Corals in southern Boso Peninsula represent the northernmost occurrence of Holocene hermatypic corals in Japan. Jomon transgression corals occur partly in situ and partly as transported fragments. Living corals are found in the Tateyama Bay.

Fossil corals are found in marine sediments on the north and southwest coast of the southern Boso Peninsula. In situ hermatypic corals form an important constituent of the oyster reef in Tateyama area, which was described in a previous section. With the exception of the above locality, all in situ coral occurrences so far reported are located in the West Coast Valley area. A.t least 12 coral localities have been given in literature over the past 40 years, but because of temporary nature of many outcrops (construction sites, excavations, etc.) only some of them could be examined by the present author. Coral localities are shown on the map in text-fig. 9. It is apparent from the map, with the probable Jomon coastline drawn in, that most outcrops lie in the inner parts of small coastal bays. The distribution appears erratic, because corals are found in one valley but not in the neighboring one, which is apparently identical.

The type locality of the Numa coral bed is in the valley of Numa. Most of the outcrop is presently covered by a pond, but an exposure about 30 cm high and 10 m long at 18m above sea level exists along the eastern edge of the pond. Stylocoeniella hanzawai and Favia speciosa, which appear to be in living position, are embedded in silty sand matrix. About ten large colonies, which were unearthed during the construction of the pond, are preserved in a local municipal monument. Included are Caulastrea tumida, Podabacia elegans lobata, and Montastrea curta; some of the colonies reach a diameter of 70 cm.

An outcrop about 1 m high and 30 m long is found at 17 m above sea level along the banks of a stream flowing into the Shioiri River at Kamisanakura. The outcrop lies about 500 m upstream from the confluence where the stream approaches the right side of the valley. Corals and shells are embedded in silty sand matrix. Lima sowerbyi, Barbatia lima, and Pretostrea imbricata are the most abundant molluscs. The coral assemblage is dominated by Echinophyllia aspera, most of which occurs as 5 to 10 cm fragments, even though a few nearly entire colonies which appear to be in living position are found. About 20 cm fragments of Goniopora sp. are not in living position, but are found in clusters, suggesting that they are nearly in situ, having probably broken off from the mother colony. Large massive colonies of Favia speciosa and Favites favosa are absent.

Two outcrops found in the Koyatsu valley are the largest found to date. The outcrops have been described in detail by Hamada (1963). The smaller of the outcrops, located in the upper part, left side of the valley, lies between 14 and 16 m above sea level, where Neogene siltstone is exposed in several places. The bedrock is overlain by angular conglomerate containing shells of Crassostrea gigas and Serpulorbis imbricatus. Occasional small Oulastrea crispata colonies are found attachgd to the rocks. Large colonies of Favia speciosa, Favites favosa, and radial fingerlike growths of Caulastrea tumida, all in living position, are attached directly to the bedrock or rest on the conglomerate and coral rubble. The corals are surrounded by greyish blue silty calcareous sand. The in situ corals are overlain by a 50 to 100 cm thick bed of silty sand containing abundant shells (especially Barbatia bicolorata, Pretostrea imbricata, and Lima sowerbyi) and fragments of predominantly platy corals: Echinophyllia aspera, Physophyllia ayleni, Pectinia lactuca, Lobophyllia japonica, and Lobophyllia robusta. Massive corals are uncommon, few Favia colonies (not in living position) occur in the upper part of the bed. Caulastrea is absent except for a few worn fragments. Most of the platy corals are present as 5 to 10 cm fragments, but several well-preserved colonies, reaching the diameter of up to 40 cm, and one 80 cm colony in living position, were found. All the corals, including the fragments, are extremely well preserved and do not show any signs of transportation.

The larger outcrop lies between 13 and 18 m above sea level and consists mainly of platy corals embedded in silty sand matrix, resembling the uppermost bed of the smaller outcrop.

The area of distribution of transported coral fragments partly overlaps but is wider than that of the in situ corals. Sand containing coral fragments and isolated coral rubble is found in fields in the valleys neighboring the Numa outcrop, suggesting that extensive coral-bearing sediments are present in the area. Worn fragments of Echinophyllia aspera, Lobophyllia japonica, and a large, well-preserved colony of Montastrea curta were found in silty sand in the center of Shiomi valley, about 500 m west of Koyatsu. Slightly worn, transported fragments of Caulastrea tumida and Favia speciosa were found in sediments in the Tomoe River valley. The sediments are silty sand located in the central part of the valley at 16 m above sea level, containing numerous Osira nipponica, apparently in situ. Several small Oulastrea crispata colonies attached to pebbles, but apparently slightly transported, were also found.

Hermatypic corals presently living in Tateyama Bay have been found to the northwest of the island of Okinoshima in depths exceeding 5 m. The corals are attached to bedrock occurring as low (1 m), long ridges formed by steeply dipping resistant beds of Neogene volcanic sand and extending from the shore to depth. The present day corals differ from the Jomon transgression corals not only in their ecology (virtually open coast versus inner bay) but also in their growth form. Although coral colonies exceeding 4 m in diameter have been reported from Tateyama Bay (Yabe and Sugiyama, 1935), the colonies the present author was able to examine were smaller than 1 m and on large parts of the surface coral tissue was absent, and living tissue occurred in patches. Smaller specimens, which can be examined in detail, revealed a repeated overlapping of coral and calcareous algae, occurring in 5 to 10 mm thick layers. Apparently, under unfavorable conditions, portions of the coral die and are overgrown by calcareous algae. When conditions improve again, the faster growing coral expands from surviving patches and covers the algae. No similar overgrowth has been observed in the Jomon transgression specimens, suggesting much more favorable conditions.

Hamada (1963) has already pointed out the similarity between the special distribution of Numa corals and distribution of recent corals in Tanabe Bay, Kii Peninsula. A similar pattern where the most luxuriant coral growth is concentrated in patches along the perimeter of bays is found in several other localities-for instance, in Kamae Bay in northeast Kyushu (Utinomi, 1971).

It is probable that the availability of hard substrate, in the form of bedrock outcrop, is the main reason for this distribution, although poor circulation and low transparency of water may contribute to keeping corals out of the central, deeper parts of the bays. The same factors were probably responsible for the distribution of corals in the southern Boso Peninsula during the Jomon transgression. At the beginning of the transgression, when the temperature became high enough, the sea level already reached at least 15 m. The centers of the drowned valley bays must have been covered with soft sediment; the only firm substrate for corals to settle and grow was available along the perimeter of the bays where bedrock was exposed. Low visibility may have been another cause preventing corals from settling in the deeper parts of the bay. As the sea level rose even further at a rate far exceeding the speed of coral growth, corals found new substrate ready on the outcrops of bedrock exposed by rising sea water along the bay edges. The deeper environment favored platy forms, such as Echinophyllia aspera, which are presently found at depths of 5 to 15 m.

Large platy corals, such as Echinophyllia, are attached by a very small area relative to the size of the entire colony. The attachment is also the oldest part of the colony and is most likely to be weakened by bioerosion (boring sponges, bivalves). Large colonies would therefore easily break off (during storms, earthquakes, etc.) and fall down the slope, breaking into smaller fragments. This accounts for the fragmented state of the fossil platy corals as well as for the lack of signs of extensive transport.

The edges of the bays represented not only a suitable habitat for the corals but also a site where corals growing over a large area accumulated in a relatively small area at the foot of the slope, producing thicker deposits with increased potential for preservation and exposure.