Population Genomics reveals Deep Evolutionary Divergence Among Peruvian Lupuna (Ceiba sp.)

Ceiba sp. (locally known as lupuna) is an economically and ecologically important forest tree widely distributed across the northeastern Peruvian Amazon. Despite its importance, genomic information on its genetic diversity and population structure remains scarse. To address this knowledge gap, we analyzed genomic variation in 30 individuals collected from two regions of Peru, Amazonas and San Martín, using genotyping-by-sequencing (GBS). High-quality DNA samples were sequenced and processed using STACKS v2, with reads aligned to the reference genome of Ceiba pentandra. After quality filtering, a dataset of 10,684 high-confidence single nucleotide polymorphisms (SNPs) was obtained for downstream analyses. Population genomic analyses included principal component analysis (PCA), Prevosti genetic distance, UPGMA clustering with bootstrap support, genetic differentiation (FST), genetic diversity statistics, and analysis of molecular variance (AMOVA). The PCA revealed a pronounced separation between samples from Amazonas and San Martín, with the first principal component explaining 82.69% of the total genetic variation. UPGMA clustering recovered two highly divergent and well-supported groups largely corresponding to geographic origin, while evidence of additional substructure was observed within the San Martín samples. Genetic differentiation between regions was exceptionally high (F_ST = 0.87), indicating strong evolutionary divergence between populations. Observed heterozygosity was considerably lower than expected heterozygosity in both regions, resulting in high inbreeding coefficients (F_IS = 0.7960 in Amazonas and 0.4594 in San Martín). AMOVA showed that 88.28% of the total genetic variation was partitioned among populations (P = 0.001). Overall, these results reveal remarkable genomic differentiation and previously hidden diversity within Peruvian Ceiba populations, suggesting the presence of distinct evolutionary lineages or potentially separate species. Our findings highlight the value of GBS-based population genomics for understanding genetic diversity and informing conservation and management strategies for tropical forest species.