Two f2 populations were produced from two indica-indica crosses and used for interval mapping of QTLs for yield and related characters (Lin et al. 1995; Lin et al. 1996). The F2 populations were grown in China National Rice Research Institute (CNRRI) in 1993 (referred to as F2 trial). RFLP analysis of 161 individual plants was made. The F3 lines of one of the populations, Tesanai/C.B., were grown at CNRRI (referred to as CNRRI F3 trial) and at International Rice Research Institute (IRRI) in 1994 (referred to as IRRI F3 trial), respectively. Eight traits, i.e., the number of productive panicles per plant (PAN), the number of filled grain per panicle (NFG), total number of grain per panicle (TNG), spikelet fertility (PERT), 1000-grain weight (TGW), filled grain weight per plant (GYD), plant height (PH) and panicle length (PAL), were scored in all the three trials. Interval mapping of QTLs was employed with a threshold of LOD = 2

Rice Genetics Newsletter Vol. 13

Table 1. Additive effect to QTLs for yield components and plant height detected in the

' CF3 = CNRRI F3 trail ; IF3= IRRI F3 trial.

Research Notes 129

by using the computer package Mapmaker/QTL (Lander and Botstein 1989). The identification of quantitative trait loci (QTLs) for yield components and plant height, and the analysis of QTL x environment interaction in the Tesanai/C.B. populations are reported herein.

Altogether 44 QTLs were identified in 18 intervals of nine chromosomes, including 3 for PAN, 5 for NFG, 6 for TNG, 3 for PERT, 7 for TGW, 5 for GYD, 8 for PH and 7 for PAL, The numbers of QTLs detected in two or three trials were 1 for PAN, 1 for NFG, 1 for TNG, none for FERT, 4 for TGW, 3 for GYD, 2 for PH and 5 for PAL. Of the 17 cases when a QTL was detected in more than one trial, the directions and magnitude of its additive effect agreed to each other in 16 and 14 cases, respectively (Table 1).

In all three trials, QTLs were frequently detected for related traits in the same intervals. In some cases, QTLs for related traits were not only simultaneously detected in the same interval, but also they were readily detected in different trials. The intervals of RG256-RG324A on chromosome 2 and RG620-RG214 on chromosome 4 provided good examples. In other cases, the traits for which QTLs were detected varied largely among different trials. In both situations, the directions of additive effect of QTLs for related traits were in agreement with few exception, no matter whether they were detected in the same trial or not. These results suggested that pleiotropism rather than close linkage of different QTLs was the major reason why QTLs for different traits were frequently detected in the same intervals.

Among the 18 intervals, 12 intervals harbored QTLs readily detected among different trials and/or demonstrated the above-depicted pleiotropism, involving 37 of the total 44 QTLs. The present investigation therefore suggested that the environmental factors have little effect on the detection of chromosome segments harboring QTLs although they might have strong influences on the detection of QTLs for any single trait. The directions of the effect of chromosomal segments harboring QTLs were also little affected by environmental factors.

Lin, H.X., H.R. Qian, J.Y. Zhuang, J. Lu, S.K. Min, Z.M. Xiong, N. Huang and K.L. Zheng, 1995. Interval

Lin, H.X., H.R. Qian, J.Y. Zhuang, J. Lu, S.K. Min, Z.M. Xiong, N. Huang and K.L. Zheng, 1996. RFLP

Lander, E.S., D. Botstein, 1989. Mapping Mendelian factors underlying quantitative traits using RFLP linkage