Genetic parameters for yield and processing characteristics were estimated using 26 experimental processing tomato hybrids in an experiment conducted in Cristalina, Goiás State, Brazil. The commercial hybrids 'Hypell-108' and 'RTP1095' were used as standards. The experimental design was a randomized complete block with three replications. Each plot consisted of two rows (30 plants each) spaced 0.50 m apart with a within row distance of 0.25 m. The following parameters were evaluated: mature fruit yield (FM), green fruit yield (FV), total fruit yield (PT), uniformity of fruit ripening (UM = (FM/PT) x 100), soluble solids [brix (°B)], paste yield (RP), fruit color (CF) and fruit firmness (FF). Significant differences were found among the 26 experimental hybrids for FM, FV, PT, UM, °B, RP and CF. Experimental hybrids (as a group) were superior to the commercial hybrids for FM, FV, PT and RP. The commercial hybrids (as a group) had superior performance only for °B, as compared to experimental hybrids. However, six experimental hybrids had °B values similar to that of 'Hypell-108'. A negative genotypic correlation (-0.40) was observed between FV and FM. A negative (-0.27) genotypic correlation was observed between CF and °B. Both associations are undesirable from the breeding for processing industry standpoint. A multivariate analysis indicated FM, UM, °B and CF as being the traits with highest relevance to discriminate hybrid performance. A multi-trait selection indicated eight experimental hybrids with superior performance in this assay. The top one ('CNPH 1346' x 'CNPH 1272') had 40% and 58% higher paste yield production as compared to 'Hypeel-108' and 'RPT1095', respectively. The groups of experimental hybrids with the inbred lines 'CNPH 1299' or 'CNPH 1346' as parental had the best performances in this assessment. These F1 hybrids could be recommended for commercial production in Central Brazil.
Lycopersicon esculentum; industrial processing; genetic parameters
