Methods

2.5.1 Literature collection We extended the literature compilations of Yamano et al. (2011) and Hongo and Yamano (2013) to cover the entire distributional range of corals in Japan. We also collected data that had been published after Yamano et al. (2011) and Hongo and Yamano (2013) constructed their database. We selected publications containing extensive species lists that presumably included occurrences at the study sites. As stated by Yamano et al. (2011), four distinct periods of coral exploration have occurred. The first was in the 1930s and was undertaken by Drs. Hisakatsu Yabe and Toshio Sugiyama at Tohoku Imperial University (Tohoku University) of Japan. A large number of specimens was collected from various regions of the Japanese coasts during the years, and some of the specimens taxonomically described by themselves are stored in Tohoku University Museum (Nakamori et al., 2016). The second period was during the 1960s and 1970s. Comprehensive surveys of marine environments and organisms were conducted as part of the designation planning of marine parks by the Environment Agency (Ministry of the Environment). Coral identification was performed mainly by Drs. Huzio Utinomi and Motoki Eguchi. The third period was during the 1980s to 1990s, when Dr. JEN Veron, who revised and integrated the recent taxonomic framework of corals, visited Japan and examined coral distributions. The last period covers the late 1990s to the present day. Japanese researchers, including the present authors, conducted field surveys following the recent taxonomic framework, and published monographs describing the coral faunas at Tanegashima (Sugihara et al. 2015) and at Kushimoto (Nomura 2016a, b). These monographs contain detailed descriptions of temperate coral species, which helped with investigation of species names in the present work. All the literature and locations were included in the CoralLiteratureDB (Tables 1 and 3) and CoralLoactionDB (Tables 1 and 4), respectively. 2.5.2 Confirmation of the species name The coral taxonomic framework that was used before 1990 differed greatly from that in current use, and the framework is still evolving (Budd et al. 2012; Huang et al. 2014; Arrigoni et al. 2016). We developed a workflow to synonymize and confirm the species names that were collected from the original literature (Fig. 2). These names are included in the CoralSpeciesDB (Tables 1 and 5). We describe the main procedure below; specific procedures are described in the remarks_CoralSpecies column of the CoralSpeciesDB. Some of the literature includes Japanese names only. We thus followed three procedures to obtain the original scientific species names from the Japanese names. In cases in which the Japanese name was used for only one species, we simply converted it to the scientific name of the species. When the Japanese name (e.g., Kikumeishi) was used for both species and genus names, we assigned the Japanese name to open nomenclature within the genus (e.g., Favia sp.). If the Japanese name was used for multiple species, we searched other literature by the same author that included both Japanese and scientific names of the species, or we looked for records of species occurrences in adjacent areas to find the scientific name of the species in order to obtain a single species name. If no information was obtained despite these efforts, we assigned the Japanese name to open nomenclature within the genus. The original species names were compared with those registered in the World Register of Marine Species (WoRMS; Hoeksema and Cairns 2019), after correction of typographic errors as necessary. WoRMS contains the most recent systematics, and the original names were revised to synonymized species names. However, corals occurring in temperate areas have not been considered adequately in WoRMS, which is based mainly on examination of tropical coral species. We thus carried out further investigations to confirm the species names, primarily following Sugihara et al. (2015), who described the coral distribution in Japanese temperate waters in detail, as well as using photographs and distributional characteristics from the original literature and our own field survey. For example, Tridacophyllia laciniata in the temperate area was revised to Pectinia paeonia according to WoRMS. However, Sugihara et al. (2015) demonstrated that Physophyllia ayleni may be confused with Pectinia paeonia in Japan. They showed that Physophyllia ayleni in Japan occurs in the temperate area, and suggested that Pectinia paeonia is not present in the temperate area of Japan, based on extensive field surveys. Therefore, in this case the confirmed species name of what had originally been called Tridacophyllia laciniata is Physophyllia ayleni. For species occurrence records from Kushimoto, we revised the species names based on Nomura (2016a, b), who described the coral distribution in Kushimoto in detail, after synonymization according to WoRMS. Although we found differences in the species names between the work of Sugihara et al. (2015) and that of Nomura (2016a, b), Nomura et al. (2016) described the correspondence of the names in the latter publications to those of Sugihara et al. (2015). We thus applied the names used by Sugihara et al. (2015) if differences were found. Other revisions were performed by referring to recent systematics (Benzoni 2006; Benzoni et al. 2010; Nomura and Suzuki 2013; Benzoni et al. 2014; Kitano et al. 2015; Nomura and Suzuki 2015; Arrigoni et al. 2016; Luzon et al. 2018) or past occurrence records (Uchida and Fukuda 1989; Veron 2000; Tachikawa 2005; ReefCheck Kushimoto 2014; Yasuda et al. 2014) as necessary. In addition, we defined three species complexes that could contain multiple and/or cryptic species. The first is the “Acropora hyacinthus complex,” which may contain A. hyacinthus, A. spicifera, and A. tanegashimensis. Sugihara et al. (2015) described how these species have been confused, and Suzuki et al. (2016) identified four genetic lineages in this complex in Japan. The second species complex is the “Acropora solitaryensis complex.” Three morphologies (arborescent table, plate, and an intermediate between the two) have been recognized in A. solitaryensis in Japan. We used the term “Acropora solitaryensis complex” for the three morphologies if we could not distinguish them from the original literature, though Suzuki and Fukami (2012) showed that the arborescent table and the intermediate forms were variations of A. solitaryensis and were distinct from the platy form species based on DNA analyses and cross-fertilization experiments. The third species complex is the “Montipora mollis complex.” Nomura et al. (2016) divided M. mollis into four species on the basis of morphology, but separate species names have not been defined. We thus used the term “Montipora mollis complex” for these possible four species. After synonymizing and confirming all the coral species data from the collected literature, we compiled the papers containing the data in CoralLiteratureDB. The location (latitude and longitude) and survey period were obtained from the description in the Materials and Methods of each reference. Where no geographic coordinates or map of the study sites was provided, we obtained approximate coordinates from GSI Maps (https://maps.gsi.go.jp/, Geospatial Information Authority of Japan, Tsukuba, Japan) and Google Maps (http://maps.google.com/, Google Inc., Mountain View, CA, USA) using the site names and descriptions provided by the authors. 2.5.3 Data verification Recorded coral names, survey years and months, occurrences, geographical locations, and county and city names were compiled into CoralOccurrenceDB. All of the datasets (CoralOccurrenceDB, CoralLiteratureDB, CralLocationD and CoralSpeciesDB) were manually checked for typographic errors.