The use of synthetic pesticides for pest and disease control in crops raises significant health and environmental concerns. As an eco-friendly alternative, biological control using biofungicides such as Trichoderma sp. has become increasingly important. Effective biofungicide production necessitates precise determination of Trichoderma sp. population levels. This study aims to identify Trichoderma sp. population levels (105, 106, and 107 CFU/ml) based on aroma data from an Electronic Nose (E-nose). The experiment began with the rejuvenation of Trichoderma sp. samples, followed by inoculation on three organic materials (sugarcane bagasse, peat, and bran). Gas sensor data were collected using the E-nose after a 7-day incubation period in glass bottles. Feature extraction, using statistical and time domain methods, was performed to identify the fungal population. Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) demonstrated that the E-nose sensor responses effectively differentiated Trichoderma sp. population levels based on the type of organic material. For the three organic materials, PCA revealed that PC1 accounted for 79.76% and PC2 for 13.49% of the variance, while LDA showed LD1 accounted for 62.54% and LD2 for 28.22% of the variance. Specifically, for the sugarcane bagasse, PCA indicated PC1 at 79.77% and PC2 at 13.49%, with LDA showing LD1 at 60.93% and LD2 at 28.58%.
Copyrights © 2025