Local adaptation and genetic differentiation between populations is improved by selective differences among populations and little or total absence of gene flow. However, determining whether population differentiation is caused by selection of certain traits is complicated by the fact that other forces such as genetic drift can also cause population differentiation in the absence of gene flow. The Quercus subsection Virentes includes seven species (Q.virginiana, Q. geminata, Q.minima, Q. brandegeei, Q. fusiformis, Q. oleoides and Q.sagrean.) distributed from southeastern US, through eastern Mexico, southern Baja California, Central America, and Cuba (Nixon & Muller 1997). Due to its wide distribution, this group presents a great variety of adaptations that could be the cause of the divergence between the populations of the species that share the same geographical site. For example, little differentiation has been found between Q. minima and Q. geminata when using SSR simple sequence repeat or chloroplast data in populations where both species co-occur. However, both species have different habitats, one being fire-dependent and the other fire-tolerant. (Kurz & Godfrey 1962; Cavender-Bares et al. 2004b). All these habitats and environmental differences in a group of so closely related species are of great evolutionary interest in determining to what degree population differentiation is caused by local adaptation (selective traits) versus neutral (stochastic) processes.  During this research, we focused on testing whether spectral information of dried samples can be used to detect divergence between genetic groups at different hierarchical levels using natural populations of seven closely related species in the Virentes subsection. Our main goals are 1) to identify spectral regions in dry leaf spectra that are important for discriminating closely related oak species; and 2) to elucidate if phenotypic variances in spectra between populations are the product of demographic history (genetic drift) or genetic variation and environmental factors (adaptation). Methodology We measured leaf spectral reflectance (400 – 2500 nm) on pressed specimens collected from 2006 through 2011 from individuals in which there were nuclear single sequence repeat (nSSR) amplification data (from Cavender-Bares et al. 2015). We ran two clustering algorithms within STRUCTURE v. 2.3.4 (Pritchard et al., 2000), first we include 428 individuals of the seven species. Then, we include only individuals of Q. oleoides given that these species have shown phenotypic variation across its distribution (Ramirez-Valiente et al. 2015). We used partial least square discriminant analysis (PLS-DA) (Chevallier et al., 2006) to assign spectral samples to different hierarchical groups (genetic groups, and species).  Results Genetic data of the nSSR showed seven genetic groups for all species in accordance with Cavender-Bares et al. 2015 and five distinct groups for Q. oleoides (Fig 2). Classification using PLS-DA models for the seven species of Virentes using raw pressed-spectra showed good performance at classifying species and correctly predicted the taxonomic identity of 295 of 302 samples in the training dataset.