MATERIALS AND METHODS

Plant Materials and Field Tests
Field experiments of this study were conducted during three successive growing seasons (2002-2004) at Kaha vegetable research station, Horticulture Research Institute (HRI), ARC which is located 30 Km North Cairo, in Qualubia governorate. Tomato varieties of TYLCV-susceptible female (Edkawy, Castle Rock, Strain-B, Peto-86 and Marmmande) which is commercially available, were obtained from Deptartment of vegetables, HRI, ARC. However, TYLCV-resistant male (Favi-9) was kindly forwarded by Volcani Center, Bet Degan, Israel as part of regional collaboration through MERC project # M21-037. Favi-9 is derived from Lycopersecon hirsutum. First season (2002), seeds of TYLCV-resistant male and TYLCV-susceptible females were sown in the field for confirmation on purity and stability before crossing process. Crossing between previously mentioned male and different females was curried out in the greenhouse to get F1 seeds of each. Seeds of parents and their F₁’s were sown in the field to be evaluated and to get F₂ seeds during 2003 growing season. Back Cross 1 (BC1) and Back Cross 2 (BC2) were carried out in the greenhouse and produced seeds as well as F2 seeds were sown in the field during 2004 growing season for evaluation.

Field Evaluation
Seeds of six population (P1, P2, F1, BC1, BC2 and F2) of each of five crosses i.e., 1-EdkawyxFavi-9, 2-Castle RockxFavi-9, 3-Strain-BxFavi-9, 4-Peto-86xFavi-9 and 5-Marmmande x Favi-9 were sown in the greenhouse to insure production of virus-free seedlings. Seedlings of P1, P2 and F1, were transplanted in three rows. Seedlings of BC1 and BC2 were transplanted in four rows. Seedlings of F2 were transplanted in 6 rows during growing season in a complete randomized block design with three replicates. Each row was 10.0 m long and 30.0 cm wide. Pesticides were not applied during the growing season to insure TYLCV-infection spreading out. Filed inspection is done once a week for virus infection using visual examination for viral symptoms (leaf curling, leaf yellowing, dwarfing and compact of vegetative growth). Score scale of Muniyappa et al. (1991) for field inspection was used. Scale degrees used are: 1 = no symptoms, 2 = mild symptoms, 3 = moderate symptoms, 4 = sever symptoms and 5 = very sever symptoms. As indicated by Vidavsky and Czosnek (1998), individuals were selected for TYLCV resistance (no symptoms without detectable viral DNA) and tolerance (no symptoms but with detectable viral DNA). Consequently, no symptoms (0-1 degrees) = highly tolerance, mild symptoms (1-2 degrees) = tolerance, moderate symptoms (2-3 degrees) = moderate susceptible, severe symptoms (3-4 degrees) = susceptible and very severe (4-5 degrees) = very susceptible.

Statistical Analysis of Genetic Components
Heterosis was estimated according to the mid-parent value method that was used by Sinha and Khanna (1975). Broad Sense Heritability (BSH) was measured using the method of Allord (1960). Potency ratio was measured according to Smith (1952) methodology. Estimation for number of genes those are controlling TYLCV-resistance character was made according to the method of Castle and Wright (1921). Mean scale for TYLCV severity was estimated as (dsxnp)/tp = mean degree of severity. ds = degree of severity, np = number of infected plants and tp = total plant examined.

Following are detailed equations used in this study:

Heritability

It was calculated using Broad Sense Heritability (BSH): It was calculated using the following equation (Allord, 1960)

(1)

The Number of Effective Factors Controlling the Measured Trait

It was calculated using: Weber’s formula that was modified by Castle and Wright (1921).

(2)

(3)

Potency ratio

The relative potency of gene set (P) was used to determine the direction of dominance according to the formula of Smith (1952):

(4)

The absence of dominance is assumed when the difference between the parents was significant and F₁-MP was not significant.

Complete dominance is assumed when potency ratio equal to ±1.0.

Partial dominance is considered when potency ratio is between + 1.0 and -1.0, but not equal zero.

Over dominance (Heterosis) is assumed when potency ratio exceeded ±1.0.

Heterosis

Variance of population mean, which are different from each other, reveals the importance of improving this character by means of recurrent selection. Moreover, coefficient of variance (CV) reveals also, population variability could be exploited in further breeding program by isolating and developing tomato with TYLCV-tolerance character.

RESULTS AND DISCUSSION

TYLCV disease severity and population reaction recorded for six populations (P1, P2, F1, BC1, BC2 and F2) of the five crosses are presented in Table 1 and Fig. 1-5. Parents (P1s and P2) had different degrees of TYLCV tolerance for all crosses (Fig. 1). It is obvious that the P1s are very susceptible whereas P2 is tolerant (Table 1 and Fig. 1). Mean severity of TYLCV for F₁’s population was (3.432, 3.594, 3.343, 3.565 and 3.346 for all crosses, respectively) (Table 1) which lied between the mid-parent value (3.049, 3.128, 3.086, 3.271 and 3.165 for all crosses, respectively) and it was nearer to mean of the susceptible parent (Table 1 and Fig. 1). The population reaction curve of F1’s, against TYLCV infection in all crosses, has the same shape with a peak at scale °4 (susceptible). They never show very susceptibility reaction (scale °5) (Fig. 2) meaning TYLCV susceptibility character is moving towards tolerance. So, it could be suggested that there are some sort of partial dominance towered the lower parent of TYLCV-tolerance character for all crosses. Reaction of F2’s population has S-shape curve in all crosses except cross # 4 (Peto-86xFavi-9). Equal percentage of individuals has susceptibility reaction that expressed by scale 3 and °4 (Fig. 3). Moreover, mean severity of F2 was greatly decreased than those of F1’s individuals in all crosses except the cross # 3 (Strain -BxFavi-9) and cross # 4 (Peto-86xFavi-9) (Table 1 and Figs 3). F2’s population-reaction expresses, around 10%, are TYLCV tolerant and around 30% are TYLCV-susceptibility (Fig. 3). Population-reaction expression toward TYLCV tolerance has moved from 0.0-10% (Fig 1 and 3). Backcross one (BC1’s) is presented in Fig. 4. BC1 individuals reacted in S-shape curve in all crosses except Crosses # 4 (Peto-86xFavi-9) (Fig. 4). They were widely distributed with a High tendency towards this parent (susceptible one) (Table 1). Backcross two (BC2’s) is presented in Fig. 5. BC2 populations had the same S-shape reaction curve, in all crosses, against TYLCV infection except cross # 1 (Edkawi x Favi-9) (Fig. 5). They were widely distributed with a high tendency towards the tolerant parent (Favi-9) for all tomato crosses (Table 1).

Statistical analysis of quantitative genetic parameters for different previous five tomato crosses is presented in Table 2. Mid-parent heterosis (relative heterosis) had positive values (12.58, 14.9, 8.33, 8.98 and 5.72) for all five crosses respectively (Table 2). Better parent heterosis (heterobeltiosis) values were (118.4, 128.7, 112.7, 126.9 and 112.9) for the five crosses, respectively (Table 2). Those results mean that, there is a hybrid vigor towards the better parent. It has been confirmed, also, by potency ratio values. Low negative values of potency ratio (-0.2596,-0.2999,-0.16981,-0.1728 and -0.1137) indicating that TYLCV tolerance behaved as partial recessive toward better parent in this trait (Table 2). These findings were in accordance with those obtained by Hassan and Abdel-Ati (1999) as well as Vidavisky and Czosnek (1998) who reported that TYLCV resistance was a recessive trait. Heritability estimate showed that broad sense for TYLCV resistance were (55.79, 83.27, 59.31, 88.38 and 75.64), for all crosses, respectively (Table 2). This is indicating that there is a minor role of the

environment on this character as general except in crosses # 1 (Favi-9 x Edkawi) and # 3 (Favi-9 xStrain-B). This indication has been confirmed by the less number of minimum gene value obtained in Table 2. These results are coincided with those obtained by Hassan et al. (1984) who estimated BSH as 85 and 62% for TYLCV tolerance derived from of L. pimpinellifolium LA 121 and LA 373, respectively.

Table 1:	Reaction of six-population individuals of five crosses against natural TYLCV infection during 2002-2004 at ARC, Research Station
Sx-= Stander error; CV = Coefficient of variance

Fig. 1:	Reaction of susceptible and tolerant tomato parents against natural TYLCV infection at ARC Research Station during 2002

Fig. 2:	Reaction of F1 population of Different 5 crosses against natural TYLCV infection at ARC Research Station during 2003

Fig. 3:	Reaction of F2 population of 5 crosses against natural TYLCV infection at ARC Research Station during 2004

Fig. 4:	Reaction of BC1 population of different 5 crosses against natural TYLCV infection at ARC Research Station during 2003

The values of inbreeding depression (ID) are presented in Table 2. Those values are -17.7, -4.82, -27.66, -18.7and -9.98 in all five crosses, respectively (Table 2). Those large amounts of inbreeding depression for TYLCV trait in all crosses have been expected because of the large degree of heterosis showed, in this study, indicating presence of non-additive gene action (Allord, 1960).

Fig. 5:	Reaction of BC2 population of different 5 crosses against natural TYLCV infection at ARC Research Station during 2004

Table 2:	Genetic components i.e., heterosis, Board Sense Heritability (BSH) potency ratio, inbreeding depression and minimum number of genes for TYLCV tolerance in all crosses

One pair of genes could be estimated in crosses # 1 (Favi-9x Edkawi) and # 3 (Favi-9x Strain-B) (Table 2). Two pair of genes could be estimated in crosses # 2 (Favi-9xCastle Rock), # 4 (Favi-9xPeto-86) and #5 (Favi-9 xMarmmande) using the equation of Castle and Wright (1921) (Table 2). One pair of genes finding is coincided different reports (Kasrow, 1989; Laterrot, 1990; Pilowsky and Cohen, 1974) which indicated that TYLCV-resistance/tolerance is controlled by one pair of genes. Two pair of genes finding agree with those obtained by Banerjee and Kalloo (1987) who estimated that 2 pairs of genes are controlling this character. Mean while, Vidavisky and Czosnek (1998) reported that TYLCV-resistance is controlled by 2 to 3 pairs of genes. Also Nainar and Pappioha (2002) found that TYLCV-resistance in L. hirsutum is controlled by 3 pairs of genes. Vidavsky and Czosnek (1998) indicated that analysis of segregation of susceptibility, tolerance and resistance during the BC1F1 to BC1F4 crosses indicated that tolerance is controlled by a dominant major gene and resistance is controlled by 2 to 3 additive recessive genes.

ACKNOWLEDGEMENT

MERC project # M21-037 is acknowledged for supporting this study.