Corals and cnidarians form symbioses with dinoflagellates across a wide range of habitats from the tropics to temperate zones. Notably, these partnerships create the foundation of coral reef ecosystems and are at risk of breaking down due to climate change. This symbiosis couples the fitness of the partners, where adaptations in one species can benefit the holobiont. However, the scales over which each partner can match their current - and future - environment are largely unknown. We investigated population genetic patterns of temperate anemones (Anthopleura spp.) and their endosymbiont Breviolum ‘muscatinei’, across an extensive geographic range to identify the spatial scales over which local adaptation is possible. Similar to previously published results, two solitary host species exhibited isolation-by-distance across hundreds of kilometers. However, symbionts exhibited genetic structure across multiple spatial scales, from geographic location to depth in the intertidal zone, and host species, suggesting that symbiont populations are more likely than their hosts to adaptively mitigate the impact of increasing temperatures.

Tentacle samples were collected across the geographic range of Anthopleura spp. along the Pacific coast of North America, 16 sites for the hosts, 7 sites for the symbionts. Hosts were genotyped using RAD-seq and subsequent alignment to the A. sola draft genome. Symbiont populations living within tentacles were enriched using FACS, sequenced using RAD-seq, and genotyped after alignment to the B. minutum draft genome. We included SNPs that were present in at least half of the surveyed populations in at least half of the individuals of each population at a coverage of greater than 5x. For the symbionts, we excluded SNPs that occurred within 500 bp of one another.

We also characterized the surrounding benthic community around each anemone or aggregation. A matrix of the surrounding benthic community is also included, one line per individual.

Genotype files for both hosts and symbionts were generated using plink. ped files contain the genotype information (bed files are binary versions), other information regarding locus location on genome assembly scaffolds, individual names and geographic groupings are available in .map, .famMissing values are present in the genotyping file as "0."