Population genetics meets ecological genomics and community ecology in Cornus Florida

Authors: PAIS, ANDREW - North Carolina State University; XIANG, QY (JENNY) - North Carolina State University; Wadl, Phillip; LEEBENS-MACK, JAMES - University Of Georgia; WHETTEN, ROSS - North Carolina State University

Submitted to: Botanical Society of America Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: 4/15/2016
Publication Date: 7/30/2016
Citation: Pais, A., Xiang, Q., Wadl, P.A., Leebens-Mack, J., Whetten, R. 2016. Population genetics meets ecological genomics and community ecology in Cornus Florida. Botanical Society of America Abstracts. 615.

Interpretive Summary: N/A

Technical Abstract: Understanding evolutionary/ecological consequences of alien pests on native forests is important to conservation. Cornus florida L. subsp. florida is an ecologically important understory tree in forests of the eastern United States but faces heavy mortality from dogwood anthracnose. Understanding genetic diversity in the species, its spatial distribution, and its relationship to ecological gradients remains vital to preserve flowering dogwoods. We integrated methods in population genomics, ecological genomics, and community ecology to gain insights into the following questions: (1) Has disease spanning three decades rendered C. florida genetically depauperate? (2) Are healthy and infected populations differentiated in any loci along ecological gradients? (3) Spatially where are changes in allele frequencies most abrupt, and what ecological gradients does allelic variation correspond? We sampled 289 plants from 86 locations and generated genotyping by sequencing (GBS) data on an Illumina NextSeq. Sequence reads were aligned to C. florida’s draft genome using Bowtie2 and further processed in the STACKS pipeline to match genotypes of individual trees to the catalogue of total loci. In addition, we quantified and removed contaminant reads by aligning sequence reads to pathogen genomes of dogwood anthracnose and powdery mildew via Bowtie2. This was done as a standard filtering step and to develop an alternative estimator of disease occurrence. Using Gradient Forests (GF) analysis, we will examine where allele turnover is high in the genetic landscape and in what loci and whether there is genetic turn over between healthy and infected populations/localities. Population genetics, FastStructure, and AMOVA analyses of the data were conducted to evaluate genetic diversity and the patterns of genetic variation. Preliminary results have revealed: low nucleotide diversity (variant: 0.2585; all: 0.0017) in the subspecies; two to three poorly differentiated genetic groups (explaining 2.11% and 2.65% genetic variation respectively) corresponding to hot continental and coastal plain ecoregions; genetic variation primarily within the genetic groups, and occurrence patterns of pathogen sequences in plants of infected populations. These results will be integrated with GF analyses to detect any localized patterns where genetic diversity may decrease as disease occurrence becomes more dominant along ecological gradients. Conclusions from current results support prior evidence of weak geographic structure, possibly due to gene flow via bird dispersal mechanisms. Although results indicate the subspecies is genetically healthy at present, there is evidence that disease has affected the genetic diversity of flowering dogwoods to an extent.