Holocene Sea Level and Tectonic History Related to the Formation of Coral Terraces at Kikai Island, Northern Ryukyu Islands

Abstract

Kikai Island in the northern Ryukyu chains, 80km from the Ryukyu Trench, has the highest uplift rate in Japan, 1.8m/ka since isotope stage 5e. Four subdivided Holocene terraces, from the highest terrace I to the lowest IV, fringe Kikai Island and record the Holocene uplift. These terraces were previously described and considered as evidence for coseismic uplift and sea level fluctuation since the culmination of the Postglacial transgression by Nakata et al. (1978) and Ota et al. (1978), on the basis of surface observation. Here, in addition to the surface data, we use subsurface data to infer the history of formation and emergence of Holocene coral reef terraces.
We used five drilling cores at the Shidooke coast, northeastern Kikai Island. These drillings reached the base of the Holocene sediment, with maximum thickness of 26m. The basal topography beneath the Holocene sediment was reconstructed. Holocene sediment is classified into five lithofacies: A (coral limestone), B and C (detrital limestone), D and E (detrital limestone with terrestrial fragments). Forty in situ corals are dated by the α-spectrometric ²³⁰Th/²³⁴U method. Isochron for Holocene sediments indicates upward reef growth from ca. 10ka to ca. 6.5ka, followed by seaward growth since that time.
The highest Terrace I is mostly underlain by non-coral sediment of Facies C. In contrast, Terrace II, the widest and with moat and crest system, is underlain by Facies A, typical coral limestone. Thus, Terrace II is a constructional terrace of coral limestone which emerged at 5.3ka. Narrow terraces III and IV are essentially cut surfaces into the steep fore reef slope of Holocene coral tract, characterized by Facies A but with some younger corals at their outer margin. The timing of their emergence is not determined at moment, but is estimated to be 3.1ka and younger than 2.6ka, respectively.
Holocene coral terraces at Kikai Island are compared with those at Huon Peninsula, Papua New Guniea, which is located in the core region of reef growth. Uplift-corrected height of the transgressive corals at Kikai is generally lower than the predicted eustatic sea level curve, implying that coral growth did not keep pace with sea level rise and probably should be interpreted as a catch-up type. In contrast, transgressive corals at Huon closely coincide with the predicted sealevel curve, which suggests that they are a keep-up type. Repeated coseismic uplifts commonly occurred at Kikai and Huon, but were more evident at Huon, where the uplift rate is higher than at Kikai. One of the significant differences is that the highest Holocene terrace at Huon is a transgressive constructional one, underlain by a transgressive reef, while that of Kikai is represented by a constructional terrace of noncoral sediments, reflecting a slower reef growth than that of Huon Peninsula.