Botanical Composition and Yields of Forages in Natural Pastures Using Principal Component Analysis and Cluster Dendrogram in South Sulawesi, Indonesia

Abstract

<p>This study aims to identify the composition and yields of forages in natural pastures using principal component analysis and cluster dendrogram in Gilireng and Majauleng subdistricts, Wajo regency, South Sulawesi, Indonesia. Pasture plants were collected using 20 randomly sampled quadrats to observe the botanical composition and yields of forages. The collected data regarding legume classification is analyzed by Multivariate Analysis of Variance (MANOVA) with PCA and dendrogram clustering by using R version 4.2.2. The botanical composition of natural pastures was majored by grasses, followed by legumes and weeds in both Subdistricts and the pasture total dry yields were 9.01 and 5.20 t/ha in Gilireng and Majauleng Subdistricts, respectively, to show the carrying capacity at 2.56 and 1.44 animal unit/ha, respectively. Principal component analysis for the biplot between the subdistricts and villages revealed the superior extensiveness of dimension 1 to dimension 2, but only a few types of outliner plants remained. Dimension 1 has a high kurtosis of 66.2%, while dimension 2 has a variance accounting for 26.5%. Furthermore, dendrogram analysis indicated that the agglomerative coefficient's combination between plants and environments was approximated at 0.84 and the dissimilarity value accounted for 30%. This indicates a high level of similarity or cohesion to the grouping between legumes and sampling locations in Wajo district, South Sulawesi. The use of Principal Component Analysis (PCA) contributed effectively and quickly to the recognition and classification of legumes present in the study</p>

Access to Document

Other files and links

• Link to publication in Scopus

Fingerprint