Rice (Oryza sativa) is one of China’s major food crops, where it accounts for approximately 18% of the world’s rice planting area, and is a staple food resource for more than half of the global population (Kraehmer et al., 2016; Arunrat et al., 2021). Conventional rice production is heavily dependent on synthetic chemical fertilizers and pesticides to ensure rice yield. However, long-term heavy application of synthetic chemical fertilizers and pesticides has contributed to environmental hazards around the world, including water contamination, soil erosion, biodiversity loss and pesticide resistance, and is also accompanied by an increase in rice production costs (Poudel et al., 2002; Brown and Funk, 2008; Shaner, 2014). Few would disagree that it is necessary to develop and promote ecologically and environmentally sustainable agricultural systems to cope with the environmental crisis caused by conventional agriculture. Organic agriculture is considered a promising solution for reducing the negative impacts of conventional agriculture on the environment due to the lower use of off-farm inputs, higher input‒output efficiency and environmental benefits (Ikemura et al., 2008; Seufert et al., 2012; Arunrat et al., 2021). Along with increasing public knowledge about health and healthy lifestyles, the demand for organic products by consumers all over the world is increasing (Hazra et al., 2018). Organic agriculture is defined as a food production system that relies on ecological processes, biodiversity and cycles adapted to local conditions to maintain soil, ecosystem and human health (Suh, 2015). Based on the ecological characteristics of paddy fields and to meet the fertilizer and pest control demands during rice planting, in addition to basic rice monoculture, a variety of organic rice production modes have been developed. These include rice-green manure rotation, which uses green manure to reduce the application of organic fertilizer (Hazra et al., 2018; Du et al., 2022), and rice integrated planting–breeding, which makes use of the aquatic environment of paddy fields to achieve mutual benefits for rice planting and waterfowl or aquatic animal breeding (Nair et al., 2014; Pirdashti et al., 2015; Xu et al., 2019).

Weed competition is one of the major factors constraining rice yield, especially in organic rice production, in which chemical herbicides have been prohibited (Hokazono et al., 2009). Weed infestation can usually reduce rice yield by 10–20%, and when the field is severely infested, the yield can be reduced by more than 50% or even to no harvest (Rathod and Somasundaram, 2017). The existence of weed seeds in the soil seed bank is the main source leading to the continuous infestation of weeds in farmland (Barberi and Lo Cascio, 2001). The composition and succession of the weed community are closely related to the soil weed seed bank, which is influenced by farming practices, such as tillage and fertilization practices, rotation and continuous cropping and weeding methods (Yin et al., 2005; Chamanabad et al., 2009; Chauhan and Opena, 2012; Santínmontanyá et al., 2013). Studies on the composition of weed seed banks and the factors affecting their formation in farmland soils could help predict the infestation of weed populations and improve decision-making for managing specific weed problems (Chauhan and Johnson, 2010; Gao et al., 2020).

Rice is the first and foremost grain crop in Jiangsu Province, and its perennial planting area and yield account for approximately 42% and 55% of the total planting area and yield of grain crops in the province, respectively (National Bureau of Statistics, 2020). With the improvement in people’s living standards and environmental awareness, the demand for organic rice has increased, and a variety of organic production modes have been developed, such as rice monoculture, rice-green manure rotation, rice-duck coculture, rice-crab coculture, rice-crayfish coculture and rice-frog coculture in Jiangsu Province. According to Du et al. (2022), the total planting area of organic rice in Jiangsu Province was 4410 ha. Among the organic production modes, rice-green manure rotation (RG), rice-duck coculture (RD), rice-crayfish coculture (RC) and rice monoculture (RM) were the four dominant organic rice production modes, and weed control in organic rice fields mainly depended on hand weeding. However, hand weeding is becoming problematic due to the shortage of labor and increasing labor costs in agricultural production (Liu and Zhou, 2022). Therefore, it is of great significance to determine the composition characteristics of the soil weed seed bank under different organic rice production modes to provide decision-making support for rational integrated weed management in organic rice production. In the present study, the soil weed seed bank under the four dominant organic production modes (RG, RM RC and RD) with different numbers of consecutive planting years (3 to 10 years) in different sites in Jiangsu Province were investigated to: (i) determine whether the composition of the weed seed bank is similar among different sites under the same organic rice production mode; (ii) identify the main farming practices leading to the differences in the composition of weed seed banks under different organic rice production modes; and (iii) establish the variation in the total seed density and diversity of the weed seed bank over consecutive years under a certain organic production mode.

Weeds are plant species associated with crops that negatively affect crop growth and yield by competing with them for the same resources, mainly sunlight, water and nutrients (Smith et al., 2010; Keller et al., 2014). Therefore, changes in farming practices for crops, such as tillage (Chauhan and Opena, 2012), fertilization (Yin et al., 2005), irrigation (Mahajan et al., 2014), weeding methods and continuous and rotation cropping (Chamanabad et al., 2009; Santínmontanyá et al., 2013), affect the composition, density and diversity of weed communities in crop fields. As mentioned above, the RG, RM, RC, and RD modes are associated with a set of relatively independent farming practices for organic rice planting. To date, there are few reports on the composition of weed communities in organic rice fields, which were only partly analyzed in a few studies on rice-duck co-culture. A study on the influence of rice–duck farming on weed community composition for four consecutive years demonstrated that the control effect of rice-duck farming on broadleaf weeds was better than that on grass weeds, and the weed density in rice fields decreased with the increase in the number of consecutive years, among which the density of E. crusgalli decreased the most slowly (Wei et al., 2006). Zhao et al. (2014) reported that A. aequalis and B. syzigachne were always the most dominant weed species in wheat field seed banks over 13 consecutive years of rice-duck and wheat rotation farming. Similarly, our results showed that the grass weeds E. crusgalli, A. aequalis and B. syzigachne were the most dominant species in the soil seed bank under the RD mode. In contrast, the broadleaf weeds M. vaginalis, A. baccifera and R. indica and the sedge weeds C. difformis, S. triangulatus and S. wallichii were the most dominant species in the soil seed bank under the RG and RC modes, respectively. The soil seed bank of the RM mode was also dominated by broadleaf weeds, however, the most dominant weed species were sedge weed C. difformis and grass weeds E. crusgalli and L. chinensis. Therefore, more attention should be given to the control of grass weeds in the RD mode; while broadleaf weeds and sedge weeds should be the primary control targets in the RG and RC modes, respectively; besides, attention should also be paid to the outbreak of broadleaf weeds during the control of the dominant weeds C. difformis, E. crusgalli and L. chinensis in the RM mode; when formulating weed management strategies for organic rice fields.

Fertilizer is one of the major input costs in cropping systems, and its application is considered crucial for determining the competition relationship between crops and weeds, and it affects the species composition and biodiversity of weeds by stimulating plant growth and modulating competition for aboveground (e.g., light and space) and belowground (soil nutrients) resources (Yin et al., 2005; Zhang et al., 2019). The ABT analysis results indicated that the application amount of organic fertilizer made the largest contribution to the difference in weed composition of the soil seed bank under different organic rice production modes. The growth of green manure in the RG mode decreased the application amount of organic fertilizer, and in the RC and RD modes, the crayfish and duck waste and forage overapplication increased the nutrient supply for rice and reduce its dependence on organic fertilizers; therefore, the application amount of organic fertilizer and nutrient sources differed among the RG, RM, RD and RC modes. A study on the impacts of long-term composted manure and straw amendments on rice-associated weeds in a rice–wheat rotation system indicated that A. baccifera, E. crusgalli, and M. vaginalis could gain competitive advantages in fertile or infertile soil, while nutrient deficiency was more favorable to the infestation of sedge weeds (Gao et al., 2022). In our study, A. baccifera, E. crusgalli, and M. vaginalis were also dominant in the soil seed bank of all four organic rice production modes (except for M. vaginalis in the RC mode), and the lower input of organic fertilizer could be one of the reasons for the higher infestation of sedge weeds in the RC mode. Water is a powerful selective agent for weed management in paddy fields, and flooding is a key factor influencing the severity of weed competition (Kim et al., 2001; Caton et al., 2002). To provide a more suitable living environment for ducks and crayfish, the amount of irrigation in the RD and RC modes was higher than that in the RG and RM modes. Ducks and crayfish can prey on weed seedlings (Pirdashti et al., 2015; Yuan et al., 2021), which reduces the inputs required for hand weeding in the RD and RC modes. Our results indicated that although hand weeding and duck and crayfish breeding directly affected weed density, their influence on the variation in weed composition in the soil seed bank was lower than that of irrigation. Correspondingly, we also found that although the weed composition was similar under the same organic rice production mode at different sampling sites, there were significant differences in the total weed seed density in the soil seed bank. Wei et al. (2006) reported that the weed species richness and diversity decreased in an organic rice-duck coculture field over four years of consecutive planting. Our results indicated that weed species diversity increased and weed community evenness and dominance decreased with the increase in the number of consecutive planting years under all four organic rice production modes. The differences in the responses of weed species diversity in the RD mode may be due to the smaller number of consecutive planting years of rice-duck coculture in the study of Wei et al. (2006). Similarly, a nine-year study by Guo et al. (2020) demonstrated that weed diversity in the first four consecutive planting years of rice-crayfish coculture (with chemical fertilizer application but without weeding) showed a downward and then upward trend.

Our results indicated the weed composition differed in the soil seed bank among the organic rice production modes of rice-green manure rotation, rice monoculture, rice-crayfish coculture, and rice-duck coculture. The most dominant weed group was broadleaf weeds in the soil seed bank of the rice-green manure rotation. The soil seed bank of the rice monoculture was also dominated by broadleaf weeds; however, the most dominant weed species were sedge and grass weeds. Sedge and grass weeds dominated the soil seed bank of rice-crayfish coculture and rice-duck coculture, respectively. Therefore, specific weed management strategies could be formulated based on the compositional differences in the weed community under different organic rice production modes. The application of organic fertilizer and irrigation had greater effects on the weed composition in the soil seed bank than hand weeding. Consequently, fertilization and irrigation strategies that alter weed composition could be used as improved weed management program components in organic rice production systems. Unlike the biodiversity loss in the long-term conventional rice planting which based on chemical fertilizer and herbicide, long-term organic rice planting was beneficial for increasing weed diversity in paddy fields. Our results also indicated that weed community evenness and dominance decreased with the increase in the number of consecutive planting years under all four organic rice production modes.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

QD contributed to the conception of the study and helped perform the analysis with constructive discussions; PG, HW, SD, and ED performed the field survey and contributed significantly to analysis and manuscript preparation; PG performed the data analyses and wrote the manuscript. All authors contributed to the article and approved the submitted version.

This study was financially supported by the Key Research & Development Program of Jiangsu Province (BE2019343), the Postdoctoral Science Foundation of China (2020M671616), the Jiangsu Natural Science Foundation (BK20220565) and the National Key Research and Development Program of China (2021YFD1700805).

We would like to thank Department of Agriculture and Rural Development of Jiangsu Province for help with the survey and data collection.

Author ED was employed by Anhui Pujiwei Modern Agriculture Group Co., Ltd.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.