*4.1. Plant Materials*

We obtained WRC accessions from the National Agriculture and Food Research Organization (NARO) Genebank in Tsukuba, Ibaraki, Japan. The selected WRC 5 (NABA) and WRC 6 (PULUIK ARANG) lines were grown in an outside paddy field or in 1/5000 Wagner pots filled with soil (Bon-sol #2; Sumitomo Chemical, Tokyo, Japan; containing 1.5 g each N, P, and K) in Itakura, Gunma, Japan (36◦13-23"N 139◦36-37"E) in 2018 to 2019. First, to compare the e ffect of the time of application of P fertilizer, 1 g of P fertilizer (as Ca(H2PO4)2) containing 0.26 g of P was applied into the pots at the seedling stage or heading stage in 2018. Second, we applied di fferent amounts of P fertilizer, 1 g (P1) and 4 g (P4), containing respectively 0.26 and 1.1 g P at the seedling stage in 2019. Developing (10 DAF) and matured (30 DAF) grains were used for further analysis. In the pot experiments, 2 plants were planted in a pot and we used 2 pots to generate 4 replicates. Rice plants grown in the outside paddy field were used for the analysis of yield-related traits. Panicle number, 1000 seed weight, and total yield per plant were analyzed in 5 replicates and panicle length and weight were analyzed in 10 replicates.

### *4.2. Determination of the PA and Pi Content*

The Pi and PA content were determined using the enzymatic method [36]. Pi and PA were extracted for 17 h in 0.66 M HCl from one grain of brown rice, and then the Pi and PA content was determined colorimetrically using the Phytic Acid (Phytate)/Total Phosphorus kit (K-PHYT; Megazyme International, Wicklow, Ireland) according to the manufacturer's instruction after neutralization. This kit measures Pi released from the extracted grain sample after treatment with phytase and alkaline phosphatase. The free Pi content was estimated from samples not treated with phytase.
