Correlation and Path Coefficient Analysis: A Study of 260 Genotypes of Tomato for Yield and Quality Traits
Gautham Suresh SP
*
Department of Vegetable Science, University of Horticultural Sciences, Bagalkot, Karnataka, India.
Shashikanth Evoor
Technical Officer to DOE, University of Horticultural Sciences, Bagalkot, Karnataka, India.
Fakrudin B
College of Horticultural Engineering & Food Technology, Devihosur, Karnataka, India.
Raveendra Jawadagi
Department of Vegetable Science, College of Horticulture, Bagalkot, Karnataka, India.
Vasudeva KR
Department of Post-Harvest Technology, College of Horticulture, Bangalore, India.
Basavarajappa MP
Department of Plant Pathology, College of Horticulture, Bagalkot, Karnataka, India.
Vinaykumar MM
College of Horticultural Engineering & Food Technology, Devihosur, Karnataka, India.
*Author to whom correspondence should be addressed.
Abstract
To enhance tomato yield through breeding programs, it is crucial to understand the complex interrelationships among various yield-contributing traits. This study employed phenotypic correlation and path coefficient analysis to dissect these relationships and identify the key selection criteria. A total of 260 tomato accessions were compiled in collaboration with the Indian Institute of Vegetable Research (IIVR), Varanasi, and UHS Bagalkot and were evaluated at the experimental block of Department of Biotechnology, College of Horticulture, Bangalore during the year 2020 by means of augmented block design for morphological attributes, yield, and fruit quality traits. Phenotypic correlation studies were conducted using a diverse set of 260 tomato germplasm accessions. Several traits exhibited highly significant and positive phenotypic associations with total yield per plant, including average fruit weight (rp= 0.598), pericarp thickness (rp= 0.553), fruit length (rp = 0.498), fruit diameter (rp= 0.489), fruit firmness (rp= 0.287), number of seeds per fruit (rp= 0.236), and number of locules per fruit (rp= 0.145). Conversely, the number of fruits per cluster showed a significant negative association with the total yield per plant (rp= -0.179). While correlation indicates association, it does not reveal direct influence or causal pathways. Therefore, path coefficient analysis was used to partition the total correlation into direct and indirect effects. This more powerful statistical technique helps breeders understand the traits that exert a large direct impact on yield. Path analysis demonstrated that average fruit weight had the maximum direct effect on fruit yield. The number of fruit clusters per plant also exhibited a strong direct effect on yield. Traits such as the number of fruits per cluster (0.2148), fruit length (0.1990), and pericarp thickness (0.1474) had moderate to low positive direct effects on yield. The analysis also revealed significant indirect effects, often channeled through average fruit weight. Pericarp thickness (0.7323), fruit diameter (0.7170), fruit length (0.6812), and fruit firmness (0.3708) showed highly positive indirect effects on yield through average fruit weight, whereas the number of fruits per cluster (-0.5365) and number of fruit clusters per plant (-0.5000) had strong negative indirect effects through this same intermediary. Based on these findings, the correlation and path analysis results are invaluable for guiding selection strategies in tomato breeding, highlighting traits such as average fruit weight and the number of fruit clusters per plant for prioritized selection pressure because of their substantial direct contributions to increasing total fruit yield.
Keywords: Tomato, Solanum lycopersicum, yield, correlation, path analysis, breeding