Reevaluating Hendy Test with Modern Cave Calcite from the Monsoon Region of China

Abstract

The Hendy Test is widely used for assessing whether isotopic equilibrium was reached in speleothems by examining the δ¹⁸O and δ¹³C correlation along a single growth layer. However, stalagmite micro-layers are typically only a few micrometers thick and taper off from the center towards the sides, making it challenging to sample within the same growth layer in practice. To address this, we selected three caves in the monsoon region of China (Shihua Cave in the north, Heshang Cave in the central, and Baojinggong Cave in the south) to verify whether the modern cave calcite has reached equilibrium fractionation with drip water. We examined the spatial variations in the δ¹⁸O and δ¹³C values of farmed calcite on glass plates, which are analogous to a single growth layer. The δ¹⁸O and δ¹³C correlations of farmed calcite from different cave sites are consistently strong, suggesting that kinetic fractionation effects are prevalent, especially at the drip sites with lower drip rates due to longer CO₂ degassing. The δ¹⁸O–δ¹³C covariations can also occur along speleothem growth axes on short time scales, while isotopic variations over longer time scales are still in response to climate change. We propose that the Hendy Test criteria might not be prerequisites to isotopic equilibrium, and a Replication Test provides a more reliable indication of the integrity of isotopic proxies in paleoclimate research.

1. Introduction

Stable isotope compositions of oxygen and carbon (δ¹⁸O and δ¹³C) in stalagmites have become a new benchmark for global paleoclimate comparison due to their high-precision absolute age scale [1,2]. Speleothem δ¹⁸O records from the monsoon region of China (MRC) have been utilized to interpret the fluctuations of the East Asian summer monsoon [3,4,5,6,7,8], which is based on the assumption that stalagmite δ¹⁸O is a reliable indicator of meteoric precipitation δ¹⁸O [9,10,11]. Although the use of speleothems in paleoenvironmental reconstructions is increasing, there are only a few studies that assess the degree to which isotopic equilibrium is achieved during speleothem formation [12,13,14]. As the climatic interpretation of speleothem isotopic proxies remains a hotly debated issue in the MRC, it is necessary to confirm that the actively growing stalagmites were formed in isotopic equilibrium with their corresponding drip water.

In 1971, Hendy detailed the different equilibrium and kinetic processes that control the δ¹⁸O and δ¹³C values in the formation of calcite speleothems [15]. There are two criteria of the Hendy Test that determine whether speleothem isotopic composition can serve as a proxy for past environmental conditions: (1) speleothem δ¹⁸O values remain constant along a single growth layer; (2) there is no progressive enrichment in δ¹³C and δ¹⁸O from center to edge along the laminae. Thus, the absence of correlation between δ¹⁸O and δ¹³C values suggests that calcite precipitation occurred under consistent isotopic equilibrium conditions. Conversely, the covariation of the δ¹⁸O and δ¹³C along the laminae indicated the kinetic effects affected the isotopic ratios.

However, it is important to note that there are limitations to the Hendy Test criteria [16]. The concept of sampling along a single growth layer has been criticized as flawed in both theory and practice. Stalagmite micro-layers are typically only a few micrometers thick and taper off from the center to the flanks, making it challenging to sample within the same layer in practice. Additionally, some studies have pointed out that the test may not be a valid control of equilibrium conditions because isotopic equilibrium could theoretically occur in the center of the speleothem while kinetic fractionation occurs at the flanks [17,18,19].

To address this, we selected three caves in the MRC (Shihua Cave in the north, Heshang Cave in the central, and Baojinggong Cave in the south) to verify whether the modern stalagmite calcite had reached equilibrium fractionation. We examined the spatial variations in the δ¹⁸O and δ¹³C values of farmed calcite on glass plates, which are analogous to a single growth layer. Based on our modern “Hendy Test” on farmed calcite, we propose that Hendy Test criteria might not be prerequisites to isotopic equilibrium. Instead, a “Replication Test” provides a more reliable indication of the integrity of isotopic equilibrium in paleoclimate research [16,20].

2. Materials and Methods

2.1. Cave Settings

A 3-year-long (from May 2011 to April 2014) on-site cave monitoring program has been carried out in the MRC [11]. A previous study has investigated the seasonal variations of precipitation δ¹⁸O at a regional scale and the transmission of isotopic signals from precipitation to different drip sites [11]. However, isotopic results of modern stalagmite calcite farmed on the glass plates have not been reported, and whether the modern stalagmite calcite reached equilibrium fractionation has not been verified for this cave monitoring program. To address this, we selected three caves in the MRC (Shihua Cave in the north, Heshang Cave in the central, and Baojinggong Cave in the south) to examine the spatial variations in δ¹⁸O and δ¹³C values of farmed calcite on the glass plates.

The locations of three monitoring caves in the MRC are illustrated in Figure 1, and the basic information for each cave is summarized in

Table 1. Summary of basic characteristics of the three monitored caves [11]. The annual precipitation and mean surface air temperature were obtained from National Meteorological Information Center, China Meteorological Administration (http://data.cma.cn).

Cave	Location	Altitude (m asl)	Annual Precipitation (mm)	Mean Surface Air T (°C)	Thickness of Bedrock (m)	Thickness of Soil (cm)	Climate	Vegetation
SH	39.78° N, 115.93° E	251	539	12.2	30–130	100	Temperate monsoon climate	Shrub and grass
HS	30.45° N, 110.42° E	294	1343	16.5	300	40	Subtropical monsoon climate	Woody perennial plant and shrub-grass
BJG	24.12° N, 113.35° E	610	1836	21.2	>170	50	Subtropical monsoon climate	Evergreen broad-leaf forest

. Shihua (SH) Cave was described in [21]; Heshang (HS) Cave was described in [22]; and Baojinggong (BJG) Cave was described in [23]. Situated within the MRC, all three caves have wet-hot summers and dry-cold winters. The summer monsoon season, which spans from May to October, accounts for over 70% of the annual precipitation for all cave sites. Both BJG and HS have a sub-tropical monsoon climate, whereas SH has a temperate monsoon climate. The annual precipitation and average annual temperature for the three cave sites increase from north to south. There are also discrepancies in cave geology among the three caves (Table 1).

2.2. Collection of Farmed Calcite

We have assessed speleothem isotopic equilibrium [25] by adapting the glass plate method, as previously reported by Mickler et al. [26]. Frosted glass plates (10 cm × 10 cm) were placed on actively growing stalagmites (5 sites in SH, 4 sites in HS, and 5 sites in BJG) for future collection of farmed calcite [11]. First, a plastic bag was placed over the growing stalagmites, and then a clay base was built on top of the bag. The plates were positioned horizontally on the clay base, with the drip water falling precisely in the center of the plate (Figure 2). After visual verification of calcite growth, glass plates were periodically removed and replaced (1–3 months). Recovered plates were rinsed with deionized water, then dried and weighed. Calcite growth and crystal habit were verified using a Leica DVM6 digital microscope (Leica Microsystems, Wetzlar, Hesse-Darmstadt, Germany).

By comparing the weights of the glass plates before placement and after collection, we selected two plates with the most significant calcite deposition on each for multi-point isotopic analysis (at least 5 points for each glass plate). The spatial variability in the δ¹³C and δ¹⁸O values of modern stalagmite calcite was evaluated by collecting the entire thickness of calcite samples using a dental drill. This is analogous to sampling an ancient speleothem along a single growth layer. Each calcite sample must contain more than 20 micrograms for further isotopic analysis. Thus, sample points with thinner calcite thickness need to be drilled into a cycle area with a larger diameter.

2.3. Stable Isotope Analysis

Drip water and contemporary calcite samples were collected and prepared for isotopic analysis, following the methods detailed in [11]. Stable oxygen isotope analyses of drip water were performed via a Picarro L2130-i water isotope analyzer (Picarro Inc., Santa Clara, CA, USA) in the Key Laboratory of Tibetan Environment Change and Land Surface Processes, Chinese Academy of Sciences. The measurement precision for δ¹⁸O was 0.1‰. We did not measure the δ¹³C value of dissolved inorganic carbon (DIC) in drip water.

Stable carbon and oxygen isotope analyses of the farmed calcite samples were performed in the Institute of Geology and Geophysics, Chinese Academy of Sciences, using a MAT-253 mass spectrometer connected to a Kiel Carbonate Device IV device (Thermo Fisher Scientific, Waltham, MA, USA) for the reaction of carbonate powder with 100% H₃PO₄ at 75 °C. All δ¹³C and δ¹⁸O values are reported in parts per mil (‰) relative to Vienna Pee Dee Belemnite (VPDB). Based on the reproducibility of the standard runs, the one-sigma errors were estimated to be 0.05‰ for δ¹³C and 0.1‰ for δ¹⁸O.

3. Results

3.1. Shihua Cave

SH Cave (39°47′ N, 115°56′ E, 251 m a.s.l.), located on the northern rim of the North China Plain, is close to the northern edge of the modern summer monsoon [21]. The drip site for SH Cave is 200 m away from the cave entrance (Figure 3). Thus, this location maintains a stable microclimate due to the minimal impact of cave ventilation. Glass plate SH-1 was placed on 8 February 2013 and collected on 8 March 2013, while SH-2 was placed on 28 March 2012 and collected on 14 June 2012. During the growth periods, mean cave air temperatures were 15.1 °C (SH-1) and 16.4 °C (SH-2); average CO₂ concentrations were 864 ppm (SH-1) and 1142 ppm (SH-2); relative humidities (RH) were 100% (SH-1) and 88.2% (SH-2); and drip rates were 0.34 drip/min (SH-1) and 0.38 drip/min (SH-2). The corresponding δ¹⁸O values of drip water (δ¹⁸O_W) during the growth period were −8.71 (SH-1) and −8.62 (SH-2).

For glass plate SH-1 (Figure 4a), δ¹³C values ranged from −9.00‰ to −5.90‰, with an average of −8.28‰; δ¹⁸O values ranged from −8.51‰ to −7.22‰, with an average of −8.05‰. For glass plate SH-2 (Figure 4b), δ¹³C values ranged from −9.00‰ to −5.90‰, with an average of −8.28‰; δ¹⁸O values ranged from −8.73‰ to −6.69‰, with an average of −7.77‰. Both δ¹³C and δ¹⁸O values exhibited the lowest levels in the center, with an enriched trend toward the edge of the glass plates. There was a strong correlation between δ¹⁸O and δ¹³C for glass plate SH-1 (R² = 0.82, N = 12), while there was a moderate correlation for glass plate SH-2 (R² = 0.58, N = 7).

3.2. Heshang Cave

HS Cave (30°27′ N, 110°25′ E, 294 m a.s.l.) is located in the Qingjiang Valley, near the central Yangtze River valley, Hubei Province, central China [22]. The drip sites for HS Cave are closer to the cave entrance, and these sites have an unstable microclimate due to the strong influence of cave ventilation (Figure 5). Glass plate HS-1 was placed on 30 May 2011 and collected on 29 June 2011, while HS-2 was placed on 30 November 2011 and collected on 30 December 2011. During the growth periods, mean cave air temperatures were 19 °C (HS-1) and 17 °C (HS-2); drip rates were 4 drip/min (HS-1) and 3 drip/min (HS-2); and low average CO₂ concentrations (<500 ppm) were low RH < 100% for both periods. The corresponding δ¹⁸O_W values during the growth period are −8.10 (HS-1) and −7.78 (HS-2).

For glass plate HS-1 (Figure 6a), δ¹³C values ranged from −13.27‰ to −10.12‰, with an average of −12.44‰; δ¹⁸O values ranged from −7.88‰ to −7.03‰, with an average of −7.56‰. For glass plate HS-2 (Figure 6b), δ¹³C values ranged from −12.02‰ to −8.65‰, with an average of −10.53‰; δ¹⁸O values ranged from −7.55‰ to −6.34‰, with an average of −7.12‰. Both δ¹³C and δ¹⁸O values exhibited the lowest levels in the center, with an enriched trend toward the edge of the glass plates. There were strong correlations between δ¹⁸O and δ¹³C for glass plates HS-1 (R² = 0.82, N = 10) and HS-2 (R² = 0.84, N = 6).

3.3. Baojinggong Cave

BJG Cave (24°07′ N, 113°21′ E, 610 m a.s.l.) is about 70 km north of the city of Guangzhou, Guangdong Province, south China [23]. We chose two drip sites close to the cave entrance for BJG Cave, but there is a gate for the entrance and the gate is closed when there are no tourists (Figure 7). Both glass plates were placed on 3 March 2012, and then collected on 2 June 2012. During the growth period, mean cave air temperature was 16.2 °C; mean CO₂ concentration was 982 ppm; mean RH was 88.4%; and drip rates were 6.25 drip/min (BJG-1) and 52.20 drip/min (BJG-2). The corresponding δ¹⁸O_W values during the growth period were −6.08 (BJG-1) and −6.02 (BJG-2).

For glass plate BJG-1 (Figure 8a), δ¹³C values ranged from −14.25‰ to −11.43‰, with an average of −13.31‰; δ¹⁸O values ranged from −5.51‰ to −4.62‰, with an average of −5.22‰. For glass plate BJG-2 (Figure 8b), δ¹³C values ranged from −15.09‰ to −12.12‰, with an average of −13.93‰; δ¹⁸O values ranged from −5.67‰ to −5.03‰, with an average of −5.41‰. Both δ¹³C and δ¹⁸O values exhibited the lowest levels in the center, with an enriched trend toward the edge of the glass plates. There were strong correlations between δ¹⁸O and δ¹³C for glass plates BJG-1 (R² = 0.93, N = 7) and BJG-2 (R² = 0.97, N = 5).

4. Discussion

4.1. Equilibrium vs. Kinetic Isotope Fractionation

We used the average temperature during the growth period of plate calcite to calculate the oxygen fractionation factor. Calcite–water oxygen isotope fractionation can be calculated using the following equations [25]:

1000 ln α_{calcite-water} = 18.03 (10³/T) − 32.42

(1)

α_{calcite-water} = (δ_calcite + 1000)/(δ_water + 1000)

(2)

ln α_{calcite-water} ≈ δ_calcite − δ_water

(3)

Note: T in Equation (1) is in Kelvin.

The summary of the basic characteristics of the monitoring drip sites for the three monitoring caves is provided in

Table 2. Summary of basic characteristics of monitoring drip sites for the three monitoring caves, and comparative analysis of predicted oxygen isotope values in calcite under equilibrium fractionation conditions with measured values from glass plates. Field measurements for the three monitoring caves were previously reported in [11].

Glass Plate	Date Placed	Date Collected	Drip Rate (drip/min)	RH (%)	CO2 Conc. (ppm)	Cave Air T (°C)	Measured δ18OC Range (VPDB, ‰)	Measured δ18OW (VSMOW, ‰)	Predicted δ18OC (VPDB, ‰)
SH-1	8 February 2013	8 March 2013	0.34	100	864	15.1	(−8.51, −7.22)	−8.71	−8.30
SH-2	28 March 2012	14 June 2012	0.38	88.2	1142	16.4	(−8.73, −6.69)	−8.62	−8.49
HS-1	30 May 2011	29 June 2011	4	<80	495	19	(−7.88, −7.03)	−7.65	−8.10
HS-2	30 November 2011	30 December 2011	3	<80	505	17	(−7.55, −6.34)	−7.76	−7.78
BJG-1	3 March 2012	2 June 2012	6.25	88.4	982	16.2	(−5.51, −4.62)	−6.08	−5.98
BJG-2	3 March 2012	2 June 2012	52.20	88.4	982	16.2	(−5.67, −5.03)	−6.12	−6.02

. We did not measure δ¹³C value of DIC in drip water; thus, we only assessed equilibrium vs. kinetic isotope fractionation for the δ¹⁸O value of farmed calcite. The spatial variability in modern calcite δ¹⁸O values exhibits consistent trends: δ¹⁸O values are lowest near the center of the glass plates, with isotopic values increasing away from the center (Figure 4, Figure 6, and Figure 8). The δ¹⁸O_C values at centers of glass plates are close to the calculated δ¹⁸O_C values based on δ¹⁸O_W and T under equilibrium fractionation, while the δ¹⁸O_C values at edges are higher than the predicted values under equilibrium. It is possible for isotopic equilibrium to occur at the centers of glass plates, but kinetic isotope fractionation occurs at the edges simultaneously.

4.2. δ¹³C and δ¹⁸O Covariation

The δ¹³C and δ¹⁸O values of the spatially sampled modern calcite on the glass plates exhibit strong positive correlations (Figure 9a). We noted the regression lines for the three monitored caves have different slopes: 2.17 for SH, 3.11 for HS, and 3.67 for BJG. The slopes for different possible isotopic effects during calcite precipitation are displayed in Figure 9b.

If oxygen isotopic equilibrium is achieved between HCO₃ and H₂O (complete oxygen isotope buffering), the calcite δ¹⁸O values will remain unchanged during progressive CO₂ degassing and CaCO₃ precipitation. The degassing process caused ¹³C enrichment, thus manifesting a trend with a vertical slope (S = ∞). In the absence of oxygen isotopic exchange between HCO₃ and H₂O (without buffering), the slope of 0.52 denotes the δ¹³C/δ¹⁸O enrichment ratio as modeled by Rayleigh distillation [26]. If the “evaporation effect” of drip water dominates, the δ¹⁸O of drip water (δ¹⁸O_w) and corresponding calcite (δ¹⁸O_C) would be elevated, while calcite δ¹³C values would not change (S = 0). The δ¹³C and δ¹⁸O covariations of our modern calcite samples displayed an intermediate slope (0.52 < S < ∞), indicating that incomplete oxygen isotopic buffering occurs between HCO₃ and H₂O. For the three monitoring caves, SH Cave, with its lower slope, might have stronger kinetic fractionation effects than the two others.

Many previous studies implied cave ventilation is the control factor of δ¹³C and δ¹⁸O covariation: the entry of low pCO₂ outside air drives rapid CO₂ degassing from drip water, which enriches the remaining DIC reservoir while preferentially removing the light isotopes [30,31]. Drip sites in HS and BJG Caves have lower CO₂ concentrations, due to their close proximity to cave entrances and strong cave ventilation. However, the δ¹³C–δ¹⁸O slopes of HS and BJG Caves are higher than that of SH Cave, which indicates more oxygen isotope buffering for HS and BJG drip sites. There is a significant correlation between the drip rates and δ¹³C–δ¹⁸O slopes: the slower drip sites have smaller slopes. Thus, the disequilibrium fractionation in oxygen isotopes is more pronounced at slower drip-rate sites than at low pCO₂ sites. For the slower drip-rate sites, the drip water spends more time on the stalagmite surface, allowing more time for both CO₂ degassing and H₂O evaporation. Long CO₂ degassing depletes the DIC reservoir, making it unable to maintain oxygen isotopic equilibrium with H₂O, particularly at the edge of the glass plates. Both DIC depletion (non-complete buffering) and H₂O evaporation tend to cause shifts to higher δ¹⁸O_C values with flatter δ¹³C–δ¹⁸O slopes [26,30].

We also observed that the crystal size of farmed calcite can indeed vary between the edges and the center of glass plates (Figure 10). On glass plate SH-1, there are dense and small crystals (5–20 μm) at the center, while sparse and large crystals (20–100 μm) are at the edge. The speed and intensity of water flow can affect the growth and morphology of calcium carbonate crystals. The stronger impact of water droplets at the center may lead to the breakage or deformation of crystals, while weaker water flows at the edge are conducive to the formation of larger and more complete crystals [32]. At the center, new drip water mixes more frequently with the existing solution on the surface of stalagmite. This mixture could maintain the DIC reservoir as non-depleted and facilitate complete oxygen isotope buffering between HCO₃ and H₂O [19,26]. At the edge, there is an extended period for dissolved CO₂ to escape from the water but without enough new drip water supply. This prolonged CO₂ degassing at the edge can lead to δ¹⁸O_C values deviating from those expected under equilibrium with the original drip solution [19,26].

4.3. Implications for Speleothem-Based Paleoclimate Studies

The Hendy Test for equilibrium conditions may be flawed due to the possibility of isotopic equilibrium and kinetic fractionation occurring simultaneously in different parts of a stalagmite. Our cave monitoring results indicate that isotopic equilibrium fractionation may take place in the stalagmite growth axis simultaneously with kinetic fractionation occurring at the flanks. Furthermore, the varying thickness of growth layers and the typical stalagmite geometry make uniform sampling difficult, often leading to the inclusion of cross-laminae calcite in samples. Many studies have adapted to along-growth-axis positive correlations between δ¹³C and δ¹⁸O values instead of along the same growth layer. However, the correlation along the growth axis might also indicate the control of climatic change on both stalagmite δ¹³C and δ¹⁸O signals rather than the kinetic isotope effect [33].

An actively growing stalagmite XMG-1 from SH Cave [20] does not exhibit any apparent δ¹³C–δ¹⁸O covariation throughout the entire dataset (Figure 11a). However, there are shorter intervals within the dataset that do show significant positive covariations. We found the 20 consecutive samples (every interval of 0.05 mm for the depth of 0–1 mm) from the top of stalagmite XMG-1 displayed strong positive δ¹³C–δ¹⁸O covariations (Figure 11b). The corresponding age of 2010–2015 CE is determined by annual layer counting for 0–1 mm depth in stalagmite XMG-1. The outside climate and cave microclimate were relatively stable during this short growth period. The δ¹³C–δ¹⁸O slope of 1.73 for the recent growth layers of XMG-1 is close to the δ¹³C–δ¹⁸O slope of 2.17 we found in the glass plate calcite. This observed positive δ¹³C–δ¹⁸O covariation in stalagmite XMG-1 may be attributed to non-equilibrium isotope effects (DIC depletion and non-complete buffering with H₂O), which we proposed to explain the isotopic variations on the SH glass plates. The XMG-1 record and the stalagmite S312 record [34] from the same cave displayed similar δ¹⁸O changes within overlapping growth periods, and the so-called “Replication Test” confirmed that kinetic fractionation is negligible over a longer time scale.

The recent investigation of two coeval stalagmites (TS9701 and TS9501) from the SH Cave [35] has revealed a similar phenomenon: on interannual to decadal scales, positive δ¹³C–δ¹⁸O covariations are attributed to kinetic isotope effects; on multidecadal to millennial time scales, climate and vegetation changes are control factors of stalagmite isotopic signals. In this study, we have identified that kinetic fractionation leads to δ¹³C–δ¹⁸O covariations on seasonal to interannual scales. However, this does not negate the effect of climate change on speleothem isotopic signals over longer time scales, particularly for speleothem records that passed Replication Test [35]. Therefore, the speleothem isotopic proxies could still be used to reconstruct paleoclimate conditions even if the Hendy Test was not passed, and the Replication Test provides a more reliable indication of the integrity of isotopic proxies in paleoclimate research.

5. Conclusions

We examined the spatial variations in δ¹⁸O and δ¹³C values of farmed calcite on glass plates, which are analogous to a single growth layer in Hendy Test. The main conclusions drawn are as follows:

(1)

The δ¹⁸O and δ¹³C correlations of farmed calcite from different cave sites are consistently strong, suggesting that kinetic fractionation effects are widely present in the MRC.

(2)

For the slower drip-rate site, the δ¹³C and δ¹⁸O covariations displayed a lower slope due to longer CO₂ degassing. DIC depletion and H₂O evaporation tend to cause shifts to higher δ¹⁸O_C values.

(3)

Spatial distribution of farmed calcite isotopic values on glass plates indicates that isotopic equilibrium fractionation may take place in the stalagmite growth axis simultaneously with kinetic fractionation occurring at the flanks.

(4)

Covariations of speleothem δ¹³C and δ¹⁸O on short time scales may result from kinetic isotope effects, but this isotopic disequilibrium does not alter isotopic variations in response to climate change over longer time scales.

(5)

Hendy Test criteria have limitations as prerequisites to isotopic equilibrium. Instead, the Replication Test provides a more reliable indication of the integrity of isotopic proxies in paleoclimate research.

However, our findings are based on a smaller sample size than would be ideal, and different caves and drip sites might exhibit varying isotopic fractionation conditions. In future studies, we plan to deploy larger glass plates or extend the growth duration to ensure sufficient calcite deposition, thereby enabling a more comprehensive analysis of a greater number of drip sites within each cave.