Molecular and Morphological Identification of Plants Consumed by Yellow Baboons in Amboseli, Kenya

Abstract

Plant species can be identified based on their morphological characteristics and molecular properties. Molecular identification utilizes specific regions in the plant genome, followed by comparison against reference databases. The flora in the Amboseli ecosystem has only been characterized based on its morphology, and no genetic data exists in the public databases. The general objective of this research was to identify the plants consumed by yellow baboons (Papio cycnocephalus) in Amboseli, Kenya, based on morphological and molecular analyses. Eighty plants (40 monocotyledons and 40 dicotyledons) were collected from Amboseli Baboon Research Project’s study site in the Amboseli ecosystem, Kenya, in three periods, namely June 2016, January 2018 and May/June 2018. Twentythree plants whose scientific names were uncertain were deposited at the University of Nairobi herbarium. DNA was extracted from all the samples using Qiagen’s DNeasy Plant Mini Kit, followed by the amplification of five barcoding genes, namely: ITS1, ITS1-Poaceae, the trnL (UAA) intron, trnL-P6, and the 18S ribosomal DNA locus. The amplicons were sequenced using the Sanger sequencing method then analysed using BLAST and phylogenetic approaches to determine the consensus identities. The amplification success rate of all the extracted DNA was generally higher in monocotyledons (93.06%) than in dicotyledons (87.67%). With regards to the candidate markers, the ITS1 locus had the highest amplification success rate (100%) followed by the trnL (UAA) intron (95%), ITS1-Poaceae (90%), the 18S rDNA locus (82.05%), and lastly, trnL-P6 (80.49%). Two-hundred and eighty-nine amplicons were sent for sequencing at Macrogen Netherlands (Europe), and highquality sequences were generated for 182 samples. Generally, more plants were identified at both the genus and species levels using GenBank® than in the BOLD database. Furthermore, more monocotyledons were identified using BLAST analysis than dicotyledons, whereas the phylogenetic analysis was more successful in the identification of dicotyledons than monocotyledons. With regard to the consensus identities, 66 out of the 80 plants were identified. Specifically, 50 plants were determined to only the genus level, while 16 samples were distinguished to the genus and species levels. The use of multiple markers - from both the nuclear and chloroplast regions - was very crucial in the overall high identification success rate achieved in this study. The data generated from this work can be used as a reference for future studies relating to the characterization of plants in the Amboseli ecosystem and by extension, in Kenya. Furthermore, because the selected plants are those that are eaten explicitly by the Amboseli baboons, the data will be used to conduct a diet metabarcoding study.