Attached is the complete raw data set from Leingang and Dixson 2019 "Phenotypic variations in the preferred host coral impact the occupancy of an obligate coral-dwelling fish." Data were inputted into an excel sheet with three different tabs. The first two tabs are comprised of data gathered during 25m belt transect surveys. Belt transects were conducted at two different locations within Fiji, Kadavu Island (n = 25) and Tavewa Island (n = 20). All corymbose, tabular, and digitate acroporid corals encountered directly under the transect tapes were inspected for the presence or absence of Gobiodon histrio, as well as the number of gobies within the corals (n = 450 corals). The first tab, "G. histrio occupancy" shows the number of G. histrio located within all acroporid corals surveyed, as well as which island the survey was conducted. The second tab, "A. nasuta morphology" shows down the morphological measurements (height, diameter, and average inter-branchial distance) as well as the color variation (blue or brown) of all Acropora nasuta surveyed. Finally, the third tab, "Trials table" lists the data from the behavioral choice experiments. Each experiment was replicated 16 times for a total of 48 fish. Adult G. histrio were placed on a dead A. nasuta coral and presented with a live A. nasuta coral 10cm away. The first set of experiments only provided one color variant of A. nasuta (either blue or brown), which allowed the gobies to either move to that color variant, remain on the dead coral, or abandon the patch reef entirely. The second experimental setup was similar to the first, except both blue and brown color variants were presented, with all corals having 10cm between them. This allowed the gobies to either select and move to a specific color variant, remain on the dead coral, or abandon the patch reef entirely Column A lists the configuration of the corals used in the experiment. Columns B-E list the number of G. histrio that chose that specific coral as their final location at the end of the trial period or abandoned the patch reefs entirely.

Habitat specialists form tight relationships with their host habitat and are able to make microscale decisions when selecting final habitat locations. The obligate coral-dwelling fish, Gobiodon histrio, is thought to make habitat choices based on the coloration and structural characteristics of Acropora nasuta, their preferred coral host. Yet, most studies on the habitat preference of G. histrio have been conducted on Australia’s Great Barrier Reef with no understanding if geographic differences in preferences exist. Here, we tested the habitat preference of G. histrio toward A. nasuta on the reefs of Kadavu and Tavewa Islands, Fiji. First, to assess the natural distribution, belt transect surveys of all acroporid corals were conducted. Transects indicated that, while G. histrio is most frequently found in A. nasuta over other acroporid corals, the coral’s structural characteristics rather than the coral’s color variation were the preferred characteristic. In contrast, the Australian G. histrio have been found to be more frequent in blue A. nasuta opposed to the brown color variation, suggesting a geographic difference in habitat preferences among the species. In addition, we conducted two in situ behavioral field experiments to determine whether G. histrio would (1) move from dead A. nasuta to a live brown or blue A. nasuta and (2) preferentially select between the brown- or blue-colored A. nasuta when placed on a dead A. nasuta. The results of the in situ experiments support the finding that Fijian G. histrio does not discriminate between A. nasuta using color but uses only structural morphologies to guide its habitat selection process. Habitat selection is a complex process, and microscale habitat preferences within a species can vary between geographic locations. This study sheds light on the need to expand research findings to incorporate large geographic regions when attempting to understand the preferences of coral reef symbionts.