Compositional Variations and Genesis of Sandy-Gravel Ferromanganese Deposits from the Yōmei Guyot (Holes 431, 431A DSDP), Emperor Ridge

: This research presents results characterizing the mineral and chemical composition of ferromanganese (Fe-Mn) deposits from Yōmei Guyot (Holes 431 and 431A), recovered during the Deep-Sea Drilling Project (DSDP) Leg 55 R/V “Glomar Challenger”. The Fe-Mn deposits are represented by sandy-gravel clasts. The mineral composition and bulk concentration of major and minor elements, as well as the distribution of rare earth elements and yttrium patterns in mineral fractions of Fe-Mn samples, showed that the deposits are composed of fragments of Fe-Mn hydrogenetic crusts and diagenetic nodules. The morphology of Fe-Mn clasts from Holes 431 and 431A DSDP, as well as a comparison with growth conditions of Fe-Mn deposits from N-W Pacific Guyots, allowed us to establish a Late Pliocene age for the formation of this Fe-Mn placer from Yōmei Guyot. Accumulations of ferromanganese clasts in a sedimentary unit led us to classify this geological body as a new mineral resource of the World Ocean.


Background
Three genetic types of ferromanganese (Fe-Mn) deposits are widely distributed and have been studied in detail in the World Ocean. These are hydrogenetic crusts, hydrogenetic and diagenetic nodules, and hydrothermal deposits, each are distinguished based on the type of aqueous fluid from which the Fe-Mn (oxyhydr)oxides precipitate [1].
Ferromanganese (cobalt-rich) crusts are composed of Fe oxyhydroxide and Mn oxide, which precipitate directly from cold, ambient ocean water onto solid rock substrates. A wide variety of metals such as Co, Ti, Mn, Ni, Pt, Zr, Nb, Te, Bi, Mo, W, Th, as well as rare earth elements and yttrium (REY) are sorbed from ocean water onto those two main mineral phases (Fe oxyhydroxide and Mn oxide) [1][2][3].
Ferromanganese nodules are composed of Fe oxyhydroxide and Mn oxide and form by hydrogenetic and diagenetic precipitation onto a hard nucleus on the surface of soft sediment. Diagenetic precipitation occurs from sediment pore fluids consisting of ocean water modified by chemical reactions within the sediment column. Diagenesis may take place under oxic or suboxic conditions, with the boundary generally defined as a 5 μM oxygen concentration. Mixed origin Pb, Sr, Zn, Mo, Zr, La, Ce, Nd, Sm, Eu, Nb, W, As, and Sb determinations using X-ray radiometric and neutron activation analysis [18].
These investigations showed that the ferromanganese deposits are composed of vernadite and 10Å manganate with minor occurrence of 7Å manganate. The average content of Mn is 21.9%, that of Fe is 12.9%, and the average sum total of Ni, Cu, and Co is 0.98%. The concentrations of other elements are also slightly different from the average composition of crusts from the Pacific Seamounts [18]. These values are higher than those in the ferromanganese core samples (431 and 431A DSDP) [17]. Closest to Yōmei Guyot, Nintoku Guyot lies at similar water depths and is characterized by the absence of the deposit cover. Due to these features, M. Melnikov [18] suggested that in the Fe-Mn deposits of the Yōmei Guyot, hydrothermal genesis was predominant and hydrogenous input had a lesser influence.
Thus, Fe-Mn deposits in hole sediments and on the surface of solid host rock at the Yōmei Guyot have a combination of hydrothermal and hydrogenous genesis [17,18]. However, more recent studies indicate the hydrogenic genesis of Fe-Mn deposits from the Pacific Ocean guyots [2,3]. Here, to clarify the genesis of the Yōmei Guyot Fe-Mn deposits, we determined the mineral and chemical compositions of 431 and 431A DSDP core samples using ICP-AES and ICP-MS analyses of both the bulk samples and the main mineral phases. The metal distributions within the mineral phases were estimated using the sequential leaching of 4 mineral phases: I-easily leachable adsorbed cations and carbonates; II-Mn-oxide; III-Fe-oxyhydroxide; and IV-residual components [19]. The distribution of trace elements [20] and REY [7] in the mineral phases of hydrothermal and hydrogenous crusts strongly reflects their genesis. In addition, the fragmented Fe-Mn deposits of Yōmei Guyot might be identified as another type of Fe-Mn marine deposit. However, a different type name has not yet been proposed distinguishing it from hydrogenous crusts of seamounts and guyots, hydrogenic and diagenetic nodules of abyssal basins, and hydrothermal crusts of underwater volcanoes. Finally, we studied the deposition factors in shelf mineral placers in order to select criteria for the genetic identification of deep-sea Fe-Mn fragmented deposits.

Geological Setting
Geological and geophysical surveys of Yōmei Guyot were carried out during the cruise of Lee8-76-NP by R/V "S.P. Lee" in 1976 [21], cruise 55 by R/V "Glomar Challenger" in 1977 [22], and cruise 9 by R/V "Geologist Peter Antropov" in 1990 [18,23]. Yōmei Guyot is located on the same basement as Nintoku Guyot: both guyots are part of the central Emperor Ridge (Figure 1). The size of the Yōmei Guyot base at 3500 m depth is 80×65 km, and the top is located at 890 m depth. The guyot has a conical shape with well-defined northwest and northeast spurs. The summit has steep southern and shallow northern slopes with a size of 40×30 km at 1500 m depth. The surface has abrasion outcrops, terraces, small hills, and hollows.  (Figure 2b). Crust growth proceeded in at least three stages [18] and had an episodic pattern. The absence of sediments at the guyot top is probably caused by high velocities of bottom water currents. This is also reflected in the smooth, polished surface of Fe-Mn deposits and volcanic rocks, as well as in the unidirectional inclination of filtering hydrobionts, which were identified by underwater photos. Four sites with a negative bottom relief were found (Figure 2a), where the thickness of the deposited detrital (mostly redeposited) material and biogenic rocks is up to 50 m. Two holes (431 and 431А DSDP) were drilled at one of these sites. In the lower parts of the guyot slopes, and on the basement slopes, the sediment thickness increases sharply with increasing ocean depth.

Materials and Methods
Thirteen samples of Fe-Mn fragments from 431 and 431A DSDP holes were studied. The sampling intervals are presented in Table 1   X-ray diffraction was carried out using a MiniFlex II diffractometer (Rigaku Corporation, Tokyo, Japan) with CuKα radiation. The voltage on the X-ray tube was 30 kV, and the tube current was 15 mA. The scanning speed was 1° 2/min. The scan angle ranged from 2.5° to 60°. The diffractometer was also equipped with a selectively absorbing Ni filter to suppress Kβ lines. The ICDD 2010 database was used for mineral identification.
ICP-AES was performed on an ICAP6500 Duo spectrometer (Thermo Electron Corporation, Waltham, MA, USA), which was calibrated using a multi-element certified ICP Multi Element Standard IV (Merck, Germany) solution. ICP-MS was carried out on an Agilent 7700× quadrupole mass spectrometer (Tokyo, Japan). The OOPE 601 and OOPE 602 standard samples of the Fe-Mn nodules were analyzed to assess the accuracy.
Acid dissolution using a mixture of concentrated HF, HNO3, and HClO4(1:5:1) was used to determine the bulk element contents of studied samples.
We used sequential leaching to study the distribution of elements in the mineral phases of Fe-Mn crusts [19,20,25,26].
The sequential leaching procedure used in our experiments included the following stages: Phase I-leaching of 1 g of powdered sample with acetate buffer (CH3COOH (1 N) + CH3COONa · 3H2O (1 N), pH = 5) for 5 h at 25°C at a 1:50 reagent ratio. This led to the extraction of Ca carbonates and related elements, and to the release of sorbed ions. The solution was filtered, mineralized in a microwave using concentrated HNO3 and analyzed by ICP-AES and ICP-MS. The residue was washed with Type I deionized water, dried, and used in further leaching.
Phase II-extraction of Mn and related elements by 0.1 M hydroxylamine hydrochloride solution (pH = 2) for 24 h at 25°C at a 1:100 sample/reagent ratio. The solution was filtered by filter paper (3-5 μm), the hydroxylamine in the solution was decomposed using concentrated HNO3 and heating (170 o C), and after evaporation, the precipitate was dissolved in 5% HNO3. Solutions were analyzed using ICP-AES and ICP-MS. To complete Mn extraction, the hydroxylamine concentration was increased to 0.5 M, as suggested in [26]. Time intervals of 3 min and 24 h were chosen to determine the optimal leaching time. In both cases, the degree of leaching of Mn was the same, but with an increase in Fe dissolution during 24 h leaching. A final Mn extraction scheme included leaching with a 0.5 M solution of hydroxylamine over 3 min. After phase I, the ratio of the sample/reagent was 1:100. The solution was filtered, the hydroxylamine in the solution was decomposed using concentrated HNO3 and heating, and after evaporation the precipitate was dissolved in 5% HNO3. Solutions were analyzed using ICP-AES and ICP-MS. After phase II, the residue was washed, dried, and left for further leaching.
Phase III-extraction of Fe-hydroxides using a mixture of 0.2 M oxalic acid and ammonium oxalate (pH = 3.5) for 12 h at 25°C at a 1:175 sample/reagent ratio. The solution was evaporated to form wet salts and concentrated HNO3 was added to destroy the organic matrix. After evaporation, the residue was dissolved in 5% HNO3. The obtained solutions were analyzed using ICP-AES and ICP-MS.
Phase IV-the solid residue from the previous stages, presented by aluminosilicates, together with the filter (filter paper) were ignited at T = 600°C in a muffle furnace. Then, the residue was digested in Teflon bombs using a mixture of HF, HNO3, and HClO4(1:5:0.5) with heat (170°C) applied to the wet salts, which were diluted by 2% HNO3 to volume 25 mL. Solutions were analyzed using ICP-AES and ICP-MS.
Analyses were performed at the Center for Collective Use of the Far Eastern Geological Institute of the Far Eastern Branch of the Russian Academy of Sciences, Vladivostok, Russia.

Results
Ferromanganese deposits recovered from the DSDP 431 and 431A holes are represented mainly by sharp-edged fragments and, more rarely, by slightly rounded fragments (Figures3 and 4). The structure is usually massive or, less often, laminar. The sediment is mostly characterized by an anthracite luster (85%), but some parts have a dull luster.
The size of the Fe-Mn fragments is in the range of sandy gravel and varies from 1 to 27 mm, gradually increasing to the core bottom: ranging from 4 to 15 mm (middle size) in the 431 core and from 3 to 20 mm in the 431A. The mineral composition of the studied samples is represented predominantly by vernadite (δ-MnO2). However, 10Å and 7Å manganates were determined in YM-12 and YM-20 samples from the 431A hole. Accessory minerals include quartz, plagioclase, and smectite.
The bulk chemical composition of 431 and 431A DSDP samples is shown in Table 2. A ternary plot of Fe:Mn:(Co + Ni + Cu)×10 (Bonatti diagram [28], Figure 5) shows that most of the studied samples fall within the hydrogenic crusts field, however, YM-12 and YM-20 samples are located beyond the field limit. Manganese, Si, K, Mg, Li, Cr, Sb, and Ba are enriched and Fe, Ti, P, Be, Se, Sr, Zr, Mo, Nb, Ta, Hf, Tl, Pb, Th, U, and REY declined in the YM-12 and YM-20 samples. The contents of other elements such as Al, Ca, Sc, V and W do not vary. We also studied the mineral and chemical composition of the YM-12 and YM-20 samples to determine the reason for differential element accumulation. The sample 431-5 from the YM-05 interval of the 431 DSDP hole has maximum sums of Co, Ni, Cu, and ΣREY of 10,963 ppm and 3042 ppm, respectively (Table 2), which are typical for hydrogenous crusts [2], therefore it was selected as a main group. We chose the 431A-12 sample from the YM-12 interval of the 431A hole as a representative sample of the contrast group. It is enriched in Mn (31.38%) and Cr (14.4 ppm), as is typical for hydrothermal deposits [2]. Karpoff et al. (1980) [17] and Melnikov (1993) [18] suggested hydrothermal influence at the Yōmei Guyot. We studied the REY distribution in the main mineral phases of the 431-5 and 431A-12 samples to assess whether this was true. The REY pattern distribution in manganese and iron mineral phases of hydrothermal-hydrogenetic deposits has a peculiarity. The Mn phase inherits the composition of the initial hydrothermal fluid and has an extremely small ΣREY. The Fe phase approaches the REY pattern distribution seen in hydrogenetic Fe-Mn crusts of guyots and has increased by an amount of REY [7].  Table 2).
The mineral composition of the 431-5 sample is composed of vernadite (δ-MnO2), and sample 431A-12 contains 10Å and 7Å manganates. Table 3 presents the chemical composition of these two samples in comparison with the Fe-Mn crusts from the different regions of the Pacific Ocean. The Mn and Fe contents in the 431-5 sample from the Yōmei Guyot are 28.89% and 15.57%, respectively, and in 431А-12, they are 34.96% and 8.76%, respectively. The differences are observed in the major elements (Si, Al, Ti, K, and P), trace elements (Li, Be, Cr, As, Sr, Nb, Mo, Ba, Ta, Tl, Pb, Th, and U), and REY. Some chemical elements have similar values (Sc, Ga, Cd) or exhibit slight variations (V, Co, Ni, Cu, and Zn). Table 4 shows the distribution (concentration (ppm) and percentage (%)) of the main elements (Ca, Mn, Fe, Si and Al) and REY in the mineral phases of the Yōmei Guyot Fe-Mn deposits.

Discussion
Most of the samples from the 431 and 431A DSDP cores of Yōmei Guyot contain vernadite (δ-MnO2), indicating a hydrogenous component, and two samples of the 431A core include hydrothermal and diagenetic 10Å and 7Å manganates [2]). This is consistent with previous studies of Fe-Mn fragments from the Yōmei Guyot core [17] and crusts and nodules from the guyot top [18]. However, X-ray diffraction methods do not allow determination of 10Å manganate minerals due to overlapping of the diffraction peaks of todorokite, asbolan, buserite-I, buserite-II, mixed-layer phases of asbolan-buserite-I, buserite-I-buserite-II, and buserite-I-"defective lithiophorite" [36]. In total, 7Å-manganate is represented by hydrothermal and diagenetic birnessite [2]. The repeated XRD results are consistent with the vernadite composition of 431-5 and with the presence of 10Å manganate and birnessite in the samples from 431A-2. As we used different fragments from the same stratigraphic hole intervals for the repeated X-ray diffraction analyses, the mineral composition indicates that the Fe-Mn fragments in the YM-05 and YM-12 intervals probably came from the same sources (Table 1). Thus, the vernadite-containing fragments reveal the hydrogenous component, and the 10Å manganate and birnessite-bearing intervals reflect either hydrothermal or diagenetic genesis.
A ternary Bonatti plot of Fe:Mn: (Co + Ni + Cu) × 10 of 431-5 and 431А-12 samples is shown in Figure 6. The 431-5 sample falls near the hydrogenous Fe-Mn crusts of the Marshall, Magellan, and Markus-Wake Seamounts, whichare located close to the Detroit Guyot of the Emperor Ridge. The 431A-12 sample is located near the hydrogenous crusts of the Okhotsk seamounts field, which are characterized by high content of allothigenic input [29]. In the Figure 6, the diagenetic Peru Basin nodules and hydrothermal crust from the Japan Sea are placed at a distance from the Yōmei Guyot samples. This proves the hydrogenous geneses of both studied samples.
However, the main (from Si, Ti to P, and LOI) elemental composition of the 431-5 and 431А-12 samples differ significantly. The 431-5 element contents are similar to those of crusts from NW Pacific Ocean and those of the 431A-12 sample are identical to diagenetic nodules of the Peru Basin ( Table 3). The latter form in a high-bioproductivity environment, that produces suboxic conditions [37] and indicates the diagenetic geneses of samples. In contrast, the Co (4654 ppm), Ni (3139 ppm), and Cu (611 ppm) contents in 431A-12 fragment is the same as that in the hydrogenous crusts of the NW Pacific Ocean (Co-3733 ppm, Ni-3495 ppm, Cu-1074 ppm), which differs significantly from the diagenetic nodules of the Peru Basin (Co -475 ppm, Ni-13010 ppm, Cu-5988 ppm). The high Co and Ni contents of the latter is attributed to the presence of Cu 2+ and Ni 2+ in available vacancies within 10Å manganate [30]. The elemental distribution in the 431А-12 sample is possibly related to the combined hydrogenous and diagenetic genesis. The pore fluids are a source of Ni, Cu and Co is sorbed from ocean water [30].
The Ba/Ti ratio in Fe-Mn deposits reflects their genesis. The Ba/Ti value of 62 is interpreted as evidence of hydrothermal input [35,[38][39][40][41] and 0.22 is typical for the hydrogenous crusts of the northern Pacific Ocean [30]. The Ba/Ti ratio of hydrogenous crusts from back-arc basins with high bioproductivity such as the Kashevarov Trough of the Okhotsk Sea increases up to 5.22 [29] and in diagenetic nodules varies from 1.97 in the Peru Basin [30] to 2.76 in the Deryugin Basin of the Okhotsk Sea [31]. The Ba/Ti ratio in the 431-5 sample of the Yōmei Guyot is 0.20 and the closest value to this is that of the hydrogenous crusts of the Pacific Guyots. In the "contrasting" 431A-12 sample, the Ba/Ti ratio is 6.91, reflecting an additional supply of Ba, probably due to a high bioproductivity.  (Table 3); 2-Fe-Mn crust from the Kashevarov Trough, the Sea of Okhotsk, [29]; 3-Fe-Mn nodules from the Peru Basin, [30]; 4-Fe-Mn crust from the Seth Guyot, Markus-Wake cluster, [33]; 5-Fe-Mn crust from the Detroit Guyot, Emperor Ridge, [34]; 6-Fe-Mn crust from the Belyaevsky Volcano, the Sea of Japan, [35]. Our unpublished data: 7-Fe-Mn crust from the Zubov Guyot, Marshall Island cluster; 8-Fe-Mn crust from the Govorov Guyot, Magellan Seamounts [32]; 9-Fe-Mn crust from the Zonne Chain, the Sea of Okhotsk, [31].
Hydrothermal deposits have an enriched Li content (from 148 to 1188 ppm) [42]. The Li content in diagenetic nodules is less (95.4-318.7 ppm) and is depleted in hydrogenous crusts (1.18-5.68 ppm) [43]. The Li content in mixed types, such as hydrothermal/hydrogenous and hydrogenous/diagenetic, varies from 2.4 to 62 ppm [42]. The Li value in the 431-5 sample is 3.20 ppm reflecting a hydrogenous component and is higher in sample 431A-12 (19.5 ppm), which probably corresponds to it having had a mixed input.
Thallium is a redox-sensitive trace element oxidized on the surface of Fe-Mn oxyhydroxides. Since hydrogenous crusts contain higher amounts of Tl (155 ppm in the North Pacific Prime Zone) in comparison with deep-water diagenetic nodules (129 ppm in the Peru Basin) [2]. The Tl content in the 431-5 sample is 192.2 ppm, slightly higher than in the hydrogenous crusts of the Marshall (163 ppm), Magellan (167 ppm), and other Pacific Guyots (Table 3). It reflects Tl precipitation in oxic conditions. The Tl content in the 431A-12 sample is significantly less (9.47 ppm) and is comparable to shallow-water diagenetic Fe-Mn nodules of the Baltic Sea, which vary from 5.35 to 19.5 ppm, hydrothermal crusts, which vary from 0.68-44.0 ppm, and mixed hydrothermal and hydrogenous Fe-Mn crusts, which vary from 7.00-9.48 ppm [44]. We suggest that Tl in the 431A-12 sample has a hydrothermal or hydrothermal/hydrogenous genesis and its accumulation probably occurred in the local suboxic conditions. Yōmei Guyot crusts formed in deep-sea conditions (1000 m or more) [17,45].
REY are sensitive to changes in redox conditions [46]. Hydrogenous Fe-Mn crusts are enriched in REY up to 1682 ppm on average, with an Nd content of 162 ppm. Diagenetic nodules have a mean REY content of about 349 ppm and Nd concentration of 58 ppm. Hydrothermal deposits contain minimal content of REY (29.8 ppm) and Nd (5.2 ppm) [1]. The distribution of PAAS normalized REY (McLennan, 1989) contents of three of these genetic types are different (Figure 7a). The ΣREY and Nd content in the 431-5 sample from the Yōmei Guyot are 3074 ppm and 266 ppm, respectively (Table 3), reflecting their hydrogenous geneses. The amounts of REY and Nd in sample 431A-12 are 451 ppm and 55.4 ppm, respectively, indicating their diagenetic formation. The PAAS normalized REY [47] contents of 431-5 and 431A-12 samples from the Yōmei Guyot show a similar fractionation of bulk composition (Figure 7b). Moreover, the REY distribution of hydrogenous crusts in Figure 7a is the same as the 431-5 sample in Figure 7b. The REY distribution excluding Ce content of the 431A-12 sample is similar to that of diagenetic nodules (Figure 7a). The Ce behavior closely corresponds to that in hydrogenous Fe-Mn nodules [1]. Thus, the REY distribution in the 431-5 sample reflects the hydrogenous geneses and the 431А-12 sample possibly corresponds to the diagenetic influence but with features typical for seamounts.  Table 2.
The chemical composition of representative samples from the 431 and 431A DSDP holes of the Yōmei Guyot indicates the hydrogenous origin of the 431-5 sample and the undeterminable hydrothermal or diagenetic genesis of the 431А-12 sample. The study of REY distribution in the mineral phases is the most useful method for studying the samples' geneses. For example, the REY distribution in the Mn phase of hydrothermal crusts differs significantly from those of the hydrogenous and diagenetic geneses and from the Fe phases of all genetic types [6,7,48,49].
The sequential leaching of 431-5 and 431A-12 samples of Yōmei Guyot helps to determine their geneses. Since Phase I is depleted by REY, PAAS normalized contents on a logarithmic scale are inconvenient to use, and we instead use normalized values on an arithmetic scale (Figure 8). The REY from Sm to Dy are enriched in samples with a maximum value of Gd, which is typical for seawater, and, accordingly, for hydrogenous Fe-Mn crusts [50]. A negative Ce anomaly in seawater is formed because Ce can take a tetravalent state and is easily hydrolyzed and scavenged onto Fe-Mn oxides/hydroxides and clay particles [51]. A negative Ce anomaly indicates oxic conditions because a positive Ce anomaly in the extraction of highly mobile REY indicates a reduced environment of precipitation [37,52]. The REY contents of leached Phase I from sample 431-5 falls near the boundary between the logarithmic high-latitude and equatorial Fe-Mn Pacific crusts, and those of the 431A-12 sample are similar to ring-like diagenetic Fe-Mn nodules from the Deryugin Basin. The REY content in Phase II is illustrated using a logarithmic scale (Figure 9), because normalized concentrations of different genetic types of Fe-Mn deposits and their sources such as seawater and hydrothermal fluid differ by orders of magnitude. We combined the REY distribution fields of high-latitude and equatorial Fe-Mn crusts from the Pacific Ocean as one logarithmic "Fe-Mn crusts from Pacific Ocean" field because of overlapping. The hydrothermal deposits are characterized by extremely low REY contents and an Eu concentration below the detection limit, reflecting their deposition from a low-temperature fluid [53]. Their formation near high-temperature fluids would lead to Euaccumulation [7]. Hydrogenous crusts are enriched in REY and have a positive Ce anomaly ( Figure 9). Graphs of most ancient phosphatized Fe-Mn crust layers and diagenetic nodules fall in-between. Phosphatized layers of Phase II show high REY contents and positive Ce and negative Y anomalies due to dissolution, restructuring, and secondary enrichment caused by environmental changes [2]. Diagenetic nodules from the Sea of Okhotsk are depleted in REY and have clearly defined positive Y and poor Ce and Gd anomalies [2]. The REY distribution of the 431-5 sample coincides with the hydrogenous crusts and nodules reflecting element sorption from bottom waters in phase II (Figure 9). The REY distribution pattern of the 431A-12 sample is similar to that of diagenetic nodules from the Sea of Okhotsk and indicates the deposition of Phase II from pore waters. The significant difference in REY content between the 431A-12 sample and the diagenetic nodules of the Sea of Okhotsk is associated with the higher growth rate of the latter. Figure 9. Shale-normalized (PAAS [47]) REY patterns of Fe-Mn deposits of differing genesis in fraction II: 1-hydrothermal Fe-Mn deposits from the Sea of Japan, [7]; 2-combined data on hydrogenetic Fe-Mn crusts from the North Pacific, [48] and the Central Pacific (our unpublished data); 3-phosphatized Fe-Mn crusts from the Central Pacific (our unpublished data); 4-diagenetic ring-like Fe-Mn nodule from the Sea of Okhotsk, [52]; 5-sample 431-5; 6-sample 431A-12.
We presented the REY contents of Phase III using a logarithmic scale (Figure 10), similarly to Phase II. The hydrogenous crusts in Phase III are enriched in REY, especially in heavy REY, due to prolonged exposure to seawater [4]. Low REY concentrations are observed in hydrothermal crusts in spite of the presence of Fe hydroxides, which are important in the REY accumulation [4,7]. In addition, due to higher growth rates, there is no increase to the heavy REY concentrations [7]. The diagenetic nodules of the Sea of Okhotsk reflect similar conditions of REY accumulation. The REY distribution of phosphatized crust layers indicates the affinity of accumulation with hydrogenous crusts, but under special conditions due the presence of positive Ce and Y anomalies. Such REY distribution is consistent with possible secondary changes of hydrogenous crusts [2]. The REY distribution of the 431-5 sample corroborates the hydrogenous formation of Phase III due to similarities between its limited REY content and the mean values in hydrogenous crusts. The 431A-12 sample pattern is otherwise the same as that of diagenetic nodules of the Sea of Okhotsk, but with a higher REY content. This makes it possible to consider Phase III of the 431A-12 sample as a diagenetic. Figure 10. Shale-normalized (PAAS after [47]) REY patterns of Fe-Mn deposits of differing genesis in fraction III (legend see in Figure 7).
The REY composition in the aluminosilicate phase differs between genetic types ( Figure 11). Distribution plots of REY in Phase IV of samples 431-5 and 431A-12 are similar, but the metal contents differ. The REY distribution pattern of both studied samples is the identical to the hydrogenous Fe-Mn crusts of the NW Pacific crusts.
Based on our results most of the Fe-Mn fragments from the 431 and 431A DSDP holes from the Yōmei Guyot are hydrogenous and fewer of them are diagenetic. We did not find any evidence for any Fe-Mn fragments having a hydrothermal origin. Figure 11. Shale-normalized (PAAS after [47]) REY patterns of Fe-Mn deposits of differing genesis in fraction IV (legend see in Figure 7).

The Initiation of Growth of the Fe-Mn Deposit.
Based on previous studies [17,18], sandy-gravel deposits with a high content of Fe-Mn fragments (80-90%) on the Yōmei Guyot formed during the Quaternary. Consequently, fragments of the youngest brown coal (Pliocene-Quaternary) crust layer should be presented in the bottom part of the core and porous (Miocene) and anthracite (Paleocene-Eocene) layers with Cretaceous relicts should cover it. These layers exist on all guyots in the equatorial Pacific Ocean [8,9]. The Fe-Mn crusts of the northern Pacific Ocean have a two-layer structure. They consist of a thick anthracite (130 mm) layer formed between the Late Oligocene and the beginning of the Late Pliocene, and an inconsistently covered brown-coal (18 mm) layer of the Upper Pliocene-Quaternary period [34,54]. This is the difference in the morphology of the coeval layers of high-latitude crusts from equatorial ones. Notably, the Pliocene-Quaternary brown-coal layer occurs throughout the northern hemisphere in the Pacific Ocean [8,9,54]. We did not discover any fragments of either the Pliocene-Quaternary layer or the most ancient Paleocene-Eocene layer characterized by a high P content (>0.7%, Table 2) in the 431 and 431A DSDP holes. The Fe-Mn deposit is represented only by fragments of the Miocene anthracite layer, reflecting the initiation of growth of the Yōmei Guyot Fe-Mn deposit earlier than the Quaternary period.
Fe-Mn crust growth on guyots takes place when the velocity of the bottom water currents varies from 10 to 30 cm/s [55]. With increasing current velocity crust growth stops and they start to erode. The Fe-Mn crusts on the Emperor Ridge guyots were eroding between 3.97-3.21 Myr [54]. We suggest that initial Fe-Mn crust destruction and post-deposition of sandy-gravel size Fe-Mn deposits in depressions on the Yōmei Guyot top occurred during this period. In addition, Pliocene (6-3 Myr) volcanic activity in the Pacific Ocean might have influenced the destruction of Fe-Mn deposits [56]. Growth of the brown-coal upper layer on Detroit Guyot initiated since 3.21 Myr [48,49], broadly corresponding to global events such as the initiation of northern hemisphere glaciation, the final closing of the Isthmus of Panama, decreasing activity of AABW, and an increase in eolian dust flux [54,[57][58][59]. Thus, the destruction of Fe-Mn crusts and the post-deposition of its fragments on the Yōmei Guyot top occurred in the Late Pliocene.

Genetic Type of the Sandy-Gravel Fe-Mn Deposit.
Investigations into the placer deposits began in the mid-to-late XX Century and focused on the shelf areas of continental borderlands [11,[60][61][62]. During this time, the metallogeny of shelf area placers was studied in detail, but the ocean area was poorly investigated. Only in 1993 were Fe-Mn sandy-gravel ocean seamount deposits proposed to exist with a MnO content up to 25% [23]. This was based on only a few publications [17,18].
Underwater placers on the coasts and continental margins are formed as a result of weathering of source rock, the release of the valuable component, its migration, and its accumulation [11,[60][61][62]. The Yōmei Guyot Fe-Mn deposit has all these properties. The source rock on this Guyot was Fe-Mn crusts and nodules. The valuable component release stage on the Yōmei Guyot corresponds to the separation of Fe-Mn material from the substrate, because Fe-Mn crusts and nodules as a rock comprises the valuable component. The main mechanism of transportation was bottom water currents, and accumulation occurred in depressions. The coarse and non-rounded fragments reflect a short migration of Fe-Mn fragments. Thus, we suggest the definition of the Fe-Mn sandy-gravel deposit of the Yōmei Guyot as an oceanic petrogenic Fe-Mn placer in the vicinity of the source rocks.

Conclusions
The 431 and 431A DSDP cores from the Yōmei Guyot contain layers bearing up to 90% Fe-Mn fragments. The average sizes of the Fe-Mn fragments decrease from 20 mm to 3 mm from the bottom of the core to the seafloor surface, reflecting a decrease in bottom water current velocities. In general, the Fe-Mn deposit is characterized as sandy-gravel.
Based on the morphology, Fe-Mn fragments are divided into two types: dense black with anthracite luster (85% of core material) and slightly porous dark gray to black with a dull luster.
The first type of fragments is composed of vernadite. The Mn content is 28.89%, Fe is 15.57%, Σ Co, Ni, Cu is 8642 ppm, and Σ REY is 3074 ppm. The chemical composition of the bulk samples and the main mineral phases reflects the hydrogenous origin of the Fe-Mn fragments. The source of these fragments was Fe-Mn crusts and pavements of the Yōmei Guyot top. The second type of fragments contains 10 Å manganate and birnessite. The average Mn content is 34.96%, Fe is 8.76%, Σ Co, Ni, Cu is 8404 ppm, and Σ REY is 451 ppm. The results show the Fe-Mn fragments of the second type have a diagenetic origin and their source is ferromanganese nodules formed in specific environments on the guyot top. In contrast to previous investigations [17,18] we did not discover a hydrothermal influence on the formation of the Fe-Mn deposits from the Yōmei Guyot.
The fragmented Fe-Mn deposits of Yōmei Guyot formed between 3.97 and 3.21 Myr. Due to the formation processes of the studied deposit, we define it as a Fe-Mn placer, a new genetic type of ocean mineral deposit.