Traits database of tidal flat macrobenthos along the Northwest Pacific coast of Japan

Project Team for Analyses of Changes in East Japan Marine Ecosystems, Japan Agency for Marine-Earth Science and Technology (JAMSTEC), Yokosuka, Japan Aitsu Marine Station, Center for Water Cycle, Marine Environment and Disaster Management, Kumamoto University, Kumamoto, Japan IDEA Consultants, Inc., Institute of Environmental Ecology, Yaizu, Japan Center for Regional Environmental Research, National Institute for Environmental Studies (NIES), Tsukuba, Japan


| INTRODUCTION
The Great East Japan Earthquake of 2011 caused extensive damage to many coastal ecosystems. Several studies have been conducted of benthic organisms before and after the earthquake and of the process of recovery of the organisms after the earthquake (e.g., Kanaya, Nakamura, Higashi, & Maki, 2013;Kitahashi et al., 2014;Yamakita, 2018;Yamakita et al., 2018;Yamakita, Fujiwara, Tsuchida, et al., 2016;Yamakita, Matsuoka, & Iwasaki, 2017). As a result, there is evidence of not only a decrease in the total biomass and numbers of organisms in the coastal area, but also of an increase or decrease in the abundance of particular species, depending on the species and location. Although it is common to evaluate the change of a community after such impacts by comparing species richness and biomass as well as by calculating community similarities, these indices will not directly reflect the identity of the species that respond to such disturbances. A better way to group species is to assign them to guilds or trophic levels in the case of food web research. More general ways of grouping individual species are needed to capture the characteristics and trends of changes in organisms after such disasters.
There have been a number of previous studies of terrestrial plant ecosystems that compared community characteristics using functional groups based on trait databases (Daz, Kattge, Cornelissen, et al., 2016). The larger numbers of higher taxonomic groups among marine versus terrestrial organisms makes it more difficult to obtain standardized data. However, at small spatial scales and for small taxonomic groups, the importance of considering individual functional groups has been pointed out, even for marine organisms (e.g., Yamada, Hori, Tanaka, Hasegawa, & Nakaoka, 2010). It is not unusual to show the importance of functional characteristics qualitatively, even in wide spatial comparisons (Wahl et al., 2011). However, there has been no comprehensive information on the various functions and taxa of benthic organisms that would enable an assessment of the entire community affected by this disaster.
Traits data make it possible to understand changes in communities at a higher level than individual species but in more detail than is likely to be provided by community indices. For example, functional diversity is a better indicator than species diversity for evaluating characteristics of a community that have responded to environmental change (Mouillot, Graham, Villéger, Mason, & Bellwood, 2013). It is also believed that functional redundancy improves the ability of the entire ecosystem to compensate for the loss of individual species. Thus, this dataset can also be used to evaluate the relationship between compensatory effects in the responses of species and communities to environmental changes (Elmqvist, Bengtsson, Angelstam, et al., 2003;Mori, Furukawa, & Sasaki, 2013).
A trait database for such purposes has already been assembled for the major vascular plants, for which there is much well-organized information about specimens and lineages (Daz et al., 2016). In the case of aquatic species, progress has been made mainly on fish (Albouy et al., 2015;Stuart-Smith et al., 2013). Recently, a database dealing with benthic organisms has been initiated.
For example, the BIOTIC database includes data from more than 680 species (Costello et al., 2015). In the case of specific taxa, the Polytraits database has been actively archiving information mainly about nematode and other meiofauna including polychaetes (Faulwetter et al., 2014). In recent years, progress has been made on a database of the functional traits of deep-sea hydrothermal organisms (sFDvent https://peerj.com/preprints/26627/). In this way, the assembling of databases and advancement of understanding of the functions of marine organisms other than fish have begun.
To contribute to the tidal flat survey along the northeastern coast of the Japanese archipelago, which was very adversely impacted by the Great East Japan Earthquake, we collected information on the functional traits of benthic organisms that inhabit mainly tidal flats. Our survey was not limited to specific taxa. As with other benthic databases that have been initiated in recent years (Costello et al., 2015;Daz et al., 2016;Faulwetter et al., 2014), this database is characterized by information on the distribution, body length, life history, and physiological characteristics of organisms across taxa.

| Temporal coverage
Best available data prior to May 1, 2017.

| Study sites
We focused on the major species observed on tidal flats along the Northwestern Pacific coast of Tohoku, Japan, based mainly on the locations sampled as a part of the Survey of the Natural Environment in the Pacific coast of the Tohoku district, part of the "Monitoring 1000" project of the Ministry of the Environment (Biodiversity Center Nature Conservation Bureau, Ministry of the Environment, 2012, 2013). We also made use of several other publications that have described monitoring of ecosystems after the earthquake (Hayasaka, Yamada, & Uchida, 2016;Kanaya, Suzuki, Maki, et al., 2012;Urabe, Suzuki, Nishita, & Makino, 2013). The study included 20 locations from Aomori Prefecture to Chiba Prefecture ( Figure 1): Takahoko Numa Lagoon, the Takase River, Osuka Coast, Kuji Coast, Tofugaura Coast, Akedo Coast, the Tsugaruishi River, the Orikasa River, the Unosumai River, the Kitakami River, Nagatsuraura Lagoon, Mangokuura Lagoon, Matsushima Bay, Gamou Lagoon, Idoura Lagoon, the Hiroura tidal flat, Torinoumi Lagoon, Matsukawaura Lagoon, the Ichinomiya River, and the Isumi River. These study sites can be characterized as western Pacific temperate ecosystems. The study sites are mainly lagoons and river-mouth tidal flats.

| List of the species
The species in the database are summarized in Figure 2. The data are based mainly on research conducted in 2011 and 2012 during the year after the earthquake (Biodiversity Center Nature Conservation Bureau, Ministry of the Environment, 2012Environment, , 2013Hayasaka et al., 2016;Kanaya et al., 2012;Urabe et al., 2013). We selected 246 species that we deemed appropriate for evaluating macrobenthos communities in the region focused mainly around Sendai Bay. For this selection, we considered the number of occurrences in the data and generality of the distribution in our study region based on expert opinion. After the selection, the availability of trait information reduced the number of species. Figure 2 shows the taxonomic distribution of the dataset. The majority of the phyla were Arthropoda, Mollusca, and Annelida; each of those phyla accounted for more than 10% of the species. The remaining species, which accounted for <15% of the total species, came from a variety of taxa. The majority of the species belonged to the classes Malacostraca, Polychaeta, Gastropoda, and Bivalvia. Although the original list of species contained several species that could not be identified, those specimens were eliminated because of the limitations of the traits data.

| Taxonomy and systematics
The species names were checked by the authors, who are experts of some of these benthic organisms. If we could not obtain sufficient information for species identification from a specimen, we recorded the taxon at a higher taxonomic level (e.g., order or class) that could be specified with certainty. Scientific names followed Okutani (2017), the latest information in Furota and Taru (2016), the WORMS database (http://www.marinespecies.org), and the Catalogue of Life database (http://www.catalogueoflife.org/), in that order of priority. However, we modified Sipuncula and Echiura into Annelida (Sipuncula) and Annelida (Echiura) from the WORMS. Polychaeta was not modified from WORMS because the polyphyletic groups in this class are still not orderd in the WORMS.

| File format
The data files are saved in comma-delimited (csv) text files with UTF-8 encoding.

| Definitions and extraction of traits data
The traits are defined in Table 2. The list includes the best available biological and ecological traits from over 40 potential traits identified at the beginning of the study. The traits were ordered as follows: taxonomic position, geographical distribution, body length, habitat, mode of reproduction, and other characteristics such as diet type and migratory type. After an initial draft of this information was produced, the BIOTIC database was published, and we refined our categories to make them consistent in most cases with the BIOTIC database. Some categories that we considered to reflect our thought on traits were retained from the first draft without considering their compatibility with the BIOTIC database (Costello et al., 2015;Nagai, Shibata, Osawa, et al., in press). Those traits included Abundance of the population, Depth of habitat in the substrate, Dispersion type, Egg type, Types of larvae, Feeding devices, Fertilization or pollination methods, Sexual expressions, Fishing pressure, and Alien species. About 20 traits in the BIOTIC database were not adapted because of data limitations and overlap with other categories such as Food type, Flexibility, Habitat, Growth rate, Dispersal potential of adults, Sociability, Regeneration per year, Toxicity, Generation time, Larval settlement period, Fecundity, Egg size, Fertilization type, Larval settling Time, Reproduction Location, Biozone, Growth form, Reproduction type, Water flow, and Wave exposure. These differences are also listed in Table 2.
We also made use of our expert knowledge and estimated the traits of some species for which there were inadequate literature data. Expert knowledge did not . Unknown values of traits were also assumed based on our expert opinion (AS: predicted value based on our empirical observations). Other values were left blank (NA). To reduce the variance of the data, almost all of the data were categorized, even in the case of data with continuous categories. Table 2 includes a list of the categorization rules. Table 3 shows the distributions among species of the traits for which we were able to obtain information. Most of basic information about body length, distribution, and habitats was complete, and more than 50% of the traits were complete for >100 species, but information about reproduction was relatively limited. Information about reproductive ecology was relatively limited, especially for the Polychaeta and rare taxa.