Quaternary radiolarian biostratigraphy in the subarctic northeastern Pacific (IODP Expedition 341 Site U1417) and synchroneity of bioevents across the North Pacific

Expedition 341 of the Integrated Ocean Drilling Program (IODP) retrieved sediment cores spanning the time interval between the Pleistocene and Miocene from the southern Gulf of Alaska. Onboard Pleistocene radiolarian biostratigraphy is hereby refined by increasing the sampling resolution. The 178 core samples from the upper 190 m CCSF-B (Composite Core Depth Scale F-B) of Site U1417 contained faunal elements similar to the northwestern Pacific; for example, the three biozones in the northwestern Pacific (i.e., Eucyrtidium matuyamai, Stylatractus universus and Botryostrobus aquilonaris) were also recognized in the Gulf of Alaska, spanning 1.80–1.13 Ma, 1.13–0.45 Ma, and the last 0.45 Myr, respectively. Based on the age model that we used in this study and the shipboard paleomagnetic reversal events, the first occurrences (FOs) of Amphimelissa setosa and Schizodiscus japonicus in the northeastern Pacific were preliminarily determined to be 1.48 and 1.30 Ma, respectively. The last occurrence (LO) of Eucyrtidium matuyamai and the FO of Lychnocanoma sakaii, both well-established bioevents in the northwestern Pacific, were dated at 0.80 and 1.13 Ma, respectively. The LO of E. matuyamai is a synchronous event at 1.05 ± 0.1 Ma in the North Pacific, while the FOs of A. setosa and S. japonicus at 1.48 and 1.30 Ma, respectively, are significantly older than what has been found elsewhere.

olarians that were present, the Pleistocene radiolarians were better preserved and more abundant (Jaeger et al., 2014).Preliminary biostratigraphic analysis at low resolution (> 10 m sampling interval) conducted during Expedition 341 revealed the existence of numerous barren intervals in cores retrieved from the site (Jaeger et al., 2014).We reexamined these core data and propose a refined Pleistocene radiolarian biostratigraphy for the northeastern Pacific using a markedly higher sampling resolution.
Therefore, widely distributed bioevents in the North Pacific were examined in this study.The last occurrences (LOs) of Amphimelissa setosa (Cleve, 1899) and Schizodiscus japonicus Matsuzaki andSuzuki, 2014 (Spongodiscus sp. in Matul et al., 2002), are all well-known, recorded at 0.07 and 0.29 Ma, respectively, in the Sea of Okhotsk (Matul et al., 2002).However, their first occurrences (FOs) in the North Pacific are poorly understood; Matul and Abelmann (2005) recorded the FO of A. setosa in the Sea of Okhotsk at 1.03 Ma, and Ikenoue et al. (2016) assigned the ages of the FOs of A. setosa and S. japonicus in the Bering Sea to be 0.99 Ma.Clarifying these bioevents in other regions is very important in facilitating discussions on bioevent synchroneity and/or diachroneity.During Expedition 341, Jaeger et al. (2014) identified several clearly defined Pleistocene paleomagnetic reversal events at Site U1417.Specifically, these included the reversal events of C1n (bottom, B), C1r.1n (top, T), C1r.1n (B), C1r.2n (middle, M) and the C2n (T), indicating 0.78, 0.98, 1.07, 1.17 and 1.78 Ma, respectively (Ogg, 2012).Assuming that the sedimentation rates are linear between reversal events, these paleomagnetic reversal events can be used to estimate the preliminary ages of the FOs of A. setosa and S. japonicus in the northeastern Pacific.After refining the radiolarian biostratigraphy for the Pleistocene, we estimated the preliminary age of the FOs for A. setosa and S. japonicus for the first time in the northeastern Pacific and discussed the synchroneity and potential suitability of these bioevents as biostratigraphic proxies.

Material and methods
We analyzed samples collected from Site U1417 on Expedition 341 of the IODP at 56 • 57.59 N and 147 • 6.59 W (Jaeger et al., 2014).The seafloor at the site, which is located in the Surveyor Fan of the Gulf of Alaska (Fig. 1; Jaeger et al., 2014), was calculated to be 4198 m below sea level (Jaeger et al., 2014).The sampling site is close to Site 178 of the Deep-Sea Drilling Project (ca.1.5 km) (Jaeger et al., 2014).According to Jaeger et al. (2014) (ver. 4) (Schlitzer, 2016).
160 m of CCSF-B, but intervals of diatom oozes are common (Jaeger et al., 2014).Between 160 and 260 m CCSF-B, the lithology is composed of gray to greenish-gray mud with interbeds of sand and silt (Jaeger et al., 2014).For radiolarian analysis, a total of 178 samples covering the upper 190 m CCSF-B from Site U1417 were freeze-dried and then treated with diluted hydrogen peroxide (15 %) and hydrochloric acid (15 %) to remove organic and calcareous matter.Undissolved residues sieved over a 45 µm screen were then mounted on 22 × 40 mm cover glasses for microscopic analysis.These samples contain such rare radiolarians that we were only able to examine an average of 296 radiolarian specimens per sample.We then calculated the relative abundance of biostratigraphically important radiolarian species to clarify the radiolarian biostratigraphy at this site for the Pleistocene.
The sedimentation rates for the depth interval between 111 and 186 m CCSF-B at Site U1417 were constrained relying only on the paleomagnetic reversal events (Jaeger et al., 2014) (Fig. 3, Table 3).Assuming that sedimentation rates are linear between the paleomagnetic reversal events reported by Jaeger et al. (2014), the sedimentation rates range from 7.0 to 10 cm kyr −1 from 111 to 150 m CCSF-B, decreasing to 5.9 cm kyr −1 from 150 to 186 m CCSF-B (Table 3).For this time interval, the LOs of E. matuyamai (1.25 ± 0.15 Ma) and A. robustum (1.25 ± 0.1 Ma) are consistent with the sedimentation rates defined based on the paleomagnetic reversal events (Fig. 3, Tables 2 and 3).Since there were no paleomagnetic reversal events for the depth interval between 0 and 111 m CCSF-B, the sedimentation rates were based only on radiolarian bioevents (this study) (Fig. 3, Table 3).Based on these bioevents, the sedimentation rate of 63.6 cm kyr −1 is considered to be exceptionally high during the last 0.03 Mky.For older intervals, the sedimentation rates ranged between 12.9 and 27.5 cm kyr −1 , except between 60 and 62 m CCSF-B, when a very low sedimentation rate of 2.0 cm kyr −1 was recorded.

Radiolarian zones
Based on the recorded radiolarian taxa (illustrated in Plate 1) and their bioevents, we were able to apply the biostratigraphic scheme proposed by Kamikuri et al. (2004) and identify three radiolarian interval zones: the Eucyrtidium matuyamai interval zone, the Stylatractus universus interval zone and the Botryostrobus aquilonaris interval zone (Fig. 2).
Eucyrtidium matuyamai interval zone, Hays (1970) Definition: The base of this zone could not be determined due to barren intervals from 186 m CCSF-B.The top of this zone is defined by the LO of E. matuyamai (Fig. 2 and Table 2).
We were therefore able to accurately recalculate the age of the LO of E. matuyamai following the sedimentation rates described above (Fig. 3, Tables 3 and 4).The revised age of the LO of E. matuyamai for the northeastern Pacific is thus 1.13 ± 0.01 Ma.
Correlation and ages: The base of this zone is defined by the FO of E. matuyamai, which is typically located near the C2n (Olduvai) paleomagnetic reversal event (e.g., Kamikuri et al., 2004Kamikuri et al., , 2007;;Matsuzaki et al., 2014b).The top of this zone is defined by the LO of E. matuyamai placed between paleomagnetic reversal event C1r.2n (M) (Cobb Mountain) and the top of C1r.1n (T) (Jaramillo).
Considering the revised age of the LO of E. matuyamai, this zone spans from 1.8 to 1.13 Ma.Stylatractus universus interval zone, Hays (1970) Definition: The base of this zone is defined by the LO of E. matuyamai and the top is defined by the LO of S. universus (Fig. 2 and Table 2).
Important datums: We identified the FO of L. sakaii in this zone as being between samples 341-U1417A, 13H-2, 50-52 cm and U1417C, 14H-1, 50-52 cm at a median depth of 113.18 m CCSF-B (Fig. 3 and Table 4).Based on the sedimentation rates shown in Fig. 3, the FO of L. sakaii is dated at 0.8 ± 0.1 Ma, in Site U1417.
Correlation and ages: The base of this zone is defined by the LO of E. matuyamai at 1.13 Ma.According to Kamikuri et al. (2004), the top of this zone is defined by the LO of S. universus (0.45 Ma), which is placed into the diatom Proboscia curvirostris zone of Barron et al. (1995), and within the C1n (Brunhes) normal polarity epoch.Therefore, this zone spans from 1.13 to 0.45 Ma.
Botryostrobus aquilonaris interval zone, Hays (1970) Definition: The base and top of this zone are defined by the LO of S. universus and the top of the core, respectively (Fig. 2 and Table 2).
Correlation and ages: The base of this zone as described above is defined by the LO of S. universus.According to Kamikuri et al. (2004), this zone corresponds to the Neodenticula seminae zone (diatom biostratigraphy) and spans the last 0.45 Myr.

Discussion
The ages of formerly known radiolarian bioevents were estimated using different geochronologic proxies, including biostratigraphy (e.g., diatoms and calcareous nannofossils), paleomagnetic reversal events and/or chemostratigraphy.In this study, for the bioevents recorded in the Botryostrobus aquilonaris interval zone, we do not have a geochronologic proxy suitable for temporal calibration of the recorded bioevents (LOs of A. acquilonium, A. setosa, L. sakaii, S. japonicus and an abundance peak of L. sakaii).However, five paleomagnetic reversal events were identified in both the Stylatractus universus and Eucyrtidium matuyamai zones (Jaeger et al., 2014) (Fig. 3).These reversal events enable us to calculate the ages of the LO of E. matuyamai as 1.13 Ma, and the FOs of A. setosa, S. japonicus, and L. sakaii as 1.48 Ma, 1.30 Ma, and 0.80 Ma, respectively.These ages estimated for the S. universus and E. matuyamai zones allow us to clarify the synchroneity and diachroneity of bioevents between the northwestern and northeastern Pacific.
The FO and LO of E. matuyamai, as well as the taxonomic description of this species, were first reported by Hays (1970), who proposed that E. matuyamai evolved from Eucyrtidium calvertense Martin (1904) near the paleomagnetic reversal event of C2n (T) (1.778 Ma; Ogg, 2012).The distribution of the ancestor E. calvertense was south of 40 • N in the North Pacific, which differs from the more northern distribution of E. matuyamai (Hays, 1970).The appearance of E. matuyamai can be attributed to allopatric speciation resulting from the invasion of water masses into the subarctic North Pacific region at the C2n (T), which resulted in isolation of some E. calvertense populations and the evolution of E. matuyamai (Hays, 1970).The FO of E. matuyamai has been placed close to C2n (T) at various locations in the northwestern Pacific (e.g., Hays, 1970;Motoyama, 1996;Kamikuri et al., 2004Kamikuri et al., , 2007;;Matsuzaki et al., 2014b), implying that levels of synchronicity were high from a stratigraphic point of view, with the event dated at ca. 1.8 Ma in most of the literature cited above.Although the core examined in the present study covers the C2n (T), the FO of E. matuyamai was not recorded at this site due to the presence of barren intervals.
The LO of E. matuyamai was first observed near the base of the C1r.1n (Jaramillo: 0.98-1.07Ma in Ogg, 2012) paleomagnetic reversal event by Hays (1970).To the best of our knowledge, this stratigraphic relationship has been corroborated in all subsequent published papers: the estimated ages exhibit little variability, mainly due to the sampling intervals used.In the northeastern Pacific (Ocean Drilling Program (ODP) Leg 145 Site 887, located close to Site U1417), Kamikuri et al. (2007) placed the LO of E. matuyamai between 1.1 and 1.4 Ma using the dating model of Barron et al. (1995), which was developed based on data from the same site (Fig. 1).The LO of E. matuyamai assigned at 1.13 Ma in this study is in agreement with the age ranges provided by Kamikuri et al. (2007).The difference in the dating accuracy between sites 887 and U1417 is likely due to sampling resolution rather than variation attributed to true diachroneity.In the northwestern Pacific, Matsuzaki et al. (2014b) chemostratigraphically calibrated the LO of E. matuyamai as 0.94 Ma at IODP Site C0001/C0002 (southern Japan).At ODP Site 1151 (northeastern Japan), the LO of E. matuyamai is placed between 1.0 and 1.1 Ma (Fig. 1; Kamikuri et al., 2004).The age range, which was estimated using the paleomagnetic reversal events as well as diatom biostratigraphy defined at the same site by Maruyama and Shiono (2003) and Kanamatsu and Niitsuma (2004), is considered to be well constrained.The LO of E. matuyamai is also recognized in the Bering Sea, a marginal sea north of the Kuril Islands.Kamikuri et al. (2007) placed the LO of E. matuyamai between 0.9 and 1.1 Ma at ODP Site 884 based on paleomag-netic stratigraphy and the diatom biostratigraphy of Barron et al. (1995).Ikenoue et al. (2016) assigned the LO of E. matuyamai to 0.95 Ma at IODP Site U1341 in the Bering Sea using a dating model based on an astronomically tuned model combined with the Si / Al ratio proposed by Takahashi et al. (2011).Although the highly constrained ages seem to be variable (0.94 Ma in the northwestern Pacific, 0.95 Ma in the Bering Sea and 1.13 Ma in the northeastern Pacific, this study), the LO of E. matuyamai is considered to be relatively synchronous across the North Pacific at 1.05 Ma with a total error range of 0.1 Myr.

Synchroneity and diachroneity of FOs of
Amphimelissa setosa and Schizodiscus japonicus One of the most important questions regarding the age of bioevents in the North Pacific is the estimation of FOs for A. setosa and S. japonicus, currently placed at 1.48 and 1.30 Ma, respectively.A. setosa is an extant species in Arctic to subarctic area of the North Atlantic Ocean (Bjørklund et al., 2015;Matul and Abelmann, 2005).It seems that A. setosa has existed in the subarctic North Atlantic Ocean since the end of Marine Isotopic Stage 11 (MIS 11) (Bjørklund et al., 2015), and this species has disappeared from the North Pacific between the MIS 4 and MIS 5 (0.06-0.07 Ma) (Kruglikova, 1976;Matul and Abelmann, 2005;Matul et al., 2002Matul et al., , 2009)).The FO of A. setosa was recorded at 1.03 Ma in the Sea of Okhotsk at Core LV28-42-4 (Matul and Abelmann, 2005) and the first continuous occurrence (FCO) of the representative species in the Bering Sea is 0.99 Ma (Ikenoue et al., 2016).The FO of A. setosa at 1.48 Ma (this study) in the northeastern Pacific (Fig. 3 and Table 4) is significantly older than in these two marginal seas.
The biostratigraphic importance of S. japonicus was first pointed out by Ling (1973).The LO of S. japonicus was dated at 0.287 Ma in the Sea of Okhotsk (Matul et al., 2002) and at 0.37 Ma in the Bering Sea (Ikenoue et al., 2016); however, the FCO of S. japonicus was determined to be 0.99 Ma in the Bering Sea (Ikenoue et al., 2016).Since the FO of S. japonicus at Site U1417 is placed at 1.30 Ma (this study), these age differences might imply the existence of a diachroneity between the Bering Sea and the southern Gulf of Alaska.
Concerning both taxa (A.setosa and S. japonicus), an earlier FO is recorded in the subarctic North Pacific than those recorded in the Sea of Okhotsk and Bering Sea, which are marginal seas.This would suggest that the FO of both taxa occurred later in the marginal seas because of a hydrographical semi-isolation from the open ocean.However, a few concerns remain.Concerning the Sea of Okhotsk, the cores studied by Matul et al. (2002Matul et al. ( , 2009) ) cover the time interval of 1.1 Ma at its maximum.Therefore, the situation for time older than 1.1 Ma is unknown.In the Bering Sea, Ikenoue et al. (2016) define their datum as FCO, which infers that sporadic occurrence of both taxa may have occurred for time intervals older than 0.99 Ma.Conversely, we do not have a report of A. setosa and S. japonicus FOs in the Japan Sea, the other marginal sea of the North Pacific.Reports of both taxa FOs in the Japan Sea would potentially elucidate this issue.

Synchroneity and diachroneity of bioevents associated with Lychnocanoma sakaii
With its characteristic three long, bladed feet, Lychnocanoma sakaii is easily identified from similar species in the Pleistocene (Sakai, 1980;Morley and Nigrini, 1995;Matsuzaki et al., 2015b).This species was widely distributed in the North Pacific, particularly in the Bering Sea (Blueford, 1983;Itaki et al., 2012;Ikenoue et al., 2016), Japan Sea (Alexandrovich, 1992;Itaki et al., 2007), Sea of Okhotsk (Matul et al., 2002), the arctic North Pacific above 50 • N (Morley and Nigrini, 1995;Kamikuri et al., 2007) and off northeastern Japan in the northwestern Pacific (Sakai, 1980;Kamikuri et al., 2004;Matsuzaki et al., 2014a).The synchroneity of the LO and abundance peak (acme in the original paper) for L. sakaii was previously reported by Matul et al. (2002) and Matsuzaki et al. (2014a).Relatively little is known regarding the FO of this taxon, but it has variously been estimated at 0.95 Ma off northeastern Japan (Kamikuri et al., 2004), 1.4-1.7 Ma in the south Bering Sea (Kamikuri et al., 2007) and 1.8-1.9Ma in the northeastern Pacific (Kamikuri et al., 2007).The FCO of this species is 0.45 Ma in the Bering Sea (Ikenoue et al., 2016).This study places the FO of this taxon from Site U1417 in the northeastern Pacific at 0.80 Ma.The differences in these ages are not considered to be due to taxonomic uncertainty, preservation state or dating accuracy.To evaluate this diachroneity, the presence or absence of the congeneric ancestor to L. sakaii is considered to play a key role in the mode of speciation for this species.However, such evaluations are difficult because most of the studies published to date did not state whether other Lychnocanoma spp., such as L. nipponica, were present or absent.

Conclusions
Examination of the Pleistocene radiolarian biostratigraphy at IODP Site U1417 (Gulf of Alaska, northeastern Pacific) verified the applicability of the northwestern Pacific radiolarian zonal scheme from the Eucyrtidium matuyamai zone (1.80-1.13Ma), through the Stylatractus universus (1.13-0.45Ma) to the Botryostrobus aquilonaris zone (0.45 Ma and later).Among the bioevents recognized in the northwestern Pacific, eight LOs were recognized at Site U1417: Lychnocanoma sakaii, Amphimelissa setosa, Schizodiscus japonicus, Axoprunum acquilonium, S. universus, E. matuyamai and Actinomma robustum.Recognition of these events indicates that they are well suited for application as biostratigraphic events across the northeastern and northwestern Pacific.This is the first study to identify the FOs of Amphimelissa setosa and Schizodiscus japonicus in the northeastern Pacific, which were preliminary dated at 1.48 and 1.30 Ma, respectively.However, in terms of synchroneity, slight differences were observed between the seven bioevents.Based on the shipboard paleomagnetic polarity reversal data, the LO of E. matuyamai was a synchronous event in the northeastern and northwestern Pacific, occurring at ca. 1.05 ± 0.1 Ma.However, the FO of A. setosa was placed at 1.48 Ma at Site U1417, which is 0.49 Myr older than the oldest age of this bioevent in the Bering Sea.Similarly, the FO of S. japonicus was dated as 1.30 Ma, 0.31 Myr older than its oldest known age in the Bering Sea.

Figure 3 .
Figure 3. Depth-age model of Site U1417 based on shipboard measurements of paleomagnetic reversal events by Jaeger et al. (2014), enabling the age estimation of the FOs of Amphimelissa setosa and Schizodiscus japonicus, the LO of Eucyrtidium matuyamai, and radiolarian bioevents defined in this study for the upper 111 m CCSF-B.

Table 1 .
Taxonomic references of the biostratigraphic taxa mentioned in this study.

Table 3 .
Jaeger et al. (2014)estimated based on the paleomagnetic reversal events ofJaeger et al. (2014)and radiolarian bioevents defined in this study.