Pedicularis dudleyi (Dudley’s Lousewort, Orobanchaceae) is an extremely rare wildflower endemic to the redwood forests of Central California. Until recently the species was known only from three extant natural populations. However, one of those populations was recently described as a novel species (P. rigginsiae D.J. Keil) based on morphological and ecological data leaving only two populations described as P. dudleyi. While little is known about the past distribution of the species, historical records have led to speculation that the species was once more widespread and may have suffered from habitat destruction as a result of widespread logging during the early twentieth century. We utilized a combination of ddRAD SNP and Sanger sequencing data to: 1) Describe the genetic diversity and population structure of P. dudleyi; 2) Test the hypothesis that the species underwent a bottleneck corresponding with increased logging of redwood forests in the early twentieth century and; 3) Test the morphological hypothesis that P. rigginsiae is distinct from P. dudleyi. Genetic diversity statistics and analyses of genetic structure suggest that both populations of P. dudleyi are highly differentiated from each other. Demographic modeling supports a scenario where the contemporary rarity of the species is explained by a recent bottleneck. Finally, recognition of P. rigginsiae as distinct from P. dudleyi is supported, increasing the conservation priority of both species.

ddRADseq libraries were constructed following a modified version of the protocol described in Peterson et al.(2012)where amplification was performed prior to size selection. 500150 ng of genomic DNA were digested using the restriction enzymes EcoRI and SphI-HF.P1 and P2 flex adapters were ligated to the digested DNA. Ligation products with similar concentrations were combined into a total of eight pools with each pool containing 1024 samples. Each pool was then divided, and PCR amplified in eight separate 25ul reactions using dual indexed PCR primers as described in Meyer and Kircher(2010). After amplification, indexed libraries were then re-pooled, quantified using qPCR and combined in equimolar ratios. The final library was size-selected for 750bp and the final Illumina library was sequenced by Novogene at the UC Davis Sequencing Center on a single lane of Illumina HiSeq X Ten(150bp, paired-end). The files included here include the final SNP data set formatted for STRUCTURE, GENPOP, and DIYABC and VCF.a

Phylogenetic data includes aligned gene sequences for the concatenated ITS and the matK-5trnK intron dataset.