Mechanism and Influencing Factors of REY Enrichment in Deep-Sea Sediments

: Deep-sea sediments with high contents of rare-earth elements and yttrium (REY) are expected to serve as a potential resource for REY, which have recently been proved to be mainly con-tributed by phosphate component. Studies have shown that the carriers of REY in deep-sea sediments include aluminosilicate, Fe-Mn oxyhydroxides, and phosphate components. The ∑ REY of the phosphate component is 1–2 orders of magnitude higher than those of the other two carriers, expressed as ∑ REY = 0.001 × [Al 2 O 3 ] − 0.002 × [MnO] + 0.056 × [P 2 O 5 ] − 32. The sediment P 2 O 5 content of 1.5% explains 89.1% of the total variance of the sediment ∑ REY content. According to global data, P has a stronger positive correlation with ∑ REY compared with Mn, Fe, Al, etc.; 45.5% of samples have a P 2 O 5 content of less than 0.25%, and ∑ REY of not higher than 400 ppm. The ∑ REY of the phosphate component reaches n × 10 4 ppm, much higher than that of marine phosphorites and lower than that of REY-phosphate minerals, which are called REY-rich phosphates in this study. The results of microscopic observation and separation by grain size indicate that the REY-rich phosphate component is mainly composed of bioapatite. When ∑ REY > 2000 ppm, the average CaO/P 2 O 5 ratio of the samples is 1.55, indicating that the phosphate composition is between carbonate fluoro-apatite and hydroxyfluorapatite. According to a knowledge map of sediment elements, the phosphate component is mainly composed of P, Ca, Sr, REY, Sc, U, and Th, and its chemical composition is relatively stable. The phosphate component has a negative Ce anomaly and positive Y anomaly, and a REY pattern similar to that of marine phosphorites and seawater. After the early diagenesis process (biogeochemistry, adsorption, desorption, transformation, and migration), the REY enrichment in the phosphate component is completed near the seawater/sediment interface. In the process of REY enrichment, the precipitation and enrichment of P is critical. According to current research progress, the REY enrichment is the result of comprehensive factors, including low sedimentation rate, high ∑ REY of the bottom seawater, a non-carbonate depositional environment, oxidation conditions, and certain bottom current conditions.

Studies reveal that phillipsite, clay minerals, Fe-Mn oxyhydroxides, and phosphate components are all REY carrier phases [10,18]. These phases prefer to accumulate in slow sedimentary environments, and so it is difficult to define the contribution of various components. Recent studies have further confirmed that the ∑REY of clay and phillipsite themselves are not high, with ∑REY generally not exceeding 256 ppm [19,20]. Under the weak influence of phosphate components (P2O5 < 0.25%), ∑REY has a certain positive correlation with Al2O3 content. However, in REY-rich deep-sea sediments, the contribution of phillipsite or clay minerals to REY enrichment is easily masked by the phosphate component.
Fe-Mn oxyhydroxides have strong adsorption capacity, and hydrogenetic Fe-Mn nodules can greatly enrich REY (∑REY, excepting Ce, generally reaches 2000 ppm) [21]. The REY enrichment of metalliferous sediment in the Southeastern Pacific was initially thought to be related to hydrothermal processes [4]. However, there is no significant correlation between ∑REY and the Fe and Mn contents in deep-sea sediments. Moreover, the ∑REY of Fe-Mn micronodules in deep-sea sediments is relatively low (246-333 ppm) [20]. The ∑REY of hydrothermal Fe-Mn deposits is 1-2 orders of magnitude lower than that of hydrogenetic Fe-Mn deposits; rapid hydrothermal processes have an obvious dilution ef-fect on REY [22]. Recent comparison of fine-grained analyses of samples from the Southeast Pacific and central North Pacific show that apatite is the main host of REY, even in metalliferous deep-sea sediments near the East Pacific Rise [23].
The results of in situ analyses of bioapatite and chemical leaching of phosphate component from deep-sea sediments showed that the ∑REY of the phosphate component was 1-2 orders of magnitude higher than that of other REY carriers [3,12,19,24]. Among the major oxides of deep-sea sediment samples, ∑REY and P2O5 always maintain a strong positive correlation. The results show that the REY patterns of deep-sea sediments are similar to those of marine phosphates. The higher the content of ΣREY, the closer their REY patterns are [6,12,25]. The P in deep-sea sediments mainly exists in the form of REYrich phosphate, and changes in the P content have a significant influence on the content and patterns of REY [12]. In summary, these studies have indicated that the high ΣREY of deep-sea sediments is closely related to the phosphate component. However, not all P forms REY-rich phosphates.
In addition to the above studies, lots of studies have discussed the mechanism of REY enrichment in deep-sea sediments from the perspectives of the marine P cycle, sedimentation rate, sediment types, redox conditions, water ∑REY, and bottom ocean current [4,6,[26][27][28]. However, there is a lack of quantitative research and systematic review of the contribution of phosphate components to REY enrichment, and in-depth discussion on the mechanism of phosphate components on REY enrichment is needed. The contribution of phosphate components to REY enrichment and the study of REY enrichment mechanisms in deep-sea sediments are of great significance for determining regions with high resource potential and for revealing the potential relationships between the geodynamic system and marine mineral resources. In this study, on the basis of compiling a large volume of chemical data and related environmental parameters of deep-sea sediment sites, the temporal and spatial distributions of REY-rich deep-sea sediments were summarized. Then the contribution of phosphate components to REY in deep-sea sediments was determined qualitatively and quantitatively, and the characteristics of phosphate components were described from chemical composition and microscopic identification results. Finally, the enrichment process of REY and the influencing factors of REY enrichment in deep-sea sediments were considered.

Types and Distribution of REY-Rich Deep-Sea Sediments
Based on the analysis of more than 2000 samples from 78 cores in the Pacific Ocean, Kato et al. (2011) found that most deep-sea sediments in the Southeast Pacific and central North Pacific have high ΣREY (400-2230 ppm for ΣREY and 70-430 ppm for ΣHREY). These REY-rich deep-sea sediments are comparable with ion-adsorption clay-type deposits of South China, and may be potential REY resources [4]. At present, the Western Pacific sediments have the most resource potential, and the highest ΣREY in the sediments off Minamitorishima Island is 7974 ppm [1]. Tanaka et al. (2020) further classified the REYrich deep-sea sediments into "ordinary REY-rich mud" (ΣREY = 400-2000 ppm), "highly REY-rich mud" (ΣREY = 2000-5000 ppm) and "extremely REY-rich mud" (ΣREY > 5000 ppm). Owing to the remarkably high content of REY (especially of HREY), extremely/highly REY-rich deep-sea sediments have attracted great attention as an unconventional and promising mineral resource for REY [6].
We collected the chemical data of 6088 samples from 177 sites in the Pacific Ocean, Indian Ocean, and Atlantic Ocean (Table S1), and calculated the ∑REY of surface sediments, the average ∑REY of surface layers within 3 m, the maximum ∑REY, sedimentation rate, bottom water oxygen concentration, and seafloor megafaunal biomass at each site (Table S2). The water depths of sites with ∑REY more than 1000 ppm are mainly concentrated between 4180 and 5160 m, indicating that REY-rich deep-sea sediments are mainly developed below the carbonate compensation depth (CCD). There are 18 cores where the average ∑REY of surface layers within 3 m exceeds 1000 ppm; these mainly developed off Minamitorishima Island of the Western Pacific Ocean, the central North Pacific Ocean, the Southeast Pacific Ocean, and the central Indian Ocean basin. The surface sediments are relatively easy to trace back to the relevant marine environment, which can provide an important reference for the REY mechanism. Current studies have confirmed that P is the basis for the REY enrichment in deep-sea sediments. However, the P enrichment in deep-sea sediments cannot be simply understood as the high biogenic potential of surface seawater. For example, the REY-rich sediments in the Western Pacific are located on the edge of a typical tropical oligotrophic sea area. The distribution of global chlorophyll content (Figure 1) is very close to that of average surface productivity [29]. According to current data, REY-rich deep-sea sediments tend to develop in relatively low chlorophyll sea areas (Figure 1), indicating relatively low biological productivity. The bottom water oxygen concentration far from the coast is generally high [29], indicating a condition of relative oxidation. Therefore, it is difficult to understand the reasons for the REY contents of sediments from current redox data. However, the sedimentation rates of these sites are quite different, and the REY-rich deep-sea sediments are obviously developed in places with relatively low sedimentation rates.
There are 16 sites off Minamitorishima Island with ∑REY of more than 4000 ppm; these are considered as the most resource potential area. At present, there is no economic feasibility evaluation of REY resources in deep-sea sediments, and there is also a lack of excavation technology. Surface sediments are always used in REY beneficiation and metallurgy experiments, and so are more representative of enrichment of REY in surface layers. However, the ∑REY of surface samples off Minamitorishima Island are relatively low, with the average ∑REY of samples within 3 mbsf (meters below sea floor) exceeding 1000 ppm from only six sites (Table S2). The ∑REY of the samples at 2.28 mbsf and 2.78 mbsf of site MR15-E01-PC05 are as high as 5619 and 6323 ppm, respectively, which increases the average ∑REY of the samples within 3 mbsf to 2299 ppm [6].
The ∑REY in Indian Ocean sediments is generally low; only the average ∑REY of samples within 3 mbsf at site DY30GC02 is >1000 ppm [7]. Samples with ∑REY of up to 1113 ppm exist at site ODP-213, but they are located at 119 mbsf [30]. Menendez et al. (2017) reported 35 samples of 32 sites recovered from the Atlantic Ocean, on a transect along ~24° N, most of which lacked surface sediment data. Samples from the Atlantic were too few and had the lowest ∑REY, with only two sites above 400 ppm and a maximum of 513 ppm [31].
The ∑REY in Pacific sediments are systematically higher than those in the Indian Ocean ( Figure 2). The data for samples with ages of older than 60 Ma are limited to just 596 sites, with ∑REY generally >1000 ppm. More REY-rich deep-sea sediments have been recovered from the Pacific Ocean, but do not show strong geographical clustering. The sequence of ΣREY enrichment in different types of deep-sea sediments is as follows: deepsea clay > hydrothermal sediment > siliceous clay > siliceous ooze > calcium ooze [32]. The petrographic characteristics of REY-rich sediments are P-rich metalliferous sediment near the East Pacific ridge and P-rich phillipsite/clay components in other places [4,9,26,30,31].

Contribution of Phosphate Component to REY Enrichment
In recent years, there is increasing evidence to show that the high ∑REY of deep-sea sediments is closely related to the phosphate component. Firstly, the results of correlation analysis show that only P content maintains a good correlation with ∑REY compared with Mn, Fe, and Al ( Figure 3). Moreover, most sites from different places show that ∑REY variation is sensitive to P content in vertical section [1,9,10,13,15,33,34]. For example, each layer of site KR13-02-PC05 shows ∑REY to have strong correlations with P2O5 and CaO content, which is quite different from the relationship with other major oxides [9]. The P and Ca contents correlate positively (Figure 4), and the CaO/P2O5 ratio is consistent with that of apatite [20,35]. The ratio of CaO/P2O5 decreases gradually as ∑REY increases from low to high. When ∑REY is >4000 ppm, the ratio of CaO/P2O5 indicates that the phosphate component is between carbonate fluorapatite and hydroxyfluorapatite ( Figure 5). Previous studies have shown that carbonate fluorapatite is the main P-bearing mineral in the oceans, and the CaO/P2O5 also indicates that it is the main phosphate mineral in the deepsea sediments [18]. With increasing ∑REY, the proportion of hydroxyfluorapatite in deep sea sediment increases. Microscopic identification has shown an abundance of bioapatite (e.g., fish teeth) in REY-rich deep-sea sediments [3,19,36].   Secondly, the REY patterns of the REY-rich deep-sea sediments are very similar to those of marine phosphorites, showing an obvious Ce negative anomaly, Y positive anomaly, and no Eu anomaly ( Figure 6A). The enrichment of REY in deep sea sediments is the result of the comprehensive action of aluminosilicates, phosphates, and other components [12,18]. The higher the ∑REY, the closer the REY patterns are to the phosphate component. Only samples with low ∑REY from site KR13-02 PC05 may have flat or positive Ce anomalies ( Figure 6A), indicating that the REY characteristics of aluminosilicates are exposed under low phosphate component contents. Regardless of whether the ∑REY is high or low, the REY pattern of phosphate leaching from deep-sea sediments shows a significant negative Ce anomaly and positive Y anomaly ( Figure 6B). These characteristics indicate the importance of phosphate components in REY-rich deep-sea sediments for REY [12].  [9]. (B) Phosphate leaching from deepsea sediments; data from [12]. NASC normalizing values are from [37]. Bottom seawater data is from [38]. Seamount phosphorites data is from [39].
Thirdly, if samples with P2O5 higher than 0.25% are excluded, the ∑REY of deep-sea sediment is not higher than 400 ppm [12]. There are 2679 samples with P2O5 ≤ 0.25%, and the linear regression equation of ∑REY and Al2O3 can be obtained as ΣREY = 0.001 × Al2O3 + 48.6, R 2 = 0.77. That is, under weak influence of the phosphate component (P2O5 < 0.25%), ∑REY has a certain correlation with the Al2O3 content. Although the contributions of phillipsite and clay minerals to REY enrichment are low, they do exist. However, this contribution is easily masked by that of the phosphate component.
Last but not least, the ∑REY of the phosphate component in deep-sea sediments is extremely high. For example, the ∑REY of in situ analyses of bioapatite samples is generally n × 10 4 ppm [3,19,24,36], and the average of ∑REY in chemical separation results of the phosphate component reach 19,260 ppm [12]. Chemical phosphate precipitation involves fine particles that are difficult to analyze directly. According to the ∑REY/P2O5 ratio of bioapatite and whole sediment samples, authigenic P has a higher ∑REY than bioapatite. In a word, the ∑REY of the phosphate component reaches n × 10 4 ppm (termed REYrich phosphate in this study), which is much higher than that of marine phosphorites and lower than that of REY-phosphate minerals (e.g., monazite). Via quantitative analysis of process mineralogy, Liu et al. (2015) found that almost all REY exists in apatite [40]. In conclusion, the ∑REY of the phosphate component is ~1-2 orders of magnitude higher than that of other carriers; changes in the P content are very important for concentration and distribution REY in deep-sea sediments.

Qualitative and Quantitative Understanding of Phosphate Components
Knowledge maps of element correlations can show the kinship of elements. Element correlation coefficient analysis was carried out on 1200 samples from 49 sites off Minamitorishima Island, and the results are shown in Table S3. A knowledge map was constructed with elements or oxides as nodes and with correlation coefficients of greater than 0.5 as edges (Figure 7).  (Figure 7).
The IBM SPSS Statistics 23 software was used to carry out multiple linear regression analysis on the compiled data; four quantitative relationship models were obtained (Table  1) [4,12]. The REY patterns are characteristics of marine phosphorites, indicating that the contribution of phosphate components to REY enrichment is consistent with our quantitative calculation results. Table 1. Multiples linear regression model between ΣREY and major oxides (data sources as in Figure 1). A small amount of data shows outliers with a low ∑REY/P2O5 ratio; these represent the "REY-poor phosphate component". For example, there are samples with P2O5 of 6.05% from site 573B, but with ∑REY of only 665 ppm (i.e., ~3000 ppm lower than the regression line; Figure 9). It is more meaningful to understand the relationship between ∑REY and REY-rich phosphate components by excluding the data of the REY-poor phosphate component. Through data screening and iterative calculation of new data, this study uses ∑REY = (0.058 ± 0.018) × [P2O5] + (0.001 ± 0.001) × [Al2O3] + (20 ± 65) as the basis for discriminating between the high REY, REY-rich phosphate, and REY-poor phosphate components ( Figure 9). On this basis, 169 data points are REY-poor phosphate (2.8% of the total). A linear regression analysis of ∑REY with P2O5 and Al2O3 was carried out after excluding the data of REY-poor phosphate; model 4 was obtained, and the R 2 is greatly improved. In this study, high REY phosphate is actually a special kind of REY-rich phosphate, which belongs to abnormal data in statistics. It can be interpreted as the extra contribution of REY-rich components (such as Fe-Mn micronodules) to REY enrichment, resulting in higher REY/P2O5 ratio.

REY Enrichment Complete near the Seawater/Sediment Interface
Seawater Nd isotopic ratios are distinct in each of the modern ocean basins because of inputs of regionally distinct Nd isotopes from the continents and the short residence time (~300 to ~600 year) of Nd in the ocean relative to the ocean mixing time (~1500 year) [41,42]. Microscopic observations show that the REY-rich deep-sea sediments are rich in bioapatite particles, mainly composed of fish teeth and bone fragments [1,19,24,43]. Bioapatite can record Nd isotopic information of bottom seawater at the time of deposition, and can even be used as a water mass tracer [41]. The hydroxyapatite of living fish teeth contains parts per billion (ppb) levels of Nd, but the hydroxyfluorapatite of fossil teeth typically contains 100-500 ppm Nd [44]. Living phosphate bones and tissues have extremely low ∑REY. The bones and teeth after biological death will undergo crystal-chemical changes in the early diagenesis process under the influence of the external environment, and the converted bioapatite will have a highly enriched ∑REY [23,45,46]. This supports the idea that REY are adsorbed or incorporated during early diagenesis while pore fluids are still in exchange with seawater. As Nd isotopes vary with depth in the water column [47,48], the isotopic ratio of Nd incorporated in vivo should vary with the depth habitat of the fish species [49].
Some studies suggest that REY enrichment by the phosphate component in the sediment column may be accomplished at a deep depth [36,46]. However, the REY content and relative content data show that REY enrichment of the phosphate component mainly occurs near the seawater/sediment interface. According to the analysis results of bioapatite particles selected from different layers, ∑REY varies greatly. The content of ∑REY in bioapatite does not increase with depth or burial time [3,24,36]. Although the ∑REY of bioapatite fluctuate, it is obviously not related to the sedimentary layers. As mentioned above, ∑REY mainly depends on the content of P in deep-sea sediments (Figure 3). Martin and Scher (2004) studied Nd isotopes and contents in fossil fish teeth and suggested that teeth preserve an early diagenetic signal that represents the Nd from the bottom water [44]. Fe-Mn oxyhydroxide, organic debris, and other particulate matter that settled to the seafloor have pre-enriched REY, which then easily released the REY at the sediment-seawater interface, providing the phosphate component for the enrichment of REY [8].
The ∑REY in seawater vary with depth and location, and the REY patterns are relatively stable. However, the REY patterns of pore water are obviously variable, especially for deep layer pore water [28]. The REY patterns of REY-rich deep-sea sediments are similar to those of bottom seawater, but quite different from those of pore water [28]. REY patterns preserved in fossil fish teeth indicate that teeth from deep-sea settings record seawater compositions [19,24,36]. These results confirm that REY enrichment during early diagenesis is still in communication with seawater.
In particular, REY is highly particle reactive and has a low content in solution. As a result, the ∑REY in pore fluids is extremely low, and is insufficient to provide a material basis for the REY enrichment of the phosphate component [44]. Both high and low temperature studies consistently illustrate that Nd isotopes are resistant to alteration during post-magmatic water-rock interactions, nor are they mobilized during metamorphic mineralization of Precambrian granitic gneisses [50,51]. One explanation for the resistance of Nd isotopes to alteration is the low ∑REY in circulating fluids relative to the contents in the minerals themselves [44]. Data from sites 1090 [44] and P10 (unpublished) in the Pacific Ocean also demonstrate that fish teeth record short-lived, rapid variations in the Nd isotopic composition of seawater. These rapid variations suggest that the isotopic signal has not been blurred by diagenetic alteration. These data imply that teeth incorporate or adsorb REY during very early diagenesis and that the content of the REY does not alter with time and/or burial.
Thus, the fact that the ∑REY do not increase with depth suggests that there is no source of REY for pore fluids. Instead, REY are introduced to the system from seawater at the seawater/sediment interface. During burial, much of REY is removed from pore fluids by fish teeth, but little is added through diagenesis.

Influencing Factors of REY Enrichment
Not all of the marine phosphate component has high ΣREY; in particular, the shallow water phosphate component generally has low ΣREY. For example, the ∑REY of continental marginal bioapatite is generally low, reflective of the relatively fast sedimentation rate. The REY enrichment by the phosphate component mainly occurs within seafloor sediments; the REY patterns are complicated and diverse, and differ from the enrichment process of REY by deep-sea sediments [46,52,53]. Moreover, the ∑REY of shallow-water phosphorite is not high. A comparison study of 75 samples from around the world shows that the ∑REY of marine phosphorites is 360 ppm on average, while the P2O5 content is 25.1% on average [39]. In addition, according to the new division mentioned above, 2.8% of the samples in deep-sea sediments also contain REY-poor phosphate. P enrichment is the basis of REY enrichment in deep-sea sediments, but not all of them can eventually form REY-rich phosphate. Compared with the material composition and formation environment of the phosphate component in shallow water, the REY enrichment mechanism of deep-sea sediments can be understood as having three stages: P enrichment, formation of REY-rich phosphate, and REY-rich phosphate re-enrichment. The relationships among P accumulation, sedimentation rate, redox conditions, ∑REY of bottom seawater, sediment type, and bottom currents on the REY enrichment of deep-sea sediments is illustrated in Figure 10. REY enrichment in deep-sea sediments firstly depends on the precipitation and enrichment of P. The P enrichment includes P adsorption by organic matter and Fe-Mn particles, as well as the chemical precipitation of phosphate from seawater and pore water. After diagenesis, most of the P is retained in the form of calcium phosphate [54]. The content of microfish tooth bone is an important indicator of REY-rich deep-sea sediments. Therefore, some studies directly suggest that bioapatite is the key to REY enrichment of deep-sea sediments [3,19]. The distribution of REY-rich deep-sea sediments on the seafloor and chlorophyll in the surface seawater shows that REY-rich deep-sea sediments mainly develop in low-and middle-latitude areas with low biological primary productivity. The trajectory of a REY-rich sediment site since 30 Ma shows that it is located in a similar marine environment, which indicates that P enrichment in REY-rich sediment does not occur in areas with high organic P deposition flux (Figure 1). The P enrichment in deep-sea sediments requires a relatively low content of other components, especially a lack of rapidly accumulating calcareous and siliceous biological components. The main components of the P-rich deep-sea sediments found to date are clay or phillipsite clay, indicating extremely slow sedimentation rate environments [33]. According to our investigation, it is true that REY-rich deep-sea sediments are developed in places with relatively low sedimentation rates (Table S2).
On the one hand, a low sedimentation rate is beneficial to the accumulation of the phosphate component in deep-sea sediments per unit volume; on the other hand, it provides sufficient time for the phosphate component to enrich the REY. If the only source of ΣREY to the fish teeth is seawater, one might expect the teeth in contact with seawater for the longest time (i.e., the slowest sedimentation rates) to have the highest ΣREY [44]. Yasukawa et al. (2016) quantitatively showed that seawater can contain the flux of REY precipitation required to explain the observed very high ΣREY (>1000 ppm) in bulk REY-rich deep-sea sediments with a sedimentation rate of less than ~0.5 m/Myr [8]. Kato et al. (2011) used a diagram of ∑REY vs. sedimentation rate to visually show that REY-rich deep-sea sediments develop in low sedimentation rate environments [4]. When the sedimentation rate is sufficiently low, the phosphate component can be exposed to the surface for a long time to obtain seawater REY [5]. Martin and Scher (2004) compared the Nd content and sedimentation rate of the fish tooth fossils in four sedimentary sites in the Pacific. There was a certain correlation between Nd content and sedimentation rate, and the tooth with the highest Nd developed at the lowest sedimentation rate [44]. Deep-sea sediments with low sedimentation rates often contain a large number of phillipsite particles, which is also consistent with a slow sedimentation rate [55,56]. However, phillipsite itself has a low ∑REY, which explains that the REY-rich deep-sea sediments are often accompanied by phillipsite.
Owing to dissolved oxygen supplementation by bottom currents, the oxygen fugacity of the deep-sea environment is relatively high. Surface deep-sea sediments have high oxygen fugacities, and the probability of remineralization is low [28]. Continental margin sediments generally display relatively narrow redox conditions (e.g., only 1 mm-thick of sediments in Buzzard bay) [57]. The high REY enrichment in deep-sea sediments requires a continuous supply of REY by the bottom seawater, and reduction conditions hinder the recharge of REY from overlying seawater to the sediment-seawater interface. Low sedimentation rate, low organic components, and relatively high oxygen fugacity are interdependent. Oxidation conditions can promote P deposition and REY enrichment [28].
The ∑REY in the ocean water column tends to increase with water depth, and the ∑REY of deep-sea water is generally several times that of surface sea water [28,38,58]. The REY-rich sediments are mainly distributed in the deep-sea environment below CCD, which prevents the rapid accumulation of dilutive components with low ∑REY (e.g., biogenic carbonate and silica) [8]. Phosphorites developed in shallow water environments are located in carbonate sedimentary environments, and phosphate roots directly replace carbonate, which will inevitably prompt REY to enter the phosphate component, resulting in a low ∑REY/P2O5 ratio of seamount phosphorites [12]. For example, Pan et al. (2004) studied the microscopic characteristics and carbon isotope characteristics of seamount phosphorites, and found that carbonates and other calcium-bearing materials had been replaced by phosphates to a varying degree. Bioapatite such as conodont developed in shallow water limestone also demonstrates that phosphates are of replacement origin, with a low ΣREY/P2O5 ratio [59].
The erosion of deep-sea sediments may play an important role in the formation of highly REY-rich sediment layers. Tanaka et al. (2020) produce a chemostratigraphic scheme for 1240 sediment samples from 49 piston sites collected within the Minamitorishima exclusive economic zone. The chemostratigraphy reveals that highly REY-enriched layers formed at least three times, and that almost all of these layers were accompanied by erosion during deposition [6]. The grain size distributions of bulk sediments from the REY-enriched layers were bimodal, with peaks at fine (~4 μm) and coarse (~40-80 μm) sizes. Calcium phosphate and phillipsite were the major components of the coarse-grained portions of these REY-enriched layers [43]. Enhancement of bottom currents can produce selective accumulation of coarse biogenic Ca phosphate (BCP) grains enriched in REY, resulting in the formation of REY-enriched layers [6,43]. However, it should be noted that if the bottom current was sufficiently strong during deposition, it could have hampered the accumulation of BCP grains or even eroded the sediments as a whole [6]. The ∑REY of living phosphate bones and teeth is low, but increases rapidly after death. The content of ∑REY in bioapatite does not increase with depth. The REY patterns of REY-rich deep-sea sediments are similar to those of bottom seawater, but quite different from those of pore water. The ∑REY in pore fluids is extremely low, and is insufficient to provide a material basis for REY enrichment. These phenomena indicate that the enrichment of REY is mainly completed near the seawater/sediment interface. 4. REY enrichment is the result of comprehensive action. Low sedimentation rate is beneficial to the accumulation of P in unit volume sediments and the recharge of REY by bottom seawater. The high ∑REY in bottom seawater provides a high ∑REY background value for the enrichment of the seafloor phosphate component. Non-carbonate depositional environments avoid the formation of REY-poor phosphates by replacing carbonate with phosphate. In addition, the sorting effect of the bottom current on deep-sea sediments further accumulates REY-rich phosphates to form REYrich deposits.

Conclusions
Supplementary Materials: The following are available online at www.mdpi.com/2075-163X/11/2/196/s1, Table S1: The content of Major element (%) and REY (ppm) of deep-see sediments complied from global Ocean, Table S2: ∑REY of the surface, average value of surface layer within 3 m and maximum value of site sample, Table S3: Element correlation coefficient of deep-sea sediments off Minamitorishima Island. Data is is calculated from [6].