Tropical Agricultural Research Vol. 17:31 - 38 (2005) Breeding and Selection of Tomato Fi Hybrids for Yield and Fruit Quality Characters L.M.H.R. Alwis, A.L.T. Perera 1, H. Fonseka 2 and B.C.N. Peiris 3 Postgraduate Institute of Agriculture University of Peradeniya Sri Lanka ABSTRACT. Tomato (Lvcopersicon esculentum Miller) of the family Solanaceae is one of the most important vegetable crops grown in Sri Lanka. To understand the actual consumer requirements, especially in terms of fruit quality characters, an extensive survey was carried out. Results revealed that for home consumption, a majority preferred acidic, round to flat- shaped, medium-sized fruits, red in colour for curries, salads and sandwiches. The Government of Sri Lanka has also identified the production of local vegetable hybrids as a matter of high priority. With this information, a half diallel genetic crossing design was carried out to produce 36 F, hybrids using nine local and foreign varieties at the Agricultural Biotechnology Centre, University of Peradeniya. These were evaluated in the Meewatura farm of the Faculty of Agriculture, University of Peradeniya. Yield and fruit quality characters including brix, acidity and fruit shape and colour were evaluated. General combining ability (GCA) and specific combining ability (SCA) were significant indicating the importance of both additive and dominance gene action. Out of the 36 Fi tomato genotypes; T068 was selected as the best hybrid for the tourist industry as it showed heterobeltiosis for yield (48%) and low acidity, while T025 was selected as the best hybrid for local consumption as it too showed heterobeltiosis for yield (76%) with high acidity. Both hybrids showed that yield potentials were much higher than the currently used varieties. rNTRODUCTION Tomato (Lycopersicon esculentum Miller) of the family Solanaceae is one of the most important vegetable crops grown in Sri Lanka. The yield potential of the currently available varieties is 20 to 30 mt/ha (HORD1,2005) while the average yield of the country is 7.65 mt/ha.' The world average yield is 26.56 mt/ha (FAO, 2005). Even though the demand for quality vegetables is increasing due to the increased demand by super markets and the Department of Agricultural Biology, Faculty of Agriculture, University of Peradeniya, Sri Lanka. Horticultural Crop Research and Development Institute, Gannoruwa, Peradeniya, Sri Lanka. Department of Crop Science, Faculty of Agriculture, University of Peradeniya. Sri Lanka. Alwis et at.. tourist industry, the prices of many vegetables show a high fluctuation due to shortage of supply as well as gluts in supply (Central Bank of Sri Lanka, 2002). The availability of quality seeds to farmers is an important factor in the production of tomatoes in order to raise their productivity and thereby increase their income (Central Bank of Sri Lanka, 2002). However, in Sri Lanka, the non availability of quality seeds and suitable varieties to meet the requirements of local households and the tourist industry are identified as the major constraints in the expansion of profitable tomato cultivation. Tomato is used in curries, salads and sandwiches both by the local population as well as by the hotels in the tourist industry. Although variations in colour, flavour, size and shape of tomatoes exist in the present varieties, no single variety can satisfy the preferences for these characters of different target groups (Villareal, 1980; Weerasinghe et ai, 2004). Therefore, quality parameters have to be set for breeding fresh market tomatoes by investigating the preferences of different consumer groups. Local hybrid seed production is also a priority area for research in Sri Lanka. Therefore, the urgent need at the moment in tomato breeding is to produce hybrid varieties with high yielding capacity combined with fruit quality traits required by the local market. The objectives of this study were to assess the consumer preference for tomato, and to develop Fi hybrids showing heterosis for yield and important fruit quality traits. MATERIALS AND METHODS (i) Survey on consumer preference A survey on consumer preference was carried out using undergraduate students of the Faculty of Agriculture, University of Peradeniya, representing almost all the districts of Sri Lanka. A questionnaire was formatted to collect information from their homes on preferences for acidity, shape, size, pericarp thickness and colour of tomatoes for different purposes such as preparation of curries, salads, soups and sandwiches. Four hundred and twenty five filled questionnaires were statistically analyzed using Categorical Data Analysis. (ii) Diallel genetic design to produce F , hybrids A half diallel crossing programme (Haulier arid Miranda, 1995; Banga and Bang, 1998; Kurtiar et ai, 2002) was carried out in the greenhouse of the Agricultural Biotechnology Centre, University of Pereadeniya, using nine local and foreign tomato genotypes during Yala 2004 (Table 1). Seeds of Tomato varieties Tharindu (T002), Rashrni (T003), Thilini (T004), Rajitha (T005), T245 (T006), KWR (T007), Ravi (T008) and T146 (T009), were obtained from the Horticultural Crop Research and Development Institute, Gannoruwa, Peradeniya, Sri Lanka, and the foreign tomato variety Garden Delight (T001) was obtained from Serendib Horticulture (Pvt) Ltd. Kalagedihena, Sri Lanka. 32 Breeding and Selection of Tomato F, Hybrids Table 1. Half-diallel genetic design used in the study T001 T002 T003 T004 T005 T006 T007 T008 T009 T00I XXX T0I2 T013 T014 T015 T0I6 T017 T018 T0I9 T002 XXX T023 T024 T025 T026 T027 T028 T029 T003 XXX T034 T035 T036 T037 T038 T039 T004 XXX T045 T046 T047 T048 T049 T005 XXX T056 T057 T058 T059 T006 XXX T067 T068 T069 T007 XXX T078 T079 T008 XXX T089 T009 XXX The seeds were first sown in plastic pots and 28-day old plants were transplanted into black polythene bags (15cm x 50cm) for emasculation and pollination purposes. All cultural practices and fertilizer applications were carried out according to the recommendations of the Department of Agriculture, Sri Lanka. Seeds of the F, hybrids were cleaned, sun dried and stored at 4°C (AVRDC, 2001; George, 2002). Thirty six Fi hybrids and nine parental varieties were grown in the research field of the Meewatura farm of the Faculty of Agriculture, University of Peradeniya for evaluation during Yala 2004. A 7x7 balanced lattice design with eight replicates and 56 blocks was used for field evaluation (Cochran and Cox, 1957). During the field evaluation, the 36 F| hybrids and their parents were observed for any signs of tolerance/susceptibility to diseases, especially bacterial wilt. Number of days to 50% flowering, fruit shape, average fruit weight, average yield per plant, brix value and acid percentage of fruit juice (Wilbur, 1983) were recorded (PGRC, 1999; 1995). Data analysis was done using SAS computer software package. The analysis of variance was carried out according to the method of Griffing and the general combining ability (GCA) and specific combining ability (SCA) were calculated according to Haulier and Miranda (1995). RESULTS AND DISCUSSION The survey on consumer preference revealed that fruits with high acidity were preferred by 57% of the consumers for all household preparations, round to flat fruit shape by 4 5 % and medium fruit size by 55% of consumers. For curries, 47% consumers preferred red-coloured fruits, while 39% preferred red fruits for sandwiches; dark red-coloured fruits were preferred by 49% consumers for salads and 32% for sandwiches respectively. Forty one per cent preferred fruits with medium pericarp thickness and 38% preferred thin pericarps (Table 2). Fruits with low acidity were preferred for salads by the tourist industry (Villareal, 1980). 33 Alwis et al.. Table 2. Household consumer preference for tomato Household Curries Salads Sandwiches Soup Overall Preparation % % % % % Acidic 80 53 58 36 57 T a s t e Sweet 4 27 10 33 18 Red 47 37 39 41 C o , o u r Dark red 25 49 32 36 „ . . . . Medium 40 45 43 36 41 Pericarp thickness 32 33 37 48 38 Medium 54 56 55 Size 25 Large 22 28 25 Round 38 52 45 S h a P e Pear 16 13 15 According to the above results, the objective of the hybrid breeding programme was to select high yielding hybrids with hybrid vigour having medium sized, round to flat shaped, red-coloured, and acidic tomato types with medium pericarp thickness for home consumption and fruits with low acidity for tourist industry. Field evaluation Effects of GCA and SCA showed highly significant values (Table 3), indicating the importance of additive genetic variance in parental performance and the possibility of exploiting dominance gene effects in hybrids. GCA indicates the additive gene effects, while SCA indicates dominance and epistatic effects. GCA is useful for selecting parents with good combining abilities in breeding programmes, while SCA indicates the possibility of obtaining superior hybrids. Table 3. Analysis of variance of yield Source DF Type III SS MS F Value Pr>F Replicates 7 7030597.4046 1004371.0578 1.61 0.1335 Blocks (Rep) 48 41324690.0799 860931.0433 1.38 0.0633 Treatments 43 133573636.1177 3106363.6306 4.98 0.0001 GCA 8 18192176.749 2274022.094 3.65 0.0001 SCA 36 10470427.568 290845.210 0.470 0.0001 Error 225 140362208.3465 623832.0370 Corrected Total 323 319134991.5895 The parent T003 had high GCA for yield and average fruit weight, while the hybrid T029 had the highest SCA for the same two characters. The parent with the highest GCA for acid percentage in the fruit was T007, while T039 was the hybrid with the highest SCA value. The hybrid T023 had the lowest SCA value for fruit acidity (Table 4). 34 Breeding and Selection of Tomato F, Hybrids Table 4. Parents and F, hybrids having the highest GCA and SCA values CCA SCA Character Best Parents Mean Best Hybrid Mean Yield (g/plant) T003 550.88 T029 2679.26 Av. Fruit Wt. (g) T003 22.61 T029 69.85 Brix (High) T001 1.16 T018 9.25 Brix (Low) T006 -0.66 T024 0.04 Acid % (High) T007 6.32 T039 38.75 Acid % (Low) T001 -9.53 T023 -4.25 Days to 50% Flw T001 -7.20 T034 ' 5.69 Although these values will be useful for breeders as a general guideline in hybridization programmes, the parents with the highest GCA need not always produce the best hybrids (Weerasinghe et ai, 2004) and the hybrids with the highest SCA values may not necessarily show high heterosis. Since analysis of variance indicated significant SCA values, the eight higher yielding hybrids and the parents were ranked as shown in Table 5. Their performances in other traits are also given. As seen very clearly, no single hybrid was superior for all the traits. Eight hybrids gave higher yields than the best performing parent (T003) showing 'top' heterosis. The hybrids T034, T036 and T035 which produced large fruits also had high yields. T068 was the highest yielding hybrid with low brix and low acidity, while T025 ranked second in yield performance with low brix and high acidity. Table 6 shows the performance of hybrids that showed heterosis for yield. Of the seven hybrids that showed heterobeltiosis, two hybrids T068 and T025 were the best. T068 showed 47.86% heterosis while T025 showed 76.28%. Although they were not superior for all the other traits as well, T068 was the highest yielding hybrid with low acidity and T025 was high yielding with high acidity. These two hybrids were therefore selected as new tomato hybrids for recommendation for commercial cultivars. T068 is recommended for the hotel/tourist industry while T025 for household use. Their yield and fruit quality characters are given in Table 7. It is significant to note that the foreign variety T001 was extremely susceptible to bacterial wilt in the field. This applied to all the hybrids produced by using this variety as a parent. None of the local varieties or their hybrids showed any susceptibility to bacterial wilt in the field. The two selected hybrids showed total resistance to bacterial wilt in the field. 35 Table 5. Means and r anks of F , hybrids and paren ts for yield and fruit quality cha rac te r s Trea tment Average Yield/Plant Average Fru i t Weight Brix Acid % Days to 5 0 % Flower Fru i t Mean R a n k Mean R a n k Mean R a n k Mean R a n k Mean R a n k Hybrids 2669.2 a b T068 2669.2 a b 1 57.1 19 3.0 34 15.17 38 50 10 2 T02S 2529.2 a b c 2 44.5 29 2.9 39 25.41 14 50 8 2 T035 2509. l 8 b c 3 90.3 5 2.8 40 23.32 21 50 29 2 T036 2430.8 a b c 4 98.0 3 3.1 29 22.79 22 51 23 2 T039 2377 .8 a b c d 5 67.4 15 2.9 38 33.44 4 51 25 3 T034 2 3 7 7 . 1 a b c d 6 98.9 2 4.0 14 19.64 29 51 33 3 T023 2352 .4 a b c d 7 56.7 20 2.9 36 17.52 34 52 38 2 T026 2233 4«bcde 8 58.5 18 2.6 43 29.70 6 50 6 2 Parents T003 2 1 4 5 . 8 a b c d c f 12 100.1 1 3.3 26 24.17 20 52 35 3 T009 1846.4 b c d e f B h i 15 69.7 14 3.3 27 21.95 25 51 24 2 T007 16 50.1 26 4.2 11 35.03 2 50 18 3 T006 , 8 0 5 jcdefghi 17 75.8 11 3.1 30 29.10 8 50 7 2 T004 1495.2 d e f 8 h i 23 78.8 8 3.7 20 18.68 30 51 28 5 TOOS 1434.7 < k f 8 l , i 24 70.4 13 3.3 25 25.14 15 49 4 3 T002 1 4 1 6 . 8 * ^ 25 30.7 33 3.1 31 19.62 27 50 19 4 T008 1403 .6*^ ' 26 24.7 34 3.0 32 7.52 39 51 30 5 Means with same letters are not significantly different at p=0.05 Fruit shape: 2 = Slightly flattened, 3 = Round, 4 = High Round, 5 = Heart snap Breeding and Selection of Tomato K, Hybrids Table 6. Yield performance of the hybrids based on heterobeltiosis F, Hybrid Mean Yield (g/plant) Better Parent Mean Yield (g/plant) Heterosis % (>BP) T068 2669.2 T006 1805.1 47.86 T025 2529.2 T005 1434.7 76.28 T035 2509.1 T003 2145.8 16.93 T039 2377.8 T003 2145.8 10.81 T034 2377.1 T003 2145.8 10.78 T023 2352.4 T003 2145.8 9.63 T026 2233.4 T006 1805.1 20.60 Table 7. Performance of the selected hybrids F, Hybrid Mean Yield (mt/ha) Av. Fruit Wt. Brix Acid % Days to 50% Flw Friut Shape Colour T068 69.7 57.1 3.0 15.17 50 Round to flat Red T025 66.1 44.5 2.9 25.41 50 Round to flat Red National Av. Yield 7.65 Round to flat CONCLUSIONS The local population prefers acidic tomatoes for all preparations including curries, salads and sandwiches, whereas the hotels catering to tourists require tomatoes with low acid content. Since, GCA values need not necessarily indicate that parents with the highest GCA will produce the best hybrids, the diallel genetic design is by far the best and most popular breeding design to evaluate hybrid performance, as all possible crosses are made. As expected, the best performing hybrids were not from parents with the highest GCA values. This research also showed that the hybrid with the highest SCA did not show highest heterosis. However significant SCA indicated the possibility of producing superior hybrids. This was proven to be true. Field testing for tolerance/resistance to bacterial wilt was shown to be very reliable. The foreign variety used and its hybrids were totally susceptible to wilt, whereas the local parents and their hybrids showed extreme tolerance/resistance. Based on consumer preference and heterosis, two superior hybrids T068 and T025 were selected for households and hotels respectively. ACKNOWLEDGEMENTS The authors thank the Council for Agricultural Research and Policy (CARP) for providing a grant to conduct the research, the Director of the Postgraduate Institute for his valuable support, Dr. B.L. Peries, Department of Crop Science of Faculty of Agriculture, 37 AI wis era/., Staff of the Agricultural Biotechnology Centre, Horticulture Research and • Development Institute, Departments of Crop Science and Agricultural Engineering of Faculty of Agriculture and all others who assisted at various times during this research. REFERENCES AVRDC (2001). Hybrid seed production in tomato. Asian Vegetable Research and Development Centre [Online]. Cited May 2005. Available from internet http://www.avrdc.org.tw/Ic. Banga, S.S. and Banga, S.K. (1998). Hybrid cultivar development. Narosa Publishing House, India. Central Bank of Sri Lanka. (2002). Annual report. Central Bank of Sri Lanka. Colmbo, Sri Lanka, pp. 83 - 96. Cochran, W.G. and Cox, G.M. (1957). Experimental designs. John Wiley and Sons Inc. Singapore. FAO (2005). Food and Agriculture Organization of the United Nations [Online]. Cited May 2005. Available from internet http://www.fao.org/waicent/DOrtal/statistics en.asp. George, R.A.T. (2002). Vegetable seed production: 2 n d Ed. CABI Publishing, UK. Haulier, A.R. and Miranda, J.B. (1995). Quantitative genetics in maize breeding: 2 n d Ed. Iowa State University Press/Ames, USA. HORDI (2005). Horticultural Crop Research and Development Institute in Sri Lanka. [Online]. Cited May 2005. Available from internet . nttp://www.gov.lk/Agriculture/Agridept/Techninformations/Vege.html Kumar, S., Kumar, R., Kumar, S., Singh, M., Banerjee, M.K., and Rai, M. (2002). Hybrid seed production of Solanaceous vegetables: A practical manual. IIVR Technical Bulletin 9, India. PGRC (1995). Descriptors for vegetables and condiments. Plant Genetic Resources Centre, Department of Agriculture, Sri Lanka. PGRC (1999). Characterization catalogue on tomato. Plant Genetic Resources Centre, Department of Agriculture, Sri Lanka. Villareal, R. L. (1980). Tomatoes in tropics. Westview Press, Inc., USA. Weerasinghe, O.R., Perera, A.L.T., Costa, W.A.J.M., Jinadase, D.M. and Vishnukanthasingham, R. (2004). Production of tomato hybrids for dry zone conditions of Sri Lanka using combining ability analysis, heterosis and DNA testing procedures. Tropical Agricultural Research. 16: 79 - 90. Wilbur, A.G. (1983). Tomato production, processing and quality evaluation: 2 n d Ed. The AVI Publishing Company, Inc. USA. 38 http://www.avrdc.org.tw/Ic http://www.fao.org/waicent/DOrtal/statistics http://www.gov.lk/Agriculture/Agridept/Techninformations/Vege.html