Introduction

Progressive Indo-Eurasian convergence induces numerous lithospheric-scale strike-slip faults and orogenic thrusts, which partly account for the Tibetan Plateau formation and its margin construction during the Cenozoic (Tapponnier et al. 2001; Royden et al. 2008). Large-scale Tibetan crustal mass is currently moving eastwards and rotating clockwise around the eastern Himalayan syntaxis (Fig. 1). Most of the motion is redirected into northeast and southeast flow around the Sichuan Basin (Clark and Royden 2000; Zhang et al. 2004b; Enkelmann et al. 2006), due to the backstop from the craton-like lithosphere of the Sichuan Basin (Copley 2008; Liu et al. 2012). Although no significant shortening has been observed across the eastern and southeastern margin of the Tibetan Plateau by geodetic and geologic studies (Burchfiel et al. 1995; Wang et al. 1998; Zhang et al. 2004b; Shen et al. 2005), the margin has experienced extensive deformation and denudation during Cenozoic time due to the far-field effect of the Indo-Eurasia convergence. There is widespread thrusting within the Longmen Mountains, and strike-slip motion along the Xianshuihe fault, the Red River fault, and other large-scale strike-slip systems (Wang et al. 1998; Hubbard and Shaw 2009; Zhang et al. 2010), resulting in significantly different topography along the eastern margin of the Tibetan Plateau (Kirby et al. 2002; Clark et al. 2005) (Fig. 1). Low-temperature thermochronology investigations have yielded valuable insights into the exhumation history of the eastern margin of the Tibetan Plateau (Fig. 1). Late Cenozoic cooling ages at the Songpan–Ganzi area reflect regional denudation of the eastern Tibetan Plateau. This regional denudation has been interpreted as a response to crustal thickening (Arne et al. 1997; Kirby et al. 2002; Li et al. 2012), lower crustal channel flow (Clark and Royden 2000; Mukherjee 2005; Godard et al. 2009a), and drainage reorganization (Richardson et al. 2008). Recently, Ouimet et al. (2010) inferred a relatively constant and uniform regional uplift at ~10 Ma in the southeastern margin of the Tibetan Plateau, by using zircon and apatite (U–Th)/He data. However, Wang et al. (2012a) argued that two phases of rapid exhumation took place at 30–25 and at 15–10 Ma ago, respectively. Most of the previous studies concentrated on the Songpan–Ganzi fold and thrust belt, but studies at further southeastern margins of the Tibetan Plateau are rare (Fig. 1).

Fig. 1
figure 1

Map of the southeastern margin of the Tibetan Plateau and the Daliang Mountains. The locations of sample transects for this study are indicated by blue stars (vertical transects) and black lines (vertical vs. depth transects). Recent thermochronological data of age–elevation profiles are indicated in yellow, white, purple, gray, deep gray, black stars, and red star are from Liu et al. (2008); Richardson et al. (2008), Li et al. (2012); Clark et al. (2005), Ouimet et al. (2010), Godard et al. (2000b), and Wang et al. (2012a, b), respectively. The ~0.25 mm/year, ca. 9 Ma indicate that rapid exhumation with rate of ~0.25 mm/year began at ca. 9 Ma. Inset shows the location of eastern margin of Tibetan Plateau with the box that outlines the study area. The Indian plate motion vector with 36–40 mm/year showed by black arrow is after Zhang et al. (2004a, b); the red arrows show the eastward growth of Tibetan Plateau. EHS—eastern Himalayan syntaxis (the red triangle), XSH F.—Xianshuihe fault, DLS F.—Daliangshan fault, ANH F.—Anninghe fault, XJ F.—Xiaojiang fault, MPS F.—Mopanshan fault, JH F.—Jinghe fault, GS G.—Gonga shan granites. a Xide transect, b Mianshan transect, c Ganluo transect, d Wushihe transect, e Muchuan transect, f Wudu transect

Full size image

The Daliang Mountains represent the boundary between the southeastern Tibetan Plateau and the Sichuan Basin (Fig. 1). This paper presents new apatite and zircon (U–Th)/He (AHe and ZHe, respectively) thermochronometry results from the Daliang Mountains, which indicate multiple periods of cooling and denudation during the Cenozoic transpressional strike-slip tectonics. The new data supports an interpretation that (1) cooling and exhumation with apparent rates of ~0.15 mm/year from ~30 to ~10 Ma were pervasive across the region and (2) Late Cenozoic eastward growth of the Tibetan Plateau controlled the rapid post ~10-Ma cooling and denudation in the Daliang Mountains. The results from the Daliang Mountains provide an example of how low-temperature thermochronometry can provide additional information on the denudation and mountain building at low-relief topographic regions, especially in absence of syn- to post-deformational strata.

Geological setting

The ~300-km-long and ~80-km-wide Daliang Mountains chain marks the southwestern boundary of the Sichuan Basin and of the southeastern margin of the Tibetan Plateau. The Daliang Mountains are located at the western margin of the Yangtze craton (South China Block) and are nearly N–S trending (Fig. 1). This mountain range is characterized by north–south-trending low- to medium-grade metamorphic rocks of Proterozoic and Archean age (Zhang et al. 1990; BGMRSP 1991; Luo 1998; Zheng et al. 2006), which comprise the basement of the Yangtze craton. The area underwent complicated tectonic evolution from a continental margin during Paleozoic–Mesozoic times to collisional orogeny since the Late Triassic and Cenozoic (Burchfiel et al. 1995; Chen et al. 2011). The Phanerozoic strata are dominated by Silurian–Triassic marine clastic–carbonate sequences, and Permian continental flood basalts that widely cover the western margin of the Yangtze craton (Xu et al. 2001; He et al. 2007). Post Late Triassic terrestrial sediments are dominated by fluvial and lacustrine red bed facies with only sparse outcrops of Cenozoic rocks. The terrestrial sediments are laterally synchronous and comparable between the Panzhihua–Xichang area and the Sichuan Basin. Moreover, the similarity and comparability in sediment, depositional facies, and paleontology indicate that both basins were linked as a paleo-Sichuan Basin during Mesozoic time (Xia 1982; Zhang et al. 1990; BGMRSP 1991; Guo et al. 1996).

The Daliang Mountains comprise a series of N to NNW sinistral strike-slip faults (e.g., Jinghe fault, Anninghe fault, Xiaojiang fault). Our field studies indicate that the Daliang Mountains are weakly deformed and the strata outcrop predominantly horizontally, or with a low inclination. There are two different phases that are generally controlling the tectonic framework of the eastern Tibetan Plateau after the Late Triassic. During the Late Triassic, the Songpan–Ganzi units, comprised by thick sequence of deep marine Triassic strata, often called the Songpan–Ganzi flysch (BGMRSP 1991), were thrust southeastward onto the Yangtze craton (Sichuan Basin), along the Longmen thrust belts, to form the western Sichuan foreland basin (Fig. 1; Chen et al. 1995; Worley and Wilson 1996; Deng et al. 2012a). The thrusting along the Jinghe fault farther south formed the Panzhihua–Xichang Basin (Fig. 1; BGMRSP 1991; Burchfiel et al. 1995). Thrusting with sinistral strike-slip occurred initially during the Late Triassic (Dirks et al. 1994; Chen et al. 1995; Worley and Wilson 1996), but was reactivated during Cenozoic time as a dextral strike-slip in the north and sinistral strike-slip in the south (Chen et al. 1994; Burchfiel et al. 1995; Wang and Burchfiel 2000). The reactivation of older structures has been interpreted as the result of the Late Cenozoic eastward or southeastward extrusion of crustal material from the Tibetan Plateau and subsequent clockwise rotation around the eastern Himalayan syntaxis (Wilson et al. 2006; Schoenbohm et al. 2006a; Royden et al. 2008). The Xianshuihe–Anninghe–Xiaojiang fault forms the natural southeastern boundary for the clockwise rotation (Wang et al. 1998; He et al. 2008). This fault marks a major strike-slip system, where the Panzhihua–Xichang area or even the southeastern margin of the Tibetan Plateau seems to be moving actively by sinistral strike-slip for more than ~60 km, relative to the western edge of the Sichuan Basin (Wang et al. 1998; Zhang et al. 2004b; Shen et al. 2005; Schoenbohm et al. 2006b). 40Ar/39Ar, U–Pb and Rb–Sr data indicate that a rapid cooling event occurred along the Xianshuihe fault zone during 12–10 Ma, coevally with the emplacement and deformation of the Gonga Shan granite (Roger et al. 1995; Zhang et al. 2004a). The Gonga Shan region accommodates extensive uplift and denudation at the southeastern margin of the Tibetan Plateau during this time (Xu and Kamp 2000; Lai et al. 2007; Clark et al. 2005; Ouimet et al. 2009, 2010).

Sample transects and methods

In the Daliang Mountains, samples were collected on three transects along one limb of an anticline, with wavelengths more than 20 km (e.g., Xide, Ganluo, Muchuan). The isotherms represented by helium age were probably perturbed by the deformation. Samples comprise Late Triassic to Early Cretaceous strata within a ~2-km-long stratigraphic profile (Table 1; Figs. 2, 3, 4). The other samples were collected from 3–1.1 km vertical transects in Precambrian granites (e.g., Mianshan granite, Wudu granite, Wushihe granite), over a short horizontal distance of ~4–9 km (Fig. 5). All the anticlines that were sampled are characterized by gentle-to-moderate deformation, whereas the granites are undeformed and lack significant foliation and/or lineation.

Table 1 Apatite and zircon (U–Th)/He results
Full size table
Fig. 2
figure 2

Simplified geologic map, structural profile, reconstructed stratigraphic column, and sample location of the Xide transect. The error in estimated stratigraphic thickness and sample location could be up to ~100 m

Full size image
Fig. 3
figure 3

Simplified geological map, structure profile, reconstructed stratigraphic column, and sample location of Ganluo transect. The error associated with the stratigraphic thickness and sample location can be up to ~ 100 m

Full size image
Fig. 4
figure 4

Simplified geological map, structure profile, reconstructed stratigraphic column, and sample locations at the Muchuan transect. The error associated with stratigraphic thickness can to be up to ~200 m

Full size image
Fig. 5
figure 5

Sample location maps for the Mianshan granite (a), Wudu granite (b) and Wushihe granite (c). Topography maps are from Google Earth

Full size image

The Muchuan and Wudu transects are located at the southeastern margin of the Daliang Mountains at elevations between 400 and 1,200 m (Fig. 4). The Ganluo and Wushihe transects are located within the Daliang Mountains at elevations between 600 and 1,300 m (Fig. 3), and the Xide and Mianshan transects are at the western margin of the Daliang Mountains at elevations between 1,700 and 3,000 m (Fig. 2).

Based on detailed orientations of the stratigraphy in the structural profiles, we can construct a stratigraphic column (Figs. 2, 3, 4). Because no faults are present and insignificant thickness variations are observed within each section, we argue that the error in estimated stratigraphic thickness and sample location is less than ~100–200 m and that the restored stratigraphic column from a single anticlinal structure can be used as a proxy for original depth. Thus, a plot of stratigraphic depth (structural depth) versus sample age can be constructed to calculate the exhumation rate over time. The other three transects in granites can be used for a comparison, using an age versus elevation relationship. All exhumation rates were derived by the linear regression of the structural depth/elevation versus ages using Isoplot (Ludwig 2003).

We used apatite and zircon (U–Th)/He thermochronometry on minerals separated from sedimentary and granitic rocks in order to gain information about the cooling history of these rocks. For an effective grain radius of ~60 μm and a cooling rate of ~10 °C/Myear, the closure temperatures of the (U–Th)/He system in apatite and zircon are ~65 °C (Wolf et al. 1998; Farley 2000), and ~180 °C (Reiners et al. 2004), with their relative helium partial retention zone (HePRZ) of ~50–80 and ~160–200 °C (Wolf et al. 1998; Stockli 2005), respectively.

Apatite and zircon grains were separated from rocks by crushing, sieving, and washing the samples, and using standard magnetic and heavy liquid separation techniques. Clear and undisturbed apatite and zircon grains without inclusions were selected using a binocular microscope. The grain dimensions were measured for the calculation of the alpha correction factor after Farley et al. (1996). Afterward individual grains (usually three replicates per sample, and each replicate was a single grain) were, respectively, packed in Nb-tubes for (U–Th)/He analysis. The helium gas was first extracted in the Patterson helium extraction line equipped with a 960-nm diode laser at the University of Tübingen. Each sample was reheated at the same conditions to ensure that all helium was extracted. The re-extracted component of helium gas showed typically <1 % of the first signal. After Helium analysis, the grain packages were sent to the University of Arizona at Tucson for U, Th, and Sm measurements using an ICP-MS. The analytical errors of the mass spectrometer measurements are generally very low and do not exceed 2 %. In contrast, the alpha correction factor and reproducibility of the sample age constitute a much larger error. We therefore report the mean (U–Th)/He age and the standard deviation of the measured aliquots as the sample error (1σ).

Results

Fifty-nine AHe ages (18 AHe samples), ranging from 4 to 50 Ma, and 63 ZHe ages (22 ZHe samples), ranging from 5 to 230 Ma, were measured (Table 1). Thirteen samples are Upper Triassic to Upper Jurassic sandstones, and twenty samples are Precambrian granites. A summary of all single-grain ages and mean ages is presented in Table 1. In most cases, apatite and zircon single-grain ages are substantially younger than depositional ages of the host rocks (or crystallization age of granite), and apatite ages are systematically younger than zircon ages from the same rock sample, or from the same stratigraphic column. ZHe ages of sediment samples with single-grain ages older than the depositional age are unreset, or partially reset (as apparent or mixed age) in the cases where some single-grain ages are younger and some are older than the deposition age.

Xide transect

Three samples were analyzed from the Xide transect located at the southeastern limb of Lianghekou anticline (Fig. 2; Table 1). The five AHe and six ZHe ages range from 4.2 ± 0.2 to 11.9 ± 0.7 Ma and from 19.6 ± 1.3 to 49.3 ± 4.0 Ma, respectively. The anticline is characterized by NE striking terrestrial sequences of the Baiguowan Formation that is Late Triassic to Early Jurassic in age (T3–J1 b), and by the Upper Cretaceous Leidashu Formation (K2 l).

AHe ages of sample SYX06 from the top of the stratigraphic column range from 4.2 ± 0.2 to 11.9 ± 0.7 Ma with a mean age of 7.4 ± 4 Ma. The sample SYX01 from the base of the column yielded a similar age with mean AHe age of 9.2 ± 0.5 Ma (Table 1). The ZHe ages in the same section show a distinct decrease with increasing depth with mean ages of 37.7 ± 10 Ma (SYX03) and 24.6 ± 4.5 Ma (SYX01; Fig. 2). It should be noted that there is a different trend in AHe and ZHe age with increasing depth, which could be correlated with a large variation in single-grain AHe ages in SYX06, or the topography influence of the AHe system that is of lower closure temperature than in ZHe system.

All cooling ages are significantly younger than the depositional age of the host rock, indicating that these samples had been exposed to temperatures >180 °C after deposition. Using the different closure temperatures and the ages of the AHe and ZHe system for our samples, we can calculate the different exhumation rates for various time intervals (Reiners et al. 2003; Reiners and Brandon 2006). The difference between mean AHe and ZHe ages for the sample SYX01 is ~16 Ma and yields an exhumation rate of ~0.21 mm/year from ca. 24.6 ± 4.5 to 9.2 ± 0.5 Ma, assuming a paleo-geothermal gradient of ~30 °C/km (Hu et al. 2000) and a surface temperature of 10 °C.

Ganluo transect

At the Ganluo transect, five samples were collected from the Late Triassic to Early Jurassic Baiguowan Formation (T3–J1 b) to Upper Jurassic Xincun Formation (J2 x) at the northeastern limb of the Yutian anticline (Fig. 3; Table 1). These samples provide thirteen AHe ages ranging from 4.6 ± 0.2 to 21.1 ± 3.1 Ma and only one ZHe age of 35.2 ± 3.4 Ma (Table 1). The NW striking anticline expose a sequence from the Emeishan basalt (P2 â) to the Niuguntang Formation (J2 n). The undeformed Neogene Xigeda Formation outcrops horizontally and overlies the gentle Yutian anticline, indicating that the deformation occurred between Early Cretaceous and Late Cenozoic time.

AHe ages from the sample at the top of the stratigraphic column (SS22) yielded a mean age of 7.9 ± 2.9 Ma, and the sample SS26 from the base yielded a mean AHe age of 6.7 ± 1.6 Ma (Table 1). Overall, these AHe data from the Ganluo transect indicate an apparent exhumation rate of 0.8+INF/−0.57 mm/year between ~9 and 5 Ma (Figs. 3, 6).

Fig. 6
figure 6

Plots of age versus structural depth relationships for a Muchan and b Ganluo transects. The bold lines define apparent exhumation rates (ER) with 95 % confidence intervals

Full size image

Furthermore, at the bottom of this transect, there is only one ZHe age of 35.2 ± 3.4 Ma from sample SS26. With only one replicate, it is difficult to be certain that the base of this transect is fully reset. Given that the ZHe age of the sample SS26 is substantially younger than the depositional age of the host rock (~175–220 Ma), we argue that the base of this transect is fully reset. Thus, the interval time of sample SS26 through the AHe and ZHe closure temperatures is ~29 Ma, suggesting an exhumation rate of ~0.13 mm/year from 35.2 ± 3.4 to 6.7 ± 1.6 Ma.

Muchuan transect

The Muchuan anticline transect is comprised by marine Permian to middle Triassic strata (P1–T2) and terrestrial Late Triassic to Early Jurassic (Xiangxi Group T3–J1 x) to Lower Cretaceous strata. Five samples were collected at the northeastern limb of the Muchuan anticline (Fig. 4; Table 1), which provided 18 AHe ages, ranging from 5.6 ± 0.7 to 50.2 ± 3.1 Ma.

Single-grain AHe ages from the top of the stratigraphic column (SQM05) show a significant range from 11.4 ± 0.9 to 50.2 ± 3.1 Ma and are interpreted to represent an exhumed AHe PRZ (Table 1; Figs. 4, 6). The samples that are stratigraphically lower show a systematic decrease in the spread of the single-grain ages and also a decrease of the mean AHe ages with increasing structural depth (Fig. 4). The mean AHe ages decrease from 26.4 ± 12 Ma to a mean age of 9.9 ± 5.5 Ma at the base of the section. The linear regression indicates an apparent exhumation rate of 0.13+0.04/−0.09 mm/year from ~30 to 9 Ma (Fig. 6).

Two samples in the lower part of the column yielded ZHe ages. At the base, the ZHe ages range from 132.0 ± 16.4 to 178.4 ± 20.1 Ma (Table 1), slightly younger than the depositional age. In contrast, the ZHe ages from sample SQM08 range from 21.1 ± 2.0 to 227.9 ± 16.7 Ma, with the two older ZHe ages predating deposition (Fig. 4). Excluding the 21.1 ± 2.0 Ma ZHe age that is unusually young, the ZHe ages within this transect display an increase in temperature with depth and reveal that none of these rocks had been exposed to temperatures >160–200 °C, high enough to reset zircon (U–Th)/He ages after deposition.

Mianshan transect

Eight samples within the Mianshan transect were collected from Precambrian granites (Fig. 5a). The ZHe ages in the Mianshan granite range from 8.1 ± 1.0 to 29.3 ± 2.8 Ma, with mean ages from 8.5 ± 0.4 to 23.2 ± 5.4 Ma (Table 1). The linear regression of the age–elevation plot shows an apparent exhumation rate of 0.15+INF/−0.12 mm/year between ~20 and 10 Ma (Fig. 7b). Due to the sample SYX14 far off the regression line, it is occurred a low probability of R 2 with 0.3.

Fig. 7
figure 7

Plots of age–elevation relationships for a, b Mianshan, c Wudu, and d Wushihe transects. The bold lines define apparent exhumation rates with 95 % confidence intervals

Full size image

The upper two samples SYX13 and SYX16 show AHe ages ranging from 19.1 ± 2.4 to 37.7 ± 3.2 Ma, and from 13.9 ± 1.1 to 44.2 ± 4.3 Ma (Fig. 7a). The samples below ~2,500 m elevation yielded mean ages of 10 ± 1.9 and 12.4 ± 1.5 Ma. The age–elevation plot suggests a negative relationship; however, the mean AHe ages are similarly within error and indicate a rapid uplift and exhumation at ~10 Ma (Fig. 7). The lowermost sample (SYX11) at the transect shows large scatter in single-grain AHe ages ranging from 5.4 ± 0.4 to 30.1 ± 3.1 Ma (Table 1). Due to the fact that the zircons of the same sample yielded a well-reproduced 11.1 ± 0.6 Ma ZHe age, we assume that the two older AHe ages were influenced by parentless helium from microinclusions. Thus, two potential causes should be considered to account for the significantly scattered AHe ages in the upper two samples.

The first one is that those samples with poor quality, e.g., microinclusions, indicating the single-grain ages >~20–30 Ma (the ZHe ages of the upper two samples), e.g., SYX13-2a and SYX16-3a, are probably influenced by the microinclusions. The second is that a long residence time in the AHe PRZ between ~10 and 20 Ma, diffusion magnified the difference in He concentration and resulted in a large variation in AHe ages, which was followed by a rapid uplift and exhumation at ~10 Ma. The lowermost sample (SYX11) with ZHe age of 11.1 ± 0.6 Ma indicates that the sample was above 160–200 °C at ~10 Ma. There is ~1500 m elevation difference across the Mianshan transect, with a 40–60 °C variation in temperature—slightly smaller than the difference between the apatite HePRZ (50–80 °C) and zircon HePRZ (160–200 °C). It means that the uppermost sample was probably located around the apatite HePRZ. Excluding the single-grain AHe ages >~30 Ma in the upper two samples, those samples still have large scatter in replicate ages (Fig. 7a). Thus, we prefer the second cause for interpretation of the variation in AHe ages, although some grains were not completely suitable for helium dating. Such an interpretation could be further resolved by apatite fission track dating.

Using the difference in the AHe and ZHe cooling ages (1.7–7.0 Ma) and the difference in the closure temperatures of the two systems, the data indicate rapid exhumation after 10 Ma with cooling at rates of ~24 °C/Ma (~0.8 mm/year), following a protracted denudation during ~20–10 Ma with rate of ~0.15 +INF/−0.12 mm/year (Fig. 7b).

Wudu and Wushihe transects

The Wushihe and Wudu transects are located along the Dadu River (Fig. 5c, b), where twelve ZHe samples were collected. The ZHe ages in the Wudu transect range from 15.4 ± 3.0 to 64.8 ± 7.4 Ma (Table 1). The uppermost sample, SS04, shows significant scatter in replicate ages. Samples below ~1,200 m show a negative age–elevation relationship (−1.89 +1.9/−41 mm/year), yet even so the scatter between single-grain ages is large (Fig. 7), indicating a rapid uplift and exhumation at ~20 Ma. It could be probably correlated to the deformation along the fault, as tilt related to deformation could result in such negative age–elevation relationship (Stockli 2005; Lee et al. 2013). The sample SS06 in the middle of the Wudu transect has single-grain ZHe ages ranging from 135 ± 14.6 to 484 ± 50.3 Ma, substantially older than the rapid cooling age and younger than the crystallization ages, which were probably influenced by the parentless helium.

The ZHe ages in the Wushihe transect yielded mean ages ranging from 2.9 ± 2.7 to 8.9 ± 3.3 Ma. Some of the single ZHe ages that are very old (from 192 ± 15.7 to 1,304 ± 140.5 Ma; Table 1; Fig. 7d) may reflect parentless helium from fluid inclusions. Moreover, one additional AHe age from a sample collected on the Dadu River, 20 km east of the main transect at ~580 m (sample SS12 in Fig. 5c), shows single-grain AHe age from 9.3 ± 1.1 to 22.1 ± 0.6 Ma, with a mean age of 15.7 ± 6 Ma. The age–elevation relationship of the Wushihe granite samples suggests an apparent exhumation rate of 0.35+INF/−0.11mm/year from ~10 to 5 Ma (Fig. 7d). However, the sample SS20 yielded very young ages of 0.4 ± 0.7 to 2.7 ± 0.6 Ma, which could be caused by local thrusting or ground-hot-water flow (Whipp and Ehlers 2007), given there is no volcanicities occurred during Late Cenozoic times across the sampled area.

Discussion

Variation of (U–Th)/He single-grain ages

Although most of the sample ages show consistency in replicates and a systematic relationship with structural depth/elevation, a few samples show a large scatter in replicate analyses (e.g., sample SQM11, SS23 and SS12, Table 1). Variations between sample replicates have been reported in an increasing number of studies and can be attributed to (1) parentless helium from microinclusion, (2) helium implantation, (3) radiation damages, and (4) zoning of U and Th (Fitzgerald et al. 2006; Flowers et al. 2007; Spiegel et al. 2009).

There are several anomalous ages that could be attributed to microinclusions, e.g., SS15_1z and SYX11_3a. It is difficult to dismiss the possible influence of poor quality of some samples that have affected the He ages. However, samples in the Muchuan and in the Mianshan transects show a systematic decrease in the spread of AHe and ZHe ages in a single grain, respectively (Figs. 6a, 7b). We thus do not consider microinclusions to be the predominant source of the observed scatter.

Spencer et al. (2004) suggested that the majority of the “too old” ages can be explained by helium implantation from surrounding U–Th-rich minerals and host sedimentary components. Spiegel et al. (2009) argued that He implantation is most pronounced in apatite with effective U concentration <5 ppm. Thus, most reliable results for AHe ages are probably attained from those samples where U concentrations are >5 ppm. We did not measure the U and Th concentration directly, but made a rough estimate of the effective U concentration in ppm by calculating the grain mass using the measured grain size and a 3.2 g/cm3 apatite density. The result shows that the effective U concentrations are relatively high (between 10 and 300 ppm; Table 1), which makes the possibility of He implantation from its surrounding unlikely. Furthermore, we do not observe a correlation between U concentration and AHe ages (Fig. 8). This is opposite to the trend of increasing age with increasing effective U concentrations resulting from radiation-induced damage to the apatite structure, as described by Shuster et al. (2006) and Flowers et al. (2007). Thus, He implantation and radiation damage trapping can be excluded as explanations for the variation in ages.

Fig. 8
figure 8

Relationship between AHe ages and effective U concentrations in different transects

Full size image

The AHe and ZHe system is thought to be an open system at temperatures between ~50–80 and 160–200 °C, respectively (Wolf et al. 1998; Reiners and Brandon 2006). That means when a sample underwent slow cooling, or prolonged residence in the PRZ, diffusion could potentially magnify the difference in He concentration. This is particularly the case when each grain has slightly different U and Th distribution and/or different grain sizes, which results in a large variation between (U–Th)/He ages of replicates of the same sample (Farley 2000; Reiners and Farley 2001; Meesters and Dunai 2002; Fitzgerald et al. 2006). Furthermore, a very slow cooling rate could amplify any kinetic, anisotropy, or zonation effects in apatite and zircon (Hourigan et al. 2005; Reich et al. 2007; Flowers et al. 2007), which result in considerable age differences (Biswas et al. 2007). Furthermore, an age signature typical for slow cooling through the PRZ is observed in several samples located at the upper part of the profiles (i.e., SQM05 at the top of Muchuan transect). These samples are interpreted to represent an exhumed PRZ and allow the estimate of burial depth and exhumation (see below).

As for considerably scattered single-grain (U–Th)/He ages, Fitzgerald et al. (2006) suggested that the “true age” lies between the minimum (U–Th)/He age and the weighted mean age. However, most of the sample ages show consistency in replicates and a systematic relationship with structural depth/elevation. Only a few samples show a large scatter in replicate analyses with a range of 10–20 Ma, smaller than the variation suggested by Fitzgerald et al. (2006) with a range of 30–100 Ma. We thus argue that the mean (U–Th)/He age is more valid than the youngest single-grain age to represent sample age. The observed variation of replicate ages can be thus attributed to the thermal history of the sample characterized by a prolonged residence time in the PRZ.

Estimates of burial depths and exhumation

There are various post-depositional (or burial) depths and exhumation magnitudes between the stratigraphic columns (T3–J1) indicated by the AHe and ZHe ages. Mean surface temperatures and proximal borehole thermal gradients represent the best available proxy for the current thermal field through which the samples cooled. More than seven measurements at the southwestern margin of the Sichuan Basin indicate a geothermal gradient of ~25–35 °C/km and a surface temperature of 20 °C (Hu et al. 2000; Xu et al. 2011). For the Muchuan transect, the partially reset ZHe ages indicate a maximum burial temperature of 160–200 °C for the T3–J1 strata. For this temperature, we estimated a maximum thickness of the overlying sedimentary column that has been removed of 4.5–6.4 km. Consequently, the Lower Cretaceous burial depth could not exceed ~2–3.5 km, which is roughly consistent with the observed paleo-AHe PRZ, which represents paleo-temperatures of 50–80 °C (Fig. 6). Given that there are ~0.8- to 1.5-km-thick Late Cretaceous deposits in the region today (BGMRSP 1991; Guo et al. 1996), we suggested that ~1- to 2-km-thick Tertiary strata must have been eroded during the Late Cenozoic, and a ~1–2 km exhumation occurred here.

The fully reset ZHe ages found at the bottom of the stratigraphic profile of the two other transects (Xide and Ganluo) suggest a minimum exhumation magnitude ranging from 4.2 to 5.6 km, assuming a paleo-geothermal gradient of 30–40 °C/km and surface temperature of 10 °C across much of the Daliang Mountains (Hu et al. 2000). The burial depths of the present surfaces (K2 and J2) are thus greater than ~1–2 and ~3–4 km in the Xide and Ganluo transects, respectively. According to the exhumation magnitude at the base of each column equal to a sum of the surface denudation, the reconstructed depth and the preserved strata overlying the columns, the estimates of ~1.8–3.5 and ~2.8–5.2 km probably represent a suitable exhumation at the Xide and Ganluo transects. This conclusion is based on (1) the reconstructed depths in each stratigraphic column have up to ~0.2–0.5 km difference than regionally averaged stratigraphic thicknesses, (2) the assumption of ~1- to 2-km-thick Tertiary section is based on the Muchuan transect, and (3) the regionally preserved thicknesses of ~0.5–1.2 km and ~0.5 km in the Lower Cretaceous and Upper Jurassic (BGMRSP 1991; Guo et al. 1996). Furthermore, the exhumed AHe PRZ in the Mianshan granite indicates an exhumation magnitude of ~1.0–2.5 km. We argue that ~2.5 and ~5 km are the best estimates of the Late Cenozoic exhumation magnitude for the Xide and Ganluo transects, respectively. Exhumation near the margin of the Sichuan Basin is characterized by much smaller amounts of denudation (~1–2 km in the Muchuan transect). It should be noted that the estimated exhumation in transects could significantly change due to new information on changes of deposits thickness (e.g., intermountain basin), deformation and thrusting along faults (e.g., growth strata), etc.

Two-phase exhumation of the Daliang Mountains

Documenting exhumation intensity, as well as its spatial and temporal variations, is important for understanding regional patterns of denudation. The Muchuan, Ganluo, Mianshan, and Xide transects show a similar exhumation rate of ~0.15 mm/year from ~30 to 10 Ma. Most of the transects (e.g., the Wushihe, Ganluo, Xide, Mianshan, and Wudu transects) show a faster rate of ~0.4–0.8 mm/year (or a negative elevation vs. age relationship) since 10 Ma (Fig. 9). The regional consistency of these data suggests that the Daliang Mountains experienced a protracted period of slow cooling and exhumation before ~10 Ma, then an accelerated cooling and exhumation during Late Miocene time.

Fig. 9
figure 9

Tectonics transect across the southeastern margin of the Tibetan Plateau and the Daliang Mountains to show the relationship between the topography, the (U–Th)/He ages, and exhumation rates. AHe and ZHe ages show the range of single-grain ages (yellow circles and star from Clark et al. (2005) and green stars are this study). a Xide transect, b Mianshan granite, c Ganluo transect, d Wushihe granite, e Muchuan transect, f Wudu granite. XSH F.—Xianshuihe fault, DLS F.—Daliangshan fault, ANH F.—Anninghe fault, XJ F.—Xiaojiang fault, JH F.—Jinghe fault, MB F.—Mabian fault, MN B.—Miannin Basin, XC B.—Xichang Basin, GL B.—Ganluo Basin, MG B.—Meigu Basin. It should be noted that there is a distinct boundary on Daliangshan fault with an apparent exhumation rate of 0.4–0.8 mm/year from ~10 to 5 Ma

Full size image

Due to the lack of Cenozoic depositional records in the region, the deformation and Cenozoic uplift in the southwestern part of the Sichuan Basin are not well constrained, although thermochronometric data have recently revealed that episodic exhumation played a key role in shaping the topography (An et al. 2008; Richardson et al. 2008; Deng et al. 2009; Li et al. 2012). The deformation of the youngest Upper Cretaceous strata demonstrates that the main deformation in the Daliang Mountains occurred post Cretaceous (Figs. 2, 3, 4), consistent with no significant change in the paleocurrent and sedimentary facies across much of the intermontane basins in the Daliang Mountains (e.g., Ganluo and Jiuxiang Basins) during Late Triassic to Jurassic times (Chen et al. 2011). Those folded precretaceous strata were overlain by undeformed Late Cenozoic strata (i.e., Neogene Xigeda Fm). Recently, Zhao et al. (2008) and Kong et al. (2009, 2012) argued that the Xigeda Formation has an age of ~1.5–8.7 Ma, according to magneto-stratigraphic research and cosmogenic 10Be and 26Al dating. In particular, there is only nonconformity to low-angle unconformity developed during Jurassic to Paleogene sedimentary deposition, across the Daliang Mountains and the southwestern Sichuan Basin (BGMRSP 1991; Guo et al. 1996; Deng et al. 2012b). Thus, we argue that the main phase of deformation in the Daliang Mountains probably occurred in the Late Miocene, post-dating the protracted period of slow cooling and exhumation.

The protracted process is consistent with an erosional response to broad regional uplift (Richardson et al. 2008; Wilson and Fowler 2011; Wang et al. 2012a; Li et al. 2012; Deng et al. 2013). It can be correlated with an increase in stream incision on the plateau margin (Clark et al. 2004; Wilson and Fowler 2011) or with a base-level fall in the adjacent Sichuan Basin (Richardson et al. 2008), where 1.5–4 km of strata was eroded by the Yangtze river (Richardson et al. 2010; Deng et al. 2013). Based on the apatite fission track and (U–Th)/He data, Richardson et al. (2008), Li et al. (2012), and Deng et al. (2013) argued that major regional erosion has not started earlier than ~40–20 Ma ago. This timing of Cenozoic erosion is consistent with our oldest ZHe age that displays cooling at 37.7 ± 10 Ma (i.e., Xide transect). It resulted to thick deposits of gypsum and mirabilite across the southwestern Sichuan Basin during Paleogene times (e.g., Shuangliu and Minshan area in the southwest of Sichuan Basin), indicating that a connection between the Sichuan Basin and the Panzhihua–Xichang area was closed and the southwestern Sichuan Basin became an interior basin (BGMRSP 1991; Guo et al. 1996). Later, the connection of the Sichuan Basin to the South China Sea or East China Sea has been established, but different hypotheses were proposed for the timing of this opening (Clark et al. 2004; Xiang et al. 2007; Richardson et al. 2010).

Late Cenozoic deformation and rapid exhumation across the Daliang Mountains

East–west compression as a far-field effect of Indo-Eurasia convergence has accommodated widespread deformation to form those nearly N–S striking structures, including anticlines (e.g., the Ganluo and Xide anticline), synclines, and faults (BGMRSP 1991; Burchfiel et al. 1995; Wang and Yin 2009); for example, the Jinping thrust belt (shown as Jinping Mts. in Fig. 1) thrusts on the footwall of Oligocene–Miocene strata during this time indicated by apatite fission track ages (Wang et al. 2012b). Kinematic studies on the southeastern margin of the Tibetan Plateau show that the Xianshuihe fault is characterized by ~60–80 km left-lateral offset and that the Xiaojiang fault show a ~48–60 km offset (Burchfiel et al. 1995; Wang et al. 1998). Southeastward displacement of the southern part of the Songpan–Ganzi fold belt (e.g., Yajiang Terrane) and in particular the Xianshuihe fault (with at least a ~10–20 km offset) was absorbed by deformation in the Daliang Mountains (Wilson et al. 2006; Wang and Yin 2009). Based on a balanced cross section, Chen and He (2008) suggested a crustal shortening of 17.8 % during Miocene–Pliocene time, with an average shortening of ~11 km in the Daliang Mountains and its western margin (e.g., the Qinghe area with a shortening of ~20 %, Wang et al. 2012b). We thus argue that rapid cooling and exhumation occurred coevally in the Daliang Mountains (e.g., Ganluo and Mianshan sections, Wushihe granite).

The estimated post ~10-Ma exhumation rates (~0.4–0.8 mm/year) across the Daliang Mountains are similar to rates of 0.25–0.5 mm/year suggested for the Songpan–Ganzi region for the same time (Clark et al. 2005; Ouimet et al. 2010). Furthermore, a similar timing in the onset of rapid denudation is observed across much of the eastern margin of the Tibetan Plateau, e.g., the Longmen Shan at 5–15 Ma (Kirby et al. 2002; Godard et al. 2009a; Wang et al. 2012a), the Gonga Shan at ~12 Ma (Roger et al. 1995; Zhang et al. 2004a), and the western Qinling at 9–4 Ma (Enkelmann et al. 2006). It suggests that this rapid exhumation is due to the eastward growth of the Tibetan Plateau and its associated uplift and erosion.

The Indo–Eurasia collision resulted in crustal material being extruded and rotated clockwise around the eastern Himalayan syntaxis (Fig. 1). The region between the syntaxis and the Xianshuihe–Xiaojiang fault system is extruded southeastward in an inhomogeneous and diachronous process during the Late Cenozoic (Wang et al. 1998; Burchfiel and Wang 2003). The plateau growth toward the southeast is mainly accommodated by large strike-slip fault zones like the Xianshuihe–Daliangshan strike-slip fault (e.g.. Tapponnier et al. 2001; Wilson et al. 2006), which is thought to be active since at least ~12–5 Ma (Roger et al. 1995; Xu and Kamp 2000; Zhang et al. 2004a). Furthermore, widespread occurrence of flower structures suggest that the deformation is mainly accommodated with transpression along strike-slip faults (e.g., Xianshuihe fault, Mopanshan fault) in the Daliang Mountains. Most of the transects in our study area that have the youngest AHe ages are located in the hanging wall of Daliangshan and Anlinghe faults (i.e., Ganluo and Xide transects, Mianshan and Wushihe granites, Figs. 1, 9), where there occurred rapid exhumation with gentle-to-moderate deformation. Due to eastward growth of the Tibetan Plateau and its compression, widespread eastward thrusting structures (e.g., the Mabian fault) and NWW to S–N striking structures around the southwestern Sichuan Basin developed (Chen et al. 2011; Wilson et al. 2006; Wang et al. 2012b). It suggests that rapid exhumation in the Wudu granite can be correlated with the deformation at the Mabian fault, which is consistent with the Late Cenozoic architecture of the southwestern Sichuan Basin (An et al. 2008; Liu et al. 2012).

Notwithstanding, there is some geophysical evidence (Xu et al. 2007; Bai et al. 2010) to support a lower crustal channel flow that caused surface uplift and exhumation across the eastern Tibetan Plateau. A consistent pattern of exhumation without anomalously uplifted areas as in the studies of Clark et al. (2005) and Enkelmann et al. (2006) is observed across the Daliang Mountains, even across much of southeastern Tibet (Wilson and Fowler 2011). Furthermore, geological studies indicated a significant amount of shortening and deformation (Chen and He 2008; Wang et al. 2012b). Thus, our results of accelerated denudation in the Daliang Mountains (post ~10 Ma) require a tectonic explanation of denudation, as related to eastward growth of the Tibetan Plateau. It is accommodated by the boundary strike-slip faults (Fig. 9) and crustal shortening accompanied with transpression as a primary driver for uplift and topography of the Daliang Mountains. It should be noticed that we did not rule out the lower crustal channel flow here, as we argued that a crustal shortening is more significant than the lower crustal channel flow for surface exhumation in the Daliang Mountains .

Pattern of erosion rates across the southeastern margin of the Tibetan Plateau shows a strong gradient in exhumation rates with the highest rates at the eastern Himalayan syntaxis and a decrease toward the southeastern margin of the plateau (Henck et al. 2011). At the southeastern plateau margin, exhumation studies reveal a general eastward increase in cooling ages and a decrease in exhumation rate and amount toward the boundary strike-slip fault in the Daliang Mountains (Fig. 9). To the east of the strike-slip system, there is a broad consistency in AHe and ZHe ages and exhumation rates. The Muchuan transect at the southwestern margin of the Sichuan Basin shows a slower exhumation rate and smaller amount (~1–2 km) than others. It indicates significant differences in apatite and zircon He ages, exhumation rates and amounts between the Daliang Mountains and the Sichuan Basin. These strike-slip faults are thus interpreted to be the northernmost part of a NW-trending boundary between the southeastern Tibetan Plateau and the Sichuan Basin (even the South China block), controlled by a crustal fragment rotating clockwise around the eastern Himalayan syntaxis.

Conclusions

A positive age–depth correlation in restored stratigraphic column and age–elevation profiles are interpreted with respect to the denudation history of the Daliang Mountains. The Daliang Mountains record a period of slow cooling and exhumation during the Cenozoic with apparent exhumation rates of ~0.15 mm/year from ~30 to 10 Ma. This protracted slow cooling was followed by accelerated exhumation with rates ~0.4–0.8 mm/year during the Late Miocene. The post ~10-Ma rapid cooling and exhumation, we have attributed to the transpressional strike-slip faulting at the Xianshuihe–Xiaojiang fault system that resulted from the southeastward growth of the Tibetan Plateau. Our results thus support the hypothesis that mountain building in the Daliang is mainly caused by crustal shortening accompanied by transpression, rather than the previously suggested lower crustal channel flow. Based on the stratigraphy of the transects and the AHe and ZHe ages, we estimate the maximum burial depths and exhumation amount to be ~3–5 km in the Daliang Mountains and ~1–2 km at the margin to the Sichuan Basin.