A comparison of growth rate of late Holocene stalagmites with atmospheric precipitation and temperature, and its implications for paleoclimatology

Quaternary Science Reviews 187 (2018) 94e111

Contents lists available at ScienceDirect

Quaternary Science Reviews journal homepage: www.elsevier.com/locate/quascirev

A comparison of growth rate of late Holocene stalagmites with atmospheric precipitation and temperature, and its implications for paleoclimatology L. Bruce Railsback Department of Geology, University of Georgia, Athens, GA, 30602-2501, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 7 November 2017 Received in revised form 31 January 2018 Accepted 4 March 2018 Available online 30 March 2018

Growth rate of stalagmites can vary with many factors of physical environment, ecology, and karst hydrogeology, to the extent that growth rates calculated from a carefully selected set of data from 80 stalagmites from around the world vary by a factor of 400 from smallest to largest. Growth rates of those 80 stalagmites nonetheless collectively show correlations to atmospheric precipitation and temperature that are non-trivial (r2 ¼ 0.12 and 0.20, respectively) and unlikely to have arisen randomly (p ¼ 0.002 and 0.00002). Those global relationships are also supported by previously published studies of individual drip sites. The general trend of growth rates is not a monotonic increase with precipitation; instead, it reaches a maximum at annual precipitation rates between 700 and 2300 mm/year, which both counters many model predictions that growth rates should increase monotonically with drip rate and complicates use of growth rate as a proxy for past precipitation. The general trend of growth rates among the 80 stalagmites is a monotonic increase with temperature. However, the low values of r2 in both of these general trends indicate that growth rate can be at best a qualitative rather than quantitative proxy of past conditions. Growth rate shows no statistically significant relationship to effective precipitation, seemingly because of the confounding effect of temperature. Growth rates of aragonite-bearing stalagmites are commonly greater than rates in stalagmites in which calcite is the only carbonate mineral, suggesting both the need for careful identification of mineralogy and the special applicability of aragonitic stalagmites in high-resolution studies. Aragonite has exceptionally great frequency in settings with low effective atmospheric precipitation, supporting previous linkages of that mineral to warm dry environments. Closely-spaced sampling used in recent paleoclimatological studies suggests that unexploited longterm low-resolution records of past climate may exist in surprisingly small slow-growing stalagmites from exceptionally cold and/or dry regions. © 2018 Elsevier Ltd. All rights reserved.

Keywords: Holocene Paleoclimatology Global Speleothems Growth rate Stalagmite Rainfall Temperature Paleoclimate Proxy

1. Introduction The rate of growth (specifically, the rate of vertical accretion or extension) of stalagmites has been used in paleoclimatological research as a proxy for past climate (e.g., Xia et al., 2001; Sletten et al., 2013; Muangsong et al., 2014; Nehme et al., 2015). Growth rate is commonly assumed to be positively related to rate of atmospheric precipitation or wetness, but multiple questions lurk behind this assumption. One might ask if growth rate is also related to temperature, overwhelming the precipitation signal or leading to

E-mail address: [email protected]. https://doi.org/10.1016/j.quascirev.2018.03.002 0277-3791/© 2018 Elsevier Ltd. All rights reserved.

a possible confounding effect. A second question might be if growth rate is positively correlated with precipitation across the entire range of the latter from arid to extremely wet climates, or if it decreases in very wet conditions. For example, the assumption of a positive relationship between atmospheric precipitation and growth rate in dry conditions finds unquestionable buttressing in the logic that zero precipitation on the land surface, and thus presumably zero drip rate in an underlying cave, must result in zero growth of a stalagmite, but many effects might stop or reverse that trend in much wetter conditions, as discussed herein. A third question might be if growth rate is better understood in terms of effective precipitation (i.e., precipitation after correction for loss of moisture to evapotranspiration). A fourth question might be if

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

growth rate is affected by mineralogy, so that growth rates might require mineralogical correction. This paper addresses these questions with a compilation of growth rates from late Holocene stalagmites that, to the extent allowed by the literature, span the globe geographically (Fig. 1 and Fig S1; Section S1) and span the observed range of temperature and atmospheric precipitation. Despite the importance of the questions posed above, one might observe that a considerable observational literature on growth rate of stalagmites has developed since the 1980s (e.g., Baker et al., 1998; Genty et al., 2001; Banner et al., 2007; Baker et al., 2008; Baldini, 2010; Mariethoz et al., 2012) and thus might ask why a new study is needed or how it could help. The answer is three-fold. First, the literature reporting stalagmites as records of past climate, and thus reporting radiometric ages to support age models from which growth rates can be derived, has grown hugely in the last ten years (Fig. S2). Secondly, the number of radiometric ages generated per stalagmite, and thus the definition and credibility of age models, has increased greatly in the last ten to fifteen years, to as many as 47 radiometric ages from a single stalagmite (e.g., Zhao et al., 2015). Thirdly, detailed studies of varying growth rate in single stalagmites, rather than comparisons between stalagmites, have added new and well-defined results relevant to variation in growth rate. Thus much more abundant data of higher quality are available today for a compilation from which to answer questions like those posed above. Multiple results arise with regard to the overall distribution of values of growth rate (Fig. 2) and the relationship of growth rate to mineralogy (Fig. 2), atmospheric precipitation (Figs. 3e5), temperature (Figs. 6e8), effective atmospheric precipitation (Fig. 9), and ecology (Fig. 10).

2. Growth rate, atmospheric precipitation, and the complexity of karst hydrology

95

Factors differing from cave to cave, or even from drip site to drip site in one cave, can include (but are not limited to) seasonality of precipitation (as in Fig. 1 of Railsback, 2017), density and type of vegetation, extent of evapotranspiration (which is considered further in Section 4.4), slope of the land surface and resultant differences in infiltration, permeability of the land surface, surface temperature (which is considered further in Section 4.3) and its effects on biological productivity, thickness of soil, thickness and lithology of bedrock, areal size of the catchment for the drip, reservoirs in the catchment, cave temperature and the extent of its annual variation, pathway of water in the cave to the drip site, extent of cave ventilation controlling PCO2 (in turn dependent on cave entrances and passages), presence or absence of a stalactite at the drip site, kinetic effects exerted by trace elements dissolved in the drip water, vertical distance of the drip, and extent of microbial activity on the growing stalagmite. Factors relevant through time at one drip site can additionally include (but are not limited to) how antecedent moisture or vegetation conditions affect infiltration, whether dual-porosity systems allow by-pass flow behavior during extremely wet periods, and development and evolution of a stalactite at the drip site. The resultant variation from all these factors is why this article reaches inferences about “the general trend of growth rates”. It is not because there is a single precise mathematical relationship or transfer function between growth rate and factors like atmospheric precipitation or temperature, about which the article is making imprecise statements. It is instead because the combined variation of the many controlling factors dictates that there can be at most a general trend to the growth rates reported here. Furthermore, the combination and interaction of all of these factors dictate that growth rate will at best be a qualitative, rather than quantitative, proxy for past environmental parameters. 3. Methods

The variation in the data (Figs. 2e4 and 6 and 7 and 9 and 10) and the small values of r2 (Figs. 4 and 6) are not surprising in light of the extreme complexity of karst hydrology (Milanovic, 1981; White, 1988; Ford and Williams, 1989, 2007; Klimchuk, 2000; Fairchild and Baker, 2012; Gilli, 2015). Many factors not considered in this study could contribute to variation in growth rates of stalagmites.

́

Fig. 1. Red bars: histogram of latitude of stalagmites used in this study and listed in Table 1. Blue lines: histogram of land area as a proportion of the surface area of the Earth. The relationship of the red bars and blue lines is discussed in Section 3.3. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

3.1. Sources, materials, and calculations Articles used as sources of data for Figs. 2e4 and 6 and 7 and 9 and 10 were found by searching the Web of Science index using the terms “stalagmite”, “climate” and “Holocene”. “Climate” was used as a search term to target articles reporting radiometrically dated stalagmites. As the search moved to progressively earlier articles, increasing rate of failure of those earlier articles to supply sufficient age data discouraged searching of papers published before 2006 (eliminating only 19% of all articles returned with the search terms). However, three earlier articles were used to fill specific geographic gaps. Each age model was examined to assure sufficient ages and continuity of deposition to yield a meaningful growth rate. Further details of the search are provided in Section S2. To minimize the possibility of an excessively uniformitarian approach, articles were selected for use only if they allowed estimation of growth rates within the last 3000 years. Exceptions were made to fill geographic gaps, but none of the data represented by filled symbols in Figs. 2e4 and 6 and 7 and 9 and 10 represent deposition earlier than 5000 years ago. One open symbol in Fig. 4 represents deposition from 5500 to 9500 years ago; it is included there (but in no other part of this work) because it is from a region of uncommon rainfall. In cases where growth rate during the stalagmite's entire history was not uniform, the growth rate during the last century was calculated, reported in Table 1, and used in the figures in this article. This is because the goal of the article is to compare growth rate with instrumental measurements of atmospheric precipitation and temperature, and those climatological measurements typically extend backward only a few decades. Thus some stalagmites from

96

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

Fig. 2. A. Histogram of growth rates of 80 stalagmites deposited during the later Holocene. The data are tabulated in Table 1 and reported in Section 4.1. B. Histogram of the same data, but with a logarithmic horizontal axis.

which data is taken for this article may have grown in the past at greater rates than the recent rates reported in Table 1, plotted in this article's figures, and compared to recent climatological parameters. The human attention to detail required in assembling the data used in this article, as discussed above, suggests that machinederived growth rates from numerous but unexamined sets of age data can lead to pitfalls masked by the seeming power of large sample sizes. 3.2. Climatological distribution of stalagmites considered The compiled data span a range of average annual temperatures from 1.1 to 28.9  C and a range of annual precipitation from 174 mm (in Oman) to at least 6100 mm (in southern Chile and in Meghalaya, in eastern India, where estimates of rainfall vary but the total is undoubtedly huge). Most of the stalagmites from which data were compiled come from regions with average annual atmospheric precipitation from 200 to 3500 mm, which spans the range of precipitation rates most frequently reported by meteorological stations around the world (Fig. 3B). The regions covered by this

compilation include several with monsoonal to dominantly summer precipitation, some with dominantly winter precipitation, a few with regular (year-round) precipitation, and at least one with equinoctially-dominated precipitation. 3.3. Possible biases in the data set (and perhaps in collection and reporting of stalagmites) Because this article draws on published data about stalagmites, the geographic and climatological distribution of those stalagmites is subject to the limitations of the literature as well as the natural limitations of stalagmite growth. Fig. 1 is a histogram of the latitudinal distribution of the stalagmites listed in Table 1, and Fig. S1 is a world map of the locations of those stalagmites. A bias at the scale of nations and continents that emerges from Fig. S1, presumably because of proximity of caves to researchers, is a disproportionate abundance of stalagmites from China and Europe (Fig. S3), and in fact the process of selection of stalagmites described in Section S2 disfavored stalagmites from those regions to avoid an even greater bias. A bias at the scale of the entire planet

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

97

Fig. 3. A. Plot of growth rates of 80 stalagmites during the later Holocene relative to average annual atmospheric precipitation above the caves in which they formed. The continuous red curve shown is the result of regression to yield a second-order polynomial from the data in the range of precipitation where data are sufficiently abundant to support regression (filled symbols); the dotted red curve shown is the result of regression to yield a second-order polynomial, forced through the origin, from all of the data. Fig. 4 demonstrates that the trend shown in this diagram results both from the absence of small growth rates, as well at the existence of large growth rates, in the interval from 700 to 2000 mm of atmospheric precipitation. The dashed blue curve suggests the general trend expected from models, as shown by Fig. 12. The data are tabulated in Table 1. B. Relative frequency of reports of amounts of average annual atmospheric precipitation from meteorological stations from around the world. Because the distribution is station-based, it under-represents exceptionally dry regions. The curve is taken from Fig. 8a of Anwer (2015). Note that, although stalagmite data from regions of high rainfall are scarce in Part A, Part B shows that they are nonetheless over-represented relative to the global distribution of rainfall by area. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

that emerges from Fig. 1 is a disproportionate abundance of stalagmites from the mid-latitudes, a disproportionate scarcity of stalagmites from low (<15 ) latitude, and a strikingly disproportionate scarcity of stalagmites from high (>60 ) latitudes (Fig. 1). This bias exists despite searches of the literature for stalagmites from low-latitude and high-latitude countries and regions. The scarcity if not absence of stalagmites at high latitudes presumably results from (a) glacial cover, (b) slow growth of stalagmites in cool conditions (as discussed further in Section 4.3), and (c) non-growth of stalagmites in caves where cave air temperature (roughly mean annual air temperature) is always less than 0  C, so that liquid water is never present. However, closely-spaced samples €uselmann et al., 2015) and for radiometric dating (as analyzed by Ha closely spaced samples for stable isotope analysis (as collected by Zhao et al., 2015) could allow paleoclimatological study of

stalagmites with growth rates much slower than the slowestgrowing stalagmites identified in this study (Fig. 11). For example, a stalagmite 5 cm tall that has grown at a rate of 0.002 mm/yr and sampled with the methods of H€ auselmann et al. (2015) and Zhao et al. (2015) would give a 25,000-year record defined by 16 230Th ages and 1000 stable isotope measurements with a resolution of 25 years. This example suggests that significant paleoclimate records may exist in cold (and/or dry) regions in slow-growing stalagmites so short that they are overlooked if not stepped on (Fig. 11). The flowstones and stalactites reported by Lauriol et al. (1997) from the northern Yukon (66.5 N, January TJan ¼ 34  C, precipitation ¼ 205 mm/yr) suggest that such stalagmites may await investigation. The scarcity of stalagmites at low (<15 ) latitudes results from reasons less clear, and it is of greater concern for the goals of this

98

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

Fig. 4. A. Plot of growth rates of 81 stalagmites during the Holocene relative to average annual atmospheric precipitation above the caves in which they formed. Filled symbols indicate stalagmite growth in the later Holocene, mostly in the most recent 3000 years (Table 1), whereas the one open symbol indicates stalagmite growth in the earlier Holocene, included because it is from a region of exceptional rainfall (Dutt et al., 2015). Five green curves indicate results from stalagmites in which growth rate, recognized as thickness of annual layers, was correlated with known variation in annual atmospheric precipitation (Cai et al., 2010; Kato and Yamada, 2016; Proctor et al., 2000; Railsback et al., 1994; Rasmussen et al., 2006). The dashed green curve represents a positive correlation for which the slope is otherwise uncertain. Dashed black curves are only the author's attempt to suggest an envelope for the data. The two blue filled circles at the far right represent stalagmites in a cave hosted by non-carbonate bedrock, making them anomalous not only with respect to amount of precipitation. The data for the filled symbols are tabulated in Table 1; the regression results shown were derived only from the data represented by filled symbols. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

study. The reasons may be either of, or a combination of, two factors. The first candidate is frequent slow or zero growth in regions so wet that water passes so rapidly through soil and bedrock that either it commonly does not achieve saturation with respect to CaCO3 or passes from cave roof to cave floor too quickly to degas to induce supersaturation with respect to CaCO3, so that commonly nothing is deposited and few stalagmites form. The second candidate is that there are many stalagmites, perhaps fast-growing ones, in wet tropical regions but fewer researchers doing paleoclimate studies in the tropics than in mid-latitude regions. The reasons are relevant to this study because the scarcity of stalagmite records at low latitudes gives a scarcity of stalagmites from regions of rainfall greater than 3500 mm/yr, making interpretation of the scarce and

scattered data challenging if not statistically untenable. This article's conclusions with regard to atmospheric precipitation will therefore largely be concerned with atmospheric precipitation less than 3500 mm/yr. If one posits that fast-growing stalagmites are underrepresented in the data because there are fewer researchers doing paleoclimate studies in the tropics to collect them, one may also consider whether exceptionally slow-growing stalagmites are under-represented in the published literature because they are collected but not reported. For example, some published research reporting data from one stalagmite nonetheless involves collection of tens of stalagmites. If some of the non-reported stalagmites are not used because their slow growth makes them difficult to sample

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

99

Fig. 5. Plot of stalagmite growth rate (vertical axis) vs. time between drips (horizontal axis) of Stalagmite C2 from Shihua Cave, Beijing, China, as derived from Fig. 2 of Cai et al. (2011). Red filled circles indicate 22 periods of roughly one month each, but four symbols overlap for zero growth at greatest time between drips. Monthly rates of atmospheric precipitation at top of plot are ranges of rates for November to March, April and May and September and October, and June to August. The dashed black lines are only the author's attempt to suggest an envelope for the data. Note that this plot has the same axes, in the same directions, as Fig. 12. This figure is discussed in Section 4.2.1. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

for high-resolution records, then the published results from which the data considered in this article come are biased to exclude slowgrowing stalagmites. 4. Results and discussion 4.1. Overall variation in growth rates Growth rates of the 80 stalagmites vary across almost three orders of magnitude, from 0.006 to 2.3 mm/yr (Fig. 2). Even across narrow ranges of environmental parameters (e.g., 800e1200 mm/ yr precipitation or 10e15  C), growth rates can vary by two orders of magnitude (Figs. 3 and 6). The distribution of growth rates is strongly biased toward smaller values, giving a median of 0.10 and a mean of 0.24 mm/yr (Fig. 2). The greatest growth rates are greater than those predicted by most models (e.g., Dreybrodt, 1981; Dreybrodt and Franke, 1987; Kaufmann, 2003; Kaufmann and Dreybrodt, 2004; Mühlinghaus et al., 2007; Dreybrodt, 2016). They are, however, compatible with growth fed by turbulent rather than laminar flow, as shown by Fig. 9 of Dreybrodt (1999). 4.2. Growth rate and atmospheric precipitation 4.2.1. Inter-stalagmite, inter-annual, and inter-seasonal data The collective results comparing 80 stalagmites indicate that the general trend of growth rate does not exhibit a monotonic relationship to atmospheric precipitation. Across the data from 80 stalagmites, growth rate increases with increasing annual atmospheric precipitation (Pa) from zero growth at zero precipitation to a maximum growth rate at Pa between 700 and 2300 mm/yr, and then decreases with increasing Pa, as shown by filled circles in Figs. 3 and 4. Thus the former data (Pa < 700e2300 mm/yr) are consistent with the model results in Fig. 12, but the latter (Pa > 700e2300 mm/yr) are counter to those model results. The non-monotonic relationship (Fig. S4, Section S3) holds whether one considers all 80 stalagmites across the entire range of atmospheric

precipitation or only the range across which data are relatively abundant (Fig. 3). The non-monotonic relationship holds whether one considers all of the data (Fig. 3) or data restricted to specific intervals of temperature (Fig. 7A and C), demonstrating that the non-monotonic relationship is not an artifact of the influence of temperature. In addition, inter-annual results from 5 individual stalagmites in regions where annual precipitation is less the 900 mm show a positive relationship between thickness of annual layers and precipitation for individual years, whereas results from individual stalagmites in regions of annual precipitation greater than 900 mm show an inverse relationship between thickness of annual layers and precipitation for individual years, as shown by green curves in Fig. 4. More generally, an inverse relationship between rate of deposition of CaCO3 (stalagmite growth or deposition on glass plates) and wetness (precipitation and/or drip rate) at inter-annual or seasonal time scales has been documented by several papers (Table 2). These observations support the inference in the previous paragraph that growth rate can decrease with increasing wetness across at least some ranges of the latter. The non-monotonic relationship evident from the data discussed above is also evident in intra-annual (seasonal) data from one drip site in Shihua Cave in Bejing, China. Cai et al. (2011) recorded depositional rates at three drips in that cave, and they recorded the drip rates feeding each. Two sites, C1 and C3, showed a monotonic but inverse relationship between drip rate and growth rate across seasons. Site C2, which had a greater range of drip rate than C1 and C3, showed a non-monotonic relationship between drip rate and growth rate, with maximum growth at a drip rate of about 0.8 drips/minute (a drip interval of about 70 s) and growth rates near zero at very slow and very fast drip rates (Fig. 5). The detailed study by Cai et al. (2011) of Site C2 with its broad range of drip rate thus showed effectively the same pattern as that evident from the multiple stalagmites reported in Figs. 3 and 4: Maximum rate of growth as the result of intermediate wetness, and slower growth with increasingly dry and with increasingly wet conditions.

100

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

Fig. 6. Plot of growth rates of 80 stalagmites during the later Holocene relative to modern annual average temperature above the caves in which they formed. Three open symbols indicate stalagmites in regions where average annual atmospheric precipitation is less than 220 mm. Dashed green curve marks one stalagmite from which authors inferred from stable isotope data that greater growth rate was correlative with higher temperature and greater atmospheric precipitation. Solid lines are the results of linear regression. Dashed gray lines show results from modeling by Kaufmann and Dreybrodt (2004) shown in their Fig. 2a: first number indicates soil CO2 concentrations in thousands of ppm (i.e., 20,000 and 50,000), and second number or numbers indicates range of drip interval in seconds (and thus inverse of drip rate). The data are tabulated in Table 1 and reported in Section 4.3. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

4.2.2. Implications for models of stalagmite growth Data demonstrating the non-monotonic relationship between growth rate and atmospheric precipitation and/or demonstrating an inverse relationship between growth rate and atmospheric precipitation (Figs. 3e5; Table 2) contradict model results that predict invariably increasing growth rates with increasing drip rates (which presumably have a positive relationship to precipitation at annual or longer time scales) (Dreybrodt, 1980, 1981; Dreybrodt and Franke, 1987; Kaufmann, 2003; Kaufmann and Dreybrodt, 2004; Mühlinghaus et al., 2007; Dreybrodt, 2016) (Fig. 12). This contradiction may arise from three phenomena that likely act together. First, runoff over saturated soils and bypass flow in underlying dual-porosity karst systems may weaken the dripsite signal of exceptionally great precipitation. Secondly, in

exceptionally wet conditions, rapid passage of percolating water through soil and bedrock above a cave may not allow time for charging with CO2 sufficient to cause saturation with respect to CaCO3 after degassing on entry into that cave (as shown by the curves for mixed and closed systems in Fig. 11 of Dreybrodt (1999)). Thirdly, rapid entry into the cave may limit the time that a drip hangs at the roof of a cave and degasses, so that the drip does not achieve saturation or supersaturation. With regard to the latter, we have come to realize the importance of prior degassing and prior calcite precipitation, or more generally prior carbonate precipitation (PCP) (Fairchild et al., 2000), but the results represented by Figs. 3 and 4 and Table 2 may suggest the importance of prior nondegassing in the opposite high-flow situation.

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

101

Fig. 7. A to C: Plots of stalagmite growth rate relative to atmospheric precipitation, with data restricted to specific intervals of temperature. These plots thus divide Fig. 3 by temperature, showing that patterns relative to precipitation do not result from confounding effects of temperature. D to F: Plots of stalagmite growth rate relative to temperature, with data restricted to specific intervals of atmospheric precipitation. These plots thus divide Fig. 6 by atmospheric precipitation, showing that patterns relative to temperature do not result from confounding effects of precipitation. Gray curves are envelopes for data, and they are solely the author's constructs. Field labeled “E?” in Part F may be vacant compared to analogous areas in D and E because of extreme effects of evapotranspiration.

4.2.3. Alternate interpretations of Figs. 3 and 4 One alternate understanding of Figs. 3 and 4 is that growth increases with atmospheric precipitation up to 1000e2000 mm yr and then plateaus at greater values of precipitation (the “Plateau Hypothesis” of Fig. 3), and that the scarcity of growth rate determinations at high precipitation rate precludes clear delineation of the plateau. This plateau could be readily explained by the effect of antecedent moisture on infiltration and/or by-pass flow behavior in a dual porosity system (which are among many hydrologic factors discussed above). However, the data of Cai et al. (2011) in Fig. 5 show that this alternate understanding is incomplete, because that figure's horizontal axis is drip rate, a post-infiltration and postgroundwater measure of water supply to a drip. Another alternate understanding of Figs. 3 and 4 is that there are many fast-growing stalagmites in wet tropical regions that are not represented because of undersampling by the paleoclimatological community. In this view, as more studies are published from the tropics, the region of Fig. 3 that is currently blank will be filled in with data and there will be no inflection point in a continually upward trend of growth rate. However, if slow-growing tropical stalagmites have been set aside in the search for fast-growing ones, as discussed in Section 3.3, then the bias of the paleoclimatological community may not be against collection of tropical stalagmites so much as against reporting of slow-growing tropical stalagmites. 4.3. Growth rate and temperature Growth rates of the 80 stalagmites listed in Table 1 show a

statistically significant positive correlation with average annual temperature (Fig. 6). The value of r2 for the correlation is 0.20, indicating that factors other than temperature additionally affect growth rate, but p for the correlation is < 0.0001, indicating that the correlation between growth rate and temperature is very unlikely to have arisen randomly. The statistical relationship becomes even stronger if one discounts the results from the three stalagmites deposited in driest conditions (represented by open symbols in Fig. 6), because all three come from warm regions but have exceptionally small growth rates. The positive relationship between temperature and growth rate holds whether one considers all of the data (Fig. 6) or data restricted to specific intervals of atmospheric precipitation (Fig. 7D to F), demonstrating that the positive relationship is not an artifact of the influence of atmospheric precipitation. It also holds whether one considers all 80 stalagmites or only the 54 reported to consist of calcite (Fig. 6), demonstrating that deposition of fast-growing aragonite in warmer climates does not account for the correlation between temperature and growth rate. The diffuse correlation of growth rate with temperature from a global set of stalagmites in Fig. 6 is complemented by results from single drip sites in a single cave by Casteel and Banner (2015). Of the six sites studied, all yielded positive relationships between temperature and depositional rate, although the strength of those relationships varied (Fig. 8). Additional evidence for the positive relationship between temperature and growth rate comes from Fig. 3 of Van Rampelbergh et al. (2015), which shows that C and O stable isotope data suggest that thicker layers in the Proserpine Stalagmite

102

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

Fig. 8. Plot of rate of deposition of calcite at six sites in Westcave in central Texas (vertical axis) relative to surface air temperature (horizontal axis) reported by Casteel and Banner (2015) and discussed in Section 4.3. Twenty-four filled symbols of each color and shape represent roughly monthly observations over 26 months. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

in Belgium were deposited in warmer conditions (this relationship is represented by a green curve in Fig. 6). Additional evidence also comes from Fig. 9 of Railsback et al. (1994), which shows that thickness of aragonite portions of annual couplets in a stalagmite from Botswana were positively correlated with summer temperature (as known from instrumental records). 4.4. Growth rate and evapotranspiration One might hypothesize that growth rate of stalagmites is proportional to the rate at which water is supplied to drips, but that the supply of water is not best understood in terms of raw atmospheric precipitation (the abscissa in Figs. 3 and 4) but instead in terms of effective precipitation (precipitation minus potential evapotranspiration, or PeE). However, a plot of growth rate against PeE shows no significant trend (Fig. 9) and is most compatible with random sampling from a log-normal distribution of drip rates (Fig. 2B). The lack of a relationship between growth rate and PeE likely results from the dependence of both on temperature, but in opposite directions. On one hand, growth rate increases with increasing temperature, as reported in Section 4.3. On the other hand, increasing temperature increases evapotranspiration, lessening PeE, which in the hypothesis above would lead to decreasing growth rate, counter to the observation in the previous sentence. One aspect of the data that may be a visible result of the effect of evapotranspiration is the absence of stalagmites with large growth rates from regions of high temperature and little rainfall, as is shown by the hachured area in Fig. 7F. Extremely negative values of

PeE may preclude the large growth rates otherwise expected at high temperature (as is also suggested by the open circles at the lower right of Fig. 6). One pattern that does emerge from Fig. 9 is that aragonite is disproportionately present in stalagmites from regions of low PeE. This is compatible with empirical observations that aragonite commonly forms in dry conditions (e.g., Murray, 1954; Pobeguin, 1955; Siegel, 1965; Thrailkill, 1971; Cabrol and Coudray, 1982). It is also compatible with the theoretical argument from Ostwald's € hnel and Garside, 1992) that aragonite, the more solstep rule (So uble of the two CaCO3 polymorphs, would be precipitated from concentrated drip waters (specifically, those high enough in aCa2þ and aCO32- to be supersaturated with respect to both aragonite and calcite). Aragonite's faster rate of growth (Section 4.6) and its frequency in hot dry regions of low PeE may combine to be another confounding influence contributing to the non-relationship between growth rate and PeE. 4.5. Growth rate and ecology Ecologists have shown that specific biomes can be broadly linked with specific ranges of temperature and atmospheric precipitation (e.g., Whittaker, 1975; Ricklefs, 2001). Superposition of those relationships on the growth-rate data considered here shows that greatest growth rates of stalagmites are not distributed across all biomes but instead are typical of biomes ranging from savanna or grassland to forest (Fig. 10). For paleoclimatological studies, this result means that high-resolution records will most commonly be

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

103

Fig. 9. A. Plot of growth rate of stalagmites relative to precipitation minus potential evapotranspiration (PeE), or effective precipitation. Values of PeE were obtained from Fig. 8 of Givnish (2002). Gray dashed curve (r2 ¼ 0.0006) is the result of regression on all data; blue dashed curve (r2 ¼ 0.04) is for stalagmites reported to consist of calcite; both values of p are compatible with random sampling. Note that the plot, when considered in vertical swaths, is compatible with sequential random sampling of a log-normal distribution (Fig. 2B) along the vertical axis, but with decreasing n of samples in a sequence from smaller to greater PeE. B. Schematic diagram illustrating that growth rate cannot follow the positive relationship with temperature observed in Fig. 6 (left side of diagram) and have a positive relationship with PeE (right side of diagram). The figure is discussed in Section 4.4. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

available from those environments, whereas colder and/or either drier (e.g., desert) or wetter (e.g., rain forest) environments will typically be represented by lower-resolution records. Examination of Fig. 10 also reveals that none of the stalagmites considered in this project came from exceptionally cold (e.g., tundra) environments, despite literature searches specifically targeting Canada, Russia, and Alaska. On the other hand, three stalagmites come from regions of rainfall so great that classifications of biomes (e.g., Whittaker, 1975; Ricklefs, 2001) do not extend there in temperature-precipitation space. Thus, although Figs. 3 and 4 show that greatest growth rates are found in regions of intermediate (700e2300 mm/yr) precipitation, stalagmites can nonetheless grow in regions of rainfall that is exceptionally great (or at least exceptionally great from the perspective of ecologists).

4.6. Growth rate and mineralogy Growth rates in stalagmites containing at least some aragonite tend to be greater than growth rates of calcite stalagmites (Figs. 2, 4 and 6). This may arise because aragonite is subject to fewer and/or less significant kinetic inhibitions than calcite, most notably the effect of dissolved Mg2þ (e.g., Berner, 1975; Morse et al., 1997).

4.7. Additional mineralogical observations Among the 80 stalagmites, stalagmites containing at least some aragonite come only from regions where annual average temperature is more than 12  C (Fig. 6). This parallels previous observations from stalagmites in the western United States (Murray, 1954), the occurrence of aragonite in modern marine sediments (Lowenstam and Weiner, 1989), and results from experimental studies (Burton and Walter, 1987; Morse et al., 1997) (although one €tl et al. (2016) reported aragonite in very cold should note that Spo conditions in a cave in the Carnic Alps of Austria). The general relationship between aragonite and warmer temperatures reported in the previous sentence, combined with the relationship between warmer temperature and greater growth rate reported above, might be taken to mean that the relationship between aragonite and greater growth rate reported above resulted only from aragonite's preferential occurrence at warmer temperatures. However, the greater growth rate of aragonitic stalagmites among stalagmites forming at uniform temperatures (i.e., in vertical swaths of Fig. 6) demonstrates that the relationship between aragonite and greater growth rate is genuine rather than artifactual. Sections S4 and S5 of Supplementary Document 1 report other mineralogical results not related to growth rate.

104

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

Fig. 10. Plot of the average annual precipitation (horizontal axis) and average temperature (vertical axis) where the 80 stalagmites considered in this study formed, with the symbols for those stalagmites color-coded according to quartiles of growth rate. In the five cases of two stalagmites from the same cave, the temperature used was increased for one by 0.3  C and decreased by 0.3  C for the other so that both symbols are apparent. Superposed on the plot is the Whittaker Biome Diagram (Whittaker, 1975), which shows that greatest growth rates are typical of biomes ranging from savanna or grassland to forest (Section 4.5). Only the ranges of temperature and precipitation considered by Whittaker (1975) have a white background, demonstrating that at least three of the stalagmites considered in this study formed in regions of exceptionally great rainfall. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

5. Conclusions

5.2. Growth rate and atmospheric precipitation

5.1. The range of growth rates of stalagmites

Three kinds of observation e data from 80 stalagmites worldwide comparing the growth rate of each with average annual precipitation (Figs. 3 and 4), data from individual stalagmites comparing thickness of annual layers with varying annual precipitation (Table 2), and data from at least one individual stalagmite comparing seasonal growth with seasonal drip rates and precipitation (Fig. 5) e combine to show that stalagmite growth rate does not increase monotonically with wetness. Instead, the general trend of growth rates has a maximum at annual precipitation rates

The growth rates of the 80 stalagmites from which data were taken range from 0.006 to 2.3 mm/yr, in a distribution strongly biased toward smaller values, giving a median of 0.10 and a mean of 0.23 mm/yr (Fig. 2). This range is large quantitatively, with a maximum 400 times the minimum, and compared to model results.

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

105

Table 1 Growth rates of later Holocene stalagmites. Precipitation (mm)

Temperature ( C)

Growth ratea (mm/yr)

Mineralogyb

Location

Cave

Stalagmite

Source

174 216 216 300 330 342 358 465 500 500 516 516 521 530 532 534 564 564 620 638 700 712 712 720 815 816 844 855 862 917 917 1000 1000 1000 1005 1005 1010 1082 1092 1100 1125 1150 1160 1179 1200 1200 1200 1260 1260 1289 1300 1353 1359 1364 1400 1400 1440 1500 1528 1600 1742 1750 1752 1855 1857 1857 1899 2036 2100 2121 2750 2790 3000

23 28.9 28.9 1.1 11 16 15.1 23 12.5 3.1 17.2 17.2 14 13 21 18.5 7.2 7.2 13 4 22.5 10.4 10.4 10.5 6.5 10.8 10.3 17 12.6 27 27 3.2 21 15 12.5 16 16.2 12.7 19.6 13 8.5 13.5 25 25 26 4 13.3 17.4 14 6.7 15.4 21.5 21.4 1.9 14 22.5 25.5 14 25 2.5 23.7 7.4 18.9 7 18.5 18.5 23 24.9 26.5 25 23.4 14.4 8

0.035 0.048 0.029 0.015 0.006 0.070 0.105 0.489 0.047 0.019 0.130 0.179 0.273 0.046 0.5 0.051 0.033 0.009 0.086 0.08 0.039 0.052 0.101 0.214 0.14 0.110 1.478 0.11 0.188 1.133 2.073 0.018 0.5 0.063 0.434 0.38 0.45 0.071 0.211 0.283 0.115 0.084 0.103 0.119 0.101 0.029 0.093 0.22 0.1 0.050 0.1 0.099 0.543 0.067 0.157 0.665 0.162 0.077 0.263 0.048 0.111 0.36 0.026 0.027 0.149 0.133 0.096 0.35 2.33 0.146 0.7 0.017 0.013

U C C C C U C A C C U U A C A&C C C C C C U C C C C C C C A A A C C C C A C C C A U ACG C&A C C C C C&A A C C C C U C A C C C C* C C C U C A&C U C C C C C C

Oman Socotra Socotra China Nevada Texas Kyrgyzstan Botswana Spain China Algeria Algeria South Africa China Namibia Spain Russia Russia China Romania Brazil Peru Peru Spain Minnesota Germany Belgium Peru Morocco Australia Australia Norway Ethiopia Lebanon France China Australia Missouri China China Japan India Madagascar Mexico Indonesia Idaho Turkey India India Italy China Florida Cuba Austria China Brazil Brazil China India Switzerland Taiwan China Brazil Scotland China China Puerto Rico Niue Guadalcanal Belize Vanuatu Japan Italy

Qunf Hoq Hoq Tianmen Lehman Carlsbad Uluu-Too Bone Molinos Kesang Gueldaman GLD1 Gueldaman GLD1 Cold Air Lianhua Dante El Refugio Kinderlinskaya Kinderlinskaya Kaiyuan Ascunsa Diva de Maura Huagapo Huagapo Kaite Spring Valley Bunker Han-sur-Lesse Shatuca Grotte de Piste KNI-51 KNI-51 Okshola Bero Jeita Villars Xianren Moondyne Devil's Ice Box Fulu Xianglong Uchimagi-do Sainji Anjohibe Rio Secreto Liang Luar Minnetonka Sofular Panigarh Dharamjali Grotta di Ernesto Yelang Briars Dos Anas Spannagel Zhuliuping Tamboril Curupira Niu Dandak Milchbach Jianfei Qingtian  Botuvera Uamh an Tartair Shennong Shennong Palco Avaiki Forestry Macal Chasm Taurius Fukugaguchi Corchia

Q5 Hq1 STM6 TM-18 WR11 BC2 Uluu-2 BC97-14 MO-7 CNKS-3 GLD1-stm2 GLD1-stm4 T7 LH9 DP1 REF-07 KC-1 KC-3 ky1 POM2 DV2 P00-H1 P09-H2 Buda-100 SVC983-1 Bu1 Proserpine Sha-2/3 GP5 KNI-51-F KNI-51-G FM3 Bero-1 JeG-stm-1 vil-stm1 YPXR5 MND-S1 DIB-1 FL4 XL21 UT-A SA-1 ANJ94-5 Itzamna LR06-B1 MC08-1 SO-1 & 10 PGH-1 DH-2 ER 76 20120824e13 BRIARS04-02 CG SPA 12 ZLP1 TMO ALHO6 N1 DAN-D MB-3 DGS-1 QT9 BTV21a SU963 SN4 SN20 PA 2b ASM1 10FC-02 MC01 Big Taurius FG01 CC26

Fleitmann et al., 2003 Van Rampelbergh et al., 2013 Van Rampelbergh et al., 2013 Cai et al., 2012 Steponaitis et al., 2015 Rasmussen et al., 2006 Wolff et al., 2017 Railsback et al., 2018 ~ oz et al., 2015 Mun Cai et al., 2017 Ruan et al., 2016 Ruan et al., 2016 Lee-Thorp et al., 2001 Dong et al., 2015 Sletten et al., 2013 Walczak et al., 2015 Baker et al., 2017 Baker et al., 2017 Wang et al., 2016 Dragusin et al., 2014 Novello et al., 2012 Kanner et al., 2013 Kanner et al., 2013 Cruz et al., 2015 Dasgupta et al., 2010 Fohlmeister et al., 2012 Van Rampelbergh et al., 2015 Bustamante et al., 2016 Wassenburg et al., 2013 Denniston et al., 2015 Denniston et al., 2015 Linge et al., 2009 Baker et al., 2010 Verheyden et al., 2008 Labuhn et al., 2015 Shen et al., 2013 Treble et al., 2005 Denniston et al., 2007 Zhu et al., 2015 Tan et al., 2015 Kato and Yamada 2016 Kotlia et al., 2015 Voarintsoa et al., 2017 Medina-Elizalde et al., 2016 Griffiths et al., 2010 Lundeen et al., 2013 € ktürk et al., 2011 Go Liang et al., 2015 Sanwal et al., 2013 Scholz et al., 2012 Zhao et al., 2015 Van Beynen et al., 2008 Fensterer et al., 2012 Vollweiler et al., 2006 Huang et al., 2016 Wortham et al., 2017 Novello et al., 2016 Zhao et al., 2016 Sinha et al., 2007 Luetscher et al. 2011 Li et al., 2015 Liu et al., 2015 Bernal et al., 2016 Baker et al., 2015 Zhang et al., 2015 Zhang et al., 2015 Rivera-Collazo et al., 2015 Rasbury and Aharon 2006 Maupin et al. (2014) Akers et al., 2016 Partin et al. (2013) Sone et al., 2015 Regattieri et al., 2014 (continued on next page)

106

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

Table 1 (continued ) Precipitation (mm)

Temperature ( C)

Growth ratea (mm/yr)

Mineralogyb

Location

Cave

Stalagmite

Source

3000 3023 3160 3435 5533 6100 6100

25 22.9 26.6 26 21 5 5

0.05 0.193 0.06 0.150 0.429 0.064 0.023

C A U C A C C

Indonesia Belize Andamans Borneo India Chile Chile

Buniayu Yok Balum Baratang Bukit Assam Mawmluh valoc Marcelo Are valoc Marcelo Are

CIAW15a YOK-I AN4 BA03 MAW-0201 MA1 MA3

Watanabe et al., 2010 Lechleitener et al. 2016 Laskar et al., 2013 Chen et al., 2016 Myers et al., 2015 Schimpf et al., 2011 Schimpf et al., 2011

a b c

For stalagmites in which growth rate has varied through time, the growth rate over the most recent decades in listed. A ¼ aragonite, C ¼ calcite, C* ¼ Calcite as a possible replacement of aragonite; G ¼ gypsum, U ¼ unreported. valo cave is the only cave in this list that is hosted by non-carbonate rock. The Marcelo Are

Fig. 11. Diagram showing spacing in time (diagonal lines) of stalagmite samples as a function of growth rate (horizontal axis) and vertical separation of samples (vertical axis). Broad €uselmann et al. (2015), and the green line is the vertical spacing of stable isotope samples by Zhao et al. (2015). Those burgundy band is the vertical spacing of age samples by Ha papers reported the closest spacing of the respective kinds of samples in a survey of all papers published in 2015 that reported paleoclimate data from stalagmites. The joined burgundy-filled and green-filled circles show the sampling by Steponaitis et al. (2015) of Stalagmite WR11, the slowest-growing stalagmite in the data set considered by this article. Vertical dashed line at 0.0005 mm/yr is the apparent growth rate below which Carolin et al. (2016) recognized a hiatus, rather than slow growth. One insight from this diagram is that sampling of stalagmites above the burgundy band and green line could yield useful long-term paleoclimate records from exceptionally slow-growing (and probably small) stalagmites from exceptionally cold and/or dry regions. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

between 700 and 2300 mm/year (Figs. 3 and 4). If one postulates that there are many fast-growing but as-yet-unsampled stalagmites in wet tropical regions, then the first of these three kinds of data fails to support the conclusion of a non-monotonic relationship, but the second and third continue to support it.

The observations described above indicate that paleoclimatological research cannot assume a monotonic relationship in which growth rate is invariably proportional to atmospheric precipitation, as discussed further in Section 5.5. They also remind one that, although hiatuses (i.e., periods of zero growth rate) have commonly

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

107

Fig. 12. Published curves of variation in stalagmite growth rate with drip interval (lower horizontal axis) or drip rate (upper horizontal axis) produced from modeling studies across the last four decades. Thicker and/or more solid curves represent results from more recent studies. Curves are from Fig. 3 of Dreybrodt (1981), Fig. 2 of Dreybrodt and Franke (1987), Fig. 2 of Kaufmann (2003), Fig. 2 of Kaufmann and Dreybrodt (2004), Fig. 3b of Mühlinghaus et al. (2007), and Fig. 1 of Dreybrodt (2016). The main point relevant to this article is that all models consistently predict greater growth rate with greater drip rate, across the entire range of drip rate.

Table 2 Articles reporting inverse relationships between growth rate and wetness. Article

Locationa

Stalagmite(s) or drip sites

Time scale

Proctor et al. (2000) Cai et al. (2010) Cai et al. (2011) Duan et al. (2012) Casteel and Banner (2015) Kato and Yamada (2016) Pu et al. (2016) Wang et al. (2016)

Scotland Thailand Northeastern China South-central China Texas, U.S.A. Japan Central China Eastern China

SU-96-7 NJ-1 C1 & C3 X6, X7, X8, & X11 WC3 UT-A D1 ky1

Interannual Interannual Seasonal Seasonal Seasonal Interannual Seasonal Seasonal

a The geographic expressions used here relate to the entire extent of the People's Republic of China. The preponderance of locations in China reflects the geographic distribution in the paleoclimatological literature, as shown by Fig. S3.

been assumed to represent dry conditions (e.g. Brook et al., 2010), slowing of deposition to zero can also result from wet conditions, as dissolutional surfaces at hiatuses suggest (Railsback et al., 2013). Further, the non-monotonic relationship of growth rate to atmospheric precipitation contradicts model results suggesting monotonic increase in growth rate with increasing drip rate.

5.3. Growth rate and temperature The combined data from 80 stalagmites show a positive relationship between temperature and growth rate (Fig. 6) that is further supported by more specific cave-monitoring studies (e.g., Fig. 8). This relationship is expected from modeling studies (e.g., Fig. 2 of Kaufmann and Dreybrodt, 2004), and Kaufmann and

108

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

proxy for relative change in past temperature. However, the scatter in Fig. 6 suggests that growth rate is at best a qualitative proxy for temperature. 5.4. Growth rate and mineralogy The combined data from 80 stalagmites show that growth rate of aragonite is commonly greater than that of calcite (Figs. 2, 4 and 6). This means that studies using growth rate as a paleoclimatological proxy, whether of precipitation or temperature, need to determine and report mineralogy throughout the stalagmite(s) used, rather than assuming the presence of only calcite. Furthermore, the issues discussed in Section S4 of Supplementary Document 1 demonstrate that credible reports of mineralogy require instrumental determination (e.g., by X-ray diffraction). 5.5. Growth rate and the paleoclimatology of wet and dry Fig. 13. A. Simple relationship between growth rate of stalagmites and average annual precipitation generalized from Fig. 3, solely as the basis for Parts B and C. B. Four scenarios of changing precipitation through time. C. The one record of stalagmite growth rate resulting from all four scenarios in Part B, as discussed in Section 5.5.

Dreybrodt (2004) generalized that “stalagmite growth rates are strongly correlated to temperature”. Growth rate has been used as an indicator of past temperature (e.g., McDermott et al., 1999; Tan et al., 2013), and recent studies have focused on controls on dripwater temperature (Rau et al., 2015). When one considers that temperature variation has generally been the foremost objective of paleoclimatological studies, whereas paleoclimatologists using stalagmites rather reluctantly accepted that stalagmites are largely a record of wetness and dryness (Fairchild and McMillan, 2007), it is surprising that growth rate is not more commonly used as a

The evidence discussed in Section 5.2 indicates that paleoclimatological research cannot assume a monotonic relationship in which growth rate is invariably proportional (or invariably inversely proportional) to atmospheric precipitation. In fact, the non-monotonic relationship between precipitation and growth rate means that, in regions of intermediate precipitation, one record of growth rate can represent multiple precipitation histories (Fig. 13). Even in regions of consistently low precipitation or consistently high precipitation, the efficacy of growth rate as a proxy for precipitation is not guaranteed because inter-annual variation in temperature may destructively interfere with any potential relationship between precipitation and growth rate (Fig. 14). Growth rate should logically be a qualitative proxy of temperature and precipitation in regions of low precipitation where warmer years (or multi-year periods) are wetter years (or periods) (Case A in Fig. 14), and likewise in regions of high precipitation where

Fig. 14. Matrix of responses of growth to variation in wetness of climate (horizontal axis) and pattern of climate change at the inter-annual to longer scale (vertical axis). In all six cases, warmer conditions would favor faster growth, but a linkage of wetness to temperature (as suggested by the vertical axis) would mean that in only two cases would changing precipitation reinforce that trend (Cases A and F). Conversely, in Cases C and D the control of changing temperature and changing precipitation on growth rate would act against each other, as discussed in Section 5.5.

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

warmer years or longer periods are drier years or periods (Case F in Fig. 14). However, its strength as a proxy for either temperature or precipitation may break down in the opposite scenarios where temperature and precipitation exert countervailing controls on growth rate of stalagmites (Cases C and D in Fig. 14). 5.6. New directions Stalagmite-based paleoclimatology has not progressed significantly into cold ecosystems (Fig. 1), but calculations suggest that recent advances in fine-scale sampling of stalagmites could allow development of long-term records from stalagmites slowergrowing and smaller than those commonly used today (Fig. 11). Acknowledgements Prof. Dr. Wolfgang Dreybrodt provided comments on an early draft of the manuscript that were critical to its improvement. The manuscript was improved by the comments by two QSR reviewers, of which Dr. Judson Partin is especially thanked for his suggestions about fast-growing stalagmites in wet tropical regions. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.quascirev.2018.03.002. References Akers, P.D., Brook, G.A., Railsback, L.B., Liang, F., Iannone, G., Webster, J.W., Reeder, P.P., Cheng, H., Edwards, R.L., 2016. An extended and higher-resolution record of climate and land use from stalagmite MC01 from Macal Chasm, Belize, revealing connections between major dry events, overall climate variability, and Maya sociopolitical changes. Palaeogeogr. Palaeoclimatol. Palaeoecol. 459, 268e288. Anwer, M., 2015. Nature of centennial global climate change from observational records. Am. J. Clim. Change 4, 337e354. Baker, A., Genty, D., Dreybrodt, W., Barnes, W.L., Mockler, N.J., Grapes, J., 1998. Testing theoretically predicted stalagmite growth rate with recent annually laminated samples: implications for past stalagmite deposition. Geochem. Cosmochim. Acta 62 (3), 393e404. https://doi.org/10.1016/S0016-7037(97) 00343-8. Baker, A., Smith, C.L., Jex, C., Fairchild, I.J., Genty, D., Fuller, L., 2008. Annually laminated speleothems: a review. Int. J. Speleol. 37, 193e206. Baker, A., Asrat, A., Fairchild, I.J., Leng, M.J., Thomas, L., Widmann, M., Jex, C.N., Dong, B., van Calsteren, P., Bryant, C., 2010. Decadal-scale rainfall variability in Ethiopia recorded in an annually laminated, Holocene-age, stalagmite. Holocene 20, 827e836. Baker, A., Hellstrom, J.C., Kelly, B.F.J., Mariethoz, G., Trouet, V., 2015. A composite annual-resolution stalagmite record of North Atlantic climate over the last three millennia. Sci. Rep. 5, 10307. https://doi.org/10.1038/srep10307. Baker, J.L., Lachniet, M.S., Chervyastsova, O., Asmerom, Y., Polyak, V.J., 2017. Holocene warming in western continental Eurasia driven by glacial retreat and greenhouse forcing. Nat. Geosci. 10, 430e435. Baldini, J.U.L., 2010. Cave atmosphere controls on stalagmite growth rate and palaeoclimate records. In: Pedley, H.M., Rogerson, M. (Eds.), Tufas and Speleothems: Unravelling the Microbial and Physical Controls. Geological Society Special Publication 36, pp. 283e294. Banner, J.L., Guilfoyle, A., James, E.W., Stern, L.A., Musgrove, M., 2007. Seasonal variations in modern speleothem calcite growth in Central Texas, USA. J. Sediment. Res. 77, 615e622. Bernal, J.P., Cruz, F.W., Stríkis, N.M., Wang, X., Deininger, M., Catunda, M.C.A., Ortegan, C., Cheng, H., Edwards, R.L., Auler, A.S., 2016. High-resolution HoloObergo cene South American monsoon history recorded by a speleothem from Botuvera cave, Brazil. Earth Planet Sci. Lett. 450, 186e196. Berner, R.A., 1975. The role of magnesium in the crystal growth of calcite and aragonite from sea water. Geochem. Cosmochim. Acta 39, 489e504. Brook, G.A., Scott, L., Railsback, L.B., Goddard, E.A., 2010. A 35 ka pollen and isotope record of environmental change along the southern margin of the Kalahari from a stalagmite and animal dung deposits in Wonderwerk cave. S. Afr.: J. Arid Environ. 74, 870e884. Burton, E.A., Walter, L.M., 1987. Relative precipitation rates of aragonite and Mg calcite from seawater: temperature or carbonate ion control? Geology 15, 111e114. stegui, J., Strikis, N., Panizo, G., Bustamante, M.G., Cruz, F.W., Vuille, M., Apae Novello, F.V., Deininger, A., Sifeddine, A., Cheng, H., Moquet, J.S., Guyot, J.L.,

109

Santos, R.V., Segura, H., Edwards, R.L., 2016. Holocene changes in monsoon precipitation in the Andes of NE Peru based on d18O speleothem records. Quat. Sci. Rev. 146, 274e287. Cabrol, P., Coudray, J., 1982. Climatic fluctuations influence the genesis and diagenesis of carbonate speleothems in southwestern France. NSS (Natl. Speleol. Soc.) Bull. 44, 112e117. Cai, B., Pumijumnong, N., Tan, M., Muangsong, C., Kong, X., Jiang, X., Nan, S., 2010. Effects of intraseasonal variation of summer monsoon rainfall on stable isotope and growth rate of a stalagmite from northwestern Thailand. J. Geophys. Res. 115, D21104. https://doi.org/10.1029/2009JD013378. Cai, B., Zhu, J., Ban, F., Tan, M., 2011. Intra-annual variation of the calcite deposition rate of drip water in Shihua Cave, Beijing, China and its implications for palaeoclimatic reconstructions. Boreas 40, 525e535. Cai, Y., Zhang, H., Cheng, H., An, Z., Edward, R.L., Wang, X., Tan, L., Liang, F., Wang, J., Kelly, M., 2012. The Holocene Indian monsoon variability over the southern Tibetan Plateau and its teleconnections. Earth Planet Sci. Lett. 335e336, 135e144. Cai, Y., Chiang, J.C.H., Breitenbach, S.F.M., Tan, L., Cheng, H., Edwards, R.L., An, Z., 2017. Holocene moisture changes in western China, Central Asia, inferred from stalagmites. Quat. Sci. Rev. 158, 15e28. Carolin, S.A., Cobb, K.M., Lynch-Stieglitz, J., Moermana, J.W., Partin, J.W., Lejau, S., Malang, J., Clark, B., Tuen, A.A., Adkins, J.F., 2016. Northern Borneo stalagmite records reveal West Pacific hydroclimate across MIS 5 and 6. Earth Planet Sci. Lett. 439, 182e193. Casteel, R.C., Banner, J.L., 2015. Temperature-driven seasonal calcite growth and drip water trace element variations in a well-ventilated Texas cave: implications for speleothem paleoclimate studies. Chem. Geol. 392, 43e58. Chen, S., Hoffman, S.S., Lund, D.C., Cobb, K.C., Emile-Geay, J., Adkins, J.F., 2016. A high-resolution speleothem record of western equatorial Pacific rainfall: implications for Holocene ENSO evolution. Earth Planet Sci. Lett. 442, 61e71. ceres, J.O., Marín-Rolda n, A., Ortega, A.I., Garralo  n, A., Cruz, J.A., Turrero, M.J., Ca nchez, L., Go  mez, P., Mun ~ oz-Garica, M.B., Edwards, R.L., Martín-Chivelt, J., Sa 2015. Long-term hydrological changes in northern Iberia (4.9e0.9 ky BP) from ~ a Karst Complex, speleothem Mg/Ca ratios and cave monitoring (Ojo Guaren Spain). Environ. Earth Sci. 74, 7741e7753. Dasgupta, S., Saar, M.O., Edwards, R.L., Shen, C.-C., Cheng, H., Alexander Jr., E.C., 2010. Three thousand years of extreme rainfall events recorded in stalagmites from Spring Valley Caverns, Minnesota. Earth Planet Sci. Lett. 300, 46e54. Denniston, R.F., DuPree, M., Dorale, J.A., Asmerom, Y., Polyak, V.J., Carpenter, S.J., 2007. Episodes of late Holocene aridity recorded by stalagmites from Devil's Icebox Cave, central Missouri, USA. Quat. Res. 68, 45e52. Denniston, R.F., Villarini, G., Gonzalez, A.N., Wyrwoll, K.-H., Polyak, V.J., Ummenhofer, C.C., Lachniet, M.S., Wanamaker Jr., A.D., Humphreys, W.F., Woods, D., Cugley, J., 2015. Extreme rainfall activity in the Australian tropics reflects changes in the El Nineo/Southern Oscillation over the last two millennia. Proc. Natl. Acad. Sci. Unit. States Am. 112, 4576e4581. Dong, J., Shen, C.-C., Kong, X., Wang, H.-C., Jiang, X., 2015. Reconciliation of hydroclimate sequences from the Chinese loess plateau and low-latitude East Asian summer monsoon regions over the past 14,500 years. Palaeogeogr. Palaeoclimatol. Palaeoecol. 435, 127e135. Dragusin, V., Staubwasser, M., Hoffman, D.L., Ersek, V., Onac, B.P., Veres, D., 2014. Constraining Holocene hydrological changes in the CarpathianeBalkan region using speleothem d18O and pollen-based temperature reconstructions. Clim. Past 10, 1363e1380. Dreybrodt, W., 1980. Deposition of calcite from thin films of natural calcareous solutions and the growth of speleothems. Chem. Geol. 29, 89e105. Dreybrodt, W., 1981. The kinetics of calcite precipitation from thin films of calcareous solutions and the growth of speleothems: revisited. Chem. Geol. 30, 237e245. Dreybrodt, W., 1999. Chemical kinetics, speleothem growth, and climate. Boreas 28, 347e356. Dreybrodt, W., Franke, H.W., 1987. Wachstumsgeschwindigkeiten und Durchmesser €hle 38, 1e6. von Kerzenstalagmiten. Die Ho Dreybrodt, W., 2016. Comment on “Stoll H. et al. (2015): interpretation of orbital scale variability in mid-latitude speleothem d18O: significance of growth rate controlled kinetic fractionation effects. Quat. Sci. Rev. 127, 215e228. Quaternary Science Reviews 142, 179e181. Duan, W., Cai, B., Tan, M., Liu, H., Zhang, Y., 2012. The growth mechanism of the aragonitic stalagmite laminae from Yunnan Xianren Cave, SW China revealed by cave monitoring. Boreas 41, 113e123. Dutt, S., Gupta, A.K., Clemens, S., Cheng, H., Singh, R.K., Kathyat, G., Edwards, R.L., 2015. Abrupt changes in Indian summer monsoon strength during 33,800 to 5500years B.P. Geophys. Res. Lett. 42, 5526e5532. https://doi.org/10.1002/ 2015GL064015. Fairchild, I.J., Baker, A., 2012. Speleothem Science: from Process to Past Environments. Wiley-Blackwell, Chichester. Fairchild, I.J., Borsato, A., Tooth, A.F., Frisia, S., Hawkesworth, C.J., Huang, Y., McDermott, F., Spiro, B., 2000. Controls on trace element (Sr-Mg) compositions of carbonate cave waters: implications for speleothem climatic records. Chem. Geol. 166, 255e269. Fairchild, I.J., McMillan, E.A., 2007. Speleothems as indicators of wet and dry periods. Int. J. Speleol. 36, 69e74. €tl, C., Pajo  n, J.M., Mangini, A., 2012. Cuban Fensterer, C., Scholz, D., Hoffman, D., Spo stalagmite suggests relationship between Caribbean precipitation and the Atlantic Multidecadal Oscillation during the past 1.3 ka. Holocene 22,

110

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111

1405e1412. Fleitmann, D., Burns, S.J., Mudelsee, M., Neff, U., Kramers, J., Mangini, A., Matter, A., 2003. Holocene forcing of the Indian monsoon recorded in a stalagmite from southern Oman. Science 300, 1737e1739. €der-Rizrau, A., Scholz, D., Spo €tl, C., Riechelmann, D.F.C., Fohlmeister, J., Schro Mudelsee, M., Wackerbarth, A., Gerdes, A., Riechelmann, S., Immenhauser, A., Richter, D.K., Mangini, A., 2012. Bunker Cave stalagmites: an archive for central European Holocene climate variability. Clim. Past 8, 1751e1764. Ford, D.C., Williams, P.W., 1989. Karst Geomorphology and Hydrology. Unwin Hyman, London and Boston. Ford, D.C., Williams, P.W., 2007. Karst Hydrogeology and Geomorphology. John Wiley & Sons, Chichester and Hoboken. Genty, D., Baker, A., Vokal, B., 2001. Intra- and inter-annual growth rate of modern stalagmites. Chem. Geol. 176, 191e212. Gilli, E., 2015. Karstology: Karsts, Caves and Springs. CRC Press, Boca Raton. Givnish, T.J., 2002. Adaptive significance of evergreen vs. deciduous leaves: solving the triple paradox. Silva Fenn. 36, 703e743. €ktürk, O.M., Fleitmann, D., Badertscher, S., Cheng, S., Edwards, R.L., Go Leuenberger, M., Fankhauser, A., Tüysüz, O., Kramers, J., 2011. Climate on the southern black sea coast during the Holocene: implications from the Sofular cave record. Quat. Sci. Rev. 30, 2433e2445. Griffiths, M.J., Drysdale, R.N., Gagan, M.K., Frisia, A., Zhao, J.-X., Ayliffe, L.K., Hantoro, W.S., Hellstrom, J.C., Fischer, M.J., Fend, Y.-X., Suwargadi, B.W., 2010. Evidence for Holocene changes in AustralianeIndonesian monsoon rainfall from stalagmite trace element and stable isotope ratios. Earth Planet Sci. Lett. 292, 27e38. €uselmann, A.D., Fleitmann, D., Cheng, H., Tabersky, D., Günther, D., Edwards, R.L., Ha 2015. Timing and nature of the penultimate deglaciation in a high alpine stalagmite from Switzerland. Quat. Sci. Rev. 126, 264e275. Huang, W., Wang, Y., Cheng, H., Edwards, R.L., Shen, C.-C., Liu, D., Shao, Q., Deng, C., Zhang, Z., Wang, Q., 2016. Multi-scale Holocene Asian monsoon variability deduced from a twin-stalagmite record in southwestern China. Quat. Res. 86, 34e44. Kanner, L.C., Burns, S.J., Cheng, H., Edwards, R.L., Vuille, M., 2013. High-resolution variability of the South American summer monsoon over the last seven millennia: insights from a speleothem record from the central Peruvian Andes. Quat. Sci. Rev. 75, 1e10. Kato, H., Yamada, T., 2016. Controlling factors in stalagmite oxygen isotopic composition and the paleoprecipitation record for the last 1,100 years in Northeast Japan. Geochem. J. 50, 1e6. Kaufmann, G., 2003. Stalagmite growth and palaeo-climate: the numerical perspective. Earth Planet Sci. Lett. 214, 251e266. Kaufmann, G., Dreybrodt, W., 2004. Stalagmite growth and palaeo-climate: an inverse approach. Earth Planet Sci. Lett. 224, 529e545. Klimchuk, A.B., 2000. Speleogenesis: Evolution of Karst Aquifers. National Speleological Society, Huntsville. Kotlia, B.S., Singh, A.K., Joshi, L.M., Dhaila, B.S., 2015. Precipitation variability in the indian central Himalaya during last ca. 4,000 years inferred from a speleothem record: impact of indian summer monsoon (ISM) and Westerlies. Quat. Int. 371, 244e253. Labuhn, I., Genty, D., Vonhof, H., Bourdin, C., Blamart, D., Douville, E., Ruan, J., Cheng, H., Edwards, R.L., Pons-Branchu, E., Pierre, M., 2015. A high-resolution fluid inclusion d18O record from a stalagmite in SW France: modern calibration and comparison with multiple proxies. Quat. Sci. Rev. 110, 152e165. Laskar, A.H., Yadava, M.G., Ramesh, R., Polyak, V.J., Asmerom, Y., 2013. A 4 kyr stalagmite oxygen isotopic record of the past Indian Summer Monsoon in the Andaman Islands. G-cubed 14, 3555e3566. Lauriol, B., Ford, D.C., Cinq-Mars, J., Morris, A.W., 1997. The chronology of speleothem deposition in northern Yukon and its relationships to permafrost. Can. J. Earth Sci. 34, 902e911. Lechleitener, F.A., Baldini, J.U.L., Breitenbach, S.F.M., Fohlmeister, J., McIntyre, C., Goswami, B., Jamieson, R.A., van der Voort, T.S., Prufer, K., Marwan, N., Culleton, B.J., Kennett, D.J., Asmerom, Y., Polyak, V., Eglinton, T.I., 2016. Hydrological and climatological controls on radiocarbon concentrations in a tropical stalagmite. Geochem. Cosmochim. Acta 194, 233e252. Lee-Thorp, J.A., Holmgren, K., Lauritzen, S.E., Linge, H., Moberg, A., Partridge, T.C., Stevenson, C., Tyson, P., 2001. Rapid climate shifts in the southern African interior throughout the mid to late Holocene. Geophys. Res. Lett. 28, 4507e4510. Li, H.-C., Zhao, M., Tsai, C.-H., Mii, H.-S., Chang, Q., Wei, K.-W., 2015. The first highresolution stalagmite record from Taiwan: climate and environmental changes during the past 1300 years. J. Asian Earth Sci. 114, 574e587. Liang, F., Brook, G.A., Kotlia, B.S., Railsback, L.B., Hardt, B., Cheng, H., Edwards, R.L., Kandasamy, S., 2015. Panigarh cave stalagmite evidence of climate change in the Indian Central Himalaya since AD 1256: monsoon breaks and winter southern jet depressions. Quat. Sci. Rev. 124, 145e161. Linge, H., Lauritzen, S.E., Andersson, C., Hansen, J.K., Skoglund, R.Ø., Sundqvist, H.S., 2009. Stable isotope records for the last 10 000 years from Okshola cave (Fauske, northern Norway) and regional comparisons. Clim. Past 5, 667e682. Liu, D., Wang, Y., Cheng, H., Edwards, R.L., Kong, X., 2015. Cyclic changes of Asian monsoon intensity during the early mid-Holocene from annually-laminated stalagmites, central China. Quat. Sci. Rev. 121, 1e10. Lowenstam, H.A., Weiner, S., 1989. On Biomineralization. Oxford University Press, New York, p. 324. €tl, C., 2011. Holocene glacier history from Luestcher, M., Hoffman, D.L., Frisia, S., Spo

alpine speleothems, Milchbach cave, Switzerland. Earth Planet Sci. Lett. 302, 95e106. Lundeen, Z., Brunelle, A., Burns, S.J., Polyak, V., Asmerom, A., 2013. A speleothem record of Holocene paleoclimate from the northern Wasatch Mountains, southeast Idaho, USA. Quat. Int. 310, 83e95. Mariethoz, G., Kelly, B.F.J., Baker, A., 2012. Quantifying the value of laminated stalagmites for paleoclimate reconstructions. Geophys. Res. Lett. 39, L05407. https://doi.org/10.1029/2012GL050986. Maupin, C.R., Partin, J.W., Shen, C.-C., Quinn, T.M., Lin, K., Taylor, F.W., Banner, J.L., Thirumalai, K., Sinclair, D.J., 2014. Persistent decadal-scale rainfall variability in the tropical South Pacific Convergence Zone through the past six centuries. Clim. Past 10, 1319e1332. McDermott, F., Frisia, S., Huang, Y., Longinelli, A., Spiro, B., Heaton, T.H.E., Hawkesworth, C.J., Borsato, A., Keppens, E., Fairchild, I.J., van der Borg, K., Verheyden, S., Selmo, E., 1999. Holocene climate variability in Europe: evidence from d18O, textural and extension-rate variations in three speleothems. Quat. Sci. Rev. 18, 1021e1038. Medina-Elizalde, M., Burns, S.J., Polanco-Martínez, J.M., Beach, T., LasesHern andez, F., Shen, C.-C., Wang, H.-C., 2016. High-resolution speleothem record of precipitation from the Yucatan Peninsula spanning the Maya Preclassic period. Global Planet. Change 138, 93e102. Milanovic, P.T., 1981. Karst Hydrogeology. Water Resources Publications, Littleton, Colorado. Morse, J.W., Wang, Q., Mai Yin Tsio, M.Y., 1997. Influences of temperature and Mg:Ca ratio on CaCO3 precipitates from seawater. Geology 25, 85e87. Muangsong, C., Cai, B., Pumijumnong, N., Hu, C., Cheng, H., 2014. An annually laminated stalagmite record of the changes in Thailand monsoon rainfall over the past 387 years and its relationship to IOD and ENSO. Quat. Int. 346, 90e97. Mühlinghaus, C., Scholz, D., Mangini, A., 2007. Modelling stalagmite growth and d13C as a function of drip interval and temperature. Geochem. Cosmochim. Acta 71, 2780e2790. ~ oz, A., Bartolome , M., Mun ~ oz, A., Sancho, C., Moreno, A., Hellstrom, J.C., Mun car, M.C., Cacho, I., 2015. Solar influence and hydrological variability during Osa the Holocene from a speleothem annual record (Molinos Cave, NE Spain). Terra. Nova 7, 300e311. Murray, J.W., 1954. The deposition of calcite and aragonite in caves. J. Geol. 62, 481e492. Myers, C.G., Oster, J.L., Sharp, W.D., Bennartz, R., Kelley, N.P., Covey, A.K., Breitenbach, S.F.M., 2015. Northeast Indian stalagmite records Pacific decadal climate change: implications for moisture transport and drought in India. Geophys. Res. Lett. 42, 4124e4132. Nehme, C., Verheyden, S., Noble, S., Farrant, A., Delannoy, J., 2015. Contribution of an accurate growth rate reconstruction of a stalagmite from the Kanaan CaveLebanon to the understanding of humidity variations in the Levant during the MIS 5. Geol. Belg. 18, 102e108. Novello, V.F., Cruz, F.W., Karmann, I., Burns, S.J., Strikis, N.M., Vuille, M., Cheng, H., Edwards, R.L., Santos, R.V., Frigo, E., Barreto, E.A.S., 2012. Multidecadal climate ́ variability in Brazil's Nordeste during the last 3000 years based on speleothem isotope records. Geophys. Res. Lett. 39, L23706,. https://doi.org/10.1029/ 2012GL053936. Novello, V.F., Vuille, M., Cruz, F.W., Strikis, N.M., de Paula, M.S., Edwards, R.L., Cheng, H., Karmann, I., Jaqueto, P.F., Trinidade, R.I.F., Hartmann, G.A., Moquet, J.S., 2016. Centennial-scale solar forcing of the South American monsoon system recorded in stalagmites. Sci. Rep. 6, 24762. https://doi.org/ 10.1038/srep24762. Partin, J.W., Quinn, T.M., Shen, C.-C., Emile-Geay, J., Taylor, F.W., Maupin, C.R., Lin, K., Jackson, C.S., Banner, J.L., Sinclair, D.J., Huh, C.-A., 2013. Multidecadal rainfall variability in South Pacific Convergence Zone as revealed by stalagmite geochemistry. Geology 41, 1143e1146. es dam la Grotto de Moulis Pobeguin, T., 1955. Sur les concretions calcaires observe ge). Acade mie des Sciences (Paris). Comptes Rendus Hebdomadaires des (Are ances 241, 1791e1793. Se Proctor, C.J., Baker, A., Barnes, W.L., Gilmour, M.A., 2000. A thousand year speleothem proxy record of North Atlantic climate from Scotland. Clim. Dynam. 16, 815e820. Pu, J., Wang, A., Shen, L., Yin, J., Yuan, D., Zhao, H., 2016. Factors controlling the growth rate, carbon and oxygen isotope variation in modern calcite precipitation in a subtropical cave, Southwest China. J. Asian Earth Sci. 119, 167e178. Railsback, L.B., 2017. Rain, riches, and empire: the relationship between nations ruling distant lands, nations of great wealth, and regions of regular moderate atmospheric precipitation. Weather, Climate, and Soci. 9, 455e469. Railsback, L.B., Brook, G.A., Chen, J., Kalin, R., Fleisher, C.J., 1994. Environmental controls on the petrology of a Late Holocene speleothem from Botswana with annual layers of aragonite and calcite. J. Sediment. Res. A64, 147e155. Railsback, L.B., Akers, P.D., Wang, L., Holdridge, G.A., Voarintsoa, N.R., 2013. Layerbounding surfaces in stalagmites as keys to better paleoclimatological histories and chronologies. Int. J. Speleol. 42, 167e180. Railsback, L.B., Brook, G.A., Liang, F., Voarintsoa, N.R.G., Cheng, H., Edwards, R.L., 2018. A Multi-proxy Stalagmite Record from Northwestern Botswana Suggesting Wetness Late in the Little Ice Age and Drying Thereafter in Response to Changing Migration of the ITCZ. in prep. Rasbury, M., Aharon, P., 2006. ENSO-controlled rainfall variability records archived in tropical stalagmites from the mid-ocean island of Niue, South Pacific. Gcubed 7, Q07010. https://doi.org/10.1029/2005GC001232. Rasmussen, J.B.T., Polyak, V.J., Asmerom, Y., 2006. Evidence for Pacific-modulated

L.B. Railsback / Quaternary Science Reviews 187 (2018) 94e111 precipitation variability during the late Holocene from the southwestern USA. Geophys. Res. Lett. 33, L08701. https://doi.org/10.1029/2006GL025714. Rau, G.C., Cuthbert, M.O., Andersen, M.S., Baker, A., Rutlidge, H., Markowska, M., Roshan, H., Marjo, C.E., Graham, P.W., Acworth, R.I., 2015. Controls on cave drip water temperature and implications for speleothem-based paleoclimate reconstructions. Quat. Sci. Rev. 127, 19e36. Regattieri, E., Zanchetta, G., Drysdale, R.N., Isola, I., Hellstrom, J.C., Dallai, L., 2014. Lateglacial to Holocene trace element record (Ba, Mg, Sr) from corchia cave (Apuan Alps, central Italy): paleoenvironmental implications. J. Quat. Sci. 29, 381e392. Ricklefs, R.E., 2001. The Economy of Nature, fifth ed. W.H. Freeman and Co., New York. Rivera-Collazo, I., Winter, A., Scholz, D., Mangini, A., Miller, T., Kushnir, Y., Black, D., 2015. Human adaptation strategies to abrupt climate change in Puerto Rico ca. 3.5 ka. Holocene 25, 627e640. Ruan, J., Kherbouche, F., Genty, D., Blamart, D., Cheng, H., Dewilde, F., Hachi, S., niger, E., Michelot, J.-L., 2016. Evidence of a prolonged drought Edwards, R.L., Re ca. 4200 yr BP correlated with prehistoric settlement abandonment from the Gueldaman GLD1 Cave, Northern Algeria. Clim. Past 12, 1e14. Sanwal, J., Kotlia, B.S., Rajendran, C., Ahmad, S.M., Rajendran, K., Sandiford, M., 2013. Climatic variability in Central Indian Himalaya during the last w1800 years: evidence from a high resolution speleothem record. Quat. Int. 304, 183e192. €tl, C., Wo €rner, G., Deininger, M., Schimpf, D., Kilian, R., Kronz, A., Simon, K., Spo Mangini, A., 2011. The significance of chemical, isotopic, and detrital components in three coeval stalagmites from the superhumid southernmost Andes (53 S) as high-resolution palaeo-climate proxies. Quat. Sci. Rev. 30, 443e459. €tl, C., Fohlmeister, J., Mudelsee, M., Miorandi, R., Scholz, D., Frisia, S., Borsato, A., Spo Mangini, A., 2012. Holocene climate variability in north-eastern Italy: potential influence of the NAO and solar activity recorded by speleothem data. Clim. Past 8, 1367e1383. Shen, C.-C., Lin, K., Duan, W., Jiang, X., Partin, J.W., Edwards, R.L., Ching, H., Tan, M., 2013. Testing the annual nature of speleothem banding. Sci. Rep. 3, 2633. Siegel, F.R., 1965. Aspects of calcium carbonate deposition in Great onyx cave, Kentucky. Scdimentology 4, 285e299. Sinha, A., Cannariato, K.G., Stott, L.D., Cheng, H., Edwards, R.L., Yadava, M.G., Ramesh, R., Singh, I.B., 2007. A 900-year (600 to 1500 A.D.) record of the Indian summer monsoon precipitation from the core monsoon zone of India. Geophys. Res. Lett. 34, L16707. https://doi.org/10.1029/2007GL030431. Sletten, H.R., Railsback, L.B., Liang, F., Brook, G.A., Marais, E., Hardt, B.F., Cheng, H., Edwards, R.L., 2013. A petrographic and geochemical record of climate change over the last 4600 years from a northern Namibia stalagmite, with evidence of abruptly wetter climate at the beginning of southern Africa's Iron Age. Palaeogeogr. Palaeoclimatol. Palaeoecol. 376, 149e162. €hnel, O., Garside, J., 1992. Precipitation. Butterworth-Heinemann, Oxford. So Sone, T., Kano, A., Kashiwagi, K., Mori, T., Okumura, T., Shen, C.-C., Hori, M., 2015. Two modes of climatic control in the Holocene stalagmite record from the Japan Sea side of the Japanese islands. Isl. Arc 24, 342e358. €tl, C., Fohlmeister, J., Cheng, H., Boch, R., 2016. Modern aragonite formation at Spo near-freezing conditions in an alpine cave, Carnic Alps, Austria. Chem. Geol. 435, 60e70. Steponaitis, E., Andrews, A., McGee, D., Quade, J., Hsieh, Y.-T., Broecker, W.S., Shuman, B.N., Burns, S.J., Cheng, H., 2015. Mid-Holocene drying of the U.S. Great Basin recorded in Nevada speleothems. Quat. Sci. Rev. 127, 174e185. Tan, L., Yi, L., Cai, Y., Shen, C.-C., Cheng, H., An, Z., 2013. Quantitative temperature reconstruction based on growth rate of annually-layered stalagmite: a case study from central China. Quat. Sci. Rev. 72, 137e145. Tan, L., Cai, Y., Cheng, H., Edwards, R.L., Shen, C.-C., Gao, Y., An, Z., 2015. Climate significance of speleothem d18O from central China on decadal timescale. J. Asian Earth Sci. 106, 150e155. Thrailkill, J., 1971. Carbonate deposition in carlsbad caverns. J. Geol. 79, 683e695. Treble, P.C., Chappell, J., Gagan, M.K., McKeegan, K.D., Harrison, T.M., 2005. In situ measurement of seasonal d18O variations and analysis of isotopic trends in a modern speleothem from southwest Australia. Earth Planet Sci. Lett. 233, 17e32. Van Beynen, P.E., Soto, L., Polk, J., 2008. Variable calcite deposition rates as proxy for paleo-precipitation determination as derived from speleothems in Central Florida, U.S.A. J. Cave Karst Stud. 70, 25e34.

111

Van Rampelbergh, M., Fleitmann, D., Verheyden, S., Cheng, H., Edwards, R.L., De Geest, P., De Vleeschouwer, D., Burns, S.J., Matter, A., Claeys, P., Keppens, E., 2013. Mid- to late Holocene Indian Ocean Monsoon variability recorded in four speleothems from Socotra Island, Yemen. Quat. Sci. Rev. 65, 129e142. Van Rampelbergh, M., Verheyden, S., Allan, M., Quinif, Y., Cheng, H., Edwards, R.L., Keppens, E., Claeys, P., 2015. A 500-year seasonally resolved d18O and d13C, layer thickness and calcite aspect record from a speleothem deposited in the Hansur-Lesse cave, Belgium. Clim. Past 11, 789e802. Verheyden, S., Nader, F.H., Cheng, H.J., Edwards, R.L., Swennen, R., 2008. Paleoclimate reconstruction in the Levant region from the geochemistry of a Holocene stalagmite from the Jeita cave, Lebanon. Quat. Res. 70, 368e381. Voarintsoa, N.R.G., Wang, L., Railsback, L.B., Brook, G.A., Liang, F., Cheng, H., Edwards, R.L., 2017. Multiple proxy analyses of a U/Th-dated stalagmite to reconstruct paleoenvironmental changes in northwestern Madagascar between 370 CE and 1300 CE. Palaeogeogr. Palaeoclimatol. Palaeoecol. 469, 138e155. € tl, C., 2006. A precisely Vollweiler, N., Scholz, D., Mühlinghaus, C., Mangini, A., Spo dated climate record for the last 9 kyr from three high alpine stalagmites, Spannagel Cave, Austria. Geophys. Res. Lett. 33, L20703,. https://doi.org/ 10.1029/2006GL027662. Walczak, I.W., Baldini, J.U.L., Baldini, L.M., McDermott, F., Marsden, S., Standish, C.D., Richards, D.A., Andreo, B., Slater, J., 2015. Reconstructing high-resolution climate using CT scanning of unsectioned stalagmites: A case study identifying the mid-Holocene onset of the Mediterranean climate in southern Iberia. Quat. Sci. Rev. 127, 117e128. Wang, Q., Zhou, H., Cheng, K., Chi, H., Shen, C.-C., Wang, C., Ma, Q., 2016. The climate reconstruction in Shandong Peninsula, northern China, during the last millennium based on stalagmite laminae together with a comparison to d18O. Clim. Past 12, 871e881. Wassenburg, J.A., Immenhaiser, A., Richter, D.K., Niedermayr, A., Riechelmann, S., €der-Ritzrau, A., Fietzke, J., Scholz, D., Jochum, K.P., Fohlmeister, J., Schro Sabaoui, A., Richelmann, D.F.C., Schneider, L., Esper, J., 2013. Moroccan speleothem and tree ring records suggest a variable positive state of the North Atlantic Oscillation during the Medieval Warm Period. Earth Planet Sci. Lett. 375, 291e302. Watanabe, Y., Matsuoka, H., Sakai, S., Ueda, J., Yamada, M., Ohsawa, S., Kiguchi, M., Satomura, T., Nakai, S., Brahmantyo, B., Maryunani, K.A., Tagami, T., Takemura, K., Yoden, S., 2010. Comparison of stable isotope time series of stalagmite and meteorological data from West Java, Indonesia. Palaeogeogr. Palaeoclimatol. Palaeoecol. 293, 90e97. White, W.B., 1988. Geomorphology and Hydrology of Karst Terrains. Oxford University Press, New York. Whittaker, R.H., 1975. Communities and Ecosystems. Macmilan, New York. Wolff, C., Plessen, B., Dudashvilli, A.S., Breitenbach, S.F.M., Cheng, H., Edwards, R.L., Strecker, M.R., 2017. Precipitation evolution of Central Asia during the last 5000 years. Holocene 27, 142e154. ~ ez, I.P., Rasbury, E.T., Wortham, B.E., Wong, C.I., Silva, L.C.R., McGee, D., Montan Cooper, K.M., Sharp, W.D., Glessner, J.J.G., Santos, R.V., 2017. Assessing response of local moisture conditions in central Brazil to variability in regional monsoon intensity using speleothem 87Sr/86Sr values. Earth Planet Sci. Lett. 463, 310e322. Xia, Q.K., Zhao, J.X., Collerson, K.D., 2001. Early-Mid Holocene climatic variations in Tasmania, Australia: multi-proxy records in a stalagmite from Lynds cave. Earth and Planet. Sci. 194, 177e187. Zhang, H., Cai, Y., Tan, L., Cheng, H., Qin, S., An, Z., Edwards, R.L., Ma, L., 2015. Large variations of d13C values in stalagmites from southeastern China during historical times: implications for anthropogenic deforestation. Boreas 44, 511e525. Zhao, K., Wang, Y., Edwards, R.L., Cheng, H., Liu, D., Kong, X., Ning, Y., 2016. Contribution of ENSO variability to the East Asian summer monsoon in the late Holocene. Palaeogeogr. Palaeoclimatol. Palaeoecol. 449, 510e519. Zhao, M., Li, H.C., Liu, Z.-H., Mii, H.S., Sun, H.L., Shen, C.-C., Kang, S.C., 2015. Changes in climate and vegetation of central Guizhou in southwest China since the last glacial reflected by stalagmite records from Yelang cave. J. Asian Earth Sci. 114, 549e561. Zhu, X., Zhang, M., Cheng, H., Wu, X., Edwards, R.L., 2015. Centennial-scale monsoon climate fluctuations from a stalagmite record during the mid-Holocene Epoch in Fulu cave of Huaping, Yunnan, China. Environ. Earth Sci. 74, 929e935.