Introduction
Oaks (Quercus L., Fagaceae) are anemophilous trees and shrubs of the Northern Hemisphere and are of high ecological, scientific and economic importance (Menitsky 2005). They occur in a wide range of habitats and climates ranging from semi-arid Mediterranean areas and subtropical rainforests to boreal (cold temperate) continental regions (Kremer and Hipp 2019). Oak taxonomy is still in flow; only recently the genus has been divided into two monophyletic subgenera and eight sections accommodating the 300–600 reported oak species (Denk et al. 2017). Discrepancies in species counts are due to the considerable morphological variation and phenotypic plasticity exhibited by oaks, the result of the complex interplay of strong ecological adaptation, convergence of morphological traits, hybridization, introgression, and reticulate evolution (Burger 1975; Van Valen 1976; Cavender-Bares et al. 2015; Simeone et al. 2016; McVay et al. 2017a, b; Hipp et al. 2019a). Consequently, oaks are a big challenge for modern taxonomists (Denk et al. 2017) and provide an ideal model for the development of a holistic taxonomy that accounts for complex ecological, biogeographic and evolutionary processes (Kremer et al. 2012), thereby improving our understanding of Earth’s biodiversity. At the same time, key questions about oak diversification and evolutionary history, especially in the Old World, are just about to be answered (Denk et al. 2017; Hipp et al. 2019; Yan et al. 2019; Jiang et al. 2019).
After three decades of research (since Whittemore and Schaal 1991), it has been widely demonstrated that oak plastid genomes are decoupled from species identity (Simeone et al. 2016, 2018; Pham et al. 2017; Vitelli et al. 2017) and that nuclear genes and regions with sufficient levels of variation for solid species delineation and phylogenetic inferences are still unavailable (Oh and Manos 2008; Hubert et al. 2014). Based on extensive phylogenomic data, Hipp (2018a) concluded that oaks can be defined as phylogenomic mosaics, meaning that every individual genome actually represents an assembly of different histories reflecting as well selection (ecological adaptation) and origin, or multiple origins (divergence, reticulation, and lineage sorting). Species-level phylogenetic reconstructions are affected by such genomic blending. Cohesive intra-specific gene flow is counterbalanced by local gene flow among individuals and populations in different parts of a species’ range, introgression, sorting of ancestral traits and divergence (Eaton et al. 2015; McVay et al. 2017b; Hipp et al. 2018b, 2019b; Crowl et al. 2020; Leroy et al. 2020).
However, detection and distinction of horizontal gene transfer, introgression and random sorting of ancestral genetic variation can be difficult. The labor and costs associated with the necessary comprehensive datasets (both at the inter- and intraspecific level) will likely prevent a wide application of phylogenomic tools beyond species trees and selected case studies (McVay et al. 2017; Hipp et al. 2019; Jiang et al. 2019). Among nuclear DNA target regions of potentially high taxonomic resolution, the spacer regions of the ribosomal DNA (35S and 5S nrDNA) are a quite popular choice for taxonomic studies. These regions have been used for phylogenetic inferences across a broad range of evolutionary lineages, have a high copy repeat number, and universal or specific PCR primers are available allowing efficient amplification in a wide range of taxa (Volkov et al. 2001, 2003; Alvarez and Wendel 2003). In particular, the internal transcribed spacers (ITS1, ITS2) of the 18S-5.8S-25S cistron (35S rDNA) still are the most widely used nuclear markers for phylogeny and systematics, including molecular taxonomy such as DNA barcoding of complex plant groups (Hollingsworth et al. 2011). However, in many groups the ITS1 and ITS2 can be nearly invariable. In western Eurasian oaks, for instance, the nuclear ribosomal 5S DNA intergenic spacer resolved species relationships to a much higher degree than ITS data (Denk and Grimm 2010).