Preliminary Reports on Comparative Weed Competitiveness of Bangladeshi Monsoon and Winter Rice Varieties under Puddled Transplanted Conditions

: Weed-competitive rice cultivars, a viable tool for integrated weed management of rice-ﬁeld weeds, may greatly reduce the weed pressure and excessive dependence on herbicide in controlling weeds. Based on this premise, ﬁeld experiments were conducted in 2018 and 2019 during the monsoon and winter rice-growing seasons at the Agronomy Field Laboratory of Bangladesh Agricultural University, Bangladesh, to evaluate the weed competitiveness of the selected rice varieties. A total of 42 monsoon and 28 winter rice varieties from Bangladesh were evaluated under “weedy” and “weed-free” conditions in the puddled transplanted system of cultivation. The ﬁeld experiments were designed with three replications in a randomized complete block design. The results revealed that weed competition greatly reduces the yield of rice, and relative yield loss was 15–68% and 20–50% in monsoon and winter season rice, respectively. The lowest relative yield losses were recorded in monsoon rice from the variety BU dhan 1 (18%) and from the winter rice BRRI hybrid dhan5 (23.7%), which exhibited high weed tolerance. The weed competitive index (WCI) greatly varied among the varieties in both seasons and the monsoon season, ranging from 0.4 to 2.8, and the highest value was recorded from the hybrid variety Dhani Gold followed by BU dhan 1. In winter season rice, the WCI varied 0.25 to 2.4 and the highest value was recorded from the variety BRRI hybrid dhan3, followed by hybrid variety Heera 6. In monsoon rice, hybrid Dhani Gold was the most productive, but BU dhan 1 was the most weed competitive variety. Among the winter-grown varieties, the hybrid Heera 6 was the most productive, and the most weed competitive. Our research conﬁrmed a high degree of variability in weed competitiveness among the 70 Bangladeshi rice varieties tested.


Introduction
In rice production, among the major pests, weeds rank first in terms of crop losses and cost involvement for their management [1,2]. The average yield losses due to weed infestation in rice may vary from 40 to 60% and in some cases go up to 94-96% depending and abundance [35,37,41]. For example, dry direct-seeded systems of rice cultivation allow more weeds than puddled transplanted systems [39,41,42]. Except for our previous work on weed competitiveness of Bangladeshi winter rice varieties under dry direct-seeded conditions [35,37], there has been no other research conducted so far to screen the weed competitive ability of monsoon and winter season rice varieties in Bangladesh.
Considering the huge pool of Bangladeshi rice varieties, this study was conducted to evaluate the magnitude of variability in weed competitive ability and yield among selected high-yielding varieties of rice grown under puddled transplanted conditions, and to identify potential, high yielding weed competitive rice varieties.

Experimental Site
The experiments were carried out at the Agronomy Field Laboratory of Bangladesh Agricultural University (24°75′ N latitude and 90°50′ E longitude) from July 2018 (July to December, monsoon season rice) to June 2019 (January to June, winter season rice). The soil in the area belongs to the Sonatola soil series of noncalcareous dark gray floodplain soil in the Old Brahmaputra floodplain agroecological zone (AEZ-9) [43]. The soil is silty loam and almost neutral in reaction (pH 6.8). The organic matter content (0.93%) and general fertility level (0.13% total N, and exchangeable P, K, and S are 16.3, 0.28 and 13.9 ppm, respectively) of the experimental field is low. The monthly average air temperature (°C), RH (%), and monthly total rainfall (mm) during the experimental period recorded are shown in Figure 1.

Treatments and Design of the Experiments
Two factorial experiments were carried out where, in both cases, factor A comprised two weeding regimes, viz., a) season-long "weed free" and b) season-long "weedy" condition. On the other hand, factor B consisted of the selected rice varieties in Bangladesh from BRRI, BINA, private seed companies, and traditional local varieties. A total of 42 varieties were selected for monsoon-grown rice (Experiment 1) and 28 varieties for wintergrown rice (Experiment 2). Characteristics (varietal identity, growth duration, and plant

Treatments and Design of the Experiments
Two factorial experiments were carried out where, in both cases, factor A comprised two weeding regimes, viz., a) season-long "weed free" and b) season-long "weedy" condition. On the other hand, factor B consisted of the selected rice varieties in Bangladesh from BRRI, BINA, private seed companies, and traditional local varieties. A total of 42 varieties were selected for monsoon-grown rice (Experiment 1) and 28 varieties for winter-grown rice (Experiment 2). Characteristics (varietal identity, growth duration, and plant height) of the selected varieties are presented in Tables 1 and 2. Both experiments were laid out in a randomized complete block design with three replications. The unit plot size was 5.0 m 2 (2 m × 2.5 m). In addition, three plots without rice were maintained to study the growth, diversity, and abundance of weeds under the field experimental settings. The weed-free plots were maintained by manual hand weeding 2-4 times immediately after the emergence of weeds.

Crop Husbandry
Quality rice seeds of each variety were collected from several institutions, (1) Bangladesh Rice Research Institute (BRRI), (2) Bangladesh Institute of Nuclear Agriculture, (3) Bangabandhu Sheikh Mujibur Rahman Agricultural University, (4) Agronomy Department of Bangladesh Agricultural University, (5) Petrochem Agro Industries Limited, and (6) Bayer Crop Science Limited. The germinated seeds of each variety were sown in the wet nursery bed. Irrigation was gently supplied to the bed and when needed (only in the case of winter rice). Nursery bed soil was fertile enough to provide nutritional support to the rice seedlings for the short period. Although no chemical fertilizer was applied, nursery beds were supplied with organic manures during preparation to ease seedling uprooting.
Thirty-day old (monsoon-grown rice) and forty-five-day old (winter-grown rice) seedlings were transplanted in the puddled land on 15 July 2018 (monsoon-grown rice) and 24 January 2019 (winter-grown rice) at 25 cm × 15 cm spacing with 3 seedlings hill −1 . The experimental plots were fertilized with the BRRI recommended dose of fertilizer, i.e., 220, 120, 75, 60, and 10 kg ha −1 as urea (nitrogen, N), triple superphosphate (TSP), muriate of potash (MoP), gypsum, and zinc sulfate, respectively. All but urea fertilizers were applied as basal; whereas, urea was applied in three equal splits at 15, 30, and 45 days after transplanting (DAT). The monsoon-grown rice was rainfed; thus, no irrigation was provided. However, in the case of the winter-grown rice, flood irrigation was given to maintain a constant water level up to 6 cm in the early stages to boost tillering, and 10-12 cm in the later stage to prevent late tillering. Finally, the field was drained 15 days ahead of harvest. No remarkable insect or pathogen infestation was identified in either experiment; hence, no crop protection measures were taken for controlling insects and pathogens.

Weed Data
At 45 DAT, a quadrat (0.5 m × 0.5 m) was placed randomly in three different places of each season-long weedy plot for collecting weed samples. Weeds were clipped at ground level, identified, and counted by species, and dried in an oven at 70 • C until the weight became constant. Weed density and biomass were expressed as number per square meter and gram per square meter, respectively. The summed dominance ratio (SDR) was computed to identify the dominant weed species using the following equations [44]: SDR of a weed species = Relative density + Relative weed biomass 2 (1) where Relative density (%) = Density of a given weed species Total weed density × 100 Relative weed biomass (%) = Biomass of a given weed species Total weed biomass × 100 The relative contribution of broadleaved, grasses, and sedges to the weed vegetation in terms of relative density and biomass were also calculated.

Crop Data
The whole plot was harvested manually when 90% of the seeds became golden yellow. The harvested crop was bundled separately as per treatment, properly tagged, and brought to the threshing floor. The crop was threshed by a pedal thresher. The grains were cleaned by winnower and sun-dried to 14% moisture content. The straw was also sun-dried. At the end, grain and straw yield plot −1 were recorded and converted to t ha −1 . The relative yield loss (%RYL) due to weed competition was also calculated as follows: Weed free yield − Season long weedy yield Weed free yield × 100 (4)

Weed Competitive Index
The weed competitive index (WCI) of the varieties was calculated to find the most weed competitive cultivar using the following formula [40]: where GYw is the yield of the individual cultivar under weedy condition, GYm is the mean yield of all cultivars under weedy conditions, WBw is the weed biomass of individual cultivar under weedy condition, and WBm is the mean weed biomass of all cultivars under weedy condition.

Statistical Analysis
Collected data were checked for homogeneity and normality and then statistically analyzed using "R Statistics" software. Two season data were analyzed separately due to different growing seasons and variety. The grain yield data were analyzed using a two-way ANOVA (weeding regimes and varieties) and for weed data one-way ANOVA (varieties). SE was used to observe the data dispersion from the mean.

Floristic Composition of Weeds
In monsoon season rice, seven weed species belonging to five different families were identified in weedy plots, and among them three were grasses, three were broadleaves, and one was sedge ( Table 3). The most dominant weed species encountered in the monsoon grown rice field was Echinochloa crus-galli (SDR 29.97) and the least dominant one was Marsilea quadrifolia (4.93). Grass weeds contributed 77% of the total biomass and 47% of total density compared to broadleaves (20% and 40%, respectively) and sedge (3% and 13%, respectively) (Figures 2a and 3a).
Under the season-long weedy condition, the highest grain yield (4.03 t ha −1 ) was found from hybrid Dhani Gold, followed by hybrid Agrodhan-12, and BRRI dhan31 followed by BR10, BR22, BRRI dhan33, BRRI dhan76, and BU dhan1 (Figure 4). The lowest grain yield (1.9 t ha −1 ) was obtained from BRRI dhan38, followed by BRRI dhan78, BRRI dhan70, and Naizersail ( Figure 4).  Seven weed species observed in monsoon season rice were also found in weedy plots of winter grown rice (Table 3). However, their level of dominance in winter grown rice differed from monsoon grown rice. Weed species Echinochloa crus-galli (SDR 24.3) remained the most dominant in winter grown rice while Cyperus difformis (1.9) was the least dominant. In winter grown rice, broadleaf weeds contributed 54% of the total biomass and 46% of total density compared to grasses (46% and 51%, respectively) and sedge (0.6% and 3%, respectively) (Figures 2b and 3b).
Considering the relative yield loss due to weed infestation, the monsoon rice varieties showed a wide diversity, which ranged from 15 to 61%. The relative yield loss was lowest in BU dhan1 (18%), which was followed by Binadhan-11 (21.2%), BRRI dhan31, and BRRI dhan57, which exhibited high weed tolerance. Variety BRRI dhan38 had the lowest tolerance to weeds with a yield penalty of 61%, closely followed by BRRI dhan38 (52.2%) and BRRI dhan78 (Figure 4).  Considering the relative yield loss due to weed infestation, the monsoon rice varieties showed a wide diversity, which ranged from 15 to 61%. The relative yield loss was lowest in BU dhan1 (18%), which was followed by Binadhan-11 (21.2%), BRRI dhan31, and BRRI dhan57, which exhibited high weed tolerance. Variety BRRI dhan38 had the lowest tolerance to weeds with a yield penalty of 61%, closely followed by BRRI dhan38 (52.2%) and BRRI dhan78 (Figure 4).

Weed Density and Biomass
The weed density and biomass differed significantly (p < 0.0001) among the rice varieties grown in both seasons (Table 4). In monsoon rice, weed density ranged from 50 to 255 (plant m −2 ) ( Figure 6). Not weed density, weed biomass is the actual parameter for weed competitiveness measurement, and in monsoon-season rice it varied from 106 to 245 (g m −2 ) in season-long weedy plots at 45 DAT. The highest weed biomass was found in variety BRRI dhan70 (245 g m −2 ), followed by BRRI dhan37 and BRRI dhan51, and the lowest one was recorded in Dhani Gold (106 g m −2 ), followed by BU dhan 1, BRRI dhan41, and BRRI dhan78. In winter rice, weeds density and biomass ranged from 65 to 290 (plant m −2 ) and 145 to 245 (g m −2 ), respectively (Figure 7). Among the winter rice varieties, the highest weed biomass was recorded in BRRI dhan81 (245 g m −2 ), followed by BRRI dhan45 (235 g m −2 ), and the lowest one was recorded in BRRI dhan84 (145 g m −2 ).

Weed Density and Biomass
The weed density and biomass differed significantly (p <.0001) among the rice varieties grown in both seasons (Table 4). In monsoon rice, weed density ranged from 50 to 255 (plant m −2 ) ( Figure 6). Not weed density, weed biomass is the actual parameter for weed competitiveness measurement, and in monsoon-season rice it varied from 106 to 245 (g m −2 ) in season-long weedy plots at 45 DAT. The highest weed biomass was found in variety BRRI dhan70 (245 g m −2 ), followed by BRRI dhan37 and BRRI dhan51, and the lowest one was recorded in Dhani Gold (106 g m −2 ), followed by BU dhan 1, BRRI dhan41, and BRRI dhan78. In winter rice, weeds density and biomass ranged from 65 to 290 (plant m −2 ) and 145 to 245 (g m −2 ), respectively (Figure 7). Among the winter rice varieties, the highest weed biomass was recorded in BRRI dhan81 (245 g m −2 ), followed by BRRI dhan45 (235 g m −2 ), and the lowest one was recorded in BRRI dhan84 (145 g m −2 ).
In winter season rice, the WCI varied from 0.25 to 2.4 and the highest value was recorded from the variety BRRI hybrid dhan3, followed by hybrid variety Heera 6, BRRI dhan69, and BRRI hybrid dhan5 (Figure 9). The intermediate WCI was found from the varieties hybrid SL-8, BRRI dhan58, BR24, and Tejgold. The lowest WCI was recorded from the cultivar BRRI dhan50, followed by BRRI dhan47 and BR16. In winter season rice, the WCI varied from 0.25 to 2.4 and the highest value was recorded from the variety BRRI hybrid dhan3, followed by hybrid variety Heera 6, BRRI dhan69, and BRRI hybrid dhan5 (Figure 9). The intermediate WCI was found from the varieties hybrid SL-8, BRRI dhan58, BR24, and Tejgold. The lowest WCI was recorded from the cultivar BRRI dhan50, followed by BRRI dhan47 and BR16.

Discussion
The WCI is the best parameter to measure the weed competitiveness of a variety [35]. In this study, we found a higher WCI value in hybrid varieties compared to the inbred, which indicates the higher weed competitiveness of hybrid varieties than inbred. Our results are supported by the previous study of Ahmed et al. [40] and Awan et al. [45]; they

Discussion
The WCI is the best parameter to measure the weed competitiveness of a variety [35]. In this study, we found a higher WCI value in hybrid varieties compared to the inbred, which indicates the higher weed competitiveness of hybrid varieties than inbred. Our results are supported by the previous study of Ahmed et al. [40] and Awan et al. [45]; they found that the hybrid cultivars had higher weed competitive ability than the inbred cultivars mainly due to higher early vigor, quick canopy coverage, higher tiller production, and higher biomass production. In general, varieties that produced higher yield in weedfree condition also had higher yield in weedy condition, resulting in a higher WCI; however, the relative yield loss did not maintain the same trends. For example, the varieties BU dhan 1 and Binadhan-11 in monsoon season produced low yield even in weed-free conditions, but RYL was low due to less yield reduction in weedy condition, indicating that these varieties have good weed competitive ability but less yield potential. Normally, the variety that gave lower relative yield loss had a higher degree of weed tolerance. Wide differences in weed competitiveness among rice varieties have already been reported [34,35,[37][38][39]46,47]. The competitive ability of a variety not only depends on the varietal characteristics, it also depends on the types and abundance of weeds in the field.
In both seasons, the same number of weed species was found in the trial field and barnyard grass, E. crus-galli, was the most dominant weed. The possible reason for the same number of weed species is that both experiments were conducted in the same experimental fields. Although the weed species were the same in the both seasons, their dominance sequence varied and the variation was mainly due to the differences of seasonal weather and microenvironmental variation.
In both seasons, the highest grain yield was observed in season-long weed-free conditions compared to season-long weedy conditions, irrespective of varieties. It was due to the absence of crop-weed competition for resources in weed-free plots as also reported earlier in dry direct-seeded [35], modified aerobic system [37] of rice and wheat [36]. With few exceptions, hybrid rice varieties performed better than inbred or local varieties in terms of yield and weed suppression. Differential weed suppressive abilities of inbred and hybrid rice varieties were also reported by Chauhan et al. [48] and Dass et al. [42], who inferred that rice-weed competition remained stronger for weeds having similar morphological characteristics with rice plants. Similar to our findings regarding hybrid varieties, it was also established that droopy-plant-type rice varieties with long leaves exhibit a high early growth rate, faster canopy covering, and high tillering ability, and encounter less competition by weeds owing to offering less space to weeds to emerge and grow [34,38,39,46].
Among monsoon varieties, BU dhan 1 showed the lowest relative yield loss, which indicates its higher weed tolerance potential. Apart from these characteristics, BU dhan 1 is also drought resistant, and pest and pathogens tolerant. Dingkuhn et al. [49] suggested that competitive varieties need to be weed competitive and possess tolerance against other biotic and abiotic stresses, and that this research area must be explored further. However, the yield potential of BU dhan 1 was lower than many other monsoon rice varieties. It was reported that the strong weed suppressive ability of a low-yielding variety under weedy conditions could not always provide a guarantee of higher yield [34,35,37]. On the other hand, high-yielding varieties could not always have high weed suppressive ability, as in this study; high-yielding inbred varieties BRRI dhan61 and BRRI dhan71 suffered the highest relative yield loss.
The results showed that the average yield potential of winter rice varieties (6.0 t ha −1 ) was higher than those of the monsoon varieties (4.8 t ha −1 ). This is because of existing weather conditions (low temperature, clear sunshine, low humidity, etc.) that influence higher dry matter assimilation by the crop. The longer growth duration of the winter rice varieties, and less disease and insect infestation during the season also trigger higher grain yield. Moreover, the varietal characteristics (leaves are stiff and straight) and the longer life cycle of winter rice varieties help them to produce a higher yield than monsoon varieties.
None of the inbred varieties were found as strong weed suppressive combined with high yield potential, except hybrid rice varieties, Dhani Gold, Heera 6, and BRRI hybrid dhan3. However, the problem with hybrid rice seeds is that farmers always have to buy seeds from the seed companies. In addition, the seed price is very high to afford by the subsistence farmers. Hence, the development of weed competitive high-yielding inbred varieties for monsoon and winter seasons of Bangladesh warrant due attention.
Environmental factors such as water availability, temperature, humidity, and the incidence of pests greatly affect the performance of a particular crop plant/variety in a given environment. Crop weed competition is very common in nature and when crops are subjected to more competition in the community, the physiological characteristics of the growth and development of crops are usually changed. Procópio et al. [50] reported that in a community when the growth and development of a crop are changed due to competition, this result changes the differences in the use of environmental resources, especially the water, which directly affects the availability of CO 2 in leaf mesophyll and leaf temperature. Therefore, when plant breeders will have a target to develop weed competitive cultivar/s, they must be considered the ecophysiological aspects of crops and weed [51]. The ecology and physiology of crops and weeds also help in weed management decisions; e.g., Echinochloa crus-galli, a major rice weed, infests rice fields worldwide and is rarely found in shaded places. This species is capable of germinating even beneath 10 cm of water and can adapt to anaerobic environments [52].

Conclusions
Although the present study identified some promising weed competitive varieties of rice from both seasons, those are not feasible in terms of yield and profitability. Therefore, identification of morpho-physiological traits related to weed competitiveness and genetic characterization of the weed competitive inbred BU dhan 1 and hybrid Heera 6 are necessary to develop strong weed competitive, high yielding rice varieties to avoid the environmental impact of indiscriminate use of herbicide for weed control. This is the first report on the comparative weed competitiveness of Bangladeshi rice varieties, considering a huge genetic pool under puddle transplanted conditions. Since a highly competitive variety may not produce high yields, the rice breeder must consider those competitive varieties for breeding toward developing both high-yielding and weed-competitive rice varieties. As multiseason and/or multilocation trials with the same varieties were not conducted in this research, it is recommended to do these trials before drawing a final recommendation.  Data Availability Statement: Data is not publicly available, although the data may be made available on request from the corresponding author.