To analyze the characterist], cs of rice varieties in the locations for selection in Toyama Prefecture, the author carried out various statistical studies on yield trial clata collected at 14 Iocations in the prefecture (Fig. 1). Tayama Prefecture has an area of 4, 252 square kilo meters, ancl it lies on the western coast of the central pal't of .Japan betlveen the Sea of Japan ancl the Nortern Japan Alps (Fig. 1). The climate of Toyama is characterized by colcl of 'Toohok pattern' in spring, hi, gh temparature to asc.end urgently with coming in Summer ancl the early autumnal colcl in autumn. This Prefecture ha
s the valous soil type called sancly shallolv or loam shallow tillage, diluvium humic and ill-clrained soil, and so fourth on. It is generally recognizecl that a high-yielding Iocation is suitable for tests. However the variance components and heritability values based on the analysis of variance failecl to reveal any relationship betlveen the heritability of yield ancl mean yield, indicatin'g that hig'h yield locations w'ere not allvays suitable selection sites in terms of heritability values (Table 1, 2). The heritability values of yield were high in Kurobe. Fuchu, Toide. Oyabe, Himi and Namerikalv, a (Table 3, Fig. 5). 'there was sig
nificant c.orrelation between the regression coefficient and primary principal component score with the year variance, indicating the presence of large val'iations among years (Fig. 6). Reg'ression analysis, principal c, omponent analysis ancl cluster anaalysis enablecl to identify highly suitable locations. Such as Nyuzen, Kurobe. Namerikawa. Toide, Fukuno, Fuchu, Takaoka. Himi and Yatsuo (Table 4, 5, 6, 7, Fig. 2, 3, 4). From these results, it is conciclerecl that the exc.ellent variants from hybricl populations are selected efficiently by utilizing the above characteristic (for height in yield ability, heritability and regression coefficient, also, for magnitude in genetic year error variance and the 1st principal component score) of locations coincided with the breeclin'g objective.
View full abstract