed at stabilizing the sink capacity and grain filling rate.These findings also suggested that in this source-limited rice crop, future strategies should be aimed at developing efficient plant type and management options ensuring the capability of not only synthesizing more biomass under aerobic environment but also to partition it more towards 'sink', that is, the developing grains.The performance of the aerobic rice grown with the drip fertigation practice scheduled at 125% PE with 100% RDF level was found to be superior for most of the source sink characters and grain yield.Our results confirm the feasibility of growing aerobic rice under drip fertigation system.

Figure 1 .
Figure 1.Details of fertilizers applied through fertigation.

Table 2 .
Grain filling duration (GFD) and its rate (GFR) as influenced by drip fertigation treatments in aerobic rice.

Table 3 .
Components of source strength due to drip fertigation in aerobic rice.

Table 4 .
Components of sink strength due to drip fertigation in aerobic rice.

Table 5 .
Variations in source-sink limitation due to drip fertigation in aerobic rice.
Figure 3. Grain yield as influenced by drip fertigation treatments in aerobic rice.