2.3 Greenhouse experiments
A series of experiments were conducted in a greenhouse with 20-30 °C night and day temperatures and 65-90% relative humidity. Each experiment was conducted in a completely randomized design with four replicates for each treatment and control. Rice seeds were sterilized with 5% H2O2 then placed in Petri dishes with moistened filter paper. Weed seeds were sterilized with 75% ethyl alcohol for 3 min, followed with 3% sodium hypochlorite for 12 min. Sterilized seeds were rinsed with distilled water and then transferred to Petri dishes with moistened filter paper for vernalization in a 4 °C refrigerator for 24 h. Dishes of rice and vernalized weed seeds were pre-germinated in an environmentally controlled chamber at a temperature of 28 °C in the dark for use in each experiment.
A preliminary experiment was conducted to investigate the effect of rice cultivar mixture proportions on four common paddy weeds (barnyardgrass,L. chinensis, C. difformis and E. prostrata ). Eight pre-germinated seeds were sown in each pot (12 cm diameter ×10 cm deep) containing 1000 g soil. Four rice seeds in a 3:1, 2:2 and 1:3 (focal cultivar: related cultivar) ratio were placed uniformly in the central area while four weed seeds of one species were sown surrounding the rice in the pot (Supplementary Figure 1). Monocultures of eight weed seeds served as controls. The weed seedlings were harvested after 4 weeks and their dry weights were recorded. This design was replicated for each focal cultivar and weed species. As a result (Supplementary Table 2), the 2:2 ratio was used for each subsequent experiment (Experiments 1- 4) as described below.
Experiment 1- the effect of rice cultivar mixtures on the performance of four weeds and their mixtures with barnyardgrass. Eight pre-germinated seeds were sown in each pot as described above. Four rice seeds for each cultivar pair were placed uniformly in the central area of each pot. Four weed seeds for each species, or two barnyardgrass with two other weed seeds, were sown surrounding the rice. Controls were monocultures of eight weed seeds for each species and their mixtures with barnyardgrass. Mixtures with barnyardgrass reflect the dominance of the species as a paddy weed and its regular co-occurrence with other weeds. Weed seedlings were harvested at the three-leaf rice stage and their shoots and roots were separated and dry weight of each was recorded.
Experiment 2- allelochemical production and grain biomass of rice cultivar mixtures in response to barnyardgrass and its mixtures with other weeds. Eight pre-germinated seeds were sown in a pot as described above. Four rice seeds for each cultivar pair were placed uniformly in the central area. Four barnyardgrass seeds or two barnyardgrass with two other weed seeds were sown surrounding the rice in each pot. Monocultures of each rice cultivar and eight weed seeds for each species and their mixtures with barnyardgrass served as controls. Allelopathic rice leaves produce allelochemicals at their early growth stages and then release them into soil through the roots that reach maximal level at the five-leaf stage and then decrease dramatically (Kong et al., 2004; Kong, Li, Hu, Xu, & Wang, 2006). This experiment focused on allelochemical production but not allelochemical release. Accordingly, rice leaves were sampled at the three-leaf stage and quantification of rice allelochemicals conducted as described below. Weed plants and rice grains were each harvested at the rice mature stage and their dry weight of each was recorded.
Experiment 3- soil microbial community in response to rice cultivar mixtures in the presence of barnyardgrass, the dominant weed competitor. Eight pre-germinated seeds were sown in each pot as described above. Four rice seeds for each cultivar pair were placed uniformly in the central area. Four barnyardgrass seeds were sown surrounding the rice in the pot. Monocultures of barnyardgrass or focal rice cultivar served as controls. At the three-leaf stage, a soil core with 2.5 cm diameter and 5 cm depth was sampled from the central area of each pot. Visible root fragments were removed and the sample freeze-dried for the analysis of phospholipid fatty acids (PLFA) as described below.
Experiment 4- the performance and resource partitioning of barnyardgrass in response to potential signaling-chemical in rice cultivar mixtures. Kin recognition in allelopathic rice cultivars involves the nitrogen-rich compound allantoin (Yang, Li, Xu, & Kong, 2018). Accordingly, allantoin and its equal quantity of NH4NO3 were applied into rice cultivar mixtures in the presence of barnyardgrass. Eight pre-germinated seeds were sown in each pot (12 cm ×10 cm) containing 1000 g soil. Four rice seeds for each cultivar pair were placed uniformly in the central area of each pot. Four barnyardgrass seeds were sown surrounding the rice in the pot and monocultures of barnyardgrass served as controls. The pots received either allantoin at a concentration of 0.32 mM (50 μg g-1 dry soil), NH4NO3at a concentration of 0.63 mM (50 µg g−1 dry soil), or distilled water, respectively. The NH4NO3 treatment represents a non-signaling nitrogen source to contrast with the signaling and high N content allantoin treatment. Barnyardgrass seedlings were harvested at three-leaf stage, separating their shoots and roots. The dry weights of each root and shoot sample were recorded and their carbon and nitrogen contents were each measured with a CN elemental analyzer (Elementar vario EL cube, Elementar Analysensysteme GmbH, Langenselbold, Germany), at a combustion temperature of 960 ℃.
Experiment 5- root placement patterns between focal rice cultivar and weeds in response to rice cultivars of different relatedness. A series of window rhizoboxes (30 cm long ×3 cm wide ×30 cm high) were made with a clear plexiglass sheet containing 2000 g soil. Each rhizobox was vertically divided into six equal parts, and three rice and three weeds for each species were sown in the rhizoboxes according to the order of two plants of related cultivars, a plant of the focal cultivar, and three weed plants from the left to the right (Supplementary Figure 2). Monocultures of six plants barnyardgrass or focal cultivars served as controls. Window rhizoboxes were placed in racks at an angle of 40° with the clear plexiglass facing down and away from the light source. Rhizoboxes were opened after 4 weeks and the root systems were scanned to a gray scale image and, divided into right and left parts, and the roots freeze-dried for biomass assessment. Each image was processed with WINRHIZO (Regent Instruments, Quebec, Canada) to obtain root length, and horizontal asymmetry in root length. Horizontal asymmetry in root biomass was also calculated from root dry weight of the left (towards rice) and right (towards weed) halves of the root system.