Abstract:
Fifty nine tomato accessions were characterized for 13 quantitative and 23
qualitative attributes at Agricultural Research Institute, Mingora, Swat, during 2006
under lath-house conditions. Seven selected tomato accessions along with one
commercial variety were crossed in half diallel fashion, during 2007 and evaluated for
inheritance pattern of important traits during 2008. The germplasm exhibited a wide
range for various parameters accompanied with substantial variation for number of
NFPC, DFFR, FrW, FL and FW. Based on descriptive statistics, the germplasm was
categorized into different groups. Significant correlations of both positive and
negative nature were recorded among various attributes both for quantitative and
qualitative traits. First four principal components with eigen values >1 contributed
70.0 and 65.3% of the variability for quantitative and qualitative traits, respectively.
Quantitative traits viz. DF, DFFR, FrW, pH of juice, LL, LW, FL and FW contributed
more positively towards PC1, while DF, PLH, ST, IL and LW contributed toward PC2.
Qualitative traits viz. MFIFC, ECMF, MFIFCI and UFS contributed more positively
with relatively greater magnitude of variance to PC1, while, FS, FF, and LP
contributed towards PC2, ECMF, MFIFC, UFS, FF and SC contributed maximally to
PC3, MFIFC, MFIFCI, UFS, SP, and LP contributed positively to PC4, respectively.
Tomato accessions were grouped into two main groups comprising two and three
clusters for quantitative traits, while, three main groups of five, two and seven sub
clusters for qualitative traits, respectively, based on Ward’s method of Euclidean
dissimilarity coefficient matrices.
Results regarding inheritance of quantitative attributes indicated that parents
P28, P30, P45 and P51 performed relatively better per se as well as in hybrid
combinations for most of the traits. Hybrids P28×P51, P45×P51 were selected on the
basis of early maturity and superiority of yield related fruit attributes. Maximum mid
parent heterosis of 53.1 % was observed for NFPC, 20.0% for NFrPC, 32.7% for FL,
10.6% for FW, 48.7 % for FrW, 34.9% for YPP, 59.0% for VL, -53.8% for IL, -2.7%
for DFFR, 73.3% for SG, -8.2% for pH of juice, 86.4% for TSS and 34.1% for RS,
respectively.
Estimates of variances due to GCA and SCA (Griffing’s approach) exhibited
preponderance of non-additive gene action for all the traits except for yield plant-1and
vine length. All the parents exhibited high GCA effects for multiple traits and hence
could be utilized for the development of superior combinations in tomato
hybridization programs. The cross combinations with high SCA common for different
characters P5(45)×P6(51), P1(E-02)×P7(54) and P2(28)×P3(30)) could be used both for
hybrid as well as pure line breeding.
Estimates of genetic parameters (Hayman’s approach) revealed preponderance
of non-additive gene action for all traits except YPP and DFFR. Unequal positive and
negative allelic frequencies among the parents were found for YPP, NFPC, FL, FrW,
VL and IL. Directional positive allelic dominance was observed for NFPC, FrW and
IL, negative directional dominance was recorded for YPP, NFrPC, DFFR and PLH,
while bidirectional dominance of genes was evinced by FL, FW, SG, pH of juice, TSS
and RS, respectively. Prevalence of recessive over dominant genes was obvious for
YPP, NFrPC, DFFR and PLH, while for all other traits, dominant and recessive genes
were exhibiting equal proportion in the parents. Symmetrical gene distribution with
positive and negative effects was recorded for NFrPC, DFFR, PLH, SG, TSS and RS,
respectively. Heritability estimates in broad sense were high for all traits, whereas,
heritability in narrow sense was high for YPP, DFFR and PLH. The graphic analysis
revealed over-dominance for all traits except YPP and DFFR, suggesting
ineffectiveness of selection in early segregating generations for traits improvement,
while the two traits with partial dominance could provide the basis for effective
selection in early segregating generations for the improvement of these parameters.