Middle Holocene humidity increase in Florida: climate or sea-level?

Middle Holocene humidity increase in Florida: climate or sea-level?

Quaternary Science Reviews 103 (2014) 170e174 Contents lists available at ScienceDirect Quaternary Science Reviews journal homepage: www.elsevier.co...

514KB Sizes 0 Downloads 0 Views

Quaternary Science Reviews 103 (2014) 170e174

Contents lists available at ScienceDirect

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

Short communication

Middle Holocene humidity increase in Florida: climate or sea-level? Timme H. Donders* Palaeoecology, Department of Physical Geography, Utrecht University, Laboratory of Palaeobotany and Palynology, Budapestlaan 4, 3584 CD Utrecht, The Netherlands

a r t i c l e i n f o

a b s t r a c t

Article history: Received 19 June 2014 Received in revised form 10 September 2014 Accepted 12 September 2014 Available online 30 September 2014

Florida climate in highly sensitive to both high and low latitude climate perturbations due to its latitudinal position surrounded by water masses that transport heat northward. A well-studied aspect is that middle Holocene conditions became significantly wetter in Florida, initiating widespread peat accumulation in the Everglades. This environmental change has been attributed to various climate forcings, such as migration of the Intertropical Convergence Zone (ITCZ), increases in tropical storm ~ o Southern Oscillation (ENSO), and intensity, position of the Bermuda High, intensification of the El Nin post glacial sea level rise (SLR). Discerning between these forcings is only possible with quantitative reconstructions from a transect of sites that are affected differentially. Application of a transfer function on a north-to-south gradient of pollen records from Florida lakes here shows that the pattern of increasing precipitation during the middle Holocene cannot be explained by SLR, but that ENSO intensification is an important contributing factor. Seasonal-resolved proxy records with improved age models are urgently needed to further solve these issues. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Florida Pollen Sea level change Holocene Palaeoclimatology ~ o-Southern Oscillation El Nin

1. Introduction The Florida peninsula has an enhanced sensitivity to long-term climatic changes as it is influenced by both high-latitude and tropical climate systems, resulting in a gradient from warmtemperate to tropical vegetation between 25  N and 31  N (Fig. 1). Surrounded by warm surface currents, Florida climate responds sensitively to both Northern Hemisphere cold events (Grimm et al., 2006) and movements of the Intertropical Convergence Zone (ITCZ) (Poore et al., 2003; Van Soelen et al., 2012). As a result, it is a key area to detect changes in low-latitude heat buildup due to slowdowns in the Atlantic meridional overturning circulation (AMOC) (Donders et al., 2011). The northern, central and southern portions of the Florida peninsula are differentially regulated by sea-surface temperatures and circulation patterns in the Atlantic and Gulf of Mexico (Enfield et al., 2001). A further key ~o climatological forcing in the SE United States is the El Nin ~ o events produce significantly Southern Oscillation (ENSO). El Nin increased winter precipitation anomalies (Vega et al., 1998; Enfield et al., 2001; Donders et al., 2013). Since most precipitation falls in summer, this additional EN-forced winter peak is a significant

* Tel.: þ31 30 2532636; fax: þ31 302531145. E-mail address: [email protected] http://dx.doi.org/10.1016/j.quascirev.2014.09.011 0277-3791/© 2014 Elsevier Ltd. All rights reserved.

ecological factor extending the hydroperiod of the Florida wetland vegetation (Donders et al., 2005b). Holocene changes in ENSO intensity e.g. (Moy et al., 2002; Riedinger et al., 2002) are therefore likely to have affected the vegetation, wetlands, and lakes of Florida. The middle Holocene wetland expansion and rise in peat formation in Southern Florida (Donders et al., 2005a; Willard and Bernhardt, 2011) is consistent with a humidity increase and a more pronounced ENSO cycle. However, these observations are not independent of sea level rise, which that reached a maximum during the middle Holocene €rnqvist et al., 2004) (Fig. 2), and a (Toscano and Macintyre, 2003; To maximum northerly position of the ITCZ around this time (Haug et al., 2001; Van Soelen et al., 2012). The impact of ENSO on Florida precipitation is more pronounced toward the south (Enfield et al., 2001; Donders et al., 2013) and, hence, a north-to-south transect of sites should show an increased southward response to the intensification of ENSO during the middle and late Holocene. To disentangle the respective impacts of ENSO, ITCZ migration, and sea-level rise (SLR) a transect of sites with quantitative palaeoprecipitation reconstructions is needed. I apply a recently developed transfer model (Donders et al., 2011) to provide a quantitative reinterpretation of previously published pollen records from Florida and examine a north-tosouth gradient of precipitation in Florida during the Holocene. Updated chronologies of the sites, based on recalibrated radiocarbon ages and new age-depth modelling, provide an improved

T.H. Donders / Quaternary Science Reviews 103 (2014) 170e174


Fig. 1. Position of Florida and the surface and core sampling sites of the pollen data used in this study. For locations of the sea level reconstructions see Toscano and Macintyre €rnqvist et al. (2004). (2003) and To

regional comparison. Seven lake sites from the interior of Florida were selected (Table 1, Fig. 1) since, relative to the coastal wetlands, these are less dependent on groundwater and sea level changes. Postglacial SLR of the broad western Florida shelf sharply decreased the distance of all sites to the sea, and any effect upon precipitation would have affected all sites to a similar degree. Alternatively, a dominantly climatological forcing would likely have been a regionally divergent. For the southernmost transect sites, Lake Tulane (LT) (Grimm et al., 1993, 2006) and Lake Annie (LA) (Watts, 1975), quantitative reconstructions are already available (Donders et al., 2011). The same transfer function is used to obtain estimates from additional localities (Fig. 1). 2. Methods Data selection is based on the available radiocarbon-dated pollen records from lakes (Table 1) that cover the most part of the Holocene. The quality of the chronology varies significantly and most sites required age calibration as original publications reported 14 C ages only. Calibration was performed with Calib 7.0 (Stuiver et al., 2013) using the IntCal13 NH calibration curve (Reimer et al., 2013). Calibrated ages are reported as median probabilities based on the 1-sigma interval. For sites with few radiocarbon ages (Buck

Lake [BL, (Watts and Stuiver, 1980; Watts et al., 1996)], Mud Lake [ML, (Watts, 1969)], Camel Lake [CL, (Watts et al., 1992)], Scott Lake [ScL, (Watts, 1971)] a simple linear interpolation was used for agedepth modelling. Radiocarbon ages from two parallel records from Sheelar Lake (ShL) were combined (Watts and Stuiver, 1980; Watts and Hansen, 1994) through pollen stratigraphic correlation, and a cubic B-spline interpolation was applied using Tilia 1.7 (Grimm, 1991e2011). The LT pollen record was originally based on a noncalibrated 14C chronology (Grimm et al., 1993). A subsequent study of LT (Grimm et al., 2006) provided an updated and highresolution chronology based on AMS radiocarbon dating of new sediment cores. The percentage data and climatic interpretation of new pollen record from LA and LT were published in Donders et al. (2011). The new age model of LA is reported in detail in Quillen et al. (2013). Reconstruction of summer precipitation is based on the transfer function for Florida described in Donders et al. (2011). The transfer function provides very robust results as it is limited to a fixed set of common regionally important taxa, and local aquatic wetland taxa were largely removed to avoid imprint of local vegetation changes and/or basin type. Highly permeable soils in the central Lake Wales Ridge of Florida indicate greatest sensitivity to the main summer precipitation peak (Grimm et al., 2006). In the 122-sample training


T.H. Donders / Quaternary Science Reviews 103 (2014) 170e174

Fig. 2. (Blue circles) Results of the reconstructed Holocene summer precipitation based on pollen assemblage data from lake sediment cores in Florida. Sites ordered left to right €rnqvist et al. (2004) (green represent the north to south gradient as shown in Fig. 1. Regional sea level reconstructions from Toscano and Macintyre (2003) (red circles) and To triangles). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

strength of the signal and the consistency between multiple reconstructions from a single lake has shown that the results are robust (Donders et al., 2011).

dataset, summer and winter precipitation have very little independent explanatory power; and because summer precipitation is ecologically the most significant factor, a PJJA (summer) reconstruction is performed. PJJA explains 9.8% of the surface sample variance. The optimal transfer function is a two-component partialleast-squares (PLS) model with an r2boot of 0.66 and root mean square error of prediction (RMSEP) of 39.7 mm/yr (see Donders et al., 2011 for details on parameter selection and model evaluation). Model performance is somewhat limited by the relatively restricted taxonomic resolution of the surface sample dataset as it is a multi-author compilation (Whitmore et al., 2005). However, the

3. Results and discussion The reconstructions (Fig. 2) clearly show a divergent climatological Holocene development between southern and northern sites in Florida, whereby the southern sites BL and LA show significant and sustained increases in PJJA (increase exceeding the error envelope) from approximately 6e5 ky BP, while SL and LT shows

Table 1 Site information. Lake

Lake type

Surface (ha)

Elevation (m)

Water depth (m)

Lat. (N)

Long. (W)

Land use


Core (m)


Scott Tulane

Sinkhole limestone Sinkhole limestone w sand cover Sinkhole limestone w sand cover Perched aquifer, clayey sand Marsh lake

120 40

51 36

2.7 22.7

27 58' 27 350

81 570 81 300

Cultivated Urban

Mesic broad leaved Scrub oak

3.4 18.5




27 120

81 250


Scrub oak





30 160

85 10


(Watts, 1971) (Grimm et al., 1993; 2006) (Watts, 1975; Quillen et al., 2013) (Watts et al., 1992)

Annie Camel Buck

Forest, reserve

Pine flatwood


Forest, reserve

Pine flatwood, scrub






27 14'

81 19


29 500

81 570

Forest, reserve

Fire adapted woodland

29 180

81 520

Forested river valley, reserve

Sand-pine scrub


Sinkhole limestone w sand cover




Sinkhole limestone w sand cover





(Watts and Stuiver, 1980; Watts and Hansen, 1988) (Watts and Stuiver, 1980; Watts and Hansen, 1994) (Bradley, 1966; Watts, 1969)

T.H. Donders / Quaternary Science Reviews 103 (2014) 170e174

intermediate increases, and the northern sites show a slight decrease (CL) or no long-term change (SL and ML). The gradual SLR based on basal peat and coral dates (Toscano and Macintyre, 2003; €rnqvist et al., 2004) show a gradual Holocene increase. The deTo gree of middle to late Holocene moisture increase is correlated with latitude, becoming clearly stronger toward the south, pointing to a climatic cause rather than a sea level-forced increase. The gradual SLR is regionally consistent, while the humidity increase, here reconstructed as PJJA, is a stepped increase confined to the south. The northernmost site (CL) even shows an opposite development with an early to middle Holocene optimum and a small and gradual decrease in the late Holocene. Previously reported quantitative interpretation of the 60 ka lake Tulane record, supported by model sensitivity experiments, showed that PJJA anomalies are caused by increased heat transport up to and through the Gulf of Mexico that reflect the persistence of the Atlantic Warm Pool (AWP) and increased Loop Current. The advection of warm surface waters in the Gulf occurs during North Atlantic cold spells and is ultimately driven by a steepened equatorpole temperature gradient and increased summer trade winds associated with a northerly summer ITCZ position (Donders et al., 2011). However, severe NH cold spells did not occur during the Holocene, and although the ITCZ was positioned father north during the middle Holocene (Poore et al., 2003; Koutavas and LynchStieglitz, 2004; Van Soelen et al., 2012), it moved south during the late Holocene (Haug et al., 2001). Hence, the continued middle through late Holocene moisture increase reconstructed here seems contradictory. Pulses of enhanced peat expansion during the middle Holocene, and hydroperiod increase during the late Holocene observed in southern Florida swamp-forest ecosystems (Donders et al., 2005a) and freshwater marshes in the Everglades (Willard and Bernhardt, 2011), are consistent with the new quantitative lake reconstructions reported here (Fig. 2). A recent report of late Holocene drying of the Everglades (Shark River Slough) seems to contrast these results (Glaser et al., 2013). However, these authors upscale an essentially very local signal to regional patterns without discussion on whether this is appropriate, and interpret the transition from Amaranthaceae to Pinus as a drying signal. Extensive surface sample data from Florida though relate high Amaranthaceae pollen to seasonal drought (Willard et al., 2001). The dominant Florida slash pine (Pinus elliottii) is adapted to moist soils and humid climates, and at Lake Tulane, high Pinus phases are consistently associated with high lake levels (Grimm et al., 2006). The climatic interpretation of Glaser et al. (2013) is therefore considered inconsistent and questionable. Besides SLR, middle and late Holocene increases in frequency and amplitude of ENSO (Moy et al., 2002) are a possible cause for the observed humidity increase (Donders et al., 2005a). ENSO primarily affects winter precipitation in Florida and, therefore, is not directly relevant to PJJA (Fig. 2). Nevertheless, increased winter precipitation lengthens the hydroperiod, which is indicated by the expansion of long-hydroperiod Taxodium swamp forest during the late Holocene (Donders et al., 2005a), and by a fundamental change in Everglades hydrology (Willard and Bernhardt, 2011). A recent independent reconstruction of limnological conditions at LA, based on diatom assemblages, also indicate a transition from a shallow to a deep lake with stable water levels after 4 ky BP (Quillen et al., 2013). The year-round wet conditions developed together with modern ENSO periodicities have so far prevented a return to drier conditions. A winter precipitation signal would provide direct evidence of this postulated change in hydrological regime. In north-central Florida, rare speleothem data for (Briars Cave, Ocala) point to a relatively minor role of ENSO relative to the NAO in long-term winter precipitation forcing in this region (Van Beynen et al.,


2007), consistent with our nearby reconstructions (sites SL&ML in Fig. 2). On the other hand, in south-central Florida, the tropical and ENSO forcing seem to dominate based on the increase in precipitation seen at LA and on the hydroperiod expansion even further south (Fakahatchee, Fig. 1). 4. Conclusions Application of the present transfer function already provides important insights in Holocene patterns of climate change in Florida based on existing data. The middle Holocene humidity increase is evidently stronger further south. As this pattern is unlikely to be caused by SLR, a climatic forcing is implied. Given the persistently wetter conditions after 3 ky BP, intensification of ENSO best explains the observations. However, for determination of leads and lags, rates of change, and identification of small-scale (submillennial to decadal) changes, strategically collected surface samples with better taxonomical resolution for improved transfer functions as well as high-resolution age models are urgently needed. These results stress the significance of Florida as a palaeoclimate indicator where high-latitude and tropical climate systems have differential effects along a north to south transect. Future changes in these systems can therefore create a locally very divergent impact and require further study. Acknowledgements This work benefitted from discussions with Debra Willard (USGS, Reston) and Evelyn Gaiser (Florida International University, Miami). The author wishes to thank Friederike Wagner-Cremer for providing comments on the manuscript, and Eric Grimm and an anonymous reviewer for their constructive reviews. This work is supported by the Department of Physical Geography at Utrecht University. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.quascirev.2014.09.011. References Bradley, W.H., 1966. Tropical Lakes, Copropel, and Oil Shale. Geol. Soc. Am. Bull. 77, 1333e1337. Donders, T.H., Wagner, F., Dilcher, D.L., Visscher, H., 2005a. Mid- to late-Holocene El ~ o-Southern Oscillation dynamics reflected in the subtropical terrestrial Nin realm. Proc. Natl. Acad. Sci. U. S. A. 102, 10904e10908. Donders, T.H., Wagner, F., Visscher, H., 2005b. Quantification strategies for humaninduced and natural hydrological changes in southern Florida wetland vegetation. Quat. Res. 64, 333e342. Donders, T.H., Punyasena, S.W., de Boer, H.J., Wagner-Cremer, F., 2013. ENSO ~ o record into signature in botanical proxy time series extends terrestrial El Nin the (sub)tropics. Geophys. Res. Lett. 40 http://dx.doi.org/10.1002/ 2013GL058038. Donders, T.H., Boer, H., Finsinger, W., Grimm, E., Dekker, S., Reichart, G., WagnerCremer, F., 2011. Impact of the Atlantic Warm Pool on precipitation and temperature in Florida during North Atlantic cold spells. Clim. Dyn. 36, 109e118. ~ ez, A.M., Trimble, P.J., 2001. The Atlantic multidecadal Enfield, D.B., Mestas-Nun oscillation and its relation to rainfall and river flows in the continental U.S. Geophys. Res. Lett. 28, 2077e2080. Glaser, P.H., Hansen, B.C.S., Donovan, J.J., Givnish, T.J., Stricker, C.A., Volin, J.C., 2013. Holocene dynamics of the Florida Everglades with respect to climate, dustfall, and tropical storms. Proc. Natl. Acad. Sci. 110 (43), 17211e17216. Grimm, E.C., 1991-2011. Tilia. 1.7.16. Grimm, E.C., Jacobson Jr., G.L., Watts, W.A., Hansen, B.C.S., Maasch, K.A., 1993. A 50,000-year record of climate oscillations from Florida and its temporal correlation with the Heinrich events. Science 261, 198e200. Grimm, E.C., Watts, W.A., Jacobson, G.L.J., Hansen, B.C.S., Almquist, H.R., Dieffenbacher-Krall, A.C., 2006. Evidence for warm wet Heinrich events in Florida. Quat. Sci. Rev. 25, 2197e2211.


T.H. Donders / Quaternary Science Reviews 103 (2014) 170e174

Haug, G.H., Hughen, K.A., Sigman, D.M., Peterson, L.C., Rohl, U., 2001. Southward migration of the intertropical convergence zone through the Holocene. Science 293, 1304e1308. Koutavas, A., Lynch-Stieglitz, J., 2004. Variability of the marine ITCZ over the eastern Pacific during the past 30,000 years: regional perspective and global context. In: Bradley, R.S., Diaz, H.F. (Eds.), The Hadley Circulation: Present, Past and Future. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 347e369. ~ o/ Moy, C.M., Seltzer, G.O., Rodbell, G.T., Anderson, D.M., 2002. Variability of El Nin Southern Oscillation activity at millennial timescales during the Holocene epoch. Nature 420, 162e165. Poore, R.Z., Dowsett, H.J., Verardo, S., Quinn, T.M., 2003. Millennial- to century-scale variability in Gulf of Mexico Holocene climate records. Paleoceanography 18, 1048. Quillen, A., Gaiser, E., Grimm, E., 2013. Diatom-based paleolimnological reconstruction of regional climate and local land-use change from a protected sinkhole lake in southern Florida, USA. J. Paleolimnol. 49, 15e30. Reimer, P., Bard, E., Bayliss, A., Beck, J., Blackwell, P., Bronk Ramsey, C., Buck, C., Cheng, H., Edwards, R., Friedrich, M., Grootes, P., Guilderson, T., Haflidason, H., , C., Heaton, T., Hoffmann, D., Hogg, A., Hughen, K., Kaiser, K., Hajdas, I., Hatte Kromer, B., Manning, S., Niu, M., Reimer, R., Richards, D., Scott, E., Southon, J., Staff, R., Turney, C., van der Plicht, J., 2013. IntCal13 and Marine13 radiocarbon age calibration curves 0e50,000 Years cal BP. Radiocarbon 55. Riedinger, M.A., Steinitz-Kannan, M., Last, W.M., Brenner, M., 2002. A ~6100 14C yr ~ o activity from the Gal record of El Nin apagos Islands. J. Paleolimnol. 27, 1e7. Stuiver, M., Reimer, P.J., Reimer, R., 2013. Calib 7.0 WWW Program and Documentation for Radiocarbon Calibration. http://calib.qub.ac.uk/calib/. €rnqvist, T.E., Gonza lez, J.L., Newsom, L.A., Van der Borg, K., De Jong, A.F.M., To Kurnik, C.W., 2004. Deciphering Holocene sea-level history on the U.S. Gulf Coast: a high-resolution record from the Mississippi Delta. Geol. Soc. Am. Bull. 116, 1026e1039. Toscano, M.A., Macintyre, I.G., 2003. Corrected western Atlantic sea-level curve for the last 11,000 years based on calibrated 14C dates from Acropora palmata framework and intertidal mangrove peat. Coral Reefs 22, 257e270. Van Beynen, P.E., Asmerom, Y., Polyak, V., Soto, L., Polk, J.S., 2007. Variable intensity of teleconnections during the late Holocene in subtropical North America from an isotopic study of speleothem from Florida. Geophys. Res. Lett. 34, L18703.

, J., Reichart, G., 2012. Mid- to Van Soelen, E., Brooks, G., Larson, R., Sinninghe Damste late-Holocene coastal environmental changes in southwest Florida, USA. Holocene 22, 929e938. Vega, A.J., Rohli, R.V., Henderson, K.G., 1998. The Gulf of Mexico mid-tropospheric ~ o an La Nin ~ a forcing. Clim. Res. 10, 115e125. response to El Nin Watts, W.A., 1975. A late Quaternary record of vegetation from Lake Annie, southcentral Florida. Geology 3, 344e346. Watts, W.A., 1971. Postglacial and Interglacial vegetation history of Southern Georgia and Central Florida. Ecology 52, 676e689. Watts, W.A., 1969. A pollen diagram from Mud Lake, Marion County, North-Central Florida. Geol. Soc. Am. Bull. 80, 631e642. Watts, W.A., Hansen, B.C.S., 1994. Pre-Holocene and Holocene pollen records of vegetation history from the Florida peninsula and their climatic implications. Palaeogeogr. Palaeoclimatol. Palaeoecol. 109, 163e176. Watts, W.A., Hansen, B.C.S., 1988. Environments of Florida in the Late Wisconsin and Holocene. In: Purdy, B. (Ed.), Wet Site Archaeology. Telford Press, New Jersey, pp. 307e323. Watts, W.A., Hansen, B.C.S., Grimm, E.C., 1992. Camel Lake: a 40 000-yr record of vegetational and Forest history from Northwest Florida. Ecology 73, 1056e1066. Watts, W.A., Stuiver, M., 1980. Late Wisconsin climate of Northern Florida and the origin of Species-Rich Deciduous Forest. Science 210, 325e327. Watts, W.A., Grimm, E.C., Hussey, T.C., 1996. Mid-Holocene Forest history of Florida and the Coastal Plain of Georgia and South Carolina. In: Sassaman, K.E., Anderson, D.G. (Eds.), Archaeology of the Mid-Holocene Southeast. University Press of Florida, Gainesville, USA. Whitmore, J., Gajewski, K., Sawada, M., Williams, J.W., Shuman, B., Bartlein, P.J., Minckley, T., Viau, A.E., Webb, T.I., Shafer, S., Anderson, P., Brubaker, L., 2005. Modern pollen data from North America and Greenland for multi-scale palaeoenvironmental applications. Quat. Sci. Rev. 24, 1828e1848. Willard, D.A., Bernhardt, C.E., 2011. Impacts of past climate and sea level change on Everglades wetlands: placing a century of anthropogenic change into a lateHolocene context. Clim. Change 107, 59e80. Willard, D.A., Weimer, L.M., Riegel, W.L., 2001. Pollen assemblages as paleoenvironmental proxies in the Florida Everglades. Rev. Palaeobot. Palynol. 113, 213e235.