KARAKTERISASI MORFOLOGI DAN ANALISIS KLASTER TIGA KULTIVAR NANGKA LOKAL (Artocarpus heterophyllus Lam.) DI KABUPATEN LAMONGAN, JAWA TIMUR, INDONESIA

MORPHOLOGICAL CHARACTERIZATION AND CLUSTERING ANALYSIS OF THREE LOCAL JACKFRUIT CULTIVARS (Artocarpus heterophyllus Lam.) IN LAMONGAN DISTRICT, EAST JAVA, INDONESIA

Kabupaten Lamongan merupakan salah satu penghasil nangka (Artocarpus heterophyllus Lam.) di Jawa Timur, Indonesia, yang telah teridentifikasi ada tiga kultivar lokal yaitu ‘Nangka Bubur’, ‘Nangka Kandel’, dan ‘Nangka Salak’. Penelitian ini bertujuan untuk mengkarakterisasi dan mengelompokkan karakter morfologi dari tiga kultivar asal Lamongan tersebut. Enam pohon per kultivar dinilai berdasarkan 41 karakter morfologi (32 kualitatif dan 9 kuantitatif). Data dianalisis menggunakan analisis komponen utama (PCA), pengelompokan k-means, pengelompokan hierarki, korelasi, dan plot heatmap. Metode pengelompokan berhasil memisahkan tiga kultivar. Dari 41 karakter, 4 karakter seragam dan memiliki varian terkecil dikeluarkan. Analisis korelasi mengidentifikasi karakter yang berkorelasi positif seperti kerapatan duri, bentuk buah, diameter duri, tekstur daging buah, bentuk daging buah, warna duri. Berdasarkan plot kepentingan variabel, PCA biplot, dan analisis heatmap, 11 karakter diagnostik didentifikasi: ketebalan mesokarp, panjang duri, warna daging buah, bentuk buah, bentuk duri, warna kulit biji, bentuk biji, tekstur daging buah, kerapatan duri, bentuk daging buah, dan perlekatan tangkai pada buah. Penemuan tersebut akan memberikan wawasan bermanfaat dalam proses perbaikan pertanian, khususnya yang berkaitan dengan nangka.

Kata Kunci: Artocarpus heterophyllus, analisis pengelompokan, analisis multivariat, PCA, pertanian berkelanjutan

Lamongan District is among jackfruit (Artocarpus heterophyllus Lam.) producers in East Java, Indonesia, where three local cultivars including `Nangka Bubur`, `Nangka Kandel`, and `Nangka Salak` have been identified. This study aimed to characterize and cluster the morphology of these 3 cultivars from Lamongan. Six trees per cultivar were assessed for 41 morphological characters (32 qualitative and 9 quantitative. Data were analyzed using principal component analysis (PCA), k-means clustering, hierarchical clustering, correlation, and heatmap plots. Clustering methods succesfully separated the tree cultivars. From the 41 characters, 4 uniform and least variance characters were excluded. Correlation analysis identified positively correlated characters such as spine density, fruit shape, spine diameter, flake texture, flake shape, and spine color. Based on variable importance plot, biplot PCA, and heatmap analysis, 11 diagnostic characters were identified: mesocarp thickness, spine length, pulp color at ripe stage, fruit shape, shape of spine, seed coat color, seed shape, flake texture, spine density, flake shape, and stalk attachment to fruit. The stated discoveries would provide helpful insight into the process of agricultural improvement, particularly concerning jackfruit.

Keywords: Artocarpus heterophyllus, clustering analysis, multivariate analysis, PCA, sustainable agriculture