3.1 Kin recognition in allelopathic rice interference with paddy weeds
When allelopathic rice cultivars were each paired with itself or other rice cultivars, the growth of four common paddy weeds (Echinochloa crusgalli, Leptochloa chinensis, Cyperus difformis andEclipta prostrata ) were inhibited. However, inhibitory effects were dependent on the genetic relatedness and proportions of rice cultivars. The mixtures with closely related cultivars (kin) had stronger weed inhibition than the mixtures with distantly related cultivars (non-kin) regardless of proportions or weed species. Significant inhibition occurred at the 2:2 proportion with an exception of the broadleaf weed E. prostrata (Supplementary Table 2).
When each mixture of allelopathic rice cultivars with other cultivars at different levels of genetic relatedness was grown in a paddy field at a 1:1 proportion, the emergence and growth of paddy weeds varied greatly in mixed culture with kin and non-kin cultivars (all weeds pooled, Figure 1). Allelopathic rice cultivars grown alone and with kin cultivars showed stronger weed inhibition than when grown with non-kin cultivars. Compared with allelopathic rice cultivars and their mixtures with kin cultivars, a significant reduction in weed inhibition occurred in the mixtures with non-kin cultivars. The results implied the role of kin recognition in mediating allelopathic rice interference with paddy weeds.
3.2 Kin recognition meditated allelopathic rice against weeds at the root level
Greenhouse experiment 1 further confirmed that rice cultivars inhibited the growth of paddy weeds and that the inhibition was dependent on the genetic relatedness of the mixture (Figure 2). Allelopathic rice cultivars grown with kin cultivars caused stronger weed inhibition than mixtures with non-kin cultivars. Regardless of weed species and the genetic relatedness of rice cultivars, there was always stronger inhibition in root growth than in shoot growth of weeds (Figure 2). Further, rhizobox experiment 5 revealed that rice and weeds showed different root placement patterns measured by horizontal asymmetry in response to interacting neighbors. (Supplementary Figure 3 & 4). Rice roots always grew towards the roots of weeds (intrusive pattern), while the roots of all weeds trended to avoid growing towards roots of rice (avoidance pattern). However, root placement patterns between rice cultivars depended on genetic relatedness. Focal rice cultivars grown by themselves or with closely related cultivars had a more root intrusion towards paddy weeds than when grown with distantly related cultivars (Figure 3). Similarly, the largest avoidance by weed roots occurred in allelopathic cultivars grown by themselves or in mixture with closely related cultivars (Figure 4). These results indicated that kin recognition in allelopathic rice could regulate belowground interactions with weeds by optimizing root placement and distribution.