Diversity and taxonomy of intertidal Bryozoa (Cheilostomata) at Akkeshi Bay, Hokkaido, Japan

We found 39 cheilostome species among more than 7000 specimens collected at 10 intertidal sites in rocky habitats along the shore of Akkeshi Bay, eastern Hokkaido Island, Japan. These species are herein described in detail and illustrated by scanning electron microscopy. Nine species (23% of total) are described as new (Electra asiatica, Callopora sarae, Conopeum nakanosum, Cauloramphus cryptoarmatus, Cauloramphus multispinosus, Cauloramphus niger, Stomachetosella decorata, Microporella luellae, and Celleporina minima), and 21 species (54%) are reported for the first time from Japan. Species richness ranged from eight to 29 species per study site. A TWINSPAN analysis showed the species fell into nine groups defined by the local pattern of distribution. A cluster analysis of study sites based on similarity of species composition showed three faunistic groups distributed geographically: in Akkeshi Lake, along the eastern‐central shore of the bay, and at the mouth of the bay. Species richness in estuarine Akkeshi Lake was low, with a species composition very different from the outer bay. Most cheilostomes were found on rock and shell substrata, but uncommonly occurred on concrete walls, algae, hydroids, tubes of polychaetes, other bryozoans, and anthropogenic debris. Of the 39 species found, 33 (85%) contained embryos during the collecting periods, 2–7 June and 3–6 July 2004. The biogeographical composition of intertidal cheilostomes at Akkeshi Bay included species with Arctic‐Boreal (28%), Boreal (59%), and Boreal‐Subtropical (13%) distributions. The overall species richness of intertidal cheilostomes was two‐thirds that documented intertidally in a comparable study at Kodiak, Alaska, a locality 15° higher in latitude. We attribute the lower richness at Akkeshi to differences in the nearshore marine environment between the two localities.


Introduction
There are several good reasons for studying intertidal bryozoans (Dick et al. 2005). Bryozoans are more readily sampled intertidally than subtidally, and intertidal sites contain a significant subset of the greater diversity of species found on adjacent subtidal shelves, at least at cool-temperate latitudes. For example, the proportion of the total nearshore fauna occurring intertidally is around 32% in the vicinity of Nanaimo, BC (O'Donoghue andO'Donoghue 1923, 1926) and around 26% for Britain Ryland 1998, 1999). Thus intertidal assemblages are speciose enough to be useful for environmental monitoring and zoogeographical comparisons. Finally, as abundant competitors for space in cryptic habitats, bryozoans are important in helping to structure shore communities (Gordon 1972), though their role in this regard has not been well studied.
The diversity and composition of intertidal bryozoan assemblages have been studied at cool-temperate (Dick and Ross 1988;Dick et al. 2005) and subtropical (Dick et al. 2006) localities in the northern Pacific Ocean. These previous studies, as well as data in the monographs of O' Donoghue andO'Donoghue (1923, 1926), have shown that local species richnesses in rock-pile habitats (see Dick and Ross (1988) for definition) in the northeastern Pacific are in the order of 70-80 total species. Maximum richnesses of cheilostome bryozoans, the main group contributing to benthic assemblages, have ranged from 28 to 33 species at single sampling sites in the northeastern Pacific and on Hawaii.
The intertidal assemblage in the cold-temperate northeastern Pacific is similar in composition over broad geographical distances, with the same genera ranking highest in the number of species they contribute to the assemblage. As might be expected, the taxonomic composition of the intertidal assemblage on Hawaii (Dick et al. 2006) is quite different from that of the northeastern Pacific. Although there are indications that the composition of intertidal assemblages may be quite similar between Hawaii and the Indo-West Pacific, no comparable intertidal studies have been conducted in the latter region.
We chose Akkeshi Bay as a study site for several reasons. Previous Russian work suggested that the intertidal bryozoan fauna in this region might be exceptionally rich. Kubanin (1997) provided a regional checklist of 128 bryozoan species documented intertidally along the Russian coasts of the Sea of Japan, Sea of Okhotsk, and Bering Sea, which by way of comparison is a higher number than reported intertidally from Britain (around 79 species). Grischenko (1993Grischenko ( , 1994 reported 63 bryozoan species from the Commander Islands. Furthermore, along the Pacific coast of Hokkaido, there is a confluence of several ocean currents (the Kuroshio from the south, the Oyashio from the north, and the Tsushima from the west) that might transport larvae from several provinces, potentially increasing species richness in the area. Finally, bryozoan alpha-level taxonomy in many parts of the world is in need of considerable revision Dick and Mawatari 2004), and it is becoming increasingly evident that intensive local studies facilitate this, since congeners can be compared and similar but distinct species distinguished without the confounding influence of geographical variation.
Reliable alpha-level taxonomy is in turn a prerequisite for drawing reliable conclusions about biogeography. The degree of relationship between the boreal bryozoan faunas of the eastern and western Pacific remains unclear, due both to insufficient sampling and poor taxonomic resolution. Species recently described in the western Pacific have been subsequently found in Alaska (e.g. Hincksina longiavicularia Gontar, 1982), and vice versa (e.g. Cauloramphus magnus Dick and Ross, 1988). It is unclear whether or not some nominal species reported from both sides are distinct species (e.g. Porella acutirostris Smitt, 1868). In some cases, species initially considered to be trans-Pacific have been recognized as separate, sibling species in the eastern and western Pacific. An example is Cauloramphus pseudospinifer originally described from the Sea of Japan ) and subsequently reported from Kodiak, Alaska (Dick and Ross 1988); Dick et al. (2005) later recognized Alaskan material as a new species, C. multiavicularia.
There has been no previous intensive local study of intertidal bryozoans in Japan, despite a history of bryozoology in this country beginning with Busk (1884) and followed soon thereafter by Ortmann's (1890) monographic study of Sagami Bay. Information on intertidal occurrences is scattered throughout the literature, including records for Hokkaido or specifically Akkeshi Bay (Silén 1941(Silén , 1942Mawatari 1956Mawatari , 1963Mawatari , 1971Mawatari , 1972Mawatari , 1973aMawatari , 1973bMawatari , 1974Mawatari and Mawatari 1973, 1981a, 1981bd'Hondt and Mawatari 1986;Ikezawa and Mawatari 1993;Suwa and Mawatari 1998;Gordon et al. 2002). At the time we began the present study, 106 bryozoan species had been reported from Akkeshi Bay, including 73 cheilsotomes, 24 cyclostomes, and nine ctenostomes, collectively comprising roughly 60% of the bryozoan species known from the coastal waters around Hokkaido Island. However, few of these records are accompanied by bathymetric, ecological, or local distributional information. This paper is based on a doctoral dissertation presented by Grischenko (2006).

The environment of Akkeshi Bay
Akkeshi Bay is situated on the northeastern coast of Hokkaido Island (Figure 1), 2u of latitude south of the boundary separating the High and Low Boreal Zones in the western Pacific (Ekaterina Strait, Iturup Island, southern Kuril Islands). The cold Oyashio current flows southward from the Bering Sea along Kamchatka and the Kuril Islands, reaching the southeastern coast of Hokkaido, including Akkeshi Bay. The warm Tsushima current, a branch of the Kuroshio current, weakly influences Akkeshi Bay from the southwest (Uchida et al. 1963).
The bay is about 13 km long by 9 km across, opening to the Pacific Ocean to the south and southeast. It is about 30 m deep at its mouth and becomes gradually shallower towards the inner part, with most of its area less than 20 m in depth (Uchida et al. 1963). Two small islands, Daikokujima and Kojima, are the emergent portions of an undersea ridge less than 5 m deep occupying the eastern half of the mouth of the bay. The bay narrows in the northeastern portion, tapering to a narrow channel connecting it with a shallow, brackish lagoon, Akkeshi Lake, into which flows the Bekambeushi River. Akkeshi Lake is about 5 by 8 km in dimensions; it is 11 m deep at the deepest point, but generally less than 2 m deep. Several tidal flats become exposed at low tide, including artificial Kakijima Island, a flat used for culture of the oyster Crassostrea gigas Thunberg.
The tidal pattern at Akkeshi Bay is semidiurnal: the height of high tides is only moderately different between day and night, but morning low waters are lower than afternoon low waters (Saigusa et al. 2000). The annual maximum range in tide level is 1.57 m. The lowest recorded water temperature in Akkeshi Bay was 21.4uC in February 2003, the highest 21.1uC in August 2004. Salinity is relatively constant, about 30 psu; it ranged from 26 psu in June 2003 to 31 psu in August 2004 (Nakamura et al. 2005). Akkeshi Bay is eutrophic; during the spring and autumn, plankton blooms occur and the water column is weakly stratified by phytoplankton (Saito and Hattori 1997).
The shoreline of Akkeshi Bay is varied. The western and northern shores are mostly sandy. The eastern shore consists of exposed capes of solid rock or boulders alternating with sandy beaches, beaches of large cobbles and boulders, and flat rocky platforms. Four main capes are located along this shore: Barasan, Aikappu, Aininkappu, and Mabiro. The capes consist of cliffs, below which lie flat, rocky platforms with crevices and cracks, and aggregated rocks, boulders, and cobbles. The shoreline of Daikokujima Island consists mostly of flat, creviced rocky reefs overlain with boulders and pebbles. In Akkeshi Lake, the tidal flats consist of muddy sand containing shells, gravel, and stones. Artificial Kakijima Island within Akkeshi Bay is surrounded by a border approximately 2-3 m wide consisting of rocks and boulders of portable size, with the interior filled in with oyster shells and sediment.

Collecting sites and collecting
Bryozoans were collected at 10 sites in the Akkeshi Bay complex (Figure 1). The sites were selected because they appeared, upon examination, to support a bryozoan fauna; some sites, such as cobble beaches and stretches of benchrock, were too unstable or exposed to support bryozoans. Two of the sites were located in Akkeshi Lake, six along the rocky eastern shore, and two on Daikokujima Island, each briefly described as follows. Collecting was conducted during cycles of extreme low tides from 2 to 7 June and from 3 to 6 July 2004. Samples were taken from along 10-50 m of shore at each site, beginning up to 1 h before low tide and continuing up to 1.5 h after low tide. We examined the undersides of boulders, pebbles and shells underneath boulders, crevices in reef flat, exposed rock faces, and algae for bryozoans. We either chipped fragments from rocks too large to transport using a hammer and chisel, or carried portable rocks, shell fragments, anthropogenic debris, and algae back to the laboratory at Akkeshi Marine Biological Station for detailed examination.

Laboratory work and data analysis
Bryozoan colonies attached to rocks, shells, and other substrata were cleaned by strong spraying with tapwater, then examined under a stereoscopic microscope. A Dremel tool (Model 395, Type 5; Racine, WI, USA) with a circular diamond-surfaced cutting bit was used to remove small pieces of rock and shell with attached bryozoans. Bryozoan-encrusted pebbles, shells, and fragments removed by Dremel were dried indoors in a warm air stream from an electric heater. Soft or fragile colonies were fixed in 70% alcohol.
For determination of relative abundance, distinct colonies attached to a substratum, as well as separate colony fragments chipped from rocks, were treated equally as ''specimens''. The number of specimens does not accurately reflect the true number of colonies examined, but likely does reflect an approximate relative abundance. Species with larger colonies would have tended to become fragmented during sampling and thus be counted as several specimens. A high number of specimens of a particular species from a particular site could thus reflect a high number of relatively small colonies, fewer relatively large colonies, or both, and in any case would be proportional to abundance in terms of either number of individuals or colony area.
For scanning electron microscopy (SEM), representative specimens of every species encountered were cleaned in sodium hypochlorite solution, rinsed with tapwater, and dried in air. Dried colonies, bleached or unbleached, were coated with Pd-Pt with a Hitachi-1039 ion sputter coater and examined with a Hitachi S-2380N scanning electron microscope at 15 kV accelerating voltage. Images were stored electronically as TIFF files at a resolution of 500 pixels per inch with Image Catcher software.
Measurements of zooidal characters were made by ocular micrometer with a Nikon SNZ-10 binocular microscope. In the text, measurements are given as ranges, with means and standard deviations in parentheses; unless stated otherwise, the sample size for each measurement was N520 zooids or structures, and usually all measurements were taken from a single colony.
For examination of the distribution of cheilostomes in the study area, a two-way indicator species analysis was performed with TWINSPAN for Windows version 2.3 (Hill 1979;Hill and Smilauer 2005). To examine community subdivisions, we performed a cluster analysis of the sampling sites, using BIODIV software (Baev and Penev 1995). This software also calculated the similarity index on which the clustering was performed, the Sørensen-Chekanovsky similarity coefficient (C S 52j/(a+b), where j is the number of species common to two samples and a and b are respectively the total number of species in the two samples).
The specimens described herein are deposited in The Natural History Museum (NHM), London. SEM illustrations are included for some type specimens loaned by the Swedish Museum of Natural History (SMNH), Stockholm, and the Zoological Institute of the Russian Academy of Science (ZIRAS), St Petersburg.
The classification of higher taxa follows the Interim Classification of Families and Genera of Cheilostomata (Working Classification for the Treatise) compiled by D. P. Gordon (unpublished). Authors of genera and higher taxa are not included in the Reference section.  Another way to interpret these data is to consider the likelihood of obtaining a specimen of a particular species during collecting. A likelihood of 1% of detecting a particular species (that is, one in 100 specimens) means that in the collection of 7033 specimens in our study, 70 specimens of this species could be expected. Sixteen of the species had a likelihood of detection of less than or equal to 1% (Stomachetosella decorata and below in Table I). Five species (Callopora sarae and below in Table I) had a likelihood of detection of less than or equal to 0.1%, meaning that between 1000 and 7000 specimens would have to be examined to detect these species.

Species distributions and community subdivisions
An analysis of cheilostome species distribution in Akkeshi Bay with TWINSPAN identified nine distributional groups (Table II). Each group includes species having a similar pattern of distribution, indicated by rows in Table II. Table II suggests that groups of species have similar ecological preferences along environmental gradients, which, however, were not measured in this study. For example, Group 1 includes species restricted to Aininkappu and Aikappu Capes projecting from the central eastern shore of the bay. We speculate that conditions here are intermediate between the exposed environment of the bay mouth and the estuarine environment of Akkeshi Lake. Group 2 contains species occurring only at the southeastern entrance to the bay, where oceanic conditions might prevail. Group 7 contains broadly distributed species, tolerant of conditions ranging from near-oceanic to estuarine.
We also conducted a cluster analysis (Figure 2), which made use of the occurrence of species at the sampling sites (Table I) to group sites of similar species composition. This analysis showed three main clusters having within-group similarity of at least 70%. One cluster contained the Minato and Kakijima Island sites located in estuarine Akkeshi Lake. This cluster was separated from all other sites, probably due to the absence of many species shared among sites in the bay proper. Another cluster contained sites located on Barasan, Aikappu, and Aininkappu Capes along the eastern shore of the bay, with the two inner sites (Barasan Cape and Aikappu Cape West) forming a group, and the outer two sites at the tips of Aikappu and Aininkappu Capes forming another group. The three sites near the bay mouth comprised the third cluster, with the two sites on Daikokjima Island forming a group to the exclusion of Mabiro Cape. The striking congruence of the main clusters with inner, middle, and outer regions of the bay complex strongly suggests that the composition of the intertidal bryozoan assemblages varies along a gradient of environmental parameters from the inner to the outer bay.

Substrata
Most cheilostome bryozoans in the study area (34 species, or 87%) occurred on rocky substrates, including solid rock, the undersides of boulders and cobbles, and pebbles (Table III). Both the undersides of some boulders and many pebbles under larger rocks were covered almost entirely by bryozoan colonies. Several species attained high abundance, either in density or cover, on undersides of boulders, e.g. Microporella luellae, Raymondcia rigida, Raymondcia klugei, Myriozoella plana, Cryptosula zavjalovensis, and Hippoporella multiavicularia. Tegella aquilirostris, Desmacystis sandalia, and Cryptosula zavjalovensis were the only species observed on the upper surface of stable boulders. Twenty-one species (54%) also encrusted shell fragments accumulated beneath and among larger rocks (Table III). Eight species (21%) were found on algae. One species (2.6%), Celleporina porosissima, was encountered only on the rhizoids of Laminaria sp.
Other substrata used by bryozoans included hydroids, polychaete tubes, other bryozoans, and anthropogenic debris. The two Celleporina species (5.1%), C. minima and C. nordenskjoldi, occurred on hydroid stolons. Parkermavella orientalis (2.6%) was noted on tubes of serpulid polychaetes. One undetermined cyclostome bryozoan, four Figure 2. Dendrogram (below) showing similarity of sampling localities based on species presence/absence, and a map (above) showing the geographical distribution of three groups of sampling localities (circled) that are less than 70% similar in species composition between groups. ctenostome bryozoans (Alcyonidium alcilobatum d' Hondt and Mawatari, 1986; Flustrellidra akkeshiensis Mawatari, 1971; F. corniculata (Smitt, 1872); and F. filispina Mawatari, 1971) and three cheilostome species (Bugula pacifica, Cryptosula zavjalovensis, and Phidolopora elongata) were substrates for another 16 (41%) bryozoan species. Large, erect colonies of Phidolopora elongata provided a substrate for 11 (28%) cheilostome species. Finally, several species were observed on anthropogenic debris scattered along the intertidal zone. Electra korobokkura was found on a piece of abraded metal plate, and eight other species (21%) (Cauloramphus niger, Cauloramphus spinifer, Tegella aquilirostris, Tricellaria occidentalis, Cribrilina annulata, Celleporella hyalina, Cryptosula zavjalovensis, and Microporella luellae) were repeatedly found on various kinds of plastic debris. Among them, C. spinifer was abundant on plastic, forming extensive encrustations occupying considerable space. Based on the range of substrata used (Table III), some of the cheilostome species of Akkeshi Bay could be classified as clearly stenotopic or eurytopic. Eleven species (28%) were stenotopic, occurring only on a single type of substratum: eight species (21%) were found only on rock (Cauloramphus multispinosus to Conopeum nakanosus); two species (5.1%), Callopora craticula and Cheilopora sincera, were found only encrusting the erect bryozoan Phidolopora elongata; and one species (2.6%), Celleporina porosissima, was found only on algae. Five species (13%) (Cauloramphus niger, C. spinifer, Cribrilina annulata, Celleporella hyalina, and Desmacystis sandalia) were eurytopic, noted on rock, shells, algae, and sometimes other substrata. The rest of the species fell somewhere in between stenotopy and eurotopy; for example, 15 species (38%) were observed on rock, shell, and sometimes other substrates, but not on algae. Table IV indicates the presence of embryos in bryozoan colonies during either of the periods of collection (2-7 June and 3-6 July 2004). Of the 39 species found in the study, 33 (85%) contained embryos; among these, the percentage of colonies with embryos was determined for 26 species, and presence or absence only was noted for seven species for which it was difficult to observe embryos in dried specimens. For the same reason, it was unclear whether four of the species contained embryos. Two species (5%), Cheilopora sincera and Bugula pacifica, lacked embryos or ovicells.  Zoogeography Table V shows a biogeographical categorization of the known distributions of the intertidal cheilostomes of Akkeshi Bay. Nearly two-thirds of the species (59%) are Boreal, known only from northern temperate waters. Over two-thirds of the Boreal species (38% of total species) are Low Boreal Asian Pacific species. This category may be artificially inflated because all but one of the nine new species described herein are known only from Akkeshi Bay, and some of these may eventually prove to have broader distributions. Over one-quarter of the species (28%) have Arctic-Boreal distributions; all but one of these

Description
Colony encrusting, yellowish when alive, forming a thin, irregular network, the largest observed about 2.5 cm across; zooids arranged in anastomosing uniserial branches that expand to lobes two to four zooids across, sometimes more ( Figure 3A-C). Zooids oblong, thin-walled, widest in middle, 0.40-0.60 mm long (0.49¡0.05 mm), 0.22-0.35 mm wide (0.28¡0.03 mm), rounded distally, narrower proximally; separated by a deep groove along lateral margins when anastomosed; lateral walls smooth, vertical to sloping when uniserial; transverse boundary between zooids indistinct. Mural rim raised, sharp. Cryptocyst a narrow, granulated sloping shelf below marginal rim ( Figure 3D). Opesia oval to elliptical, 0.25-0.35 mm long (0.31¡0.03 mm), 0.14-0.20 mm wide (0.16¡0.02 mm), occupying 50-70% of zooidal length. Frontal membrane thin, transparent. Operculum ( Figure 3D) semicircular, calcified, of sharply contrasting white colour. Gymnocyst smooth, narrow distally and laterally, elongate and tapering proximally, semicircular in transverse section, sometimes with weak transverse rugae. Proximal to opesia, gymnocyst often rises into a short, calcified spinous process ( Figure 3D). Zooids interconnect via a few minute pores in the basal half of distal wall. Avicularia, lateral spines, ovicells, and hibernacules absent. Ancestrula not seen. Nikulina (2006) recently described this species from Akkeshi Bay. Among congeners, Electra korobokkura most closely resembles Electra arctica (Borg, 1931) in growth form, as both can form narrowly multiserial branches in portions of colonies. For comparison with E. korobokkura (Figure 3), we illustrate a subtidal specimen of what we consider to be E. arctica, from the western Kamchatka shelf, Sea of Okhotsk ( Figure 4). This specimen illustrates all the characters considered diagnostic by Hansen (1962), who briefly reviewed E. arctica: well-developed gymnocyst ( Figure 4B-D); enlarged base of proximal spine ( Figure 4C, D); well-developed cryptocyst with a crenulated border ( Figure 4D); more heavy calcification than in other Electra species; operculum heavily calcified, with the proximal border straight (in our specimen, there is intra-colony variation, with some opercula having a concave proximal border) ( Figure 4D); the opesial opening restricted or reduced by a closure plate ( Figure 4C) in some zooids; and a tendency to form kenozooids (Figures 4B-D) of reduced size among autozooids. Compared to E. arctica, E. korobokkura has a greater tendency to form uniserial branches (compare Figure 3 with Figure 4). Compared to Electra arctica, zooids of Electra korobokkura appear less heavily calcified; the proximal gymnocyst has the transverse rugae weaker or lacking; the cryptocyst is narrower and much less heavily granulated, leaving a larger opesial opening in relation to overall zooid size; the operculum is less heavily calcified; and the basal chamber of the proximal spine is smaller. Furthermore, the closure plates of kenozooids have a smooth surface in E. korobokkura (Nikulina 2006, Figures 2, 3B, 4) whereas in E. arctica the closure plates have a broad, granulated component of cryptocystal calcification surrounding the opening ( Figure 4C, D). Zooids of the E. arctica specimen we illustrate (Figure 4) are 0.42-0.73 mm long (0.57¡0.07 mm) by 0.22-0.48 mm wide (0.34¡0.06 mm), and the opesia is 0.20-0.33 mm long (0.26¡0.03 mm) by 0.12-0.21 mm wide (0.18¡0.02 mm). Thus, while the zooids of E. arctica are larger than those of E. korobokkura, the opesia of the former is roughly the same size or even smaller than that of the latter.

Diagnosis
Colony coherent; opesia large; gymnocyst well developed; cryptocyst moderately well developed, the entire rim crenulate; chitinous spine well developed; base of spine variable in size; chitinized operculum contrasting sharply in colour with the transparent frontal membrane; colony uniserial only near ancestrula, rapidly giving rise to fan-shaped coherent sheets.

Etymology
The species name refers to this species' distribution in the northwestern (Asiatic) Pacific.

Remarks
Among congeners, E. asiatica most closely resembles E. baltica (Borg, 1931) in having a similarly coherent colony, a large opesia, and a chitinized operculum contrasting sharply in colour with the transparent frontal membrane. Both species commonly have, proximal to the opesia, a knob terminating in a pointed chitinous spine. However, unlike E. baltica, chains of zooids of E. asiatica radiate and anastomose only close to the ancestrula, and subsequently form only fan-shaped multiserial sheets that do not diverge again into uniserial chains. In E. asiatica, the proximal gymnocyst is more extensive and the opesia is proportionately smaller than in E. baltica. Owing to the general similarity between these two species, previous investigators have likely interpreted the differences in zooid size and colony form as intraspecific variation in a single species, E. baltica. The records of E. baltica by Kubanin (1976Kubanin ( , 1997 from numerous localities in the Far Eastern seas appear to be misidentifications of E. asiatica; a specimen from the intertidal zone of Ptichiy Island ( Figure 5H) proved to be this species.
At Akkeshi, E. asiatica is often found growing close to E. korobokkura on the same substratum, particularly on dead bivalve shells. Electra asiatica has larger zooids than E. korobokkura (compare Figure 5C-E, H with Figure 3D at the same scale), a difference readily apparent in specimens close to one another. Nikulina (2006) illustrated and discussed this size difference, but identified the species with larger zooids as E. arctica.

Distribution
We consider Kubanin's (1976Kubanin's ( , 1997 E. baltica to be synonymous with E. asiatica. To the extent that all of Kubanin's records represent correct identifications of the latter, E. asiatica is widely distributed in the northwestern Pacific: southeastern Kamchatka (Kamchatsky, Kronotsky, and Avacha Gulfs); northern and northeastern coast of the Sea of Okhotsk (southwestern Kamchatka; Penzhinskaya, Gizhiginskaya, Yamskaya, and Taujskaya Inlets; Okhotsk; Ayan; Zavyalov Island); southern and southeastern coast of Sakhalin Island (Terpeniya and Aniva Gulfs); northern Sea of Japan (southwestern coast of Sakhalin Island and Moneron Island). Akkeshi Bay is the southernmost known locality.

Diagnosis
Zooids large; opesia about 70-80% of zooidal length; cryptocyst widest proximally, finely tuberculate; gymnocyst hidden; all zooids with a triangular, cone-shaped kenozooid at each distal corner; kenozooids with small, distally facing opesial opening in central depression, sometimes transformed into an avicularium with triangular, vertically orientated mandible; two to three spines occasionally present around mural rim. Ovicells absent. Multiporous septula in lateral and transverse wall.

Etymology
The species name refers to Nakanose Bank in Akkeshi Bay, where the holotype specimen was collected subtidally.

Description
Colony encrusting, multiserial, sheet-like, up to 3.5 cm across, greyish when alive, with purple to reddish marginal zone one or two developing zooids deep. Zooids ( Figure 6A, B, D) oval, hexagonal, or subquadrangular, widest in middle, 0.67-0.95 mm long (0.81¡0.09 mm), 0.40-0.50 mm wide (0.45¡0.02 mm), separated by fine sutures between raised lateral walls. Opesia oval, elliptical, or rounded-rectangular; 0.55-0.73 mm long (0.63¡0.04 mm), 0.30-0.43 mm wide (0.36¡0.03 mm); occupying 70-80% of zooidal length. Cryptocyst ( Figure 6A, F) extending around entire opesia; steep and narrow laterally, flatter and wider proximally, reduced distally, finely tuberculate below mural rim. Lateral and proximal gymnocyst hidden between adjacent zooids. Distal gymnocystal wall raised slightly above level of lateral wall into a smooth, narrow, sharpedged crescentic lip separated from cryptocyst by a transverse, slit-like cavity. At each distal corner is a raised, cone-shaped, subtriangular kenozooid ( Figure 6F, H); sharp or blunt on top, smooth except for depressed, finely granulated cryptocystal area distolaterally, surrounding a small, circular to oval opesial opening. Kenozooids at first have this cryptocystal area covered with transparent membrane, but later often become closed to form a pair of knobs, rarely vestigial or lacking. Sometimes kenozooids are transformed into avicularia ( Figure 6E) with a distolateral rostral face; mandible triangular with sharp, hooked, vertically orientated tip; such avicularia are rare, developing only in mature regions of the colony. Mural rim often with two or three spines ( Figure 6C): a straight, blunt hollow tubular spine on one or both sides of orifice, in line with proximal orificial margin, jointed at base; another finer, curved, acute spine more proximally on one side, slightly arching over opesia. Like the avicularia, the spines tend to appear in mature regions of the colony; many young zooids lack them altogether. Ovicells lacking, but raised mural rim distally with transverse, slit-like cavity below ( Figure 6F) may represent kenozooidal ooecium. Four oval multiporous septula ( Figure 6G) in each lateral wall and two septula in transverse wall. Ancestrula not observed.

Remarks
In zooidal characters, C. nakanosum is very similar to Conopeum reticulum (L., 1767), the type species of the genus, especially in having triangular kenozooids at the distal angles of autozooids. However, these two species differ in several characters. In C. reticulum, the kenozooids develop predominantly on older zooids, whereas in C. nakanosum they occur on all zooids, at all astogenetic stages. With increasing calcification in C. reticulum, the pair of kenozooids occasionally comes into contact along the entire distal end of the zooid, forming a single, dumbbell-shaped unit. The kenozooids of C. nakanosum are always separated by a raised, crescentic gymnocystal lip. The opesial opening in the central depression of the kenozooids is terminal on the knob and facing upward in C. reticulum, whereas it always faces distally or distolaterally in C. nakanosum. Finally, zooid size in C. nakanosum exceeds that of C. reticulum, and the ranges do not overlap: zooids are 0.67-0.95 mm long by 0.40-0.50 mm wide in the former, compared to 0.4-0.6 mm long by 0.2-0.3 mm wide in the latter (Hayward and Ryland 1998).
There has been some confusion over the nature of the distal kenozooids in Conopeum, and both Osburn (1950) and Mawatari (1974) were sceptical that these represented vestigial avicularia or ever formed true avicualaria. However, there is no doubt that in C. nakanosum the distolateral kenozooids can be replaced by avicularia. This does not occur on every zooid, but usually only among a few zooids in a mature region of a colony. It remains unclear whether the possible occurrence of avicularia in place of the kenozooids should be incorporated into the generic diagnosis of Conopeum. Another anascan, Crassimarginatella leucocypha Marcus, 1937, described from Brazil and subsequently reported by Mawatari (1952) from the Kii Peninsula, Sea of Japan, is morphologically similar to C. nakanosum in having angular kenozooids that can transform into an avicularium (see Marcus 1937, p 46, Plate 8, Figure 20A). The apparent lack of ovicells in the former suggests Conopeum rather than the calloporid Crassimarginatella. The kenozooids in this species are more extensive than in C. nakanosum, with a more central opesial opening, and there has been no mention of spines.

Distribution
Akkeshi Bay is the only known locality for C. nakanosum. Mawatari's (1956) short description and illustration of C. reticulum from the Kuril Islands (Alaid and Paramushir) and from Hokkaido (Akkeshi Bay and Kushiro) suggest he might have had C. nakanosum instead.

Distribution
This is a circumpolar, Arctic-Boreal species. Gontar and Denisenko (1989) summarized numerous previous records from the Arctic. In the eastern Atlantic, C. craticula extends southward to Scotland and northern England in Britain (Hayward and Ryland 1998), but its occurrence in the western Atlantic is uncertain (Winston et al. 2000). In the eastern Pacific it extends as far south as Ketchikan (Dick et al. 2005). In the western Pacific, it has been recorded from Saint Lawrence Island, Avacha Inlet (Kluge 1961), Commander Islands (Grischenko 1997, Sea of Okhotsk, Kuril Islands , northern Sea of Japan along Primorye and western Sakhalin Islands , and from Kushiro, Akkeshi, and Hakodate on Hokkaido, northern Japan Mawatari 1980, 1981b).

Diagnosis
Opesia 60-85% of zooidal length; 17-25 spines with enlarged, cylindrical bases around mural rim, including three pairs of thick, vertical distal spines (the most proximal pair by far heavier than the rest), 12-17 thin proximal spines arched over opesia, and often one or two additional thin spines present on outer distolateral gymnocyst; avicularia small, on lateral or proximal gymnocyst, usually single, rarely paired, associated with distal half of ovicell in mature zooids; ovicell with a wide transverse tabula; transverse ridge often enlarged into a median knob; proximal ovicell margin raised as a wide lip; zooids interconnect by multiporous septula.

Etymology
The species is named in honour of Sarah Taranto, who collected the holotype specimen.

Remarks
Callopora sarae is similar to C. corniculifera (Hincks, 1882), a Boreal species reported from both sides of the Pacific (see Soule et al. 1995, p 66, Plate 16A, B), in having small lateral avicularia. However, C. corniculifera has larger zooids (0.65-0.70 mm long by 0.45-0.50 mm wide); roughly half the number of spines (10-12); and a smaller avicularium situated along the lateral margin only, in line with the proximal orificial margin. Callopora decidua Dick and Ross, 1988, described from Kodiak, Alaska, also has a high spine number, 14-21, including three pairs of heavy distal spines, as well as small, single or paired lateral avicularia; however, in this species, proximal avicularia are rare unless preceded by an ovicell, and when they occur, the rostrum is long and curved.
Callopora longispinosa , distributed in the northwestern Pacific, is similar to C. sarae in having two or three pairs of long, stout distal spines, occasionally 1.5-2 times or more the length of the zooid (Grischenko 1997); however, this species differs from C. sarae in having fewer proximal spines (four to six); avicularia rare and located on the proximal gymnocyst only; and the spherical ovicells weakly calcified and not associated with avicularia.

Distribution
Akkeshi Bay is the only known locality.

Diagnosis
Zooids closely appressed; gymnocyst negligible; cryptocyst strongly developed, widest proximally, entirely and uniformly coarsely granulate; 8-10 spines, generally short, not meeting in opesial midline, including two to four straight orificial and four to six thin, curved opesial spines; avicularia commonly paired, originating between orificial and opesial spines, approximately lateral to proximal orificial border, with short, thin pedicel abruptly expanding into curved chamber; non-pedunculate avicularia can arise from pore chambers at colony margin, sometimes in abundance.

Etymology
The species name is from the Greek kruptos (hidden) and Latin armatus (armed), referring to the occurrence of small marginal avicularia, which are unusual for the genus.

Remarks
Cauloramphus cryptoarmatus resembles eastern-Pacific C. variegatus (Hincks, 1881) and C. tortilis Dick et al., 2005 and Korean C. korensis Seo, 2001 in having closely appressed zooids with a broad, cryptocystal mural rim uniformly covered with coarse granules. However, in C. variegatus, the bases of the spines are yellowish brown to black in colour (Dick and Ross 1988), and their number, shape, and arrangement are different: there are three pairs of blunt, cylindrical distal spines and two to six much finer, acuminate proximal spines curved over the opesia, with the tips often converging near the midline. Cauloramphus tortilis has more and longer spines (9-15 heavy, elongate spines, with strong bases conspicuous after bleaching), and the pedunculate avicularia are as long as the longest spines, with a peduncle that is twisted around the spine near which it originates. Cauloramphus korensis is similar in having a coarsely granulated cryptocyst widest proximally (Seo 2001). However, C. korensis has dark spines that are brown or violet in colour, fewer total spines (five to nine) and fewer orificial spines (one to three), and the proximalmost opesial spines originating close to the edge of the opesia (see Seo 2001, Figure 1A) rather than at the zooidal margin. Furthermore, the pedunculate avicularium of C. korensis is quite curved in the rostral direction, whereas that of C. cryptoarmatus is fusiform. Finally, the kenozooidal ooecium of C. korensis is larger and more conspicuous than that of C. cryptoarmatus. The non-pedunculate marginal avicularia observed in some colonies of C. cryptoarmatus have not been reported in the other species just mentioned.

Remarks
Specimens from Hokkaido agree well with the original description (Dick and Ross 1988) of C. magnus from Kodiak Island, Gulf of Alaska. Both zooidal and opesial sizes are similar, but the spine number is greater at Akkeshi (11-18 compared to 11-14); a population from Ketchikan, Alaska (Dick et al. 2005) also had more spines (12-18) than material from the type locality.

Distribution
This is a Boreal Pacific species, reported in the eastern Pacific from Kodiak Island in the western Gulf of Alaska (Dick and Ross 1988) and Ketchikan in southern southeast Alaska (Dick et al. 2005). In the western Pacific, it has been reported from the Commander Islands (Grischenko 1997, though this material needs reexamination by SEM. We also collected this species at Oshoro Bay (Hokkaido), which is the southernmost known locality for C. magnus.

Diagnosis
Zooids large, oval, demarcated by groove; opesia large, up to 85% of zooidal length; cryptocyst narrow, granulated; gymnocyst reduced, smooth; 20-26 spines around mural rim; distal spines long, heavy, straight or nearly so; proximal spines very thin, sharp, curved and arched over opesia to meet nearly horizontally at midline; avicularia not observed and may be absent in this species; ancestrula tatiform.

Etymology
The species name refers to the unusually large number of spines.

Remarks
Cauloramphus multispinosus is distinguishable from congeners by the large size of zooids; large number of thin, curved spines (up to 26) forming a tight, neat basket; and the apparent absence of avicularia. Northeastern Pacific C. spectabilis Dick and Ross, 1988 has up to 24 spines, but differs in having four pairs of heavy orificial spines, thicker opesial spines, longpedunculate avicularia, and smaller zooids, not exceeding 0.65 mm in length. Cauloramphus pseudospinifer Androsova, 1958 has up to 23 spines, but compared to C. multispinosus, there is less difference in thickness between orificial and opesial spines; the opesial spines are thicker; and there are thick, pedunculate avicularia (see Dick et al. 2005, Figure 3G, H).

Distribution
Cauloramphus multispinosus is at present known from Akkeshi and Oshoro Bays, Hokkaido.

Diagnosis
Zooids closely appressed; mural rim narrow, rounded, largely cryptocystal, covered with conical granules; gymnocyst reduced, evident proximally; 12-19 spines, including four to six vertical orificial spines and 8-13 curved opesial spines angled over opesia; spines light with dark bases in marginal zooids, dark brown to black in older zooids; avicularia single, numerous near colony periphery, longer than distal spines, with short peduncle rapidly expanding into chamber with terminal rostral face; mandible triangular, the tip rounded; kenozooidal ooecium much larger than in other species, granulated.

Etymology
The species name is from the Latin niger (dark coloured, black), referring to the colour of the spines.

Description
Colony unilaminar, encrusting, coherent, forming irregularly circular patches up to 3 cm across; brown, dark yellow, or grey when alive, with reddish to pink marginal zone one or two developing zooids deep. Zooids ( Figure 12A-F) oval to rounded-hexagonal, occasionally tapering proximally, 0.47-0.65 mm long (0.56¡0.04 mm), 0.30-0.45 mm wide (0.35¡0.03 mm), closely appressed, demarcated by a shallow groove. Mural rim ( Figure 12E) narrow, rounded, largely cryptocystal, covered with conical granules. Gymnocyst reduced, smooth, occasionally elongate proximally between adjacent zooids, distinct from granulated mural rim. Opesia ( Figure 12E) oval, widest in middle or sometimes proximally, 0.27-0.40 mm long (0.35¡0.03 mm), 0.17-0.28 mm wide (0.22¡0.02 mm), with crenulate outline due to granulation, occupying 60-80% of zooidal length. Around mural rim, 12-19 spines ( Figure 12B-D), light yellowish with dark bases in marginal zooids, rapidly changing to entirely dark brown or black in older zooids, contrasting sharply with yellowish to greyish zooidal walls; two to three pairs of orificial spines long, hollow, thick, blunt, vertically orientated, with enlarged cylindrical bases, occasionally most distal pair reduced in size; 8-13 opesial spines thin, sharp, arched over opesia, occasionally meeting in midline; distance between adjacent opesial spines two to three times or more their basal thickness; in some proximally broadened zooids, one to four most proximal spines are vertically orientated and slightly longer and thicker than the others. Avicularia ( Figure 12C, D) single, arising from distolateral gymnocyst between orificial and opesial spines, of same length as or slightly longer than orificial spines; short, thick peduncle rapidly expanding to thick, laterally compressed chamber with terminal rostral face; mandible triangular, the tip rounded, pointing in any direction, but most frequently laterally; avicularia numerous near periphery of colony but may be entirely absent in older areas. Kenozooidal ooecium ( Figure 12E) distal to orifice is prominent, caplike, with granulose surface, occupying entire space between bases of distalmost spines. Six pore chambers in each lateral wall and two in distal wall. Ancestrula ( Figure 12F) tatiform, oval, with entirely calcified basal wall, 0.39 mm long by 0.30 mm wide; large, oval opesia 0.28 mm long by 0.20 mm wide; 10 spines around opesial margin. Ancestrula buds triplet of small zooids distally and distolaterally, each with six or seven spines, including two pairs of hollow distal spines with enlarged bases and two or three thin proximal spines curving over opesia.

Remarks
Characters that distinguish C. niger from other reported species of this genus are the unusually large kenozooidal ooecium and the dark brown spines. The only other known species of Cauloramphus with similarly dark spines is C. brunnea Canu and Bassler, 1930, originally described from the Galapagos Islands. However, whereas zooids of C. niger are closely appressed, those of C. brunnea are separated by a broad groove, so that the opesiae are about as far apart as their width. In addition, the avicularium of the latter species is long-pedunculate, narrowly fusiform, and attached proximal to the middle of the zooid, lateral to the zone of opesial spines. The light yellowish spines with dark bases on marginal zooids of C. niger are similar to spines of C. variegatus (Hincks, 1881); however, the spines of the latter are never entirely dark brown, and other characters differ as well.

Distribution
Cauloramphus niger is known at present only from Akkeshi Bay.

Remarks
Dick and Ross (1988) discussed diagnostic characters and geographic variation in spine number for C. spinifer.

Distribution
This is a circumboreal species extending into the Arctic. In the eastern Atlantic, it is widely distributed in cool-temperate waters, from the White Sea (Gostilovskaya 1978) and Barents Sea (Kluge 1962) southward to the Shetland Isles and northern coast of France. In the eastern Pacific, C. spinifer is known from Kodiak Island, Gulf of Alaska; previous records from farther south need to be re-examined (Dick and Ross 1988). In the western Pacific it has been reported from the Gulf of Anadyr and the vicinity of St Lawrence Island in the Bering Sea; Commander Islands; Sea of Okhotsk; Kuril Islands; Sakhalin Island; and Primorye and Gulf of Peter the Great in the northern Sea of Japan Kubanin 1997;Grischenko 1997. From Japan, C. spinifer is known from the Pacific coast of Hokkaido, including Akkeshi, Kushiro, Mori, Muroran, and Shirikishinai, southward to middle Honshu Mawatari 1981a, 1981b). We have determined material from Oshoro Bay, western coast of Hokkaido, Sea of Japan, identified by Kubota and Mawatari (1985a) as C. spinifer, actually to be C. magnus.

Remarks
Colonies often aggregated, covering considerable areas on benchrock faces and beneath overhangs; those in sheltered microhabitats are often loosely attached, the margin sometimes raised in folds and frills.

Distribution
This is a Boreal Pacific species. In the eastern Pacific there are records from Kodiak Island (Dick and Ross 1988) southward to Ketchikan (Dick et al. 2005), British Columbia (O'Donoghue and O'Donoghue 1923), Puget Sound (McCain and Ross 1974, and Santa Barbara, California (Osburn 1950). In the western Pacific, T. aquilirostris is known from the Commander Islands (Grischenko 1997; and from the Kuril Islands, Aniva and Terpeniya Gulfs of southeastern Sakhalin Island, the Sea of Okhotsk, and Primorye in the Sea of Japan (Kubanin 1997). In Japan, this species has been found at Akkeshi, Kushiro, and Hakodate on Hokkaido, and also at Kominato and Misaki, Pacific coast of Honshu Mawatari 1980, 1981b).  Figure 163; Gontar 1980, p 5;Mawatari and Mawatari 1980, p 94, Figure 33;1981b, p 46;Dick and Ross 1988, p 43, Plate 3B;Kubanin 1997, p 121;Grischenko 1997, p 157;2004, p   pair located just proximal to lateral avicularia and in line with orifice, straighter, thicker and more erect than the rest; proximal four to six spines curved, acuminate, arched over opesia. Proximal gymnocyst covered by relatively large frontal avicularium with raised rostrum; mandible triangular, with complete cross-bar, raised at an angle to frontal plane, directed proximally or proximolaterally, not extending laterally around aperture; abutting and covering distal half of ovicell of preceding zooid, in this case pointing distolaterally. Two small paired avicularia located on mural rim lateral to orifice; mandibles triangular, directed distally and raised 45u from frontal plane. Ovicells ( Figure 15D) hyperstomial, hemispherical, imperforate, 0.13-0.19 mm long (0.16¡0.02 mm), 0.16-0.21 mm wide (0.19¡0.02 mm), with thick transverse rib across top that may be thickened into a blunt median knob. Proximal margin raised, with lunate tabula between margin and ridge. Four pore chambers in each lateral wall and two in basal half of distal wall. Ancestrula not observed.

Distribution
Tegella arctica is a circumpolar, Arctic-Boreal species. Kluge (1962Kluge ( , 1975 and Gontar and Denisenko (1989) summarized a number of previous Arctic records. In North America it is known from Point Barrow southward to Frederic Sound, southeastern Alaska (Osburn 1950;Dick and Ross 1988). In the western Pacific, there are records from the Bering Sea near St Lawrence Island, along western Kamchatka (Kluge 1961;Kubanin 1997), the Commander Islands (Grischenko 1997, the Shantar Archipelago (Kluge 1961), and the Kuril Islands ) on the Sea of Okhotsk side. In Japan it has previously been found at Akkeshi and Hakodate, Pacific coast of Hokkaido Mawatari 1980, 1981b).

Description
Colony ( Figure 16A) erect, flexible, dichotomously branched, spiralled around a central axis with basal sides facing outwards. Our specimens small, up to 5.5 cm high, yellow in colour when alive, attached by a stalk composed of rhizoids originating from zooids at base of colony. Branching dichotomous, with axil at bifurcations formed by inner zooid of each of first pair on either side of branch point (branching pattern type 3; Ryland 1998 after Harmer 1923). Zooids ( Figure 16B) in biserial series, elongate, narrow, 0.50-1.03 mm long (0.67¡0.13 mm), tapering proximally, truncate distally, with thin, flexible, transparent, weakly calcified walls. Opesia occupies nearly entire frontal surface, leaving only a small zone of proximal gymnocyst. One relatively short, spinous projection located on inner distal zooidal margin; two elongate spinous projections on outer distal margin; distalmost projection strongly thickened, pointing distolaterally, curved away from zooid. Avicularia attached by a flexible joint near base of zooid, on lateral wall close to opesial margin, 0.21-0.27 mm long, with hooked rostrum; not occurring on all zooids. Neither ovicells nor ancestrula present in our material.

Remarks
Androsova (1977) described a new subspecies, Bugula pacifica nana, from Aniva Gulf, southern Sakhalin Island, Sea of Okhotsk. In comparison with the nominal subspecies, B. pacifica pacifica, this subspecies forms smaller colonies, zooids, and zooidal structures (see Androsova, 1977, p 795). Kubanin (1984aKubanin ( , 1984bKubanin ( , 1997 reported B. pacifica nana from Primorye, Peter the Great Gulf, southern coast of Sakhalin Island, and concluded that it is distributed only in Low Boreal Asiatic waters. He suggested that B. pacifica pacifica is primarily distributed in the Boreal eastern Pacific, extending to the fringes of the western Pacific, including the Commander Islands. The finding of B. pacifica pacifica in Akkeshi Bay shows that this form has a trans-Pacific distribution, although its occurrence in the western Pacific as a relatively recently introduced population cannot be ruled out.

Distribution
Bugula pacifica is a Boreal Pacific species distributed from the Channel Islands off Southern California  northward to Ketchikan, Alaska (Dick et al. 2005) and the Pribilof Islands, Bering Sea (Robertson 1905;Osburn 1950). On the Asian side, it has previously been reported from the Commander Islands (Grischenko 1997); Akkeshi Bay is the southernmost known locality in the western Pacific.

Family CANDIDAE d'Orbigny, 1851
Genus Tricellaria Fleming, 1828 Tricellaria occidentalis (Trask, 1857) ( Figure 16C-F) Menipea occidentalis Trask 1857, p 102, Plate 4, Figure 4. Menipea occidentalis: Jelly 1889, p 173; Robertson 1905, p 254, Plate 6, Figures 22-25;Yanagi and Okada 1918, p 409;O'Donoghue and O'Donoghue 1923, p 159;1925, p 99;Okada and Mawatari 1936, p 59;1937, p  Description Colony ( Figure 16C) bushy, tightly arborescent, composed of dense branches curved and rolled inward. Colony up to 3 cm high, yellow to tan in colour when alive, attached by a bundle of rhizoids ( Figure 16D) originating from zooids situated low in colony. Branching dichotomous, regular, branching pattern type 9, with proximal extremity of zooid F and G not in contact Ryland 1998 after Harmer 1923). Most internodes have three zooids, but distal ones with ovicells have five to nine zooids. Internodes connected by strong, tubular chitinous joints, brown in colour; flexible nodes crossing well proximal to opesia of both outer and inner zooids. Zooids in biserial series, alternate, elongate, narrowing proximally, 0.45-0.78 mm long (0.63¡0.10 mm), varying significantly in size depending upon their location in internode and in colony. Opesia oval or elliptical, 0.22-0.28 mm long (0.25¡0.02 mm), 0.11-0.14 mm wide (0.12¡0.01 mm), occupying 30-50% of zooidal length; cryptocyst negligible. Zooids ( Figure 16E) typically have six hollow, tubular jointed spines; most proximal pair straight, located near middle of opesia and tilted slightly inward; next pair straight or turned slightly outward; outer distalmost spine strongly calcified, elongate (up to 0.65 mm long), originating from dorsal side of distal zooidal margin; inner distalmost spine often strongly reduced or lacking. In some zooids the external proximal spine is bifid. Axial zooids at bifurcations have six spines: two pairs along distolateral margins of opesia, straight or tilted slightly outwards, and a medialmost pair of tubular, hollow, elongate spines originating from dorsal side of distal wall from closely set, heavily calcified cylindrical bases; these latter two spines are often asymmetrically placed with respect to the midline, with one occupying the midline position and longer and heavier than the other. Scutum attached to inner border of opesia, proximal to middle of opesia, varing in form from spine-like to a broad flabellate process with two to four or more lobes. Non-axial zooids have large lateral avicularium with hooked rostrum; mandible triangular, with hooked tip. Frontal avicularia absent. Ovicells ( Figure 16F) globular, smooth, wider than long, 0.16-0.21 mm long (0.18¡0.01 mm), 0.20-0.24 mm wide (0.21¡0.01 mm), with around 10 small, circular or oval pores. Some zooids situated low in colony have proximally directed kenozooids (rhizoids or radicle fibres) originating from a slight, flattened disc on dorsal or lateral wall, close to nodal joints. Ancestrula not observed.

Remarks
Many characters of T. occidentalis, such as presence or absence of lateral avicularia, number of zooids per internode, and size and shape of scuta, may vary considerably even within a single colony Gordon and Mawatari 1992). This variation, noted by Robertson (1905), Yanagi and Okada (1918), Okada (1929), and Osburn (1950), and superbly illustrated by Mawatari (1951), has caused taxonomic difficulties and resulted in descriptions of several varieties of this species. In specimens from Akkeshi, the scutum varies in form from spine-like to a broad flabellate process with two to four or more lobes, which is within the range of variation of T. occidentalis (see also Soule el al. 1995). Dyrynda et al. (2000) concluded that Japanese and some other populations previously reported under the name Tricellaria occidentalis Trask are actually T. inopinata d' Hondt and Occhipinti Ambrogi, 1985, a fouling species recently described from the Lagoon of Venice. According to Dyrynda et al. (2000), T. inopinata is of Pacific origin, is widely distributed along the coasts of western North America and Japan, occurs in Australia and New Zealand, and has been introduced to Britain and the Mediterranean. They note that the original source region of the species in the Pacific, before anthropogenic introductions to other areas, is unknown. Dyrynda et al. (2000) regarded T. inopinata as consistently morphologically distinguishable from T. occidentalis Trask, which they considered to be restricted to the western coast of North America. However, they considered many of the Pacific records of T. occidentalis var. catalinensis Robertson, 1905 to be T. inopinata. Our material has one of the distalmost pair of spines on axial zooids located in the midline and better developed than the other; zooids occasionally with a bifid spine; and a quite variable scutum, sometimes broad with a spiky margin. According to Dyrynda et al. (2000) these are all characters distinguishing T. inopinata from T. occidentalis. Unfortunately, these authors did not provide a detailed description and illustrations of the latter, noting that the type description of T. occidentalis was inadequate, that the type material was lost, and that they had been able to examine only limited quantities of mostly very old material. Here we retain the name T. occidentalis, pending a detailed redescription of T. occidentalis Trask as distinct from T. inopinata d'Hondt and Occhipinti Ambrogi and clarification of the range of the former.

Description
Colony encrusting, coherent, forming irregularly circular sheets up to 2.7 cm across; lightyellow, red, or pink when alive; unilaminar, but often with scattered frontally budded dwarf zooids. Autozooids ( Figure 17A rising medially to a pointed suboral umbo, sometimes incompletely fused. Secondary orifice oval, 0.09-0.13 mm long (0.10¡0.01 mm), 0.15-0.20 mm wide (0.17¡0.01 mm), flanked by a pair of hollow, erect spines distolaterally, with a shorter, tapering hollow spine in midline, or sometimes two, occasionally fused. Ovicellate zooids absent in primary layer. Frontally budded dwarf ovicellate zooids ( Figure 17A, C), 0.22-0.33 mm long (0.29¡0.02 mm), 0.19-0.25 mm wide (0.22¡0.02 mm), occupy central region of colony, orientated in any direction; frontal wall reduced, consisting of three or four fused costae, including the pair of heavier subapertural costae comprising proximal lip of longitudinally compressed orifice and an additional one or two others, with a few intercostal pores between them. Ovicell of dwarf zooids derived from two pairs of spines, one pair broadened and thickened to form proximal border of ovicell, the other fused to form cap-like distal wall of ovicell; ovicells of dwarf zooids perforated on top with one to four pores that sometimes merge with one another; pseudopores at tips of thick proximal ovicellar costae appear as additional small perforations. Avicularia lacking. Zooids with two distal and two distolateral basal pore chambers. Ancestrula ( Figure 17D) identical in form to autozooid, 0.35 mm long, 0.23 mm wide, with five orificial spines. Ancestrula buds triplet of zooids distally.

Remarks
In the material examined, ovicells were found only on dwarf zooids, usually concentrated densely in the centre of the colony. Dick et al. (2005) noted that differences in the fusion of distal orificial spines, form of the ovicell, and ancestrular budding patterns exist in populations across the range of this putatively circumpolar species and that nominal C. annulata may comprise a complex of cryptic species. However, resolution of this issue will require monographic treatment and is beyond the scope of the present study.

Distribution
This is considered a circumpolar Arctic-Boreal species (Kluge 1962(Kluge , 1975Gontar and Denisenko 1989). In the eastern Pacific, it has previously been reported from Kodiak, Cordova, Yakutat, and Ketchikan in Alaska (Robertson 1900;Osburn 1950;Dick and Ross 1988;Dick et al. 2005), and from British Columbia (O'Donoghue andO'Donoghue 1923, 1926). On the Asian side, it has been reported from the Commander Islands, southeastern Kamchatka, the Kuril Islands, Sakhalin Island, and Primorye on Peter the Great Gulf (Kluge 1961;Grischenko 1997Kubanin 1997). In Japan it is known from Akkeshi, Muroran, and Shirikishinai on the Pacific coast of Hokkaido (Mawatari and Matawari 1981b), and from Oshoro Bay on the Sea of Japan side (Kubota and Mawatari 1985b).

Remarks
There are some small differences between our material and that from Kamekawa, Hokkaido, the type locality . In specimens from Akkeshi, zooids are slightly larger; the ooecial fenestra tends to be larger and more variably shaped; some zooids have a moderately developed frontal umbo; and the strap-like frontal carina can have a slit or fenestra at the proximal end. Our observations add to the range of variation known for this species.

Distribution
Hokkaido Island, Japan. This species was originally described from a tidal flat at Kamekawa (43u039N, 140u359E) on the Sea of Japan side of Hokkaido; the record at Akkeshi extends the known range to the Pacific side.

Description
Colony encrusting, coherent, unilaminar when young ( Figure 19A), multilaminar, often several layers thick, with age; forming irregularly circular patches up to 2.5 cm across, white to tan in colour when alive. Primary layer consists of autozooids only; male polymorphs, ovicellate female zooids, and autozooids budded frontally in secondary layers. Zooids elongate-elliptical, spindle-shaped, widest in middle, rounded distally, tapering proximally, 0.45-0.78 mm long (0.61¡0.09 mm), 0.20-0.30 mm wide (0.25¡0.03 mm), separated by a deep groove, with slit-like lacunae and incipient zooeciules evident between young, marginal zooids. Autozooids distinct only in primary layer ( Figure 19A) in marginal budding zone and in young colonies; proximal ends of zooids submerged under distal parts of preceding zooids. Frontal wall hemicylindrical, smooth, translucent, convex, rising distally into lunate suboral umbo. Autozooidal orifice ( Figure 19B, D, E), including sinus, roughly circular, longer than broad, 0.11-0.15 mm long (0.13¡0.01 mm), 0.10-0.13 mm wide (0.11¡0.01 mm), with condylar shelves bearing small condyles pointing distally, distomedially, or medially, often with a notch between condyle and orificial rim; between condylar shelves is a deep, broadly U-shaped proximal sinus. Orifice surrounded laterally and distally by a sharp, thin, raised peristomial rim. Orifice of male zooids ( Figure 19C-E) similar in shape to that of autozooidal orifice, but about one-half to one-third the length and width; proximal sinus sometimes appears proportionately longer and narrower in male zooids. Orifice of female zooids ( Figure 19C, D) semicircular, broad, with a concave proximal margin. Ovicell ( Figure 19C-E) hyperstomial, spherical, smooth, variable in shape, size and orientation, 0.17-0.23 mm long (0.20¡0.01 mm), 0.18-0.25 mm wide (0.22¡0.02 mm), covered with a variable number of pores, some occluded. Spines and avicularia lacking. Six pore chambers along basal side of lateral wall and two to three in distal vertical wall. Ancestrula ( Figure 19F) similar in form to autozooid, though smaller and shorter. Early astogeny a spiral budding pattern, with first zooid budding distolaterally from ancestrula and each following periancestrular bud arising from angle between ancestrula and preceding zooid.

Remarks
Celleporella hyalina (L., 1767) has been considered a cosmopolitan species (Osburn 1952;Kluge 1962), distributed around the world from the Arctic to tropical latitudes. Recent studies, however, have suggested that nominal C. hyalina likely involves a worldwide complex of similar, perhaps cryptic species (see discussion by Dick et al. 2005, p 3726). Until further data are available from morphological, molecular, and reproductive compatibility studies, we simply refer the material from Akkeshi to the Celleporella hyalina (L., 1767) species complex. Variation among some of our specimens (e.g. compare number and distribution of ovicellar pores between Figure 19D and 19E) may be indicative of more than one representative of this species complex at Akkeshhi.

Description
Colony encrusting, unilaminar, coherent, more or less circular, up to 3 cm across, bright yellow, orange, or tan when alive. Zooids ( Figure 20A, B) hexagonal, rectangular, or oval, often tapering proximally, 0.47-0.77 mm long (0.58¡0.08 mm), 0.20-0.32 mm wide (0.25¡0.03 mm), separated by a groove, with a faint suture line between adjacent vertical walls. Frontal wall convex, coarsely granulated, imperforate centrally, with 5-11 conspicuous areolar pores along each lateral margin, separated by short ridges. Primary orifice ( Figure 20C) semicircular, broader than long, 0.09-0.12 mm long (0.11¡0.01 mm), 0.11-0.15 mm wide (0.12¡0.01 mm); proximal margin varying from slightly concave to straight or slightly convex, with an inconspicuous, low, flattened condyle near each proximal corner. Peristome ( Figure 20B, C) cormidial, consisting distally of upturned proximal margin of succeeding zooid; this is confluent with sharp, raised lateral flanges that meet avicularian chamber proximally; in ovicellate zooids lateral flanges are confluent with upturned proximal margin of ovicell. Secondary orifice semicircular, or approaching quadrate in non-ovicellate zooids, often markedly quadrate in ovicellate zooids. A median suboral avicularium ( Figure 20C) lies on internal slope of peristomial rim, orientated almost vertically, mandible triangular with rounded apex or almost semicircular, cross-bar complete; avicularian chamber ( Figure 20D, E) raised from frontal wall, broad, roughly granulated, often umbonate, occupying up to one-half of frontal surface, developing from an areolar pore on each side. Ovicell ( Figure 20D, E) hyperstomial, hemispherical, prominent, 0.19-0.24 mm long (0.21¡0.02 mm), 0.20-0.29 mm wide (0.24¡0.02 mm), imperforate, surface coarsely granulated like frontal wall, sometimes bearing a small, salient central umbo. Interzooidal communication via two multiporous septula in basal half of distal wall of zooid and a single multiporous septulum in each distolateral wall. Ancestrula ( Figure 20F) of uncertain shape, obscured by surrounding zooids, with semicircular orifice bearing eight short, hollow spines along lateral margins; surrounded by triplet of small zooids distally and distolaterally and two larger zooids proximally; periancestrular zooids similar to later autozooids, but have one to three hollow ephemeral spines along distal margin of orifice. Dick et al. (2005) discussed the taxonomic status of P. acutirostris in some detail, concluding that what has been considered a circumpolar, Arctic-Boreal species may comprise a complex of closely related species in the northern hemisphere, including P. major Hincks, 1884 andP. columbiana O'Donoghue and. Until this species complex is better understood, we simply refer the specimens from Akkeshi to P. acutirostris Smitt, 1868. Compared to zooids of specimens at Ketchikan, Alaska, zooids at Akkeshi are somewhat longer; the frontal wall and ovicell more coarsely granulated and appear more heavily calcified; the low but conspicuous lyrula seen in the Ketchikan population is absent; and both the ovicell and frontal wall tend to be umbonate, which is not the case at Ketchikan. However, the differences in morphology between Alaskan populations (Dick and Ross 1988;Dick et al. 2005) might represent geographical variation among conspecific populations.

Distribution
Smitt (1868) originally described P. acutirostris from Spitzbergen. Subsequently, it has come to be considered a circumpolar, Arctic-Boreal species (see Kluge 1962and Gontar and Denisenko 1989 for many distributional records), extending as far south as Cape Cod in the northwestern Atlantic (Osburn 1912), the Lofoten Islands in the northeastern Atlantic (Nordgaard 1918), and southern California in the northeastern Pacific (Osburn 1952). In the northwestern Pacific, P. acutirostris has been recorded from the southern Chukchi Peninsula, Commander Islands, eastern Kamchatka, Shantar archipelago, Sakhalin Island, Kuril Islands, and Primorye (Kluge 1961;Grischenko 1997;Kubanin 1997); in Japan it is previously known from Akkeshi and Hakodate, Hokkaido, southward to middle Honshu (Mawatari and Mawatari 1981b).

Description
Colony encrusting, unilaminar, coherent, tightly attached to substratum, irregularly circular, up to 2.5 cm across; yellowish, greyish, or off-white when alive. Zooids ( Figure 21B-E) hexagonal or rectangular, sometimes tapering proximally, 0.35-0.60 mm long (0.48¡0.07 mm), 0.23-0.38 mm wide (0.31¡0.04 mm), delineated by a deep groove with a suture line when young, and by deep, undulating lateral groove with age; transverse boundaries indistinct. Frontal wall imperforate centrally, coarsely granulated, with four to seven circular areolar pores along each lateral margin; inflated and convex in young zooids; increasingly thick with age, quite convex, roughly granulose, sometimes with a prominent umbo in suboral or proximolateral region. Primary orifice semicircular, deeply submerged, difficult to measure, but about 0.10-0.13 mm long by 0.12-0.15 mm wide; lyrula low and broad, occasionally lacking in some zooids, which have a slightly convex proximal border; condyles low, tapering distally, located near proximal corners of orifice. Peristome deep, secondary orifice pyriform in outline, 0.11-0.16 mm long (0.14¡0.02 mm), 0.12-0.16 mm wide (0.15¡0.01 mm), cormidial, its distal half formed by indented proximal margin of succeeding zooid; this is confluent with thick, rounded lateral flanges that meet proximally with sides of suboral avicularian chamber; in ovicellate zooids lateral flanges connect with proximolateral corners of ovicell; suture lines separate contributions of secondary calcification from a zooid and its neighbour that make up the secondary orifice. A large, oval median suboral avicularium lies within peristome, below secondary orifice, orientated vertically or tilted slightly proximally; mandible semicircular, cross-bar complete; avicularian chamber broader than long, lunate, occupying frontal surface from margin to margin, with two to five small pores around chamber margin or in chamber itself; in young zooids chamber is convex, inflated, finely granulated, but with age it becomes immersed and rugose. Ovicell ( Figure 21C-E) hyperstomial, spherical, 0.16-0.21 mm long (0.19¡0.01 mm), 0.20-0.25 mm wide (0.23¡0.02 mm), initially finely granulated, rapidly immersed with thick, roughly granulose calcification from surrounding zooids, becoming flush with colony surface. Oral spines lacking. Two multiporous septula in transverse wall and four circular multiporous septula in each lateral wall. Ancestrular complex ( Figure 21F) comprises five zooids radiating from common centre; in the specimen illustrated, the ancestrula proper is completely obscured by periancestrular zooids.

Remarks
The structure of the frontal wall varies considerably in P. belli according to substratum. In general, the frontal wall is inflated in young zooids and becomes gradually mucronate with age. Colonies on flat surfaces tend to retain the inflated appearance of the frontal wall, without an umbo, whereas those on irregular substrata may comprise entirely heavily calcified zooids with a strongly mucronate frontal wall.

Distribution
This is a circumpolar, Arctic-Boreal, species. Kluge (1962Kluge ( , 1975 and Gontar and Denisenko (1989) gave many distributional records for the Arctic. In the northern Pacific, P. belli has been reported from the Commander Islands (Grischenko 1997;Kubanin 1997), eastern Kamchatka (Kluge 1961), the Kuril Islands , Sakhalin Island, and Primorye . Akkeshi Bay is the southernmost known locality of P. belli in the Asian Pacific.

Remarks
Gontar (1982) described a new subspecies, D. sandalia concinna, on the basis of material from Kunahsir among the southern Kuril Islands. Characters distinguishing this subspecies from the nominal subspecies include strongly calcified ridges on the proximal part of the frontal shield; the occasional presence of a small additional avicularium in the same area; and smaller zooidal dimensions. Grischenko (1997) noted a similar condition of the frontal shield in specimens from the Commander Islands, but did not observe the additional frontal avicularia. Colonies of D. sandalia from Akkeshi Bay show considerable variation in the degree of calcification of the frontal shield and avicularian chamber, and in the development of buttresses. Whereas in some colonies zooids have an inflated, relatively smooth frontal shield and reduced, shortened buttresses, in others zooids have a highly mucronate frontal, reinforced by strong buttresses. A parallel variation in characters also occurs during individual zooidal development. Except for the sporadic presence of an additional frontal avicularium, the characters considered diagnostic for D. sandalia concinna appear to comprise astogenetic, ecophenotypic, or intra-population variation. It thus remains unclear whether D. sandalia concinna represents a valid subspecies.

Distribution
This is a Boreal Pacific species, originally described from Yakutat, Alaska (Robertson 1900) and subsequently reported from the Queen Charlotte archipelago (Osburn 1950), Kodiak Island (Dick and Ross 1988), the Commander Islands (Gordon and Grischenko 1994;Grischenko 1997Kubanin 1997), and Kunashir and Shikotan among the southern Kuril Islands . Our record from Akkeshi is the first report of D. sandalia from Japan and represents the southernmost limit of its known distribution in the western North Pacific.

Description
Colony encrusting, coherent, unilaminar, but sporadically building up frontally budded layer of irregularly orientated zooids; irregularly circular, largest observed 3.8 cm in maximum dimension; bright yellow when alive, with lemon-yellow membranous growing edge one or two zooids deep. Zooids ( Figure 23B, C, E) oval to irregularly hexagonal, rounded distally, 0.40-0.63 mm long (0.49¡0.06 mm), 0.27-0.43 mm wide (0.34¡0.04 mm), separated by shallow groove, with appressed adjacent vertical walls forming a thick line of calcification flanked by rows of areolar pores. Frontal wall convex, vitreous; smooth, or rugose with coarse granulation; imperforate centrally, with 7-10 conspicuous areolar pores along each lateral margin, separated by short buttresses; with large conical or nodular suboral umbo variable in size, occasionally with one to three additional protuberances scattered elsewhere on frontal surface. Primary orifice ( Figure 23B) subcircular, typically slightly longer than broad, 0.10-0.14 mm long (0.12¡0.01 mm), 0.10-0.13 mm wide (0.11¡0.01 mm), with a low, narrow, truncate lyrula and long, pointed condyles directed proximomedially. Young zooids ( Figure 23A) have two short, ephemeral distal spines. Peristome formed by narrow, raised lip proximally and laterally in young zooids, or rarely by two lateral lappets separated by sinus; primary orifice becomes sunken with increased secondary calcification. Zooids with or without a large avicularium ( Figure 23A-C), about 0.12-0.15 mm long, lateral to orifice, abutting peristome, the rostrum raised distally, with an acute, slightly long-triangular mandible directed medially, approximately in line with proximal margin of orifice; cross-bar thin, complete; sometimes the lateral-oral avicularium is paired. The lateral-oral avicularium may be replaced by a larger avicularium ( Figure 23D, E) of similar shape occupying central or proximal area of frontal wall and pointing distally, distolaterally, or sometimes proximally. Small avicularia ( Figure 23D, F), about 0.07-0.10 mm long, pyriform or oval in shape, with semicircular mandible directed laterally or proximally, can also occur anywhere along the proximal or lateral margins proximal to the orifice, with or without the larger, acute types; sometimes an oval avicularium overlaps the margin of an ovicell. Ovicell ( Figure 23F, G) spherical, broader than long, 0.20-0.28 mm long (0.24¡0.02 mm), 0.23-0.33 mm wide (0.28¡0.02 mm), overhanging the orifice; smooth, flattened on top and bearing a single large, circular or transversely elliptical pore; recumbent and sunken in frontal wall of distal zooid, with contributions of ectocystal calcification from that zooid and laterally flanking zooids delineated by raised suture lines; ovicell lacking ornamentation, or with one to three conical, tuberculate processes, one per sector of secondary calcification. Interzooidal communication via uniporous septula. Ancestrula not observed; ancestrular region we observed ( Figure 23H) appears entirely covered by a frontally budded layer of irregularly orientated zooids.

Remarks
The stable character that distinguishes this species from any other Parasmittina reported from the northwestern Pacific, including P. jeffreysii (Norman, 1903), P. trispinosa (Johnston, 1838), and P. macroavicularia , is the presence of only a single, large pore in the ovicell. Gontar (1982) described and illustrated two types of avicularia in her original description, but only briefly mentioned their arrangement, which we found to be quite variable. Most zooids located marginally or peripherally lack avicularia or have only the lateral-oral avicularium with a triangular mandible pointed medially. In contrast, zooids situated near the colony centre tend to have several different combinations of avicularia, such as: (1) a triangular lateral-oral avicularium on both sides of orifice, mandibles pointing medially; (2) an oval avicularium on each side of the orifice; (3) both triangular and oval avicularia lateral to orifice; (4) a single triangular avicularium located proximally on the frontal wall, with the mandible directed distally, laterally, or proximally; (5) a single oval avicularium lateral to orifice; (6) one or two oval avicularia in the proximal half of the frontal wall; or (7) one oval avicularium lateral to the orifice and another more proximally.

Description
Colony encrusting, unilaminar, coherent, forming circular patches up to 3 cm across, reddish to orange when alive. Zooids ( Figure 24A, B) hexagonal, oval, or irregularly rectangular, mostly with indistinct meandering boundaries, 0.43-0.78 mm long (0.59¡0.07 mm), 0.33-0.45 mm wide (0.38¡0.03 mm), separated by shallow undulating groove with suture line at bottom. Frontal wall slightly to moderately convex, finely granulated, covered with numerous large pores; with age, pores became infundibular. Frontal wall rises into a small or prominent conical median umbo ( Figure 24B, D) in suboral or central region. Primary orifice ( Figure 24C) round or slightly longer than broad, with wide, variably tall median lyrula; small, sharp lateral condyles evident in some marginal zooids, but usually appear to be lacking ( Figure 24A, C). Oral spines lacking outside zone of astogenetic change. Secondary orifice irregularly circular, oval, or pyriform in outline, longer than wide, often constricted proximally; surrounded by thin, smooth, angled rim, 0.14-0.19 mm long (0.16¡0.01 mm), 0.13-0.18 mm wide (0.15¡0.01 mm); cormidial ( Figure 24D), formed distally by proximal margin of succeeding zooid and laterally by extensions of secondary calcification from adjacent zooid on each side. Median suboral avicularium ( Figure 24D, E) oval or pyriform in outline, with short-spatulate mandible and complete cross-bar; rising above proximal orificial denticle and lying mostly within peristome, tilted proximally from perpendicular. Avicularian chamber small, semicircular, completely immersed in suboral umbo, often flanked laterally by a minute pore on each side. Ovicell ( Figure 24C-E) hyperstomial, spherical, imperforate, broader than long, 0.14-0.20 mm long (0.17¡0.01 mm), 0.18-0.23 mm wide (0.20¡0.01 mm), rapidly immersed, the top flush with colony surface, finely granulated; outer layer cormidial like secondary orifice, with calcification from the succeeding and lateral zooids, the contributions delineated by sutures on surface. Lateral wall of zooids with two distal and six lateral basal pore chambers. Ancestrula ( Figure 24F) modified tatiform, about 0.42 mm long by 0.33 mm wide; rapidly obscured by periancestrular zooids; opesia a little less than half the length of ancestrula, with an undetermined number of opesial spines. Ancestrula buds a small zooid distally and a pair distolaterally; from the latter arise two proximal periancestral zooids. Zooids in zone of astogenetic change with two to four ephemeral oral spines. Soule et al. (1995) established the genus Raymondcia for species resembling Smittina, but with the secondary orifice and ovicell cormidial, composed of distal and two lateral segments, characters present in our material. Accordingly, we use the new combination R. rigida (Lorenz, 1886) herein. Raymondcia rigida is similar to R. macginitiei Soule et al., 1995 in having a pyriform secondary orifice and the avicularium orientated nearly vertically within the peristiome; however, in the latter species, the orificial denticle is wider than the width of the suboral avicularium, whereas in R. rigida it is narrow, about the same width as the avicularium. Although the frontal wall is inflated in R. macginitiei, it is distinctly umbonate in R. rigida, which also has considerably larger frontal pores. Finally, the most deeply immersed ovicells of R. rigida are still a little convex and recognizable, whereas those of R. macginitiei became totally immersed.

Distribution
This has been considered a circumpolar Arctic-Boreal species; Kluge (1962Kluge ( , 1975 and Gontar and Denisenko (1989) provided many distributional records. Most of these records need to be re-examined (see Dick and Ross 1988), due to possible confusion with R. bella (Busk, 1860), originally described from Shetland, Britain (Hayward and Ryland 1999). On the Asiatic side, the nominal species has been recorded from eastern Kamchatka, the Commander Islands, the Shantar Archipelago, the Kuril Islands, southern Sakhalin Island, Primorye, and Peter the Great Gulf Kluge 1961;Grischenko 1997). There is a record of nominal R. rigida from Muroran, Hokkaido, Japan (Mawatari and Mawatari 1981b).

Description
Colony encrusting, unilaminar, coherent, forming circular patches up to 4 cm across; orange, dark red or light violet when alive. Zooids ( Figure 25B) hexagonal, oval, or pyriform, 0.47-0.65 mm long (0.57¡0.05 mm), 0.30-0.45 mm wide (0.38¡0.04 mm), usually with very irregular boundaries, separated by undulating suture line. Frontal wall slightly convex to inflated, finely granulated, with five or six areolar pores along each lateral margin and additional large, infundibuliform pores in the central region. With age, outer pore openings became enlarged and frontal wall appears reticulate, with rounded ridges between pores. In some colonies, frontal wall rises to a prominent, conical median suboral umbo ( Figure 25C). Orifice circular to irregularly oval, 0.13-0.17 mm long (0.15¡ 0.02 mm), 0.14-0.19 mm wide (0.17¡0.01 mm), with a thin marginal rim; condyles and lyrula lacking. Secondary orifice cormidial, usually comprising four sectors, including contributions of frontal wall from the distal and lateral zooids, with distinct sutures between sectors. With increasing secondary calcification, the low peristome becomes irregular. A small, oval median suboral avicularium ( Figure 25B, D, E) abuts the primary orifice; with increasing secondary calcification it comes to lie within peristome, orientated perpendicularly, or nearly so, to colony surface, hidden from frontal view; mandible semicircular, with complete cross-bar. Avicularian chamber small, crescentic, completely immersed with age; avicularium occasionally lacking. Spines and ovicells lacking. Zooids with two distal and four lateral basal pore chambers. Ancestrula not observed; obscured by ancestrular complex ( Figure 25F) of heavily calcified zooids smaller than astogenetically mature zooids, often overgrown by layer of irregularly orientated, frontally budded zooids.

Remarks
The taxonomic position of this species is unclear. It does not belong in Porella, as originally placed, because that genus is characterized by having an ovicell and an umbonuloid frontal shield with marginal areolae only, although spines and lyrula may be lacking and the condyles greatly reduced (Hayward and Ryland 1999). The cormidial secondary orifice, with contributions from the lateral and distal zooids, is very similar to that of Raymondcia ; Raymondcia also has a median suboral avicularium. However, characters of Raymondcia as the genus is now defined include lyrula, condyles, and ovicell. Although the definitions of some genera (e.g. Porella) encompass variation in the presence or degree of development of a lyrula and condyles, inferred loss of the ovicell is problematic; we are not aware of any lepraliomorph genus in which some species have ovicells and others brood internally. Nonetheless, the overall resemblance of Raymondcia klugei (Gontar) to Raymondcia rigida (Lorenz) is remarkable. Since we were unable to observe embryos internally in any colonies of R. klugei, it is unknown whether this species really lacks ovicells and broods internally. As an alternative explanation for the absence of ovicells in our specimens, it might be that conditions in the intertidal zone of Akkeshi Bay are suitable for growth, but not for reproduction. Such a situation would occur, for example, if colonies reproduced only in relatively deep water, but recruits could survive intertidally.

Distribution
This species was originally described from Ivanovskogo Cape, Kunashir Island, southern Kuril Islands. Akkeshi Bay is the second known locality.

Remarks
Gordon and d 'Hondt (1997) established the genus Parkermavella for Schizomavella-like species that differ from Schizomavella in having an imperforate frontal shield and only marginal areolae. Characters of Parkermavella include a proximal oral sinus; articulated oral spines distally; one or more adventitious avicularia near the orifice or elsewhere on the frontal surface; and a prominent or subimmersed ovicell with smooth ectooecial calcification, many perforations that may be rimmed, and secondary calcification sometimes encroaching around the distal margin. The species of Androsova and Gontar treated here lacks frontal perforation and has most of the other characters of Parkermavella, and therefore belongs in that genus. Gontar (1982) originally described this species as subspecies orientalis of Schizomavella auriculata (Hassall, 1842). However, the nominal subspecies never has more than a single median avicularium associated with the orifice, and also differs in orifice shape and in having numerous frontal pores; it is distributed in the northeastern Atlantic from Scotland to Gibraltar (Hayward and Ryland 1999). On the basis of these differences in morphology and range, we here elevate Gontar's subspecies to species rank as P. orientalis (Androsova and Gontar, 1982).
Parkermavella orientalis is very similar to S. triavicularia Soule, Soule, and Chaney, 1995, described from the Santa Barbara Channel, which likewise has a single median suboral avicularium in non-ovicellate zooids and an additional pair of lateral oral avicularia in ovicellate zooids. Parkermavella orientalis differs from the latter in several characters: (1) it lacks frontal pores, with well-developed areolar pores instead; (2) developing zooids near the colony margin have one or two ephemeral distal oral spines, whereas S. triavicularia has three spines; (3) the median suboral avicularium is closer to the suboral sinus than in S. triavicularia; and (4) ovicellate zooids have dimorphic suboral avicularia, with the lateral avicularia larger than the median one, and with more elongate mandibles; the median and lateral avicularia are similar in size and form in mature zooids of S. triavicularia.
The remarkable overall similarity of S. triavicularia and P. orientalis dispels misgivings one might have in accepting genera with and genera without a perforate frontal shield in the same family. Schizomavella triavicularia, which is evenly perforated, is otherwise so similar to S. orientalis, which is not, that there is little doubt the two are closely related; the similarity extends to both species having a pair of small pores flanking the avicularian chamber. Although convergence is a possibility, we conclude that the secondary loss of frontal perforation in Parkermavella is a more likely explanation. The small pores that flank the avicularian chamber are actually primary perforations in the frontal shield, as indicated by their presence in the forming shield in marginal zooids ( Figure 26A), and in this sense, the frontal pores can be viewed as having been severely reduced in number, rather than lost entirely.

Description
Colony encrusting, unilaminar, occasionally with bilaminar overgrowth of one portion of colony by another, forming extensive irregularly circular patches up to 5.5 cm across, red to bright orange when alive. Zooids ( Figure 27A, B) rectangular to hexagonal, 0.52-0.73 mm long (0.63¡0.05 mm), 0.30-0.43 mm wide (0.38¡0.03 mm), separated by shallow grooves. Frontal wall weakly to moderately convex, uniformly perforated from margin to margin with small circular pores except in suboral area, with seven to nine larger areolar pores along each lateral margin. Frontal pores become infundibular with development of calcification in mature zooids. Usually frontal wall rises into a small conical median umbo proximal to orifice. Orifice ( Figure 27B, E) broader than long, 0.11-0.15 mm long (0.13¡0.01 mm), 0.15-0.17 mm wide (0.16¡0.01 mm); sinus broadly U-shaped, flattened on bottom; conspicuous condylar shelves bearing blunt condyles. Oral avicularia ( Figure 27B) situated lateral or proximolateral to orifice, close to condylar shelf; mandible elongate-triangular, with acute tip, directed distolaterally to distally, cross-bar complete; chamber comparatively small, narrow, smooth, with one to three minute pores laterally around base. Oral avicularia usually single ( Figure 27B), often absent ( Figure 27C), occasionally paired. Additionally, some colonies have zooids with large frontal avicularia ( Figure 27D, E) similar in form to oral avicularia, but with a highly raised, smooth chamber. Position of frontal avicularia variable; they can lie close to orifice, but a little more proximal than oral avicularium, on opposite side; just proximal to oral avicularium on same side; along zooidal lateral margin; or centrally on frontal surface. Large avicularia close to orifice point distolaterally; those in central or lateral region of frontal wall point distally or laterally. Large frontal avicularia developed predominantly in mature zones of colony, among zooids with complete ovicells. Ovicell ( Figure 27C, D) hemispherical, prominent, 0.30-0.35 mm long (0.32¡0.02 mm), 0.29-0.36 mm wide (0.33¡0.02 mm), lying on frontal wall of daughter zooid and partially overhanging zooidal orifice; rugose, with heavily calcified, finely granulated radiating ribs, evenly perforated by numerous small pores, with larger round to slit-like pores around base. Occasionally ovicell has a small, knob-like central umbo. Ovicellate zooids can be sparsely distributed among non-fertile zooids or concentrated as a reproductive band within colony. Zooids intercommunicate via three to five distal and six lateral basal pore chambers. Ancestrula ( Figure 27F) oval, imperforate, 0.33 mm long, 0.28 mm wide, with eight spines around D-shaped orifice (0.13 mm long, 0.15 mm wide) with straight proximal margin. Ancestrula buds three small zooids distally; surrounded by seven zooids.

Distribution
In the western Pacific, S. japonica (reported as S. unicornis) extends from China (Liu et al. 2001) northward to Hokkaido Island, where it has previously been recorded at Akkeshi, Muroran, and Shirikishinai on the Pacific side (Mawatari and Mawatari 1981b) and at Oshoro Bay on the Sea of Japan (Kubota and Mawatari 1985b). Schizoporella japonica was introduced on Pacific oysters (Crassostrea gigas) from Japan to the Pacific coast of North America during the 20th century; it is now widely distributed from San Francisco, California to southeastern Alaska (Powell 1970;McCain and Ross 1974;Ross and McCain 1976;Dick et al. 2005). As mentioned by Dick et al. (2005), the actual range of S. japonica may be much more extensive, since this species could have been introduced on oysters to other parts of the world as well.

Diagnosis
Frontal wall moderately convex, finely granulated, with a small central umbo; covered uniformly with small pores when young, with large marginal pores when mature. Primary orifice semicircular, with straight or slightly convex proximal margin. Secondary orifice cormidial, with raised, complete peristome having a short mid-proximal sinus. Ovicell hyperstomial, broad, rapidly submerged with secondary calcification, granulose, without pores, with a small, central nodule.

Etymology
The species name refers to the frontal wall decorated with nodular umbones in mature zooids.
In the northern hemisphere, only one species of Stomachetosella, S. distincta Osburn, 1952, described from the Beaufort Sea near Point Barrow, Alaska, has a similarly umbonate ovicell and frontal wall. However, the distribution of frontal pores is different in S. distincta; its primary orifice has a sinus; and zooid size is significantly larger (0.65-0.85 mm long, 0.45-0.65 mm wide) and not overlapping with Stomachetosella decorata n. sp.
The form of the secondary orifice of S. decorata is almost identical to that of S. sinuosa (Busk, 1860), previously reported from Akkeshi Bay (Mawatari and Mawatari 1981b). The latter also has a sinuate, peristomial secondary orifice and immersed hyperstomial ovicells; however, the primary orifice is sinuate, rather than straight or slightly convex, and the ovicell has a central pore.

Remarks
This species has appeared in the northern literature under the names M. plana and M. crustacea. Osburn (1952) discussed the controversy surrounding the names, noting that some authors had preferred M. crustacea Smitt, 1868 because of the inadequacy of Dawson's (1859) original description of M. plana. Osburn argued convincingly that there is little doubt that M. plana represents the same species as M. crustacea, and that the former thus has priority.

Description
Colony encrusting, unilaminar, coherent, forming delicate crusts of very irregular form with meandering margins, up to about 3 cm in maximum dimension, light yellow when alive. Zooids ( Figure 30A, B, E) hexagonal, ovoid, or irregularly rectangular, widest in middle, 0.40-0.60 mm long (0.53¡0.06 mm), 0.25-0.43 mm wide (0.33¡0.05 mm), rounded distally, demarcated by a groove and fine suture line. All zooids of one type; non-fertile and fertile zooids similar in size and in size and form of orifice. Frontal wall convex, vitreous, texture smooth or granulated, uniformly perforated with circular pores, except for suboral area; with age, pores become infundibular and the frontal wall appears reticulate; proximal to orifice is a heterozooidal chamber ( Figure 30C) appearing as a raised, crescentic arch or umbo, with a median distal cavity covered by membrane. Primary orifice ( Figure 30C) roughly circular or broader than long, 0.10-0.14 mm long (0.12¡0.01 mm), 0.10-0.15 mm wide (0.13¡0.01 mm); with a high, semicircular anter separated from narrower, shorter poster by a pair of small condyles pointing proximomedially; proximal orificial margin straight. Peristome low; orifice flanked laterally by low, sharp, curved peristomial flanges extending distally from umbo, connecting with proximal corners of ovicell or raised proximal margin of distal zooid, sometimes present as distinct lateral lappets. Ovicell ( Figure 30D, E) globose, prominent, 0.19-0.25 mm long (0.22¡0.02 mm), 0.20-0.27 mm wide (0.23¡0.02 mm), recumbent on frontal wall of distal zooid and partially overhanging orifice; imperforate, with smooth or slightly nodular central area and short, finely tuberculate radiating ribs around periphery, with series of slit-like pores around distal margin. Spines and avicularia lacking. Zooids intercommunicate via multiporous septula. Ancestrula ( Figure 30F) similar to later zooids but elongate-oval, 0.45 mm long, 0.23 mm wide, inflated frontal wall finely granulated and perforated with numerous pores; ancestrular orifice roughly circular, 0.10 mm long, 0.09 mm wide, with reduced peristomial rim and small suboral opening; ancestrula buds quartet of smaller zooids, two distolaterally and two laterally; surrounded by seven zooids.

Remarks
Pacificincola perforata (Okada and Mawatari, 1937) was originally described from Onagawa Bay on the Pacific side of northern Honshu. It was subsequently reported (as Hippodiplosia insculpta) from the Peter the Great Gulf (Kubanin 1975) and (as Hippoporina perforata) from the Hong Kong vicinity ). Liu and Liu (1999) used P. perforata from the coastal waters of China as the type species for the new genus Pacificincola and a new family, Pacificincolidae. Pacificincola perforata most resembles its congener P. insculpta (Hincks, 1882) in having a similar, crescentic proximal suboral arch with a central cavity covered by membrane, and prominent spherical ovicells with radiating ribs. However, P. perforata differs from the latter in the following characters: (1) non-fertile and fertile zooids of P. perforata are all of the same type, whereas they are dimorphic in P. insculpta; (2) zooids of P. perforata are mostly distinctly hexagonal, widest in the middle, whereas zooids of P. insculpta are widest at the orifice, and tapering or truncate; (3) the lateral peristomial lobes, peculiar to P. perforata, are absent in P. insculpta, where the secondary orifice is a continuous rim; (4) the condyles are minute in P. perforata, but larger, prominent and shelf-like, in P. insculpta; (5) ovicells of P. perforata have reduced peripheral ribbing, in contrast to the well-developed ribs on ovicells of P. insculpta; and (6) the ancestrula of P. perforata is similar to later autozooids and buds four daughter zooids, whereas that of P. insculpta is tatiform and buds three daughter zooids (see Nielsen 1981, p 108, Figures 14, 15); and (7) colonies of P. perforata are always encrusting and tightly attaching to the substrate, whereas those of P. insculpta are initially unilaminar but then rise in irregular, bilaminar lobes and frills.

Remarks
Despite the apparent absence of lateral oral avicularia in our material, other characters confirm the identification of Cheilopora sincera. These characters include the convex, uniformly perforated frontal wall; cormidial secondary orifice with a well-developed median suboral tooth; and narrow slits at the base of the ovicell proximolaterally. In the absence of avicularia and the rounded-quadrangular outline of the secondary orifice, especially in mature zooids, Akkeshi specimens resemble Cheilopora inermis (Busk, 1880), previously recorded on the Pacific side of Hokkaido near Shirikishinai (Mawatari and Mawatari 1981b). However, the frontal shield of C. inermis is always flat (see Kluge 1962, p 563, Figure 398), whereas that of C. sincera is convex, especially in ovicellate zooids; the proximal margin of the peristome is only gently convex in C. inermis, whereas in C. sincera the median suboral tooth is usually prominent. Kluge (1952) described a new variety, C. sincera orientalis, based on material collected from the Chukchi Sea near Bering Strait. Subsequently, he reported the same form from Avacha Inlet, eastern Kamchatka, elevated to the rank of species as C. orientalis (Kluge 1961). Kluge's primary justifications for elevating that taxon to species rank were the form of the suboral tooth (a wide, inflated process instead of a sharp tooth) and the occurrence of bilaminar colony form. However,  described and illustrated the form and size of the suboral process in C. sincera as a significantly variable character that can be sharp, blunt, straight, cone-like, long or short, wide or narrow, or sometimes forked or entirely absent. Furthermore, colonies of C. sincera from the Ust' Khayryuzovo area, western Kamchatka shelf, Sea of Okhotsk, can be quite variable in form (A. V. Grischenko, unpublished data). Some colonies encrusting the internal concave surface of dead shells of Chlamys sp. bivalves can be loosely attached with elevated margins, or even produce small, bilamellar frills and folds. Hence, C. orientalis falls within the range of variation of C. sincera, and we here consider it as a junior synonym of C. sincera.

Distribution
This is a circumpolar, Arctic-Boreal species; Kluge (1962Kluge ( , 1975 and Gontar and Denisenko (1989) provide detailed records for the Arctic. In the northwestern Pacific, C. sincera has been reported from eastern Kamchatka, the southern Kuril Islands, western and southern Sakhalin Island, the Shantar Archipelago (Kluge 1961), and Primorye ; Akkeshi Bay is the southernmost known locality on the Asian side. Osburn (1952) expressed the opinion that C. praelucida (Hincks), described from British Columbia, is likely synonymous with C. sincera (Smitt), but he hesitated to make a formal synonymy without examining type material. O'Donoghue and O'Donoghue (1926) also reported C. praelucida from British Columbia.

Remarks
This species and its convoluted synonymy were discussed by Dick and Ross (1988) and Dick et al. (2005). Cryptosula zavjalovensis is patchily abundant in Akkeshi Bay. It is one of the main components of the fouling community on the vertical cement surfaces of the pier of the Akkeshi MBS, where layers of zooids form a thick crust, with free spaces or slits between layers. These spaces provide a habitat for other benthic invertebrates, mainly spirorbid worms, but also small isopod and amphipod crustaceans and errant polychaetes. The external surface of living colonies of C. zavjalovensis remains largely free from encrusting epizoonts, perhaps due to allelopathic activity of the substance responsible for the unpleasant, pungent odour.

Distribution
A Boreal Pacific species, this is one of the most common and widespread bryozoans found intertidally around the north Pacific Rim, extending from the mid-intertidal zone to a depth about 40 m . On the North American side it has been reported from the eastern Aleutian Islands, Bering Sea, northern Gulf of Alaska, Kodiak Island, and Ketchikan, Alaska (O'Donoghue 1925;Dick and Ross 1988;Dick et al. 2005). On the Asian side there are records from numerous localities, including the Gulf of Anadyr, Kamchatka, Commander Islands, Sea of Okhotsk, Kuril Islands, Sakhalin Island, Primorye, and northern sector of the Sea of Japan Kluge 1961;Kubanin 1997;Grischenko 1997. In Japan, C. zavjalovensis has been previously reported from Shirikishinai, southern Hokkaido (Mawatari and Mawatari 1981b), and Mutsu Bay, northern Honshu (Okada 1929).

Description
Colony encrusting, unilaminar, coherent, more or less circular, up to 2 cm in diameter, pale yellow to whitish when alive. Zooids ( Figure 33A, B) rounded-hexagonal to oval, 0.50-0.73 mm long (0.58¡0.06 mm), 0.35-0.53 mm wide (0.40¡0.04 mm), separated by a deep groove between smooth, inward-sloping, lateral walls. Frontal wall moderately convex, thin, translucent, smooth and imperforate in central area proximal to ascopore; stellate pores in a single marginal row in proximal half of zooid, two or three marginal rows in distal half, and two rows between ascopore and orifice. Orifice ( Figure 33C) semicircular, curving slightly inward at proximolateral corners; broader than long, 0.10-0.12 mm long (0.11¡0.01 mm), 0.12-0.16 mm wide (0.14¡0.01 mm), with straight proximal margin. Non-ovicellate marginal zooids have three short, hollow spines ( Figure 33A-C) closely set around distal curvature of orifice; ovicellate zooids also have three spines, the lateral two are more widely separated than in zooids that will not produce an ovicell and remain close to proximal corners of ovicell. Ascopore separated from proximal border of orifice by a distance about equivalent to length of zooidal orifice; crescentic, with a denticulate edge, located on a small, oval, elevated prominence which sometimes coincides with highest point of frontal wall, but more often lies on its distal slope. Ovicell ( Figure 33D) hemispherical, prominent, conspicuous, 0.22-0.28 mm long (0.25¡0.01 mm), 0.24-0.30 mm wide (0.27¡0.02 mm), smooth, without ribs, imperforate except for large marginal pores inside raised border. No avicularia. Zooids communicate via two distal and two distolateral basal pore chambers. Ancestrula ( Figure 33E) tatiform, oval, 0.33 mm long, 0.25 mm wide, with smooth, narrow gymnocyst and large, oval opesia, 0.22 mm long, 0.16 mm wide, surrounded by 10 short, jointed hollow spines. The ancestrula becomes modified ( Figure 33F) by formation of a frontal shield and loss of spines proximal to level of orifice, thereby coming to resemble periancestrular zooids. Ancestrular orifice irregularly oval, 0.09 mm long, 0.11 mm wide, with slightly concave proximal margin. Frontal wall convex, smooth, with only one row of stellate pores along periphery; slit-like ascopore on small oval prominence at highest point of frontal wall. Ancestrula buds triplet of zooids distally, eventually surrounded by six zooids; periancestrular zooids with three to five oral spines.

Remarks
Specimens from Akkeshi Bay agree well with the original description and illustrations of F. orientalis (Liu et al. 2003). Many previous records of Fenestrulina from the northwestern Pacific (e.g. Androsova 1959;Kluge 1961;Mawatari and Mawatari 1981b;Kubanin 1997;Grischenko 1997) were nominal F. malusii (Audouin, 1826), which has been long considered a widely distributed, cosmopolitan species. However, Soule et al. (1995) determined that F. malusii does not occur in the eastern Pacific, at least, and that previous records of that species there actually comprised a complex of two genera and several species. According to Hayward and McKinney (2002), F. malusii occurs throughout the Mediterranean and northwards to the British Isles and western Norway. The date of original description of this species is ambiguous, and we were not able to resolve this ambiguity. In their monograph, Liu et al. (2001) included ''Fenestrulina orientalis Liu et Liu, 2001'' andlisted ''Liu &Liu, 2001: 13, pl. 5, figs. 3-5'' as the citation in the synonymy. However, Liu et al. (2001) did not list Liu and Liu (2001) among the monograph's references. Furthermore, in a separate paper whose title indicated ''seven new species'' of microporellids, Liu et al. (2003) reported ''Fenestrulina orientalis Liu sp. nov.'' in the portion of the paper written in Chinese, and ''Fenestrulina orientalis sp. nov.'' in the portion of the paper containing an English translation; this paper contained no reference to Liu and Liu (2001), which indeed we have not been able to locate. We thus consider the authorship and date of Fenestrulina orientalis to be Sun, 2003. Distribution Liu et al. (2003) originally described Fenestrulina orientalis from China, but other authors had previously reported it from the Yellow Sea (as Microporella malusii or Fenestrulina malusii) (Androsova 1959). Some records of F. malusii from Japan, e.g. from Akkeshi, Muroran, Shirikishinai, northern to middle Honshu (Mawatari and Mawatari 1981b) and Mutsu Bay (Okada 1929), may represent F. orientalis. distolaterally or laterally, cross-bar complete. Ascopore crescentic, with denticulate edge, located on elevated oval prominence, just distal to prominent conical umbo. Ovicell raised, hemispherical, imperforate centrally, often umbonate, with radial ribs and pores around base. Ancestrula tatiform with 12 spines, budding a pair of zooids distolaterally.

Etymology
Named in honour of Luella Taranto, who greatly helped with collecting during July 2004.
Microporella luellae is similar in many characters to M. neocribroides Dick and Ross, 1988, previously reported intertidally from Densin-Hama, Muroran, southern Hokakido. The two widely separated oral spines distinguish M. luellae and M. neocribroides from all other congeners reported around Hokkaido. However, in M. luellae, the ascopore is always crescentic, with a denticulate edge, and is located on an elevated oval prominence. Except for developing and heavily calcified zooids, the proximal side of this prominence is developed into a small, conical umbo. In M. neocribroides, the ascopore is transversely elliptical and covered with a cribriform plate with 10-20 round pores (Suwa and Mawatari 1998, p 899, Figure 2;Dick et al. 2005, p 3753, Figure 19D); only occasionally is there an umbonate process proximal to the ascopore. Another difference is that the ovicell is often umbonate in M. luellae, but rarely so in M. neocribroides.

Distribution
Microporella luellae is presently known only from Akkeshi Bay.

Remarks
Our material generally agrees with the original description, except that in contrast to Suwa and Mawatari (1998), we did not observe blunt condyles occasionally present at the proximolateral corners of the orifice. Also, ancestrulae of M. trigonellata from Akkeshi Bay are larger than indicated for the type material. Suwa and Mawatari (1998) recently described this species from Deshin-Hama, Muroran, on the Pacific coast of Hokkaido; Akkeshi Bay is the second known locality.

Diagnosis
Colony small, pisiform, up to 6 mm in diameter. Colony surface very irregular, with numerous knob-like lobes. Zooids erect, cylindrical, haphazardly orientated. Distal pores encircle an oval orifice with a V-shaped sinus. Orifice flanked by paired lateral avicularia on columnar chambers that are straight or curved inward, narrowing terminally, with a small, oval rostrum angled to plane of orifice, the semicircular mandible pointing laterally. Zooids with completed avicularia have a narrow, transverse, slit-like fold in proximal peristomial lip. Vicarious avicularia rare; with broadly spatulate mandible. Ovicell hyperstomial, globose, recumbent on neighbouring zooids; tabula small; circular, semicircular, or roughly triangular, with slit-like, radially arranged pores.

Etymology
The species name refers to the small size of colonies, which can be mature at a few millimetres in diameter.

Remarks
Celleporina minima n. sp. is similar to the Arctic-Boreal species C. nordenskjoldi in having a pisiform colony, columnar lateral avicularia, and globular, recumbent ovicells. However, colonies of C. minima are always smaller and can have ovicellate zooids at a diameter of just a few millimetres. Avicularian columns of C. minima tend to be curved inward, narrow terminally, with the rostrum angled to the orificial plane and the mandible directed laterally. The avicularian columns of C. nordenskjoldi are long and stout, do not narrow terminally, and are usually straight or only slightly curved inward; the mandible is directed distolaterally. Though the ovicells are quite similar in form in these two species, the tabula is much smaller in C. minima, its diameter one-third to one-half the width of the proximal orifical margin; the tabula in C. nordenskjoldi is usually nearly as wide as the proximal margin of the ovicell.

Distribution
Akkeshi Bay is the only known locality.

Distribution
Celleporina nordenskjoldi is distributed along the Eurasian and American sectors of the Arctic region and extends into the Boreal Pacific, but appears to be absent in the Canadian Arctic and northern Atlantic. It has been reported along the Eurasian Arctic in the Laptev, East-Siberian, Chukchi Seas (Kluge 1962(Kluge , 1975Gontar and Denisenko 1989), and from the Beaufort Sea near Point Barrow, Alaska (Osburn 1952;Morris 1979). In the northwestern Pacific, C. nordenskjoldi has been recorded from eastern Kamchatka, the Commander Islands, the Kuril Islands, the Shantar Archipelago, Sakhalin Island, Primorye, Peter the Great Gulf (Kluge 1961;Kubanin 1997;Grischenko 1997), and Akkeshi and Muroran along the Pacific coast of Hokkaido, Japan (Mawatari and Mawatari 1981b).

Description
Colony encrusting, discoid, domed; a typical colony measured 1.5 by 1.2 by 0.2 cm; with multilayered arrangement of zooids, one over the other; light orange when alive. Colony surface undulating, with many knob-like lobes. Zooids at colony margin ( Figure 38A) decumbent. As colony rises from substrate, zooids ( Figure 38B) become erect, cylindrical, 0.28-0.38 mm across (0.33¡0.03 mm), haphazardly orientated. All zooids have numerous spinous processes within the zooidal cavity ( Figure 38A, B). Frontal wall of decumbent zooids smooth, convex, with a few marginal pores and imperforate central area; in erect zooids, pores carried towards orifice by tubular extensions and form ring around peristome. Primary orifice ( Figure 38B) oval, 0.12-0.18 mm long (0.15¡0.01 mm), 0.11-0.16 mm wide (0.14¡0.01 mm), submerged with age by peristome, with flattened condyles on condylar shelves and deep, U-shaped proximal sinus. Orifice of non-ovicellate zooids encircled by narrow peristomial lip; peristome enclosed by thickened proximal orificial lip and columnar lateral orificial avicularia, one on each side lateral to orifice, rarely single or absent; avicularian column stout, curved inward, often submerged with age; rostrum circular, terminal, angled almost perpendicular to orificial plane, with a complete cross-bar; semicircular mandible directed in a lateral direction. Vicarious avicularia ( Figure 38D) scattered over colony surface, numerous in central colony region, 0.13-0.29 mm long, with spatulate mandible; cross-bar complete, rostral opesia large, sometimes with extensive palatal shelf and high rostral rim. Ovicell ( Figure 38C, D) globose, 0.24-0.33 mm long (0.29¡0.02 mm), 0.29-0.39 mm wide (0.34¡0.03 mm), initially prominent, subimmersed with age; tabula semicircular, crescentic, or triangular, extensive, occupying nearly entire frontal surface of ovicell, bordered with radially arranged slit-like pores; secondary calcification results in a distinct border surrounding tabula. Ancestrula and early astogeny not observed. Okada (1923) was apparently the first person to collect this species, but described his material from the Korea Strait as a new variety, irregulatum, of Myriozoum marionense Busk, 1884. Harmer (1957 recognized Okada's form as different from M. marionense, and on the basis of specimens from Japan described it as a new Celleporina. A key diagnostic character is the common occurrence in young zooids of several rows of pores distal to the orifice, which was well illustrated by Ikezawa and Mawatari (1993, Figure 2A, C) in their redescription of C. porosissima; these authors also described the early astogeny. The deepwater specimens described by Okada (1923) formed colonies with erect, cylindrical, anastomosing branches, whereas Harmer (1957) described colonies as ''small pleurilaminar crusts'', and Ikezawa and Mawatari (1993) described them as ''encrusting, discoidal, domed''. Although Harmer (1957) attributed this difference in growth form to his specimens possibly having been in an initial stage of growth, the difference raises the question whether Celleporinna porosissima is indeed synonymous with Okada's (1923) Myriozoum marionense var. irregulatum.

Distribution
Originally collected from depths of 90-200 m in the Korea Strait (if Harmer's 1957 synonymy is valid), this species has been reported from the vicinity of Shimoda, Shizuoka Prefecture as colonies attached to seaweeds (Okada 1934), and from Oshoro Bay, Hokkaido, as colonies on fronds of Laminaria religiosa (Ikezawa and Mawatari 1993). Akkeshi Bay is the northernmost known locality.

Description
Colony encrusting, unilaminar, coherent, more or less circular, largest observed 1.2 cm across, beige to orange in colour when alive. Zooids ( Figure 39A, B) hexagonal to oval, rounded distally, 0.35-0.55 mm long (0.45¡0.05 mm), 0.27-0.38 mm wide (0.33¡0.03 mm); newly budded zooids ( Figure 39A) demarcated by groove and fine suture line; boundaries indistinct between heavily calcified zooids. Frontal wall moderately to markedly convex, tessellate, dimpled, imperforate except for two or three areolar pores along each lateral margin. Primary orifice ( Figure 39A, C) elongate-semicircular, occupying about one-third length of frontal wall, 0.12-0.15 mm long (0.13¡0.01 mm), 0.10-0.14 mm wide (0.12¡0.01 mm), with straight to slightly concave proximal margin; condyles well developed as rounded-triangular projections located about one-third of distance from proximal to distal margin, separating semicircular anter from shorter, broader poster. Five or six long, hollow oral spines ( Figure 39A) with enlarged bases occupy distal curvature of primary orifice of marginal zooids; with age, three or four distal spines become covered by secondary calcification and only most proximal pair remains lateral to orifice, in mature zooids lying close to proximolateral corners of ovicell. Primary orifice evident only in newly budded zooids, becoming rapidly submerged and surrounded by shallow, sloping peristome. Secondary orifice oval to irregularly circular; cormidial, bounded proximally by avicularian chamber, laterally and distally by contributions of calcification from lateral and distal zooids. A relatively large median suboral avicularium ( Figure 39B, D) lies within peristome, just below secondary orifice, often orientated slightly obliquely to longitudinal axis and usually tilted proximally; rostrum oval, cross-bar complete, mandible semicircular, rostral margin proximal to hinge bar even or with small prominences. Avicularian chamber fusiform along long axis, evident in young zooids, completely submersed with increased calcification. Some zooids have hypertrophied suboral avicularium ( Figure 39C), about 0.19 mm long, comparable to orifice in size, with short-spatulate mandible tilted proximally, almost parallel with frontal plane. Oval frontal avicularia ( Figure 39F) with semicircular mandible sometimes occur in proximal half of strongly calcified zooids; oval adventitious avicularia 0.17-0.21 mm long, with semispatulate mandible ( Figure 39F), occupy older regions of colony. Ovicell ( Figure 39D, E) hyperstomial, globose, 0.17-0.21 mm long (0.19¡0.01 mm), 0.20-0.25 mm wide (0.22¡0.01 mm), proximal margin straight or nearly so; rapidly submerged and often covered with contributions of secondary calcification from distal and lateral zooids, indicated by fine suture lines; with a smooth proximal tabula having a minute median pore rhombic, or irregularly oval, rounded distally, 0.32-0.53 mm long (0.45¡0.06 mm), 0.27-0.45 mm wide (0.36¡0.04 mm), demarcated by shallow groove with fine suture line. Frontal wall moderately to markedly convex, translucent, smooth, imperforate except for three or four areolar pores along each lateral margin, rising distally to suboral umbo that varies from a small, low nodule to a tall, conical projection, sometimes rounded in mature zooids. Primary orifice ( Figure 40A) hat-shaped, slightly longer than broad, 0.10-0.14 mm long (0.12¡0.01 mm), 0.10-0.13 mm wide (0.11¡0.01 mm), with straight to slightly concave proximal margin; blunt condyles are swellings in internal rim around anter, separating long-semicircular anter from short, broad poster. Four hollow ephemeral oral spines ( Figure 40A), more proximal pair with enlarged bases, located along distal margin of orifice of marginal zooids. Primary orifice evident only in developing zooids near colony margin; with development of secondary calcification, it becomes submerged and surrounded by shallow, sloping peristome. Secondary orifice ( Figure 40B) similar in shape to primary orifice, long-semicircular in outline; cormidial, bounded proximally by avicularian chamber, distally and laterally with contributions of calcification from distal and lateral zooids. A circular suboral avicularium ( Figure 40B, C) lies within peristome, below secondary orifice, orientated perpendicularly to colony surface or tilted slightly proximally; mandible semicircular, cross-bar complete; avicularian chamber small, broader than long, smooth, inflated in young zooids and submersed by umbo in older zooids. One to five small, circular adventitious avicularia ( Figure 40E) with semicircular mandible occupy frontal surface of many zooids; these tend to occur on proximal half of frontal wall in immature zooids and lateral slopes of suboral umbo in ovicellate zooids; often interzooidal in position, surrounding secondary orifices. Ovicell ( Figure 40C-E) hyperstomial, globose, broad, smooth, 0.15-0.20 mm long (0.17¡0.01 mm), 0.16-0.23 mm wide (0.19¡0.01 mm), with a large, circular pore close to slightly concave proximal margin of ectooecium; rapidly submerged ( Figure 40D, E), with contributions of secondary calcification from frontal walls of distal and lateral zooids delineated by fine sutures. Zooids interconnect by multiporous septula. Ancestrula ( Figure 40F) 0.07 mm long, 0.10 mm wide, orifice semicircular with straight proximal margin, surrounded by 10 spines; ancestrula buds three zooids distally and distolaterally, with another three larger zooids proximolaterally and proximally; periancestrular zooids with suboral avicularia only, and with five or six spines around distal curvature of orifice.

Remarks
The hat-shaped primary orifice surrounded by four spines with strong bases, the vertically orientated suboral avicularium, and the presence of numerous adventitious avicularia characterize this species. Specimens from Akkeshi Bay agree well with the original description of H. multiavicularia, except for fewer and differently arranged adventitious avicularia, which occasionally cover the entire frontal surface and number up to 10 per zooid in the type material.  also did not mention a pore near the proximal margin of the ovicell; however, this is difficult to observe without SEM.
This species is close to H. fastigatoavicularis (Kluge, 1955), which is common subtidally in Akkeshi Bay (Mawatari and Mawatari 1981b) in having a similarly shaped secondary orifice, and in the presence of suboral and numerous adventitious avicularia. However, all avicularia, including the suboral avicularium, of H. fastigatoavicularis are very small and lie in the plane of the frontal wall. Hippoporella fastigatoavicularis also has more regularly hexagonal zooids, with small umbones flanking the orifice laterally, in addition to the conical suboral umbo.
Another congener, H. kurilensis (Gontar, 1979), has marginal zooids with very similar morphology. In contrast to four spines in H. multiavicularia, H. kurilensis has five or six oral spines, with a pair of them remaining in ovicellate zooids. The frontal wall in H. kurilensis is convex and tessellated; adventitious avicularia are normally lacking, only occasionally present in heavily calcified zooids from the central region of the colony; and the relatively large suboral avicularium is orientated slightly obliquely to the longitudinal axis and usually tilted proximally. Complete ovicells of H. kurilensis are often not entirely covered by secondary calcification.

Distribution
This species was originally described from coastal waters off southern Sakhalin Island in the northern part of the Sea of Japan; it has also been recorded from Primorye Kluge et al. 1959;Kluge 1961). Gontar ( , 1992 reported it from Paramushir, Shikotan, and Zelenyy among the Kuril Islands. Hippoporella multiavicularia also occurs in the shelf zone of the Commander Islands (A. V. Grischenko, unpublished data).

Description
Colony ( Figure 41A) erect, orange in colour when alive, consisting of fenestrate bilaminar sheets, complexly folded and fused into rigid three-dimensional meshwork, largest observed 116864.5 cm in dimensions, rising from rounded base attached to substratum by kenozooids. Branches range from three to 11 zooids wide between fenestrulae. All feeding zooids open on one surface of a sheet; opposite surface consists of kenozooids. Fenestrulae elongate-oval or irregularly rhombic in shape, 0.7-1.9 long by 0.3-0.9 mm wide. Zooids ( Figure 41C) hexagonal, rhombic, or oval, 0.48-0.70 mm long (0.59¡0.07 mm), 0.24-0.38 mm wide (0.31¡0.04 mm). Newly budded zooids delineated by raised vertical walls; in older regions of colony, boundaries between zooids occluded by secondary calcification or represented by fine sutures. Frontal wall slightly convex to inflated in young zooids, markedly convex in mature zooids, finely granulated, imperforate except for two or three pores along each proximolateral margin. Primary orifice ( Figure 41D) long-semicircular, with beaded rim, 0.10-0.13 mm long (0.11¡0.01 mm), 0.09-0.13 mm wide (0.11¡0.01 mm); proximal margin straight, with shallow but distinct U-shaped sinus flanked by condylar shelves bearing low, blunt condyles. With age, primary orifice submerged in peristome; secondary orifice with median pseudosinus. Two short, hollow, ephemeral oral spines with enlarged bases frequently present at distolateral edges of primary orifice in marginal zooids. Large frontal avicularium ( Figure 41C-E), 0.14-0.20 mm long, with raised, beak-shaped rostrum, occupies central part of frontal wall in many zooids; cross-bar complete, rostral opesia triangular to oval in shape; mandible longtriangular, with acute tip, directed proximally or proximolaterally; avicularian chamber broad, conical, very prominent in immature zooids, inflated in ovicellate zooids ( Figure 41D) by general thickening of frontal wall owing to secondary calcification, coarsely granulated, with two or three small pores flanking its base. Single transversely orientated avicularium occupies proximal abfrontal axils of fenestrulae ( Figure 41B), 0.15-0.18 mm long, with short, equilaterally triangular mandible; one or two pairs of circular pores flank rostrum. Ovicell ( Figure 41F) hyperstomial, spherical, smooth, often forming a hood overhanging orifice, 0.18-0.23 mm long (0.20¡0.01 mm), 0.20-0.25 mm wide (0.22¡0.01mm), with incompletely calcified ectooecium having fine concentric striae; proximal edge of ovicell often with slight median denticle. Ovicell rapidly immersed ( Figure 41E) by secondary calcification from distal and lateral zooids covering most of its surface, with fine sutures delineating calcification from different zooids. Dorsal surface of colony ( Figure 41B) shows outlines of kenozooids of irregular form and size, recognizable by sutures indicating raised vertical walls; kenozooids inflated, roughly granulated, with a few sparse circular pores on the surface; avicularia lacking dorsally except for those in axils of fenestrulae. Ancestrula and early astogeny not observed.

Remarks
Specimens from Akkeshi Bay are similar in many characters to the type specimen of P. elongata (SMNH-1316, many fragments) ( Figure 41G, H). Both have rhombic zooids with a submerged, rounded primary orifice having a shallow median sinus and encircled by peristome bearing a proximal pseudosinus. Large frontal avicularia are similar in form and position; the ovicells have, on the proximal margin, a median denticle that is often incompletely closed. There are some differences; specimens from Akkeshi Bay have 3-11 zooids comprising branches between fenestrulae, the type of P. elongata three to seven. This may be ecophenotypic variation, as depth-dependent variation in branch width occurs in P. elongata from the Commander Islands region; colonies from depths of 1-25 m have twice the number of zooids per branch than those from depths of 65-165 m (A. V. Grischenko, unpublished data). Another is that the abfrontal avicularian mandibles are larger with a long-triangular mandible in the type of P. elongata, versus smaller, with an equilateral mandible in specimens from Akkeshi.
Several studies of intertidal bryzoans on suitable rocky shores (usually semi-protected, usually a habitat of layered boulders) in the North Pacific (Table VI) have indicated both ranges of single-site diversity and values of total local diversity. Estimates of diversity will increase with increasing sampling effort, and this is evident in Table VI: sampling effort was similar for Ketchikan and Hawaii, and these studies both found a cheilostome diversity of around 30 species. Likewise, sampling effort was similar for our study at Akkeshi and that at Kodiak; however, the cheilostome diversity measured at Akkeshi was only two-thirds that at Kodiak (39 and 57 species, respectively). Considering the similar, substantial sampling efforts in these two localities, the difference in diversity is likely not an artefact, and raises the question why the diversity at Akkeshi is lower.
In fact, we expected on biogeographic grounds that Akkeshi Bay would have a higher diversity than Kodiak. Akkeshi is situated just 2u of latitude south of the boundary separating the High and Low Boreal Zones in the western Pacific (Ekaterina Strait, Iturup Island, south Kuril Islands). The cold Oyashio current flows southward from the Bering Sea along Kamchatka and the Kuril Islands, reaching the southeastern coast of Hokkaido, including Akkeshi Bay. The warm Tsushima current, a branch of the Kuroshio current, weakly influences Akkeshi Bay from the southwest (Uchida et al. 1963). The transitional biogeographic position of Akkeshi Bay and also the influences of both cold and warm currents should be conducive for the dispersal there of species from the Arctic, Boreal, and Subtropical regions, and hence for enrichment of the fauna. Aside from this argument, one might expect higher bryozoan diversity at Akkeshi than at Kodiak due simply to the latitudinal gradient of increasing diversity from the poles to the tropics that is apparent for many animal phyla (Fisher 1960), including marine bryozoans (Dick and Ross 1988). Akkeshi (43uN) is considerably farther south than Kodiak (58uN).
We speculate that the unexpectedly low diversity of intertidal bryozoans between Akkeshi and Kodiak is due to differences in the environments of the two study areas and correlated ecological differences. Specifically, the difference in diversity may be due to differences in (1) the diversity of the bryozoan source community available for colonization of the intertidal, (2) the extent of the intertidal zone available for colonization by bryozoans, and (3) the diversity of habitats and microhabitats inhabited by bryozoans.  Dick and Ross (1988) are here condensed to 11 sites, because in two cases, two or three sampling stations were located at different intertidal heights at the same site. d Includes a rough estimate of the number of ctenostome and cyclostome species present, without identification to species. e Includes at least five ctenostome and five cyclostome species observed, but not identified to species.
Most intertidal bryozoan species are not restricted to the intertidal zone, but also occur subtidally and reach the upper limit of their depth range in the intertidal (Dick et al. 2005). Although intertidal bryozoan populations reproduce and can directly repopulate the intertidal zone, it might also be the case that species that are collected rarely intertidally represent relatively temporary incursions into the intertidal zone of species whose optimal habitat is subtidal. If this were the case, then the diversity of intertidal bryozoans at a site might be directly related to the species diversity on the adjacent subtidal shelf. We believe that the nearshore shelf areas at Akkeshi Bay and Kodiak may differ significantly in bryozoan diversity, although the subtidal bryozoan assemblages have not been adequately studied at these localities and confirmatory data are thus lacking.
The highest diversity of marine bryozoans occurs in the 0-100 m depth interval, that is, in nearshore shelf areas, although diversity is also high at 200 m (Ryland 1970;Gordon 1999). The width of the adjacent shelf zone is much narrower at Akkeshi than at Kodiak. At Akkeshi, the 200 m depth contour lies about 30 km from the mouth of the bay. At Kodiak, the 200 m depth contour lies about 100 km from the eastern entrance of Narrow Strait, where bryozoans were studied intertidally (Dick and Ross 1988).
Furthermore, few marine bryozoan species are capable of living on sandy or muddy bottoms; most species require a hard substratum such as rocks or shells on which to attach. The bottom of Akkeshi Bay and the adjacent shelf is mostly sandy and muddy; encrusting bryozoans can exist there only on scattered ''islands'' of substratum comprising isolated banks, scattered boulders, anthropogenic debris, and mollusc shells. At Kodiak, both the nearshore zone close to the intertidal study area of Dick and Ross (1988) and the vast adjacent shelf have mostly a rocky bottom. The Kodiak shelf, with a much greater area of substratum available to encrusting bryozoans, should thus also have much higher overall subtidal species diversity through a simple species-area relationship (MacArthur and Wilson 1967), and consequently more species available for incursion into the intertidal than Akkeshi Bay.
There is some evidence to support this speculation. At Akkeshi, three (8%) of the 39 species found intertidally were extremely rare, collected as one to three colonies per species. At Kodiak, with sampling effort similar to that at Akkeshi, 13 (23%) of the 57 species found intertidally were collected only as one to three colonies each (Dick and Ross 1988). This higher proportion of extremely rare species in the intertidal would be expected if subtidal diversity were much higher and if intertidal diversity were positively influenced by occasional incursions into the intertidal of some proportion of the overall complement of subtidal species.
Another difference between Kodiak and Akkeshi lies in the former locality having a more extensive intertidal zone available for bryozoans. The annual tidal range at Kodiak is 4.18 m, compared to 1.57 m at Akkeshi. This means that both the overall area of the intertidal zone and the areas of particular sub-zones, such as the infralittoral fringe, are considerably greater at Kodiak than at Akkeshi. This greater area should entail a greater number of species, again through a species-area relationship.
In addition to being less extensive, the intertidal zone at Akkeshi probably includes a lower diversity of habitats and possibly microhabitats than at Kodiak, which would also theoretically lower overall species diversity. Our cluster analysis showed the sampling sites forming three main species assemblages (outer Akkeshi Bay, inner Akkeshi Bay, and Akkeshi Lake) that were less than 70% similar to one another. The distribution of these assemblages from the outer to the inner bay suggests that some factor or combination of factors (e.g. degree of wave exposure, salinity, degree of sedimentation) varying from the outer to the inner bay influences the species composition of the assemblages. A cluster analysis of collecting sites at Kodiak (Dick and Ross 1988) showed a more complex pattern that suggested that degree of wave exposure, shore composition (reef flat versus layered boulders), and possibly proximity to the outlets of estuaries influenced species composition. The shoreline at Kodiak is less steep, and two of the sampling sites located in the infralittoral fringe comprised extensive rocky reef-flats exposed for tens of metres from the water's edge at low tide, with a complex structure of tidepools and overlying boulders. Extensive rocky reef flats of this type are not present in Akkeshi Bay.
Among the 39 cheilostome species found at Akkeshi, nine (23%) proved to be undescribed. This proportion of new species is not exceptionally high compared to other intensive studies conducted in local areas in the last two decades: 25% new among 59 species at Kodiak (Dick and Ross 1988); 35% new among 119 species in the eastern Pacific ; 26% new among 31 species at Ketchikan (Dick et al. 2005); 22% new among 92 species at Vanuatu (Tilbrook et al. 2001); 40% new among 178 species in the Solomon Islands (Tilbrook 2006); 31% new among 31 species on the Island of Hawaii (Dick et al. 2006). Interestingly, 80 years ago O' Donoghue andO'Donoghue (1923, 1926) found 26% new among 204 species they reported from the vicinity of Nanaimo, BC, so the rate of description of new species from intensive local studies in the Pacific does not appear to be reaching an asymptote. To put these numbers into context, we note that Hayward and McKinney (2002) reported only 5% new among the 106 species included in their study of the northern Adriatic; this low value reflects a greater cumulative taxonomic effort in the Mediterranean. The high proportion of new species in recent studies in the Pacific is attributable to generally low previous taxonomic effort throughout much of the region, an increase in taxonomic resolution due to the now-routine use of scanning electron microscopy, and a paradigm shift in how taxonomists view geographic variation Dick and Mawatari 2004).
In addition to changing our estimate of total bryozoan species diversity, intensive studies in local areas around the Pacific are also identifying speciose clades whose diversity had gone unrecognized in part due to the lumping of multiple, distinct species as geographical variants of putatively widely distributed species. A number of genera in the Pacific are proving much more speciose than was indicated a few decades ago; examples include Parasmittina (Soule and Soule 1973;Tilbrook 2006) and Rhynchozoon (Hayward 1988;Ryland and Hayward 1992;Tilbrook 2006). The genus Cauloramphus provides another example. As of the early 1980s, ten recent species of this genus had been reported from all of the Pacific Ocean, Sea of Japan, Bering Sea, Sea of Okhotsk, and Arctic Ocean (Osburn 1950;Kluge 1962Kluge , 1975Kubanin 1975;Mawatari and Mawatari 1981a). Subsequently, another six new species were reported from the Pacific, including two at Kodiak (Dick and Ross 1988), another two at Ketchikan (Dick et al., 2005), and one each from California and Korea Seo 2001). Our discovery of five species of Cauloramphus intertidally in the relatively small area of Akkeshi Bay, including three new species (C. cryptoarmatus, C. niger, and C. multispinosus) suggests that there are still many undescribed species of Cauloramphus in the North Pacific and that this genus may prove to be one of the most speciose among anascan genera in the region.