Introduction

Rice, a staple food of most of the Indian population, covers an area of 44.4 m ha with a 122 m t production1. In India, the most common method for cultivating rice is by transplanting seedlings from nursery to main field, but there has been a shift from conventional transplanting to directly sown rice (DSR) cultivation system, because of unavailability of labour, water shortage and mounting cultivation costs2. In conventional system of transplanted puddled rice (TPR), there are many hurdles like high water consumption (1000–2000 mm) for puddling and keeping field constantly flooded3; a high energy demanding (5630–8448 MJ ha−1)4 and over 15–20% elevated labour requirement5 over direct-seeded rice (DSR) resulting in production unit of low profit6,7, making it prohibitively expensive for various farming persons, particularly farmers of small and marginal land holdings in Southeast Asia8. Oppositely, DSR cultivation has emerged as a game changer with multiple advantages like labour saving of 40–45%9, water saving of 30–40%10, energy saving of 60–70%11, irrigation water saving of 20%12, faster and convenient planting with 7–10 days earlier maturity of rice crop13, yield advantage of 10–20%14, carbon use efficiency15, input efficient16 and higher economic returns with substantially higher benefit cost ratio of 69%17. Of the 161 M ha area under rice crop worldwide, area under direct-seeded rice constitutes 33 m ha18.

Optimization of sowing window for DSR has remained a challenge, particularly in the areas following double cropping of rice with any long duration rabi crop. Optimum sowing date influences the emergence19, growth20,21, yield22 and economics23 of DSR. Direct drum sowing of rice supplemented with herbicide application and manual weeding (hand) led to decrease in weed infestation and less dry matter accumulation, in turn giving higher grain as well as straw yield and net income24. Such studies are required for DSR in this part of the world to work out the optimum sowing dates. However, the biggest obstacle in DSR cultivation is the potential biotic harm that weeds constitute which is primary biological challenge to the direct seeded rice production25,26. Weed insurgence in DSR is made worse by concurrent crop and weed development, as well as the lack of flooding condition during the early crop growth period27. Ninety percent reduction in yield under DSR has been attributed to immense crop-weed competition for different growth factors, due to the absence of early head start and suppression impact of flooded water on weed species25,26. In India, weeds alone have been contributing a huge loss of 11 billion US dollars to ten popularly grown crops, with rice accounting about 21.4% and 13.8% of this cost in directly sown (DSR) and transplanted crop, respectively.

Different weed management strategies like manual/hand weeding, mechanical weeding and chemical aided weed eradication are being employed for controlling weeds efficiently under direct sown rice. However, manual weeding by hands has numerous demerits such as immense labour use, labour non-availability and problem in crop weed differentiation; thus, making it less practicable and expensive9. Chemical weed management has been more popular over time due to its advantages over other weed control strategies, such as simplicity of application and quick and efficient control. Numerous PE (pre-emergence) herbicides, like oxadiazon, pendimethalin, oxadiargyl, pretilachlor28 and PoE (post-emergence) herbicides like, bispyribac-sodium, penoxsulam, fenoxaprop, azimsulfuron, cyhalofop-butyl, 2, 4-D, metsulfuron-methyl, triafamone + ethoxysulfuron29 are preferred and employed in DSR and conventional transplanted rice in India. However, in that particular situation, a brief period of time for herbicide treatment to accommodate different crop environments is essential for weed control30. Under immense infestation of weeds in DSR, the early (10th-12th days after sowing) and late post (25th-30th days after sowing) application of herbicides is quite necessary. Though different studies are available regarding the influence of varying sowing timeframes and weed managing strategies on the rice growth, yield, economics and weed population density globally, but such studies are limited in this part of the world, for the reason of DSR being the recent introduction as the rice establishment method. Keeping all things in view, a research trial was laid at experimental field of Division of Agronomy, Faculty of Agriculture, SKUAST Kashmir to test whether differential sowing dates and weed managing strategies influence the crop growth and yield in rice and at the same time the weeds associated with DSR.

Materials and methods

Details of experimental site

The experiment was conducted at the research farm/ field of Agronomy Division, Faculty of Agriculture, SKUAST-Kashmir during Kharif, 2018 and 2019, which is located at an altitude of 1590 m amsl between latitude of 34° 21′ N and longitude of 74° 23′ E. Required permission was obtained from the concerned authority to conduct the research trial and collect the plants from the trial. The climate of the experimental zone falls in the temperate category, having freezing winter temperature and hot summer. In this region, rice crop completes its growth cycle in a very short time span of about 140–150 days. Soil sampling was done by analysing samples from 20 cm surface layer of soil for assessing the initial physico-chemical properties of soil (Table 1). After harvesting the rice crop, the soil samples were taken individually from every plot, dried and were subjected to chemical analysis for the estimation of final available NPK in the soil (Table 10). For estimation of available Nitrogen in the soil, Modified alkaline per manganate method31 was used. For determination of available phosphorus in the soil, extraction with 0.5 M NaHCO332 was done using Systronics Spectrophotometer. For estimation of available potassium in the soil, extraction with neutral normal NH4OAC was done using Systronics Flame Photometer33.

Table 1 Physico-chemical status of the soil at experimental site.
Full size table

Weather parameters

The weather conditions of the experimental area were changing frequently during the entire growth period of rice crop and the average data of different weather factors prevailing during 2018 and 2019 were graphically represented (Figs. 1 and 2). The mean maximum and minimum temperatures for both consecutive years during entire growing season was recorded as 32.46 °C; 27.94 °C and 17.24 °C and 12.92 °C in 2018 and 2019, respectively. The amount of total precipitation received during the two consecutive years, respectively was 168 mm and 402 mm approximately. The mean of total sunshine hours received in different SMWs (standard meteorological weeks) was 172.38 and 154.89 during 2018 and 2019, respectively. The mean morning (max.) relative humidity was 90.45% and 77.58%, however evening (min.) relative humidity was having values 62.21% and 54.16% in 2018 and 2019, respectively.

Figure 1
figure 1

Graphical representation of average weather data in relation with crop growing season, 2018.

Full size image
Figure 2
figure 2

Graphical representation of average weather data in relation with crop growing season, 2019.

Full size image

Experimental design and treatments used

Current investigation was based on two variable components (2 different sowing times and 6 weed managing practices), having three replications each and the Split plot design was practiced for laying the trial. Two variable sowing times represented main plot treatments i.e., D1: 10th of May and D2: 3rd of June, however, 6 weed managing approaches viz. W1: weedy check (untreated check), W2: 4 mechanized conoweedings at 15, 30, 45 & 60 Days after sowing (equivalent to weed free), W3: Bensulfuron-methyl + Pretilachlor @ 60 and 600 g a.i. ha−1 as pre-emergence (BSM + Pretilachlor, PE), W4: Oxyfluorfen @ 750 g a.i. ha−1 as pre-emergence (Oxyfluorfen, PE), W5: Bensulfuron-methyl + Pretilachlor @ 60 and 600 g a.i. ha−1 as pre-emergence followed by 2, 4-D @ 0.75 kg a.i. ha−1 as post-emergence (30–35 DAS) (BSM + Pretilachlor PE fb 2, 4-D PoE) and W6: Oxyfluorfen @ 750 g a.i. ha−1 as pre-emergence followed by 2, 4-D @ 0.75 kg a.i. ha−1 as post-emergence (30–35 DAS) (Oxyfluorfen PE fb 2, 4-D PoE) represented sub-plot treatments. The dimensions for each treatment plot were kept as 5 m × 3 m (15 m2) with a buffer zone of 0.5 m to separate sub-plots for limiting all sorts of herbicide cross-contamination. Mechanized conoweedings were repeated for keeping plots weed-less in treatment W2 i.e., 4 conoweedings at 15, 30, 45 and 60 DAS, while as weeds were undisturbed and allowed to grow in case of treatment W1 i.e., weedy check (control). Different herbicide categories were applied as per their suitable application method, like for application of granular herbicides, they were thoroughly mixed with sand and then applied uniformly to respective treatment plots, however for the incorporation of liquid herbicides, a flat fan nozzle was fitted to a knapsack sprayer operated at a pressure of 5.6 kg cm−2 and discharge rate, 500 ml min−1.

Crop management practices

The entire experimental area was tilled once with a tractor-mounted disc-plough, subsequently followed by 3 more ploughings by a tiller and a manual leveling to achieve soil with fine tilth. Before the irrigation of experimental field, bunds around separate treatment plots were formed in dry soil. All the treatment plots were puddled uniformly with a power tiller after applying irrigation in order to level the field properly. The excess irrigation water was removed out from the experimental field by establishing proper drainage outlets and the total plot was then laid into main and sub-plot treatments as per the experimental plan. Shalimar Rice-3 variety seeds were utilized in the experiment which is yielding about 7.0–7.5 t/ha grains on an average and completes its life cycle in a period of around 135–140 days. The seeds of selected rice cultivar (@ 80 kg/ha) were incubated for approximately 48 h after soaking them in normal water for 48 h for uniform sprouting. Seeds in all treatment plots were sown in rows aided by 6-rowed drum seeder keeping a required spacing of 20 × 10 cm row-row and plant-plant, respectively in north-east direction. For prevention of diseases in rice crop, 1 kg seed was treated with 0.6 g of tricyclazole wettable powder (WP) before sowing. FYM @ 10t/ha was incorporated in the soil in a fully decomposed state at the time of field preparation. The major nutrients viz. N (120 kg ha−1), P2O5 (60 kg ha−1) and K2O (30 kg ha−1) were provided to crop by applying urea, DAP and MOP, respectively. Nitrogen application was covered in three splits (half quantity as a basal dose and rest half in two equal splits), however phosphorus and potassium were added in a single basic dose.

Biometric crop monitoring

The plants were marked for taking data of crop dry matter accumulation at various phenological stages and leaf area meter was used for measuring leaf area (cm2/plant) (Systronics, 211) from same plants. For this, whole leaves were taken out from the stem and cleaned thoroughly with normal tap water and then de-ionized water followed by drying with the help of tissue paper. Leaf area index is defined as the ratio of leaf surface to the ground area occupied by the plant and was calculated by the formula as per34:

$$LAI = \frac{Total Leaf Area}{Ground Area}$$
(1)

Crop growth rate (CGR) is termed as an index of the amount of growing material per unit area per unit time and was recorded in gm−2 day−1 by the formula as per35:

$$CGR = \frac{1}{A}\times \frac{W2 - W1}{t2 - t1}$$
(2)

where, W1 and W2 = dry matter produced per plant (g) at time t1 and t2, respectively and A is the area (spacing).

Relative growth rate (RGR) is termed as an index of the daily increase in dry weight per unit dry weight of the plant and was recorded in gg−1 day−1 by the formula as per36:

$$RGR = \frac{Log e W2 - Log e W1}{t2 - t1}$$
(3)

where, W1 and W2 = Dry weight of plant at time t1 and t2.

Net assimilation rate (NAR) is expressed as the rate of increase in the whole plant dry weight per unit leaf area per unit time and is recorded in gcm−2 day−1 by the formula as per37:

$$NAR =\frac{W2-W1}{T2-T1}\times \frac{Loge L2 - LogeL1 }{L2-L1}$$
(4)

where W1 and W2 = Total dry weight of plant at time t1 and t2, respectively and L1 and L2 = Total leaf area of plant at time t1 and t2, respectively.

The total yield (grain as well as straw) from net plot was calculated in kg and converted to t ha−1.

Weed measurements

The data on weed population/density and dry matter accumulation was recorded at 30, 45, 60 days after sowing as well as harvest. For taking data, spots were picked randomly in each sub-plot to place quadrants, followed by cutting the weeds right from ground and washing with normal tap water to make it dirt free. After washing, the counting and categorization of weeds was done by dividing them into 3 main groups viz., grasses, sedges and broad-leaved weeds (BLWs). The drying of samples was carried out under shade in normal air and then oven drying at 70 °C temperature and 10% relative humidity to attain a constant weight.

The weed control efficiency directly influences the crop performance and latter thereby shows an inverse association with the weed index. Weed index was given by the formula:

$$Weed\,\,index\,\left( \% \right) = \frac{{\left( {YWF - YT} \right)}}{YWF} \times 100$$
(5)

where, YWF and YT depicts yield of crop in weed free plot and treated plot respectively.

Dominance and marginal analysis

The economic analysis of the experimental data followed the methodology outlined in CIMMYT (1988). To achieve this, gross income and total expenditure were computed to assess the net field benefits. These net field benefits were determined by deducting the total variable costs from the total income derived from each treatment. The costs associated with inputs and outputs for each treatment were converted to US$ ha−1. To assess the variability between costs and net field benefits, a marginal analysis was conducted. This marginal net field benefit (MNB) is also referred to as the marginal rate of return (MRR). MRR was formulated by using the formulae given by38 as percent ratio of marginal net field benefit to marginal cost (MC).

$$\text{MRR}=\frac{{\text{MNB}}}{{\text{MC}}} \times 100$$

Statistical analysis

The data recorded was subjected to standard analysis statistically by performing ANOVA (SAS Software packages, SAS EG v4.3, SAS Institute Carry, North Carolina, US), followed by calculation of critical difference (CD) to test the significant variance among different treatment means at a significance level of 5%. The data on dry weed biomass and weed density were utilized in the analysis only after performing square-root transformation. As the data of various tested weed parameters such as yield, weed population per square-meter and dry weed biomass during two study years was showing a non-significant difference statistically, therefore entire data sets of two years was pooled.

Results

Leaf area index

There was a remarked effect of sowing dates and weed managing approaches over leaf area index of DSR for both experimentation years. The progressive enhancement in leaf area index was observed up to 50% flowering in both the sowings i.e., 10th May (D1) and 3rd June (D2), thereafter leaf area index decreased sharply (Fig. 3). This may be probably explained by the fact that in the progressing reproductive period of the crop, the leaves at lower most branches are continuously growing under shade which led to their senescence and eventual death. Sowing on 10th May (D1) registered more leaf area index significantly in comparison to 3rd June (D2) sowing throughout the crop growth season. All the tested weed managing techniques performed significantly better than the weedy check in generating a higher LAI. However, among the tested weed management practices, conoweeding (equivalent to weed free) and sequential use of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) were at par in recording higher LAI at all phenological stages while as, significantly lowest LAI was recorded in weedy check (control) at all the phenological phases of the crop.

Figure 3
figure 3

Leaf area index of DSR under varying sowing dates and weed management practices.

Full size image

Crop growth analysis

Crop growth rate

The acquired data revealed that rate of crop growth increased on advancement of crop age up to 50% flowering stage and then showed a declining trend (Fig. 4). Among the sowing dates, sowing date first (D1) i.e., 10th of May reflected higher crop growth rate in comparison to second sowing (D2) i.e., June 3rd at all successive crop phenological phases. Different weed managing practices with respect to crop growth rate showed significant results at different phenophases of DSR. On comparing different weed managing techniques, conoweeding (equivalent to weed free) significantly showed a maximum crop growth rate in comparison to other treatments as well as weedy check at different crop phenological stages. However, among all tested herbicidal treatments, sequential incorporation of Oxyfluorfen (PE) fb 2, 4-D (PoE) was at par with sequential use of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) in getting remarkably high crop growth rate than other herbicides.

Figure 4
figure 4

Crop growth rate of DSR under varying sowing dates and weed management practices.

Full size image

Relative growth rate

There was generally a decline in relative crop growth rate towards the advancing crop age upto harvest (Fig. 5). Among the sowing dates, significant variations were observed in relation to relative growth rate of crop at each succeeding stage of crop growth. Earlier sowing, 10th May (D1) registered a significant surplus in relative growth rate than late sowing (D2). Different weed management practices had significant impacts on growing crop at various phenophases with respect to relative growth rate. Among all tested weed managing approaches, conoweeding (equivalent to weed free) registered notably enhanced relative growth rate with respect to other tested treatments as well as weedy check. Among the herbicidal treatments done, use of BSM + Pretilachlor PE and oxyfluorfen PE were at par in recording higher relative growth rate as compared to other herbicidal treatments.

Figure 5
figure 5

Relative growth rate of DSR under varying sowing dates and weed management practices.

Full size image

Net assimilation rate

The acquired data revealed that in general the net assimilation rate enhanced with the age advancement in crop till 50% flowering to milking stage and then showed a declining trend (Fig. 6). Among the sowing dates, net assimilation rate observed significant results with respect to changing growth stages of the crop. Sowing date first (D1) i.e., 10th of May reflected higher net assimilation rate in comparison to second sowing (D2) i.e., 3rd June. Different weed management practices with respect to net assimilation rate of direct seeded rice showed significant results at succeeding phenological phases of crop. Among the weed managing approaches, notably peak net assimilation rate was seen in case of conoweeding treatment (equivalent to weed free) in contrary to weedy check. Among the tested herbicides, sequential application of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) and Oxyfluorfen (PE) fb 2, 4-D (PoE) were at par in recording significantly higher net assimilation rate in comparison to weedy check and other tested herbicides.

Figure 6
figure 6

Relative growth rate of DSR under varying sowing dates and weed management practices.

Full size image

Grain, straw, biological yield and harvest index

The acquired data showed a notable impact of both weed managing techniques as well as changing sowing time over grain and straw yield (Table 2). After analyzing data, it got clear that there was remarked rise in grain yield in first sowing i.e., 10th May than second sowing i.e., 3rd June. The biological yield was having the values 16.33 t ha−1 and 14.01 t ha−1 in respect of two sowing times; 10th of May and 3rd of June, respectively. There was an overall enhanced harvest index of 44.59% and 43.05% during two sowing times, 10th May and 3rd June, respectively. The amount of grain, straw and biological yield plus harvest index were remarkably influenced by different weed managing approaches. Conoweeding (equivalent to weed free) recorded higher grain yield (7.79 t ha−1), straw yield (9.33 t ha−1), biological yield (17.31 t ha−1) and harvest index (46.06%). Sequential use of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) recorded the grain yield of (6.95 t ha−1), straw yield (8.58 t ha−1), biological yield (15.53 t ha−1) and harvest index (44.75%) which was significantly surpassing control and other tested weed managing approaches. Oxyfluorfen (PE) fb 2, 4-D (PoE) was in turn behaving more likely as that of sequential incorporation of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) and produced significantly enhanced grain, straw, and biological yield than other tested herbicides and weedy check also, however, control was found to be yielding least amount of grain yield (4.25 t ha−1), straw yield (6.86 t ha−1), biological yield (11.11 t ha−1) as well as harvest index (38.19%).

Table 2 Yield and harvest index of direct seeded rice in relation to changing sowing time and weed managing approaches (2 year pooled data).
Full size table

Weed studies

Weed flora associated with the crop

All three weed categories (grassy, sedges and broad leaved weeds) were associated with main crop and among grassy weeds, Echinochloa crusgalli, Echinochloa colona, Cynadon dactylon were most dominating, among broad leaved weeds Ammania baccifera, Marsilea quadrifolia, Monochoria vaginalis, Alisma plantago-aquatic and Potamogeton distinctus, however, the most predominant among sedges include Cyperus iria, Cyperus difformis and Fimbristylis littoralis. Data from all the tested treatments revealed that maximum weed density and dry weight was seen after 60 DAS. As compared to weedy check, all the tested weed managing techniques proved superior in controlling weed growth and development through the crop growing period (Tables 3, 4, 5, 6).

Table 3 Density (No.m−2) of grassy weeds in Direct Seeded Rice in relation to changing sowing time and weed managing approaches (2 year pooled data).
Full size table
Table 4 Dry weight (g m−2) of grassy weeds in direct seeded rice in relation to changing sowing time and weed managing approaches (2 year pooled data).
Full size table
Table 5 Density (No.m−2) of sedges in direct seeded rice in relation to changing sowing time and weed managing approaches (2 year pooled data).
Full size table
Table 6 Dry weight (g m−2) of sedges in Direct Seeded Rice in relation to changing sowing time and weed managing approaches (2 year pooled data).
Full size table

Weed population per square-meter and weed dry matter (grasses, sedges and BLWs)

Different sowing dates and weed managing techniques significantly impacted density (Tables 3, 5, 7) as well as dry matter of all weed categories (Tables 4, 6 and 8) at different growth stages of DSR. 3rd June (D2) sowing recorded significantly higher population per square-meter and dry matter of all category weeds at 30, 45, 60 DAS and at harvest than the early 10th May (D1) sowing. With respect to weed managing techniques, population and dry matter of all category weeds varied significantly. Weedy check treatment recorded notably higher population and dry matter of all category weeds at 30, 45, 60 DAS and at harvest and was followed by oxyfluorfen @ 750 g a.i./ha (PE) which in turn recorded remarkably higher population and dry matter of grasses, sedges and broad leaf weeds than rest of the weed managing approaches. However, lowest population and dry matter of all category weeds was recorded in conoweeding (equivalent to weed free) at 30, 45, 60 DAS and at harvest. Among the herbicides tested, sequential use of BSM + pretilachlor (60 and 600 g a.i. ha−1 PE) fb 2,4-D @ 0.75 kg a.i. ha−1 (PoE, 30–35 days after sowing) though at par with application of oxyfluorfen @ 750 g a.i. ha−1 (PE) fb 2,4-D @ 0.75 kg a.i. ha−1 (PoE, 30–35 days after sowing) registered a considerably lesser population and dry matter of all category weeds at 30, 45, 60 DAS and harvest.

Table 7 Density (no. m−2) of BLWs in direct seeded rice in relation to changing sowing time and weed managing approaches (2 year pooled data).
Full size table
Table 8 Dry weight (g m−2) of BLWs in Direct Seeded Rice in relation to changing sowing time and weed managing approaches (2 year pooled data).
Full size table

Weed index

The results revealed that sowing dates are not showing any significant impact on weed index of the crop. However, 10th May (D1) sowing recorded numerically lower weed index as compared to sowing on June 3rd (D2) (Table 9). Of all the tested weed managing approaches, apart from conoweeding (equivalent to weed free), sequential incorporation of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) recorded lowest weed index while as weedy check treatment received highest weed index than all other weed management practices.

Table 9 Weed index (%) of direct seeded rice with respect to changing sowing dates and weed managing approaches (2 year pooled data).
Full size table

Climatic variablity and weed dynamics

While the temperature was seen to vary between the two years of study, the effect of temperature was quite apparent on weed density and dry weed biomass. Under elevated temperatures during the year 2018, the herbicides showed higher efficacy probably due to enhanced herbicide absorption, penetration and translocation within weed plants or due to accelerated herbicide metabolism leading to lesser weed density and dry weight (Tables 10, 11).

Table 10 Density (No. m−2) of grasses, sedges and broadleaf weeds as influenced by sowing dates and weed management practices in DSR during the year 2018 and 2019.
Full size table
Table 11 Dry weed biomass (g m−2) of grasses, sedges and broadleaf weeds as influenced by sowing dates and weed management practices in DSR during the year 2018 and 2019.
Full size table

Final available soil nutrient analysis (NPK)

The data recorded on final available soil NPK revealed that 10th May (D1) sowing significantly recorded the lowest final available soil NPK whereas 3rd June (D2) sowing recorded the highest final available soil NPK (Table 12). Of all the tested weed managing approaches, conoweeding (equivalent to weed free) recorded remarkedly lower final available soil NPK over weedy check. Among the herbicidal treatments sequential incorporation of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) recorded lowest final available soil NPK closely followed by Oxyfluorfen (PE) fb 2, 4-D (PoE). The highest final available soil NPK was recorded in weedy check. Among tested herbicides, application of Oxyfluorfen, PE recorded highest final available soil NPK.

Table 12 Final available soil N, P2O5 and K2O (kg ha−1) in direct seeded rice with respect to changing sowing dates and weed managing approaches (2 year pooled data).
Full size table

Dominance and marginal analysis

The adoptability of a technology or practice used is determined by the net monetary gain from it. Weed management practices gave higher net field benefits over control (weedy check). The maximum net field benefit of 796.51 USD ha−1 was recorded in earlier sowing i.e., 10th May coupled with sequential application of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) (D1W5) treatment followed by application of Oxyfluorfen (PE) fb 2, 4-D (PoE) and sowing on 10th May (D1W6) with net field benefit of 749.82 USD ha−1 (Table 13). Dominance analysis involved organizing the treatments based on ascending costs. Treatments that had higher costs but lower net field benefits compared to the preceding treatment were labeled as "dominated" and marked with a "D." These dominated treatments were excluded from the calculation of the marginal rate of return. The results obtained through dominance analysis are given in Table 12. The treatments D2W4, D2W5, D2W6 and D2W2 were found dominated as their net field benefit did not increased with an increase in total cost. Maximum marginal rate of returns (18,108%) was recorded in earlier sowing i.e., 10th May coupled with sequential application of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) (D1W5) as depicted from Table 13.

Table 13 Dominance and marginal analysis.
Full size table

Dominance of relative importance index

For the Total Cost (TC) of cultivation of all treatment, the dominance or η2 value was 0.003, though positive and not significant. The value of 0.003 implied that the total cost explains only 0.3% of the variance in the Benefit Cost Ratio. Again, Partial η2 partial) value of 0.007 indicates that, when considering the effect of the total cost relative to the total variance explained by all factors, total cost contributes 0.7%. However, the F-value is very low; indicating that total cost is not statistically significant in explaining the variability of the benefit cost ratio. Equally, the net returns which is the economic gain from all treatment had a dominance or η2 value of 0.567. This means that net returns explain 56.7% of the variance in the benefit cost ratio, which is a substantial proportion. Furthermore, the η2partial) value was 0.585. However, when considering the effect of net returns relative to the total variance explained by all factors, net returns contribute 58.5%. The high F-value 12.71 (***), indicates that net returns are highly statistically significant in explaining the variability of the Benefit Cost Ratio. Lastly, the Residuals had an η2 value of 0.401, meaning that 40.1% of the variance in the Benefit Cost Ratio is unexplained by the factors considered (total cost and net returns) (Table 14).

Table 14 Dominance of relative importance index.
Full size table

Conclusively from the analysis the net returns have a dominant role in explaining the variance in the Benefit Cost Ratio, with η2 value of 0.567. This indicates that net returns are relatively economically important and have a substantial impact on the Benefit Cost Ratio. Total Cost has a very minor and statistically insignificant role in explaining the variance in the Benefit Cost Ratio, with η2 value of 0.003. A significant portion of the variance (40.1%) is unexplained by the factors considered, indicating that other factors not included in this model might also be influencing the Benefit Cost Ratio. These results suggest that, for improving the Benefit Cost Ratio, focusing on strategies to enhance net returns would be more effective than focusing on reducing total costs.

Discussion

Effect of sowing dates and weed management strategies on growth characteristics of direct seeded rice

For attaining the optimized yield and control of weed species, adjustment of time of sowing and incorporation of effective weed managing approaches are prerequisite. Our study revealed that different sowing times supplemented with various weed managing practices has significant impact on overall response of rice sown through direct drum seeding measured in terms of growth characteristics, yield, weed population per square-meter, dry weed biomass and economics. Earlier sowing (10th May) recorded significantly more leaf area index in comparison to late sowing (3rd June) due to the fact that earlier sowing experienced comparatively a longer vegetative phase and more favourable weather conditions during reproductive stage39. The longer vegetative growth phase increased the cumulative heat unit accumulation and other agrometeorological indices resulting in improved leaf area index in earlier sowing. Late sowing experienced short vegetative growth phase which led to reduced leaf area index40. In comparison to weedy check (control), all the weed managing strategies registered higher leaf area index. High leaf area index of rice crop after conoweeding (equivalent to weed free) and sequential incorporation of bensulfuron methyl + pretilachlor (PE) fb 2, 4-D (PoE) may be due to less crop-weed competition which lead to efficient crop development during early crop growth phase, which improved availability of all the growth factors especially availability of nutrients and light, favouring enhanced leaf area index41,42.

Higher crop growth rate (CGR), relative growth rate (RGR) and net assimilation rate (NAR) of directly sown drum seeded rice crop was observed in earlier 10th May sowing as compared to late sowing on 3rd June, at each phenological stages of crop growing period. This could be due to better weather conditions and favourable growth factors available to the crop during early sowing43,44. Different weed management practices with respect to crop growth rate, relative growth rate and net assimilation rate showed significant results at various phenophases of direct drum seeded rice. Significantly higher CGR, RGR and NAR was recorded in conoweeding (equivalent to weed free) in comparison to other weed managing treatments plus weedy check treatment at successive phenophases of rice crop due to periodic removal of weeds resulting in higher availability of growth factors to the crop. Sequential application of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) recorded considerably higher value of CGR, RGR and NAR in contrary to rest of the treatments. This could be due to less weed interference for light, water, space and nutrients which might have maintained high fertility status and growth factors resulting in higher uptake of essential nutrient elements and necessary crop growth factors and suppressing weed species by crop. Lowest CGR, RGR and NAR in weedy check treatment might have happened probably owing to the fact that crop plants could not utilize all the biotic factors favourably on account of high population pressure of weeds45.

Effect of sowing dates and weed management practices on yield of direct seeded rice

In case of early sown crop, due to minimal crop-weed competition and precise use of all inputs, majority of yield governing factors perform well resulting in enhanced grain as well as straw yield. Same facts were registered by46,47 who reported that the yield of a crop growing in any specific environmental condition is the outcome of yield factors produced at different phases of growth and development. There is a significant influence of different weed managing approaches over grain, straw and biological yield as well as harvest index of crop. Increased yield in conoweeding treatment and sequential use of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) was the result of enhancement in all yield parameters, decrease in crop-weed interference and broad spectrum weed control48. This enhancement thereby can be attributed to better growth and development including enhanced total dry matter, photosynthate production in different reproductive regions and expanded leaf area of the crop. As a result, less crop-weed competition helped the crop to grow better and enhanced yield attributing traits due to which the balance shifted more in favour of crop than weeds49. Weedy check had the lesser grain yield and harvest index than other treatments because of the severe competition by different categories of weeds that reduced the nutrient accumulation and caused an imbalanced relation between source and sink with poor performance of yield attributing traits and a much higher weed index as reported by50,51.

Effect of sowing dates and weed management strategies on weed species population

Earlier sowing remarkably reduced weed density and dry weed biomass of all category weeds at 30, 45, 60 DAS and at harvest in comparison to late sowing on 3rd June. Early sowing may have seen a decreased weed population and dry matter due to more favourable environment for proper germination and good initial plant stand aided by a better amount of foliage, which helped plants in efficient smothering of weeds and allowed crop to utilize available resources in a better manner as due to wide foliage cover over ground there is lesser light transmission towards ground facilitating more weed suppression by inhibiting weed seed germination and growth52. In addition of being much less toxic to rice seedlings, the herbicides also suppressed the weeds effectively. Moreover, the improved efficacy and ion lasting effect of tested herbicides prevent weed seeds from germination and also caused the already germinated weeds to exhaust their energy reserves at a quicker pace by enhanced respiration, senescence of leaves, reduced leaf area and decreased photosynthesis53. The enhanced weed flora flux and dry weed biomass in case of weedy check can be probably due to conducive soil moisture conditions and some environmental factors at sowing time which favours germination of weed seeds and successive growth particularly when there is lack of good weed managing approaches54. Sowing time notably did not impact weed index of DSR crop. However, earlier sowing on 10th May sowing recorded numerically lower weed index than late sowing on 3rd June sowing. This may be attributed to greater yields in 10th May sowing because of efficient nutrient uptake, more dry matter piling and also by in line effects of other growth supplementing factors in comparison to sowing on 3rd June. Also, stiff competition impressed by different categories of weeds in late sowing led to higher weed index55. Of all the weed management practices, apart from conoweeding (equivalent to weed free), sequential use of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) registered lowest weed index while as weedy check treatment recorded highest weed index than all other weed management practices because of lesser weed population which in turn led to enhanced nutrient uptake and thus a considerable increase in grain yield, hence lower weed index. The probable reason could be that under weed free treatment, there was limited competition for light, space, nutrients and water as weeds were continuously removed while as in weedy check plots, there was severe competition for light, space, nutrients and water which resulted in the more yield loss in weedy check plots56. When compared to weed free plots, yield reduction was scarce, due to presence of weeds in a negligible amount; the major yield loss was seen in case of control/check plot (weedy) plots, provided other limiting factors to be missing. This could be possibly due to the interference by uncontrolled weed flushes and competence for necessary resources like nutrients, moisture and light with main crop, as visible from lesser growth and yield factors in weedy plots57.

Effect of sowing dates and weed management practices on final available soil NPK status

Significantly lowest final available soil NPK was recorded in earlier sowing whereas late sowing on 3rd June sowing recorded the highest final available soil NPK. As the nutrient uptake was better in earlier sown direct drum seeded rice, available NPK in soil declined at harvest. From the tested weed managing practices, conoweeding (equivalent to weed free) recorded considerably lower final available soil NPK over weedy check treatment because nutrient assimilation by crop plants in absence of weed species. Sequential use of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) registered lowest final available soil NPK closely followed by Oxyfluorfen (PE) fb 2, 4-D (PoE) among the herbicides tested. The highest final available soil NPK was recorded in weedy check treatment. This could have occurred as a result of enhanced weed flush and root-shoot biomass of crop that is left as debris in soil fed upon by soil microorganisms and in turn increasing the mineralization process. Increased nutrient availability in weedy check plots can also be attributed to one more possible fact that the overall uptake by crops in weedy check plots became lesser than rest of treatments as a result of interference impressed by weeds and crop58.

Effect of sowing dates and weed management strategies on relative economics of direct seeded rice

At last, it is the economics (B:C ratio) of different treatments which ultimately differentiates among various treatments or treatment combinations, whether it is significantly more effective than other or not. Sequential application of Bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) on first sowing date (D1) i.e., 10th of May registered higher net returns and maximum B:C ratio, however least returns as well as B:C ratio was reported in weedy check plots sown late on 3rd June. This could be as a result of low grain and straw yield in late sowing leading to lesser economic returns and lower benefit cost ratio59. Both gross returns and net monetary returns were higher in conoweeding (equivalent to weed free) treatment on 10th of May i.e., first sowing with a lower B:C ratio. This could be as a consequence of high expenses needed for 4 repetitive conoweedings to create a weed-free condition for crop in entire growing period. A higher B:C ratio was reported in treatment involving sequential use of bensulfuron-methyl + pretilachlor (60 and 600 g a.i. ha−1 PE) fb 2, 4-D @ 0.75 kg a.i. ha−1 (PoE, 30–35 days after sowing) applied on first sowing (10th May) which in addition of creating a weed-free condition for crop at all critical phases of growth enhances the capacity of rice crop to compete well, hence registered highest B:C ratio among tested weed managing approaches and sowing dates. However, weedy check registered the least B:C ratio ain all tested treatments due to lesser grain and straw yield60,61.

Conclusions

Appropriate sowing date along with efficient weed management in DSR is necessity as direct seeded rice could prove fruitful in establishing plants well in the current global scenario of water shortage and escalating labor wages. The experimental results showed that earlier sowing of DSR (10th May) was more effective than late sowing (3rd June) with respect to growth characteristics, yield, weed population per unit area, dry weed biomass and economics. Of all tested weed managing treatments, 4 mechanized conoweedings at 15, 30, 45 and 60 DAS (equivalent to weed free) registered remarkably better results with respect to growth, yield, better weed control and economics and was statistically equating with sequential use of bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE). These findings led to the conclusion that 10th May sowing proved to be optimum date of sowing and four mechanized conoweeedings at 15, 30, 45 and 60 days after sowing or sequential use of bensulfuron methyl + Pretilachlor (PE) fb 2, 4-D (PoE) came out to be as a promising weed managing strategy for obtaining economically viable yields and efficient weed management in directly sown/seeded rice (DSR) under temperate ecology of Kashmir.