Response of Magnetic Minerals to the Mid-Brunhes Climate Event Recorded in Deep-Sea Sediments of the West Philippine Sea

Abstract

The Mid-Brunhes Event (MBE) was one of the most important global climate events since 800 ka. The deep-sea palaeoceanographic changes in the Western Pacific might have been more sensitive to the MBE and they have not been well documented yet. In this study, we investigated a deep-sea core collected from the Western Philippine Sea and then obtained the magnetism record since around 900 ka. The MBE signal in the deposition process was derived from the concentration-dependent (χ, χ_ARM and SIRM) and the grain-size-dependent magnetic parameters (χ_ARM/χ and χ_ARM/SIRM). Across the MBE, the content of magnetic minerals and the fluctuation of the grain size decreased. Both the content and grain-size indices calculated from magnetic parameters are well correlated with the glacial–interglacial alternations and display a major shift at the MBE. The decreased grain size of magnetic minerals may be associated with the flourishing biota in the tropical Pacific during interglacial intervals. The accumulation of magnetic minerals in the Western Philippine Sea was mainly regulated by the sediment accumulation rate, which may be related to the shoaling of the carbonate compensation depth. Overall, this study expanded the environmental magnetism record for the MBE, suggested a new possible influence of this critical climatic event on the deep-sea deposition process in Western Philippine Sea and inferred the interactions between various environmental systems on glacial–interglacial timescales.

1. Introduction

The Philippine Sea, located in the Western Pacific, is a key area for material and energy exchanges. It plays an important role in regional source-to-sink processes and global climate changes [1,2,3]. Reconstructing the paleoenvironment in the Philippine Sea on geological timescales (since around 900 ka) can provide us with useful insights for revealing the deep-sea process in the Western Pacific and its linkage to global climate changes.

The Mid-Brunhes Event (MBE) was one of the most important global climate events since 800 ka [4]. It was characterized by remarkably increased amplitudes of frequent glacial–interglacial cycles and carbonate changes since ~430 ka [5,6]. According to the ice core records, the Antarctic temperature was cooler and the concentration of atmospheric CO₂ was lower during the interglacial intervals before the MBE [7,8]. It was also revealed by deep-sea benthic δ¹⁸O that the sea level was higher in the interglacial period after the MBE [9]. The MBE started with warm and high productivity events, and it was most prominent in the southern Arctic region. However, it seems that the concentration of atmospheric CH₄ and the Asian monsoon were less affected by the MBE [10,11,12]. Thus, the exploration of the mechanism of the MBE is becoming an increasingly significant issue [4,13,14,15].

There are many controversies about the corresponding mechanism of the MBE [4,13,14,15]. It was pointed out that the meteorite impact event that occurred at 480 ka in Antarctica may have led to the occurrence of the MBE [16]. Some researchers believe that the MBE may have something to do with the latitudinal shift of the Southern Hemisphere westerlies which led to the CO₂ respiring from the Southern Ocean [17] and/or a slowdown of Antarctic Bottom Water (AABW) formation 14. However, the relationships between orbital forcing, greenhouse gases and ice volume are complex [9]. In addition, it is not well known how the global carbon cycle changed before and after the MBE. Thus, more studies need to be carried out to better understand the mechanism of the MBE and the earth’s climate response to the MBE.

In this study, we investigated a 3.5 m deep-sea sediment (Core I8) collected from the Western Philippine Sea, Western Pacific (Figure 1). The chronology of the section has been well set up by Xu et al. (2022) [18]. Reconstructing the accumulate history of the magnetic minerals in Core I8 can potentially unlock a wealth of new information from the deep-sea sediment. The MBE signal in the deposition process was derived from the magnetism record of Core I8 since around 900 ka. The purpose of this study is twofold: (1) to deduce the deep-sea palaeoceanographic processes according to the changes of magnetic properties; and (2) to discuss the linkage of the deep-sea process in the Western Philippine Sea to the global climate changes.

2. Materials and Methods

2.1. The Study Core (I8)

The study core, I8 (12°75′ N, 131°29′ E, water depth 5405 m), was collected from the western part of the Philippine Sea (Figure 1) using a gravity corer. The core was drilled to a depth of 3.5 m. The sediments of core I8 are generally light brown to brown in color and mainly contain silty clay. For paleomagnetic and paleoenvironment studies, we obtained 172 samples in 2 cm intervals by non-magnetic plastic cubes (2 × 2 × 2 cm³). Prior to this, Xu et al. (2022) [18] determined two magnetic zones in core I8 (Figure 2a,b) and established a reliable chronostratigraphic framework by magnetostratigraphy and tuning the Ba intensities of the I8 core to the global ice volume (Figure 2d,e). The sediment accumulation rates (SARs) for each section were then calculated (Figure 2f). Average SAR was estimated as 0.41 ± 0.15 cm/ka over the depositional interval. As seen in Figure 2f, the SAR has shown a general downward trend since ~887 ka with a great jump at about 450 ka and it has remained at a low level after 400 ka.

2.2. Magnetic Measurements

Magnetic susceptibility (χ_lf and χ_hf) was measured using an AGICO MFK1-FA multi-frequency Kappabridge magnetic susceptibility meter at frequencies of 976 Hz and 15,616 Hz, respectively. By using a 2G enterprises SQUID magnetometer with inline AF coils, anhysteretic remanent magnetization (ARM) was imparted to all the samples with a peak alternating field (AF) of 100 mT and a direct biasing field of 0.05 mT. Isothermal remanent magnetization (IRM) was induced with a 2-G Enterprises model 660 pulse magnetizer successively in pulsed fields of 1 T, which is regarded here as saturated isothermal remanent magnetization (SIRM). Then, the IRM of 0.3 T was applied in the opposite direction, denoted IRM_−0.3T. The percentage of frequency-dependent susceptibility (χ_fd%), hard isothermal remanent magnetization (HIRM) and S-ratio were calculated using the following equations [21]. All the magnetic measurements were performed in the State Key Laboratory of Marine Geology, Tongji University.

$${\chi }_{\mathrm{fd}}=\frac{{\chi }_{\mathrm{lf}}-{\chi }_{\mathrm{hf}}}{{\chi }_{\mathrm{lf}}}\times 100\%,$$

(1)

$$\mathrm{HIRM}=\frac{\mathrm{SIRM}-{\mathrm{IRM}}_{-0.3\mathrm{T}}}{2},$$

(2)

$$S_{-300}=\frac{SIRM-IR{M}_{-0.3T}}{2\times SIRM}\times 100\%,$$

(3)

$${\mathrm{S}}_{-\mathrm{Ratio}}=\frac{{\mathrm{IRM}}_{-0.3\mathrm{T}}}{\mathrm{SIRM}}.$$

(4)

3. Results and Analyses

3.1. Changes in Magnetic Parameters

The low-frequency magnetic susceptibility (χ_lf) of core I8 varies from 9.32 × 10⁻⁷ m³/kg to 1.52 × 10⁻⁶ m³/kg with an average value of 1.21 × 10⁻⁶ m³/kg and a standard deviation of 1.63 × 10⁻⁷ m³/kg. The high-frequency magnetic susceptibility (χ_hf) varies from 8.89 × 10⁻⁷ m³/kg to 1.45 × 10⁻⁶ m³/kg with an average value of 1.15 × 10⁻⁶ m³/kg and a standard deviation of 1.56 × 10⁻⁷ m³/kg (

Table 1. Results of magnetic parameters.

	Min	Max	Average	Standard Deviation	Principal Component Analysis
					Content	Grain Size
ꭓlf (m3/kg)	9.32 × 10−7	1.52 × 10−6	1.21 × 10−6	1.63 × 10−7	0.973	∕
ꭓhf (m3/kg)	8.89 × 10−7	1.45 × 10−6	1.15 × 10−6	1.56 × 10−7	0.972	∕
SIRM(Am2/kg)	1.57 × 10−2	2.68 × 10−2	2.05 × 10−2	2.73 × 10−3	0.648	∕
χARM (m3/kg)	3.85 × 10−6	6.91 × 10−6	5.25 × 10−6	7.07 × 10−7	0.908	∕
ꭓfd (%)	3.44	4.81	4.26	0.24	∕	0.455
χARM/χ	3.70	4.89	4.37	0.31	∕	0.945
χARM /SIRM(m/A)	2.11 × 10−4	3.03 × 10−4	2.56 × 10−4	2.05 × 10−4	∕	0.937
S-Ratio	0.90	1.00	0.97	0.02	∕	∕

). The susceptibility of ARM (χ_ARM) is usually sensitive to single-domain (SD) grains [22,23], and SIRM can be used to infer magnetic particles excluding the influence of superparamagnetic (SP) grains [24]. The values of χ_ARM and SIRM are 3.85 × 10⁻⁶−6.91 × 10⁻⁶ m³/kg and 1.57 × 10⁻²−2.68 × 10⁻² Am²/kg, respectively (Table 1). Their average values are 5.25 × 10⁻⁶ m³/kg and 2.05 × 10⁻² Am²/kg and both of them display a similar pattern to χ curves (Figure 3a–d). The χ, χ_ARM and SIRM generally reflect the concentration of magnetic minerals, especially ferrimagnetic minerals [25]. The concentration-dependent magnetic parameters (χ, χ_ARM and SIRM) have similar variations throughout the core, which suggests that these proxies were not greatly influenced by mineralogical variations. They all declined greatly at 480 ka, indicating a significant decrease in magnetic minerals.

Overall, the χ_fd% has a relatively low value, indicating a small amount of super-paramagnetic material. The minimum and maximum values of χ_fd% are 3.44% and 4.81%, and the average value is about 4.26%. The standard deviation of χ_fd% is 4.26% (Table 1). Unlike the χ_lf and the χ_hf, the χ_fd% does not show an obvious downward trend at around 480 ka. However, the variation of χ_fd% has a greater amplitude after ~480 ka (Figure 3e). The χ_ARM/χ varies between 3.70 and 4.37 with an average value of 4.37, and χ_ARM/SIRM varies between 2.11 × 10⁻⁴ m/A and 2.56 × 10⁻⁴ m/A with an average value of 4.37 m/A. Their standard deviations are 0.31 and 2.05 × 10⁻⁴ m/A, respectively (Table 1). Both the χ_ARM/χ and χ_ARM/SIRM are commonly used as grain size indicators for ferrimagnetic minerals, as they decrease with increasing grain size [25]. The main discrepancy is that the first proxy is more sensitive to size change but may be biased by magnetic susceptibility from super-paramagnetic and paramagnetic material. The latter is less sensitive than the first, but it only contains remanence, avoiding the effect of super-paramagnetic and paramagnetic grains. In addition, the χ_ARM/χ and χ_ARM/SIRM agree well with less deflection by super-paramagnetic and paramagnetic grains, both of them are main proxies of magnetic grain size (Figure 3f,g).

The S_-ratio reflects the relative proportion of low- to high-coercivity minerals in samples. The larger the ratio is, the more low-coercivity magnetic minerals there are, and this kind of mineralogy change can indicate source variation. The S_-ratio of Core I8 varies between 0.9–1 with an average value of 0.97 and the standard deviation is 0.02, indicating the dominance of low-coercivity magnetic minerals throughout the whole sequence (Figure 3h).

Principal component analysis was used to obtain the more representative concentration- and grain-size-dependent magnetic parameters. The results show that χ and χ_ARM can better reflect the magnetic mineral content and that the χ_ARM/χ and χ_ARM/SIRM are more relevant to the magnetic mineral grain size (Table 1). Based on the principal component analysis, we detrended, normalized and then averaged the χ_lf and SIRM data to build the magnetic mineral content index, and perform the same operation on χ_ARM/χ and χ_ARM/SIRM to obtain the magnetic mineral grain-size index.

3.2. Relationships between Magnetic Parameters

Based on the general variation in the concentration-dependent (χ, χ_ARM and SIRM) and the grain-size-dependent magnetic parameters (χ_ARM/χ and χ_ARM/SIRM), the magnetic series can be divided into two stages (Figure 3). In Stage 1 (before 480 ka), the magnetic mineral content was at a high level and accompanied by large vibrations. Meanwhile, the grain-size-dependent magnetic parameters have a higher frequency of variation during this time. Stage 2 (since 480 ka) was characterized by a sudden drop of magnetic mineral content and it has remained at a low level since then. At the same time, there was a slowdown in the frequency of changes in grain size. The difference between the two stages may be related to the MBE.

In order to determine the variation of magnetic parameters and its linkage to the MBE, we analyzed the correlation coefficients of different parameters for both stages (before and after 480 ka). As shown in their scatter plots (Figure 4), the differences associated with both the concentration- and grain-size-dependent parameters are significant before and after MBE. Specifically, in the case of concentration-dependent parameters (χ_lf vs. SIRM), the differences in the slopes and in regression intercepts are both evident (Figure 4a). As seen in Figure 4b, it is the same case with the grain-size-dependent parameters (χ_ARM/χ vs. χ_ARM/SIRM). For the case of χ_ARM vs. χ_ARM/χ, the differences in regression intercepts are evident with a small difference in the slopes (Figure 4c). Both the slopes and regression intercepts are significantly different in the case of SIRM vs. χ_ARM/SIRM (Figure 4d). Additionally, for other paired parameters, such as χ_ARM vs. SIRM and χ_fd% vs. χ_ARM/SIRM, they are not shown in this work because of the poor fitting. The distinct regression intercepts indicate that the basic background of magnetic particles would be evident, while a difference in regression slopes may indicate different magnetic properties in different time intervals. Hence, it is inferred that magnetic properties changed significantly at ~480 ka and the MBE was a key factor for deep-sea environmental changes in the West Philippine Sea.

3.3. Power Spectral Analyses

In order to obtain the environmental significance hidden in these magnetic parameters, we performed spectral analysis on the proxies in this study and other widely used environmental records. The power density spectra of Benthic δ¹⁸O stack (LR04) [26] was calculated by Acycle software (https://acycle.org/, accessed on 1 November 2022) in MATLAB [27,28], as were the mean grain size (MGS) of the Chinese Loess Plateau [10], the magnetic parameters (ꭓ_lf, ꭓ_ARM, ꭓ_ARM/ꭓ, ꭓ_ARM/SIRM), and both the content and grain-size indices of magnetic minerals.

LR04, one of the most widely used proxies for global climate changes, was first calculated for its power spectrum. There appeared all the well-known peaks corresponding to eccentricity (100 ka), obliquity (40 ka) and precession (23 ka) when considering the interval since 900 ka (Figure 5a). The eccentricity cycle shows the glacial–interglacial variation related to the deep-water temperature and sea level change [29]. In the studies of tropical oceans, the obliquity and precession cycle are usually used as indicators of monsoonal changing [30,31]. MGS has been widely used to indicate the intensity of the Asian monsoon, and it shows significant peaks at both 41 ka and 23 ka (Figure 5a).

The magnetic mineral parameters related to the content (ꭓ, SIRM; Figure 5b) show obvious periods of 100 ka, 41 ka and 23 ka in the spectrum. Of course, such a phenomenon also appears in the more representative magnetic mineral content index (Figure 5d). However, the grain-size-dependent magnetic parameters (ꭓ_ARM/ꭓ, ꭓ_ARM/SIRM; Figure 5c) and the calculated magnetic mineral grain-size index only show a period of 100 ka in the spectrum.

4. Discussion

It has been reported that the so-called MBE, representing an enlargement of glacial–interglacial alternations since ~430 ka, has had a major influence on key climatic processes. To better understand the potential linkages between regional paleoenvironmental processes and global climate changes, the magnetic properties of core I8 were further compared with various proxies (Figure 6). All these proxies exhibited in Figure 6 have been broadly employed for paleoenvironmental inferences in different types of sediments [32,33,34,35,36].

As discussed in the Section 3.3, both the content and grain-size indices of magnetic minerals have evident cycles (Figure 5 and Figure 6a,b). Moreover, their changing patterns are similar to those of the Northern Hemisphere glaciation, as reflected in the stacked benthic δ¹⁸O record of deep-sea sediments [26] (LR04, Figure 6b). The similarity between the magnetic mineral index of core I8 and the LR04 record, indicating that they were both regulated by eccentricity period (~100 ka, Figure 5), indicates the force of glacial–interglacial periods. We know that the grain-size-dependent magnetic parameters (ꭓ_ARM/ꭓ, ꭓ_ARM/SIRM; Figure 5c) and the calculated grain-size index only show a period of 100 ka in the spectrum. However, the content index also exhibits a more evident cycle in precession bands (~23 ka, Figure 6a). It generally agrees with changes in the Asian monsoon as reflected in the stack grain size (MGS) of the China Loess Plateau [10] (Figure 6c), especially during interglacial intervals. Spectral analysis confirms this consistency between them (Figure 5), thus highlighting an integrated forcing of Asian monsoon variability on regional sedimentary processes, perhaps related to provenance.

The magnetic mineral content index was usually higher during the interglacial periods, which indicates a higher amount of magnetic minerals in the sediments (Figure 6a,b). Meanwhile, the higher magnetic mineral grain-size index during the interglacial periods represents the thinning of magnetic minerals. In general, during the interglacial periods, the magnetic minerals in the sediments of core I8 mainly show an increase in content and decrease in grain size. Nevertheless, for the entire deposition phase, the indices of magnetic minerals underwent a significant change in the MBE. After this crucial global climate event, the change frequency and amplitude of the magnetic mineral grain-size index gradually slows down and the magnetic minerals become finer. Due to the climate becoming warmer in the interglacial period after the MBE, magnetic mineral grain-size thinning may be associated with flourishing biota in the tropical Pacific during interglacial intervals [37,38,39,40,41].

The SAR of Core I8 dropped greatly after the MBE (Figure 2f), which is consistent with the content-dependent magnetic parameters (Figure 6a–d). The low SAR during the post-MBE period may be the key reason for the reduction of magnetic mineral content. The lower sedimentation rate led to the large reduction of magnetic minerals after the MBE. As one of the most important global climate events, the MBE had a great influence on the carbonate compensation depth (CCD) [42,43]. By comparing the paleoenvironmental proxies, Xu et al. (2022) [18] proposed that the CCD was about 200 m shallower than in the post-MBE time. The CCD shoaling was reflected in the bulk carbon δ¹³C of core C6 [18] (Figure 6e) and the weaker AABW [44,45] recorded at ODP 1088 [46] (Figure 6f), inferring a weakening of deep-sea effective sedimentation leading to the decrease in SAR and magnetic mineral content.

5. Conclusions

Based on a well-dated record of deep-sea sediment in the Western Philippine Sea, environmental magnetic information for the middle Pleistocene was derived to deduce the deep-sea palaeoceanographic processes. Our main findings are summarized as follows: (1) The magnetic series can be divided into two stages. Before the MBE, the magnetic mineral content was at a high level and accompanied by large vibrations. Meanwhile, the grain-size-dependent magnetic parameters have a higher frequency of variation during this time. After the MBE, there was a sudden drop in magnetic mineral content and it has remained at a low level since then. Furthermore, there was a slowdown in the frequency of changes in grain size. (2) Both content and grain-size indices calculated by the magnetic parameters can be well correlated to the glacial–interglacial alternations and display a major shift at the MBE. (3) The decrease in magnetic mineral grain size may be associated with flourishing biota in the tropical Pacific during interglacial intervals. (4) The accumulation of magnetic minerals in the Western Philippine Sea was mainly regulated by the sediment accumulation rate, which may be related to the carbonate compensation depth. According to these observations, this study expanded the environmental magnetism record of the MBE, suggested a new possible influence of the MBE on the deposition process in the Western Philippine Sea and inferred interactions between various environmental systems on glacial–interglacial timescales.