PR sites, the water depths at these sites have been deeper than the CCD since the Eocene; thus, dilution of the hydrothermal component by biogenic carbonate has not occurred since that time (Fig. 3b; Supplementary Fig. S18). However, the IC7 signal that was maintained at a higher level than that in other areas, together with the Fe n concentrations, showed a stepwise decrease to negative values during the Miocene (Fig. 3f; Supplementary Fig. S18). If plate migration and change in distance from the mid-ocean ridges caused this reduction, the hydrothermal deposition should have gradually and consistently decreased upward from the core bottom. Therefore, we consider the possibility that relatively rapid reduction can be attributed to the disappearance of the hydrothermal source owing to subduction of the ridge during the mid-Miocene37 (Supplementary Fig. S18). However, other regional factors such as changes in ocean currents resulting in migration of hydrothermal plumes could also affect the IC7 signal intensity.Rare-earth element geochemistry. To further clarify the geochemical interpretations of IC1 and IC4, REY-rich mud samples strongly influenced by either IC1 or IC4 were selected, and their REY-patterns normalised to the post-Archean average Australian shale38 were compared (Fig. 4). Because the variables analysed by ICA do not contain explicit information of each lanthanoid element and Y content, this approach can provide an additional geochemical constraint. The bulk REY-patterns of representative high-IC4 sediments obtained from the eastern South GW 4064 chemical information Pacific and central North Pacific are characterised by distinct negative Ce anomalies, positive Y anomalies, and relative enrichment in heavy rare-earth to light rare-earth elements (Fig. 4g ). These features are almost the same as those of biogenic Ca-phosphate22?5 (Fig. 4m), suggesting that the bulk-sediment REY compositions of high-IC4 sediments were predominantly controlled by highly REY-enriched biogenic Ca-phosphate. The representative high-IC1 sediments are distributed in the central South Pacific, central North Pacific, and eastern Indian oceans. The Indian Ocean’s REY-rich mud contains no age-diagnostic fossils3 and is thus not shown in Fig. 3. In contrast to the case of high-IC4 sediments, the bulk REY patterns of high-IC1 sediments were characterised commonly by smaller negative Ce anomalies, almost no Y anomaly, and relatively flat REY patterns (Fig. 4a ). These features can be reasonably explained4 by superimposing the REY pattern of slowly precipitating hydrogenous Fe n oxide ( nO2) showing pronounced positive Ce and negative Y anomaliesScientific RepoRts | 6:29603 | DOI: 10.1038/srepwww.nature.com/scientificreports/(Fig. 4m) on that of biogenic order EPZ004777 Ca-phosphate in pelagic clay inheriting distinct negative Ce and positive Y anomalies from the seawater22?5 (Fig. 4m). It is noteworthy that the bulk REY abundance in IC1-type REY-rich mud also appeared to be largely attributed to biogenic Ca-phosphate, considering its very high REY concentration (REY + Ce > 20,000 ppm)25 and negative Ce anomalies in the bulk REY compositions. Therefore, the characteristic data distribution in the IC space was confirmed by a different geochemical index using elements not explicitly included in the ICA calculation. Despite the very high concentration of REY in biogenic Ca-phosphate, these elements are not incorporated into Ca-phosphate in significant amounts during the primary skeleton-forming process.PR sites, the water depths at these sites have been deeper than the CCD since the Eocene; thus, dilution of the hydrothermal component by biogenic carbonate has not occurred since that time (Fig. 3b; Supplementary Fig. S18). However, the IC7 signal that was maintained at a higher level than that in other areas, together with the Fe n concentrations, showed a stepwise decrease to negative values during the Miocene (Fig. 3f; Supplementary Fig. S18). If plate migration and change in distance from the mid-ocean ridges caused this reduction, the hydrothermal deposition should have gradually and consistently decreased upward from the core bottom. Therefore, we consider the possibility that relatively rapid reduction can be attributed to the disappearance of the hydrothermal source owing to subduction of the ridge during the mid-Miocene37 (Supplementary Fig. S18). However, other regional factors such as changes in ocean currents resulting in migration of hydrothermal plumes could also affect the IC7 signal intensity.Rare-earth element geochemistry. To further clarify the geochemical interpretations of IC1 and IC4, REY-rich mud samples strongly influenced by either IC1 or IC4 were selected, and their REY-patterns normalised to the post-Archean average Australian shale38 were compared (Fig. 4). Because the variables analysed by ICA do not contain explicit information of each lanthanoid element and Y content, this approach can provide an additional geochemical constraint. The bulk REY-patterns of representative high-IC4 sediments obtained from the eastern South Pacific and central North Pacific are characterised by distinct negative Ce anomalies, positive Y anomalies, and relative enrichment in heavy rare-earth to light rare-earth elements (Fig. 4g ). These features are almost the same as those of biogenic Ca-phosphate22?5 (Fig. 4m), suggesting that the bulk-sediment REY compositions of high-IC4 sediments were predominantly controlled by highly REY-enriched biogenic Ca-phosphate. The representative high-IC1 sediments are distributed in the central South Pacific, central North Pacific, and eastern Indian oceans. The Indian Ocean’s REY-rich mud contains no age-diagnostic fossils3 and is thus not shown in Fig. 3. In contrast to the case of high-IC4 sediments, the bulk REY patterns of high-IC1 sediments were characterised commonly by smaller negative Ce anomalies, almost no Y anomaly, and relatively flat REY patterns (Fig. 4a ). These features can be reasonably explained4 by superimposing the REY pattern of slowly precipitating hydrogenous Fe n oxide ( nO2) showing pronounced positive Ce and negative Y anomaliesScientific RepoRts | 6:29603 | DOI: 10.1038/srepwww.nature.com/scientificreports/(Fig. 4m) on that of biogenic Ca-phosphate in pelagic clay inheriting distinct negative Ce and positive Y anomalies from the seawater22?5 (Fig. 4m). It is noteworthy that the bulk REY abundance in IC1-type REY-rich mud also appeared to be largely attributed to biogenic Ca-phosphate, considering its very high REY concentration (REY + Ce > 20,000 ppm)25 and negative Ce anomalies in the bulk REY compositions. Therefore, the characteristic data distribution in the IC space was confirmed by a different geochemical index using elements not explicitly included in the ICA calculation. Despite the very high concentration of REY in biogenic Ca-phosphate, these elements are not incorporated into Ca-phosphate in significant amounts during the primary skeleton-forming process.
