Contrasting patterns of local adaptation along climatic gradients between a sympatric parasitic and autotrophic tree species

Sympatric tree species are subject to similar climatic drivers, posing a question as to whether they display comparable adaptive responses. However, no study has explicitly examined local adaptation of co-occurring parasitic and autotrophic plant species to the abiotic environment. Here we test the hypotheses that a generalist parasitic tree would display a weaker signal of selection and genomic variation would associate with fewer climatic variables (particularly precipitation) but have similar spatial patterns to a sympatric autotrophic tree species. To test these hypotheses, we collected samples from 17 sites across the range of two tree species, the hemi-parasite Nuytsia floribunda (n=264) and sympatric autotroph Melaleuca rhaphiophylla (n=272). We obtained 5,531 high-quality genome-wide single nucleotide polymorphisms (SNPs) for M. rhaphiophylla and 6,727 SNPs for N. floribunda using DArTseq™ genome scan technology. Population differentiation and environmental association approaches were used to identify signals of selection. Generalized dissimilarly modelling was used to detect climatic and spatial patterns of local adaptation across climatic gradients. Overall, 322 SNPs were identified as putatively adaptive for the autotroph, while only 57 SNPs were identified for the parasitic species. We found genomic variation to associate with different sets of bioclimatic variables for each species, with precipitation relatively less important for the parasite. Spatial patterns of predicted adaptive variability were different and indicate that co-occurring species with disparate life history traits may not respond equally to selective pressures (i.e. temperature and precipitation). Together, these findings provide insight into local adaptation of sympatric parasitic and autotrophic tree species to abiotic environments.

Methods
Genomic data was generated from individual genotyping of 272 Melaleuca rhaphiophylla individuals and 264 Nuytsia floribunda individuals sampled at 17 populations across each species range. Climatic data was downloaded with 1 km cell resolution from the Worldclim 2.0 database (Fick & Hijmans, 2017; Hijmans et al. 2005).