Mineral Weathering and Element Migration in Granite Weathering Pits (Gnammas): A Case Study in Eastern China

Abstract

This paper investigates weathering pits at three granite sites located on mountain tops, in a former river bed, and on the coastline of an island, respectively, from north to south in eastern China and aims to characterize weathering pit formation in the above settings in terms of mineral weathering and elemental transport. In these settings, the main elements, and mineral groups in the debris in the weathering pits and the fragments of the rock surface directly adjacent to the pits were analyzed. The chemical index of alteration (CIA), the quartz/feldspar (Q/F) ratio and the Na/K (Na₂O/K₂O) ratio were applied to identify the chemical origin of the weathering pits and assess the difference in the chemical weathering processes of the weathering pits in the different settings; the mass transfer coefficient was used to measure the characteristics of element migration during weathering pit formation at the three sites. The result of CIA, Q/F, and Na/K analysis shows that debris in a weathering pit suffered from higher chemical weathering intensity than nearby rock surfaces, indicating that the weathering pits of the study sites originated from chemical weathering. However, the differences in the CIA values of weathering pits in different areas are only the result of different chemical weathering durations and cannot be used to identify the climate types of the areas. The calculation of element mass transfer indicates that only Na and K are continuously leached during the formation process of weathering pits regardless of whether in valleys, mountains or on the coast. Other elements may or may not be the external source for the formation of weathering pits resulting in different natural tendencies for element mass transfer in weathering pits. Seawater can also be a factor contributing to the different patterns of element migration in weathering pits in coastal and inland areas. In addition, the environment of river valleys is more conducive to weathering pit formation than mountain tops.

1. Introduction

Weathering is the alteration of rock surfaces exposed to atmospheric conditions and is the starting point of many dynamic systems and a dominant force in shaping many landscapes [1,2,3]. Weathering pits, also known as rock basins, rock holes or gnammas, are depressions on flat rock surfaces that vary in shape and size from a few centimeters to several meters wide. The formation processes of weathering pits have been studied worldwide [4]. The most widely accepted theory states that near-horizontal rock surfaces are susceptible to chemical weathering by small water-filled depressions, which react more rapidly than other rock surfaces outside the depressions. In this process, water plays an active role either directly or indirectly in this process to promote chemical changes [4,5,6,7,8,9,10,11,12,13,14,15,16,17]. Except for chemical weathering caused by standing water, salt weathering, microgelivation, wind sculpture and biological weathering have also been attributed to the formation of weathering pits [8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24]. However, the causes of weathering pits are still not well known [2]. For example, many studies have shown that weathering pits often occur at intersections of joints; however, in some places, massive granites form weathering pits whereas nearby well-jointed and faulted granites do not [2], and it is unclear how weathering pits develop differently in coastal and inland environments, or in river valleys and mountain tops. More than 100 years have passed since the first academic paper dedicated to the study of granite weathering pits, and the study of weathering pits has never ceased [12,17,25].

Weathering pits are commonly found in granite terrain, but they can also be seen in other lithologies. They have been observed in hot and cold regions such as deserts and polar regions, as well as in coastal and inland environments. In China, weathering pits are a common feature on mountain tops, river valley beds, and coasts [12,13,14,26]. There is, however, controversy over the origins of weathering pits in China, which has continued until recently [26]. Weathering pits have been mistaken for stream potholes and glacier potholes in some studies in China [26,27,28,29]. China is a vast region with different climates in different geological and geomorphological settings. The chemical weathering processes that occur in weathering pits in those settings are unlikely to be identical. This study investigates the chemical characteristics of weathering pits found at three granitic sites in eastern China. These sites include mountain tops, a former river bed, and the coastline of an island, respectively, aiming to find evidence supporting the origins of chemical weathering in weathering pits in eastern China and to determine the differences in chemical weathering processes that occur in weathering pit formation processes at the various topographic locations mentioned above. The study also evaluates the efficiency of CIA values, the quartz/feldspar (Q/F) ratio and the Na₂O/K₂O ratio in identifying the intensity of chemical weathering in weathering pit formation.

2. Materials and Methods

2.1. Study Areas

Weathering pits formed across three different sites from north to south in eastern China were investigated (Figure 1).

Study Site A. Laoshan Mountain (N36°10′, E120°37′) is situated in the eastern part of Qingdao City, southeastern Shandong Peninsula, China, and is the main mountain of the peninsula, with its highest peak, Laoding (Giant Peak), at 1132.7 m above sea level (Figure 1a). Loashan mountain is characterized by a warm-temperate continental monsoon with strong oceanic influences and has a temperature range of −2.3 to 25.3 °C, and the annual precipitation is 849.9 mm, with most of the precipitation falling between June and September [30]. During the Cretaceous, the Yanshanian granitic complex of Loashan intruded into the Proterozoic metamorphic rocks that formed the basement of the mountain area [31] (Figure 1d-1). Laoshan Mountain is composed primarily of grayish-white or flesh-red alkaline granites of Yanshan age with medium-coarse grains [31]. This study examined weathering pits in two areas: Laoding (about 1000 m above sea level) and Lingyangu (about 330 to 400 m above sea level) (Figure 1a, Figure 2a-1,a-2). Near the eastern edge of Laoshan Mountain, deeply weathered deposits of boulders of various sizes, sand and clay were discovered on the banks of Baiyun Reservoir (Figure 1a and Figure 2a-3).

Study Site B. Pingshan Mountain (N25°50.193′, E119°32.429′) is a granite low mountain located on the central coast of Fujian Province, China (Figure 1b). The climate in the area where the mountain is located is subtropical and marine monsoon, which is warm and humid throughout the year, with four distinct seasons. The average annual temperature is 19.6 °C; generally, the highest temperature occurs in July, with an average of 28.3 °C, and the lowest temperature occurs in January, with an average of 10.3 °C. The average annual precipitation is 1346 mm, with 130 days of rainfall per year, most of which occur from March to September. During this time, the monsoon or rainy season lasts from April to June [32]. Miarolitic K feldspar granite of the late Yanshan period constitutes the mountain’s granite body, whose major minerals are quartz and potassium feldspar [33]. An NNE-trending Mesozoic fault lying along the tectonic boundary between the Pacific plate and the Eurasian plate is situated on the Fujian coast and controls the formation of the Yanshanian granites as well as the miarolitic K feldspar granite (Figure 1d-2) [33]. Weathering pits occur on the exposed rock surface of the mountaintop at an elevation of about 100 m (Sampling site 2 in Figure 1b). The Sanxi River flows from the Sanxi River Reservoir through the northern foot of Pingshan Mountain (Figure 1b). After the reservoir dam was built in 1959, the riverbed downstream of the dam was exposed. As a result, river potholes were exposed in groups in the riverbeds, and weathering pits were visible on the erosion terrace of the river channel (Sampling site 1 in Figure 1b and Figure 2b-1) [17]. There is a granite weathering crust profile exposed along the north slope of the Sanxi River valley (Sampling Site 3 of Figure 1b and Figure 2b-3).

Study Site C. Miaowan Island (21°52′ N, 114°01′ E) is located at the mouth of the Pearl River (Figure 1c). There are two peaks on the island, each of which has an elevation of 228 m and 157 m, respectively (Figure 1c). The annual average rainfall in the study area is 1583 mm, with approximately 80% between May and September. The mean temperature ranges from 15.8 °C in January to 27.7 °C in July, and the prevailing wind directions are SE (90°–120°) and N (0°), respectively [34]. The island mainly consists of coarse-grained Yanshanian biotite and porphyritic granite, with small outcrops of migmatitic granite and basalt formed during the third episode of the Yanshan Movement [35]. Weathering pits were found on the rock surface of a saddle about 50 m above sea level between the two peaks (see Sampling sites 2 in Figure 1c and Figure 2c-2). On the northern shore of the saddle, there is a platform about 9 m above sea level where several large weathering pits have formed (Sampling site 1 in Figure 1c and Figure 2c-1). A granite weathering crust profile was exposed on the north slope of Miaowan Island (Sampling site 3 in Figure 1c and Figure 2c-3).

2.2. Materials and Methods

This study used samples taken in the field between 2012 and 2013. The weathering pits containing the highest number of fragments taken from each study site were selected for study. This was performed under the assumption that these pits have suffered sufficient chemical weathering in the local environment. Therefore, the chemical weathering processes occurring in these pits are maximally representative of the chemical weathering processes occurring in all the pits of the study sites. In the study, three types of samples were collected: (1) fragments remaining within weathering pits, (2) rock fragments knocked down with a hammer from rock surfaces adjacent to (no more than half a meter away from) the pits and (3) debris contained in local weathering crusts or deposits (Figure 2a-3,b-3,c-3). In total, 26 samples were collected from three sites (Figure 1), including 12 debris samples from weathering pit accumulations, 11 fragment samples knocked down from the rock surface near the pits and 3 samples from local weathering crusts and deposits (Table 1, Table 2 and Table 3). Geochemical analyses of the selected samples were carried out at the Guangzhou Institute of Geochemistry (GIG), Chinese Academy of Sciences. Major element oxides (Al₂O₃, CaO, Fe₂O₃, K₂O, MgO, MnO, Na₂O, P₂O₅, SiO₂, TiO₂) were obtained using a Rigaku ZSK100e XRF on a fused glass bead. Loss-of-ignition (LOI) measurements were undertaken on dried sample powder by heating in a pre-ignition silica crucible to 1000 °C for 1 h and recording the percentage of weight loss. Calibration lines used in quantification were produced by bivariate regression of data from 36 reference materials covering a wide range of silicate compositions. Analytic uncertainties range from 1% to 5% [36].

The chemical index of alteration (CIA), a widely used index to measure the degree of weathering, was used in this study based on the major elements (oxides) of the samples. The formula for calculating CIA based on molecular proportions is as follows [37]:

CIA = [Al₂O₃/(Al₂O₃ + CaO* + Na₂O + K₂O)] × 100

(1)

where CaO* must be siliciclastic CaO as well as phosphate and carbonate corrections required for non-siliciclastic CaO, e.g., from biogenic apatite and/or carbonates, such as calcite and dolomite. The phosphate correction is calculated as CaO* = CaO − (10/3 × P₂O₅) [38]. After phosphate correction, no further CaO correction is required if CaO/NaO ≤ 1, and the CaO content is calculated as NaO if CaO/NaO > 1 [39]. The Na/K ratio is also a measure of the degree of weathering of plagioclase in the samples and can be used to characterize the chemical weathering of accumulation [40].

Since the leaching of active elements leads to an increase in stable elements during chemical weathering (residual enrichment), the true elemental geochemical behavior is often not reflected by the absolute content of individual elements in parent rocks [41]. The mass transfer coefficient [42], which is the ratio of the mobile component j to the additional inert component i, is used as follows:

$${\tau }_j=\frac{C_{j.w}·C_{i,p}}{C_{j,p}·C_{i,w}}-1$$

(2)

where C_j,w and C_j,p represent the contents of mobile elements in the sample and unweathered rocks, respectively. C_i,w and C_i,p represent the stable elements in the sample and unweathered rocks, respectively. τ_j is defined as the mass transfer coefficient. A value of τ_j = 0 denotes no mobility, whereas τ_j = −1 denotes complete mobility, and τ_j > 0 denotes external additions. By calculating the variation ratio of an element relative to some stable elements (e.g., K, Ti, Al, etc.), Al was chosen as the immobile reference element in this study. Stable elements in the parent bedrock of the 3 study sites are necessary for calculating the mass transfer transformations between the parent rock and the samples taken from the field. However, we do not have such samples. Fortunately, the granitic rocks of the three study sites have previously been well studied [33,35,43]. Therefore, we can cite the data of the main elements in the local bedrock from published papers for our calculations (see later).

X-ray powder diffraction (XRD) was utilized to study the major minerals and clay minerals in samples collected from the field at GIG. Each sample was pulverized (200 mesh) and oven dried at 80 °C prior to mineral analysis, which was performed using a Bruker D8 Advance X-ray diffractometer operated at 40 kV and 30 mA and scanned step-wisely at 4° min⁻¹ from 3° to 85° (2Ѳ) with a copper (Cu) X-ray tube and receiving slit = 1 mm. The relative percentage compositions of the minerals were determined semi-quantitatively by employing the area under the curve for the main peaks of every mineral. The instrument’s analytical precision for the results was ± 3%. The ratio of quartz to feldspar content (Q/F) was calculated for each sample.

3. Results

Table 1 shows the mass percentages of major-element oxides (SiO₂, TiO₂, Al₂O₃, Fe₂O₃, MgO, CaO, MnO, Na₂O, K₂O and P₂O₅) and LOI for the field samples collected from the three study sites (Figure 1). Table 2 shows the percentages of major minerals and clay minerals of the samples. Table 3 shows the values of CIA, Q/F and Na₂O/K₂O for each sample.

Table 1. Major element compositions for samples taken from the three study sites.

Sample No.	Al2O3	CaO	Fe2O3	K2O	MgO	MnO	Na2O	P2O5	SiO2	TiO2	L.O.I	Total
Study site A (Loashan Mountain)
LD-1-BY	13.29	0.09	1.42	4.59	0.1	0.06	4.3	0.02	74.86	0.19	0.63	99.54
LD-1-CJW	10.83	0.98	2.08	2.81	0.62	0.05	2.46	0.05	77.24	0.52	1.93	99.58
LD-2-BY	12.22	0.03	1.05	4.24	0.01	0.09	4.08	0.13	77.17	0.08	0.54	99.64
LD-2-CJW	10.78	0.22	2.12	3.54	0.36	0.1	2.8	0.05	78.29	0.29	1.02	99.55
LD-3-BY	13.66	0.04	1.3	4.77	0.04	0.04	4.34	0.01	74.61	0.21	0.66	99.67
LD-3-CJW	12.29	0.3	2.68	3.64	0.51	0.07	2.75	0.06	75.02	0.37	1.84	99.55
LYG-BY	14.12	0.46	1.36	4.98	0.18	0.07	4.07	0.05	73.49	0.22	0.53	99.54
LYG-CJW	12.24	1.63	3.34	3.87	1.03	0.07	2.65	0.14	70.85	0.4	3.44	99.66
YNP-BY	13.77	0.29	0.8	5.24	0.1	0.04	3.78	0.03	74.79	0.11	0.58	99.54
YNPG-CJW	13.03	0.38	1.26	4.89	0.21	0.03	3.29	0.07	75.29	0.17	0.93	99.55
HL-1-1-BY	13.97	0.35	1.27	5.09	0.13	0.05	3.98	0.05	73.94	0.2	0.53	99.54
HL-1-1-CJW	13.67	0.56	1.93	4.27	0.5	0.04	3.28	0.1	73.34	0.38	1.51	99.57
BYSK-2	23.20	0.11	9.89	2.01	1.22	0.04	0.25	0.20	46.19	1.41	15.49	100
Study site B (Pingshan Mountain/the Sanxi River)
TLJ-15-BY	12.4	0.16	0.68	4.58	0.07	0.02	4.64	0.02	76.33	0.11	0.42	99.44
TLJ-15-CJW	6.67	0.07	1.17	2.69	0.05	0.06	0.74	0.02	86.66	0.28	1.03	99.45
TLJ-16-BY	12.17	0.15	0.82	4.89	0.07	0.04	3.06	0.02	77.68	0.11	0.42	99.44
TLJ-16-2CJW	5.09	0.08	1.08	1.55	0.06	0.03	0.67	0.03	89.89	0.15	0.85	99.48
TLJ-16-5-CJW	6.51	0.23	1.03	1.72	0.11	0.04	0.34	0.05	87.49	0.26	1.8	99.58
SXHG-BY	12.52	0.18	0.9	4.54	0.09	0.05	3.71	0.01	77.08	0.15	0.29	99.54
SXHG-CJW	9.33	0.15	1.34	2.42	0.15	0.05	0.78	0.03	82.67	0.23	2.47	99.6
SXHG-R	18.21	0.03	4.79	0.21	0.12	0.12	0.04	0.01	68.91	0.2	7.51	100.15
Study site C (Miaowan Island)
MW-WP-BY	12.66	0.44	0.68	4.8	0.1	0.03	3.16	0.02	77.34	0.06	0.43	99.72
MW-WP-CJW	12.86	0.36	0.64	4.84	0.09	0.02	3.36	0.02	76.92	0.07	0.55	99.72
MW-2-BY	12.84	0.29	0.91	5.1	0.06	0.06	3.16	0.01	76.87	0.05	0.35	99.72
MW-2-CJW	10.51	0.23	1.43	3.83	0.15	0.04	2.14	0.03	79.12	0.12	2.2	99.78
MW-R	16.07	0.1	3.1	1.56	0.27	0.02	0.16	0.01	71.02	0.2	7.24	99.75

Table 1. Major element compositions for samples taken from the three study sites.

Sample No.	Al2O3	CaO	Fe2O3	K2O	MgO	MnO	Na2O	P2O5	SiO2	TiO2	L.O.I	Total
Study site A (Loashan Mountain)
LD-1-BY	13.29	0.09	1.42	4.59	0.1	0.06	4.3	0.02	74.86	0.19	0.63	99.54
LD-1-CJW	10.83	0.98	2.08	2.81	0.62	0.05	2.46	0.05	77.24	0.52	1.93	99.58
LD-2-BY	12.22	0.03	1.05	4.24	0.01	0.09	4.08	0.13	77.17	0.08	0.54	99.64
LD-2-CJW	10.78	0.22	2.12	3.54	0.36	0.1	2.8	0.05	78.29	0.29	1.02	99.55
LD-3-BY	13.66	0.04	1.3	4.77	0.04	0.04	4.34	0.01	74.61	0.21	0.66	99.67
LD-3-CJW	12.29	0.3	2.68	3.64	0.51	0.07	2.75	0.06	75.02	0.37	1.84	99.55
LYG-BY	14.12	0.46	1.36	4.98	0.18	0.07	4.07	0.05	73.49	0.22	0.53	99.54
LYG-CJW	12.24	1.63	3.34	3.87	1.03	0.07	2.65	0.14	70.85	0.4	3.44	99.66
YNP-BY	13.77	0.29	0.8	5.24	0.1	0.04	3.78	0.03	74.79	0.11	0.58	99.54
YNPG-CJW	13.03	0.38	1.26	4.89	0.21	0.03	3.29	0.07	75.29	0.17	0.93	99.55
HL-1-1-BY	13.97	0.35	1.27	5.09	0.13	0.05	3.98	0.05	73.94	0.2	0.53	99.54
HL-1-1-CJW	13.67	0.56	1.93	4.27	0.5	0.04	3.28	0.1	73.34	0.38	1.51	99.57
BYSK-2	23.20	0.11	9.89	2.01	1.22	0.04	0.25	0.20	46.19	1.41	15.49	100
Study site B (Pingshan Mountain/the Sanxi River)
TLJ-15-BY	12.4	0.16	0.68	4.58	0.07	0.02	4.64	0.02	76.33	0.11	0.42	99.44
TLJ-15-CJW	6.67	0.07	1.17	2.69	0.05	0.06	0.74	0.02	86.66	0.28	1.03	99.45
TLJ-16-BY	12.17	0.15	0.82	4.89	0.07	0.04	3.06	0.02	77.68	0.11	0.42	99.44
TLJ-16-2CJW	5.09	0.08	1.08	1.55	0.06	0.03	0.67	0.03	89.89	0.15	0.85	99.48
TLJ-16-5-CJW	6.51	0.23	1.03	1.72	0.11	0.04	0.34	0.05	87.49	0.26	1.8	99.58
SXHG-BY	12.52	0.18	0.9	4.54	0.09	0.05	3.71	0.01	77.08	0.15	0.29	99.54
SXHG-CJW	9.33	0.15	1.34	2.42	0.15	0.05	0.78	0.03	82.67	0.23	2.47	99.6
SXHG-R	18.21	0.03	4.79	0.21	0.12	0.12	0.04	0.01	68.91	0.2	7.51	100.15
Study site C (Miaowan Island)
MW-WP-BY	12.66	0.44	0.68	4.8	0.1	0.03	3.16	0.02	77.34	0.06	0.43	99.72
MW-WP-CJW	12.86	0.36	0.64	4.84	0.09	0.02	3.36	0.02	76.92	0.07	0.55	99.72
MW-2-BY	12.84	0.29	0.91	5.1	0.06	0.06	3.16	0.01	76.87	0.05	0.35	99.72
MW-2-CJW	10.51	0.23	1.43	3.83	0.15	0.04	2.14	0.03	79.12	0.12	2.2	99.78
MW-R	16.07	0.1	3.1	1.56	0.27	0.02	0.16	0.01	71.02	0.2	7.24	99.75

Table 2. Percentage of minerals of weathering pits and parent rocks of the three study sites (%).

Study Site	Sample No.	Quartz	Feldspar	Illite	Montmorillonite	Others
	Study site A (Loashan Mountain)
A	LD-1-BY	20.2	65.4		14.4
	LD-1-CJW	26.2	48.5	4.3	1.1	3.9 (Amphibole)
	LD-2-BY	29.4	61.8		8.7
	LD-2-CJW	27.8	50		22.2
	LD-3-BY	24.5	67.5		8
	LD-3-CJW	34.6	58.3		7.2
	LD-4-BY	18.1	61		20.8
	LD-4-CJW	36.7	63.3
	LYG-BY	18.3	77.9		3.8
	LYG- CJW	27.9	67.1		5
	YNP-BY	17.9	65		17.1
	YNP-CJW	20.5	67.3		12.2
	HL-1-1-BY	18.6	66.8		14.6
	HL-1-1-CJW	29.6	55.2		15.2
	Study site B (Pingshan Mountain/the Sanxi River)
B	SXHG-CJW	78.4	21.6
	SXHG-BY	16.3	74.7	3.5	5.5
	TLJ-16-BY	32	64.4	3.6
	TLJ-16-2-CJW	76	14		9.9
	TLJ-16-5-CJW	76	14		9.9
	TLJ-15-BY	41.1	55.2	3.7
	TLJ-15-CJW	79.6	20.4
	Study site C (Miaowan Island)
C	MW-2-BY	16.5	63.4	9.2	10.9
	MW-2-CJW	30	48.8	7.8	13.4
	MW-WP1-BY	22.8	54.7	14.5	8.1
	MW-WP1-CJW	33.8	59			7.2 (Mica)

Table 2. Percentage of minerals of weathering pits and parent rocks of the three study sites (%).

Study Site	Sample No.	Quartz	Feldspar	Illite	Montmorillonite	Others
	Study site A (Loashan Mountain)
A	LD-1-BY	20.2	65.4		14.4
	LD-1-CJW	26.2	48.5	4.3	1.1	3.9 (Amphibole)
	LD-2-BY	29.4	61.8		8.7
	LD-2-CJW	27.8	50		22.2
	LD-3-BY	24.5	67.5		8
	LD-3-CJW	34.6	58.3		7.2
	LD-4-BY	18.1	61		20.8
	LD-4-CJW	36.7	63.3
	LYG-BY	18.3	77.9		3.8
	LYG- CJW	27.9	67.1		5
	YNP-BY	17.9	65		17.1
	YNP-CJW	20.5	67.3		12.2
	HL-1-1-BY	18.6	66.8		14.6
	HL-1-1-CJW	29.6	55.2		15.2
	Study site B (Pingshan Mountain/the Sanxi River)
B	SXHG-CJW	78.4	21.6
	SXHG-BY	16.3	74.7	3.5	5.5
	TLJ-16-BY	32	64.4	3.6
	TLJ-16-2-CJW	76	14		9.9
	TLJ-16-5-CJW	76	14		9.9
	TLJ-15-BY	41.1	55.2	3.7
	TLJ-15-CJW	79.6	20.4
	Study site C (Miaowan Island)
C	MW-2-BY	16.5	63.4	9.2	10.9
	MW-2-CJW	30	48.8	7.8	13.4
	MW-WP1-BY	22.8	54.7	14.5	8.1
	MW-WP1-CJW	33.8	59			7.2 (Mica)

Table 3. CIA, Q/F and Na₂O/K₂O values of weathering pits of the three study sites.

Study Site	Sample No.	CIA	Q/F	Na2O/K2O
	Study site A (Laoshan Mountain)
A	LD-1-BY	52.13	0.31	0.94
	LD-1-CJW	54.98	0.54	0.88
	LD-2-BY	51.84	0.48	0.96
	LD-2-CJW	54.99	0.56	0.79
	LD-3-BY	52.48	0.36	0.91
	LD-3-CJW	57.67	0.59	0.76
	LYG-BY	52.21	0.24	0.82
	LYG- CJW	51.55	0.42	0.68
	YNP-BY	52.56	0.27	0.72
	YNP-CJW	53.34	0.31	0.67
	HL-1-1-BY	52.42	0.28	0.78
	HL-1-1-CJW	55.35	0.54	0.77
	BYSK-2	84.73	-	0.06
	Study site B (Pingshan Mountain/the Sanxi River)
B	TLJ-15-BY	49.16	0.75	1.01
	TLJ-15-CJW	61.29	3.9	0.28
	TLJ-16-BY	53.58	0.5	0.63
	TLJ-16-2-CJW	64.08	3.12	0.43
	TLJ-16-5-CJW	70.37	5.43	0.20
	SXHG-144F-CJW	72.04	3.63	0.25
	SXHG-143F-CJW	69.53	1.41	0.32
	SXHG-BY	52.52	0.22	0.82
	SXHG-R	98.26	-	0.19
C	Study site C (Miaowan Island)
	MW-2-BY	53.35	0.26	0.19
	MW-2-CJW	56.84	0.61	0.62
	MW-WP-BY	52.35	0.42	0.66
	MW-WP-CJW	53.07	0.57	0.69
	MW-R	88.28	-	0.10

Table 3. CIA, Q/F and Na₂O/K₂O values of weathering pits of the three study sites.

Study Site	Sample No.	CIA	Q/F	Na2O/K2O
	Study site A (Laoshan Mountain)
A	LD-1-BY	52.13	0.31	0.94
	LD-1-CJW	54.98	0.54	0.88
	LD-2-BY	51.84	0.48	0.96
	LD-2-CJW	54.99	0.56	0.79
	LD-3-BY	52.48	0.36	0.91
	LD-3-CJW	57.67	0.59	0.76
	LYG-BY	52.21	0.24	0.82
	LYG- CJW	51.55	0.42	0.68
	YNP-BY	52.56	0.27	0.72
	YNP-CJW	53.34	0.31	0.67
	HL-1-1-BY	52.42	0.28	0.78
	HL-1-1-CJW	55.35	0.54	0.77
	BYSK-2	84.73	-	0.06
	Study site B (Pingshan Mountain/the Sanxi River)
B	TLJ-15-BY	49.16	0.75	1.01
	TLJ-15-CJW	61.29	3.9	0.28
	TLJ-16-BY	53.58	0.5	0.63
	TLJ-16-2-CJW	64.08	3.12	0.43
	TLJ-16-5-CJW	70.37	5.43	0.20
	SXHG-144F-CJW	72.04	3.63	0.25
	SXHG-143F-CJW	69.53	1.41	0.32
	SXHG-BY	52.52	0.22	0.82
	SXHG-R	98.26	-	0.19
C	Study site C (Miaowan Island)
	MW-2-BY	53.35	0.26	0.19
	MW-2-CJW	56.84	0.61	0.62
	MW-WP-BY	52.35	0.42	0.66
	MW-WP-CJW	53.07	0.57	0.69
	MW-R	88.28	-	0.10

For all study sites (study sites A-C), CJW in the sample number represents the samples taken from particles remaining in the weathering pits, and BY denotes the samples taken from surface rocks near the weathering pits. In study site A (Figure 1a), HL-1-1, YNP, and LYG represent samples collected from rocks at sampling point 1 in Laoshan Mountain, LD represents samples collected from rocks at sampling point 2 (Laoding of Laoshan Mountain), and BYSK-2 is the sample taken from deeply weathered deposits near the Baiyun Reservoir of Laoshan Mountain (sampling site 3 in Figure 1a; Figure 2a-3). In study site B (Figure 1b), SXHG represents samples taken at sampling point 1 (the Sanxi River valley), TLJ represents samples taken at sampling point 2 (the top of Pingshan Mountain), and SXHG-R is the sample taken from the weathered crust, located on the northern slope of Sanxi River valley (sampling site 3 in Figure 1b; Figure 2b-3). On Miaowan Island (Figure 1c), MW-WP and MW-2 are the samples taken from rocks at sampling points 1 and 2, respectively, and MW-R represents the sample at sampling point 3, which is a weathering crust exposed on the northern slope of Miaowan Island (Figure 1c and Figure 2c-3).

Table 1 shows that the major chemical compounds in the granite of the three study sites are mostly SiO2, which accounts for about 77% of the total on average. This indicates typical granite compositions [44] with subordinate Al₂O₃, K₂O, Na₂O and Fe₂O₃, and minor CaO, MgO, TiO₂, P₂O₅ and MnO. On average, MnO contributes less than 0.5% to the total.

According to Table 2, all samples taken from the three study sites contain quartz and feldspar, but their assemblage of clay minerals is irregular. Except for sample LD-4-CJW, almost all samples collected from Laoshan mountain (study site A) contain montmorillonite. Only one sample (LD-1-CJW) taken from the same mountain contains illite and amphibole (Table 2). In Pingshan Mountain and the Sanxi River valley (study Site B), only one sample (SXHG-BY) contains both illite and montmorillonite, whereas the rest contain either illite or montmorillonite (Table 2). Samples collected from Miaowan Island contain illite and montmorillonite, except for sample MW-WP1-CJW, which contains mica, which is absent from all other samples (Table 2).

When the samples taken from local weathering crusts and deposits are excluded, Table 3 indicates that the CIA values range between 51.48 and 57.6 for samples taken at study site A (Loashan Mountain), 49.16 and 72.04 for samples taken at study site B (Pingshan Mountain and the Sanxi River valley) and 52.35 and 56.84 for samples taken at study site C (Miaowan Island), respectively; the values of the Q/F ratio range from 0.27 to 0.59 for study site A, from 0.22 to 5.43 for study site B and from 0.26 to 0.61 for study site C, respectively; and the values of the Na₂O/K₂O ratio change from 0.68 and 0.94 for the samples taken from study site A, from 0.2 to 1.01 for study site B and from 0.19 to 0.69 for study site C (Table 3). The samples of weathering crusts and deposits of the three study sites had much higher CIA and much lower Na₂O/K₂O ratio values ranging from 84.73 to 98.26 and 0.06 to 0.19, respectively (Table 3).

4. Discussion

4.1. Difference in CIA Values inside and outside Weathering Pits

CIA has been widely used as a geochemical indicator to determine the degree of chemical weathering and climatic conditions; CIA values between 50 and 65 reflect weak chemical weathering in cold and dry climates; CIA values between 65 and 85 reflect moderate chemical weathering in warm and humid conditions, while CIA values between 85 and 100 reflect strong chemical weathering in hot and humid tropical and subtropical conditions [45]. Plagioclase is rich in sodium, while potassium feldspar, illite and mica are rich in potassium [46]. As plagioclase has a much higher chemical weathering rate than potassium feldspar, the Na₂O/K₂O ratio in weathered samples is inversely proportional to the CIA values [41]. The scatter diagram of CIA versus Na₂O/K₂O can show the trend of chemical weathering intensity from the surrounding rock face through the debris in weathering pits to the weathering crust (Figure 3). However, according to the chart detail, all surface rock samples have Na₂O/K₂O values greater than 0.6, whereas more than half of the debris samples taken from weathering pits have values greater than 0.6 as well. In other words, all rock surface samples and more than half of the pit samples fall within circle A in Figure 3. The red horizontal line in Figure 3 indicates that the samples in circles B and C have overlapping values at lower values of the Na₂O/K₂O ratio. Jenny (1931) was the first to use the total K₂O to Na₂O ratio to identify the degree of chemical weathering in rock [47]. However, Ruxton (1968) believed that this index did not reflect weathering changes [48]. Figure 3 suggests that in the case of this study, CIA values are more effective than the Na₂O/K₂O ratio in identifying chemical weathering degrees between weathering pits and nearby rock surfaces.

The CIA values for weathered crusts (and deposits) on Miaowan Island and in the Pingshan and Laoshan Mountains were 88.28, 98.26 and 84.73, respectively (Table 3), and they are higher than those of weathering pits and surrounding rock surfaces in the same area because of the significant mass loss of more soluble major elements during chemical weathering and pedogenesis of the crust and deposits [45]. Obviously, the CIA values of weathering pits and nearby rock surfaces are not representative of the degree of chemical weathering caused by local climatic or weathering conditions. Processes that increase the contact time of water–rock interactions increase chemical weathering rates in granitic environments [49]. Weathering pits formed within a 60-year period have been observed by Wang et al. [17]. Therefore, the difference in CIA values between the weathered debris in the pits and the locally weathered crusts may be due to the difference in the duration of chemical weathering. In other words, despite being formed under the same climatic conditions, the chemical weathering time of the weathering pits is shorter than that of the local weathering crusts. Then, the CIA values of weathering pits in different climatic zones cannot be used as an indicator to identify climate types. Additionally, it indicates that weathering pits can form in these areas within a short period of time, independent of any glaciation.

On the other hand, in this study, all fragments within the weathering pits had CIA values higher than those of the adjacent rock faces. This indicates the origin of the weathering pits from chemical weathering. However, rock surfaces near weathering pits cannot be regarded as parent rocks as they have a higher CIA value than true parent rocks. For example, in study site B (Laoshan Mountain), where the climate is warm-temperate, the CIA value of the rock surface averaged 53.46 (Table 3), whereas an average CIA value of 50.37 was reported for local parent rock [43], indicating that chemical weathering occurs on the rock surface next to the pits as well.

4.2. Differences in Mineral Composition inside and outside Weathering Pits

According to Goldich, minerals are arranged in order of decreasing vulnerability to alteration: a mineral crystallizing early is less resistant to chemical weathering, while a mineral crystallizing later is more resistant; thus, quartz is situated at the end of this sequence [50]. Granite is composed mainly of quartz and feldspar with minor amounts of mica, amphiboles, and other minerals. Chemical weathering is most effective when water is present [51]. Quartz does not dissolve in water and is more resistant to chemical weathering than other granite-forming minerals such as feldspar and mica, which are either completely or partially dissolved to form clay [51]. As shown in Table 2, the contents of quartz are higher in the samples taken from debris in weathering pits (the samples denoted by CJW) than those taken from the nearby rock surface (the samples denoted by BY) suggesting that chemical weathering is heavier in a weathering pit than in surrounding rock surfaces next to the pit. If the formation of weathering pits is the result of chemical weathering, the percentage of feldspar in debris from weathering pits should decrease. In contrast, quartz is relatively more abundant [52]. By comparing the Q/F ratios of the weathering pits with the surface rocks surrounding the pits, one can illustrate the difference in the chemical weathering degree between the two. Regardless of whether the debris comes from inland or coast, and from north or south, the Q/F ratio values in weathering pits are always higher than those of adjacent rock surfaces (Table 3). Again, this strongly indicates that the weathering pits were formed as a result of chemical weathering.

The following chemical equations describe the processes in which water interacts with minerals to create various chemical reactions and transform between feldspars and related clay minerals [53]:

                       potassium feldspar                           kaolinite

K[AlSi₃O₈] + H⁺ + OH⁻ → Al₄Si₄O₁₀(OH)₈ + H₄SiO₄ + K⁺

(3)

               potassium feldspar                     illite

2K[AlSi₃O₈] + H⁺ + OH⁻ → KAl₂Si₄O₁₀(OH)₂·nH₂O + H₄SiO₄ + K⁺

(4)

                                 plagioclase                    montmorillonite

(Na, Ca) [AlSi₃O₈] + Mg²⁺ + H⁺OH⁻ → (Ca, Al, Mg) Si₄O₁₀(OH)₂·nH₂O + H₄SiO₃ + K⁺

(5)

                                                       illite                     kaolinite

KAl₂Si₄O₁₀(OH)₂·nH₂O + H⁺ → Al₄Si₄O₁₀(OH)₈ + K⁺

(6)

                                kaolinite                        aluminite trihydrate

Al₄Si₄O₁₀(OH)₈ + OH⁻ + H₂O → Al₂O₃·3H₂O + H₄SiO₄

(7)

The above reactions denote the alteration from potassium to illite to kaolinite and from plagioclase to montmorillonite.

If the rock surface adjacent to the pit has also been chemically weathered, both the samples of the weathering pit and the surrounding rock surface should contain clay minerals as a result of chemical weathering, based on the reaction formulas described above. Feldspar mineral in Laoshan granite is primarily potassium feldspar, but plagioclase is also locally present [43], so all but one sample contained montmorillonite (Table 2). However, the clay mineral assemblages are more complicated in the other two study sites. For instance, as shown in Table 2, most of the samples, including debris in weathering pits and nearby rock surfaces, taken from Miaowan Island (study site C) have illite, whereas only one sample taken from Laoshan Mountain has illite. The granite bedrock of Miaowan Island contains plagioclase (40.3%) and potassium feldspar (16.3%). In contrast, Laoshan granite contains potassium feldspar up to 65–75 percent [35,43]. During the hydrolysis process of chemical weathering, potassium feldspar produces kaolinite or illite and plagioclase produces monazite (montmorillonite and saponite) [53]. Laoshan Mountain and Miaowan Island have different parent rocks, which results in different clay minerals in their weathering pits. This is also because the role of feldspar minerals must depend not only on their total abundance but also on their present content and stages of chemical weathering or formation [54]. Additionally, some samples from Sites B and C contain clay minerals whereas others do not (Table 2), suggesting that the differences in clay mineral variation between debris in weathering pits and surrounding rocks are not as regular as indicated by the CIA and Q/F (Table 3). Chemical weathering can lead to the transformation of clay minerals rather than elemental changes, complicating the relationship between element-based parameters and chemical weathering intensity [55]. Further research is necessary to understand this complexity.

4.3. Major Element Migrations during the Formation of Weathering Pits

During chemical weathering, changes in the chemical composition of rocks are mainly manifested by the hydration of silicate minerals and the leaching of labile components such as alkali metals (K and Na) and alkaline earth metals (Ca and Mg) [56]. Figure 4 illustrates the mass transfer coefficients for three sample groups based on their chemical weathering intensity relative to their parent rock. The chemical weathering intensities of the three groups are classified according to CIA values (Figure 3). As expected, the strongest transfer takes place in the strongly weathered group and the weakest transfer in the weakly weathered group (Figure 4). In other words, the formation of weathering pits was controlled by chemical weathering. However, more information about mass transfer between weathering pits and nearby rock surfaces in chemical weathering processes is needed.

The chemical weathering process disaggregates rock, but it does not remove rock material from its original location. When weathering pits are formed as a result of chemical weathering occurring on the rock surface where the pits are situated, it is possible to determine more details about the mass transfer characteristics between the rock surface and the debris in weathering pits by the mass transfer coefficient τ_j between them. If the increase in the Q/F ratio or CIA values of the debris in weathering pits is a consequence of the chemical weathering of feldspar, element migration, particularly alkali and alkaline earth metal elements, should be expected in the rock system between the debris in weathering pits and the nearby rock surface. According to Formula 2, element migration in the rock system between the pit and the rock surfaces of different study sites is shown in Figure 5. As shown in Figure 5, τ_j > 0 for most elements in weathering pits indicates external additions [57], which may result from receiving external materials in the pits (see later). Only elements of Na and K always have τ_j < 0 (Figure 5) in all samples, indicating that Na and K are constantly leached during weathering pit formation, regardless of whether it takes place in valleys, mountains or coastal regions. This is also the reason why the Q/F ratio and CIA values in this study can be used to assess chemical weathering intensities.

Some elements, such as Ca and Mn, show a variation in τ_j between negative and positive values in different samples (Figure 5). The contents of Mg and Ca in the pit fragments of Laoshan Mountain are several times more abundant than those in the surrounding rock surface (Table 1). The MgO content varies between 0.01% and 0.18%, and the CaO content varies between 0.09% and 0.46%, respectively, in the surrounding surface rock. However, the contents of the same elements in the weathering pit debris range from 0.21% to 1.03% and 0.22% to 0.98%, respectively (Table 1). Experimental studies have demonstrated a non-stoichiometric release of Sr > Ca > Na from plagioclase due to the preferential leaching of surfaces and exsolution lamellae [58]. Granites in Laoshan Mountain are not enriched in plagioclase feldspar [43], indicating that most of the enrichment of Ca (as well as Mg) in the weathering pits of Loashan Mountain was not produced by chemical weathering processes or released from the parent rock.

When the mass transfer coefficient of an element is equal to negative 1 (τ_j = −1), the element is completely mobilized [42]. As the Sanxi River cut into Pingshan Mountain to form the Sanxi River valley, the rock surface of the mountaintop may be older than that of the river valley [13]. Although the valley and the mountain share the same climatic zone, the largest negative value of Na and K is detected in the weathering pits in the valley (Figure 5). This suggests that river valleys have a more conducive environment for chemical weathering than mountain tops. This is most likely due to the fact that river valleys are generally wetter than mountain tops.

Studies have shown that atmospheric settlements in the area of Qingdao city, where Laoshan Mountain is located, contain large amounts of K, Na, Ca and Mg [59]. The deposition fluxes of dissolved K, Na, Ca and Mg in Qingdao rainwater are 10 times higher than the world average deposition fluxes [59]. Therefore, Ca and Mg show an abrupt increase in τ_j values because of external additions from the atmosphere, whereas Na and K exhibit negative τ_j values due to the lack of Ca and Mg but an enrichment in Na and K in the rocks of Laoshan granite (Table 1).

Only the samples collected from Miaowan Island had a negative value of τ_j for Mn (Figure 5). The reason for this may be that seawater samples collected from the island coast have high pH values. The seawater in the northern South China Sea is alkaline with a pH value of 8.15 [60]. In addition, moisture around an island may also be alkaline due to the effects of wave spray and sea fog [61,62]. At a pH value of 8 for seawater or 5–7 for fresh surface water, Mn should be soluble [63]. As a result, the pattern of Mn migration in the weathering pits of Miaowan Island differs from that of other study sites.

5. Conclusions

An investigation of mineral weathering and element migration in weathering pits formed at different topographic locations at three selected sites in eastern China is presented in this study. The above analysis leads to the following conclusions:

(1)

CIA values for debris within weathering pits are higher than those for adjacent rock surfaces, indicating that the weathering pits resulted from chemical weathering. However, the CIA values of weathering pits in different climatic zones cannot be used as an indicator to identify local climate types. CIA values of different weathering pits differ only because of differences in chemical weathering times.

(2)

When assessing the chemical characteristics of weathering pits, the rock surfaces adjacent to the pits cannot be regarded as the parent rocks of the pits. This is because the rock surface next to the pits is also weathered and generally has higher CIA values than local parent rock.

(3)

The Q/F ratios of debris in weathering pits and surrounding rock surfaces may also indicate the chemical origin of the weathering pits. It is because the ratio is generally lower at the surface of the rock and higher in the debris in the pits. CIA is more effective than Na₂O/K₂O in indicating the degree of chemical weathering in weathering pits.

(4)

The magnitude of mass transfer intensities of the rock samples relative to their parent rocks coincides with that of CIA values of the samples, also indicating the chemical origin of the weathering pits.

(5)

Based on the mass transfer between debris in weathering pits and nearby rock surfaces, only elements Na and K always have τ_j < 0. This indicates that Na and K are continuously leached during the formation of weathering pits, regardless of whether they are situated in valleys, mountains or coastal regions.

(6)

Clay mineral assemblages are more complicated and less effective than the major minerals in assessing the degree of chemical weathering in weathering pits.

(7)

Chemical weathering conditions in coastal areas differ from those in inland areas, and migration patterns of elements in weathering pits are also different.

(8)

As a result of wet conditions in river valleys, chemical weathering processes are more likely to occur in the valley than on the top of mountains.