A 12,000-yr pollen record off Cape Hatteras — Pollen sources and mechanisms of pollen dispersion F. Naughton a,b,c,d, ⁎, L. Keigwin c , D. Peteet e,f , S. Costas g , S. Desprat h , D. Oliveira a,b,h , A. de Vernal d , A. Voelker a,d , F. Abrantes a,b a Portuguese Institute for the Ocean and Atmosphere (IPMA), Av. Brasília 6, 1449-006 Lisboa, Portugal b Center of Marine Sciences (CCMAR), Algarve University, Campus de Gambelas 8005-139 Faro, Portugal c Woods Hole Oceanographic Institute (WHOI), Woods Hole, MA 02543, USA d University of Quebec at Montreal, GEOTOP Center, CP 8888, succ. Centre-Ville, Montréal, Québec H3C 3P8, Canada e NASA/Goddard Institute for Space Studies, 2880 Broadway, NY, NY 10025, USA f Lamont Doherty Earth Observatory, Palisades, NY 10964, USA g CIMA, Algarve University, Campus de Gambelas 8005-139 Faro, Portugal h EPHE, UMR-CNRS 5805 EPOC, Université de Bordeaux, Allée Geoffroy St Hilaire, 33615 Pessac, France abstract article info Article history: Received 17 December 2014 Received in revised form 28 May 2015 Accepted 7 June 2015 Available online 9 June 2015 Keywords: Eastern North America Cape Hatteras Marine pollen signature Vegetation changes Holocene Land–sea pollen transfer Integrating both marine and terrestrial signals from the same sediment core is one of the primary challenges for understanding the role of ocean–atmosphere coupling throughout past climate changes. It is therefore vital to understand how the pollen signal of a given marine record reflects the vegetation changes of the neighboring continent. The comparison between the pollen record of marine core JPC32 (KNR178JPC32) and available terres- trial pollen sequences from eastern North America over the last 12,170 years indicates that the pollen signature off Cape Hatteras gives an integrated image of the regional vegetation encompassing the Pee Dee river, Chesapeake and Delaware hydrographic basins and is reliable in reconstructing the past climate of the adjacent continent. Extremely high quantities of pollen grains included in the marine sediments off Cape Hatteras were transferred from the continent to the sea, at intervals 10,100–8800 cal yr BP, 8300–7500 cal yr BP, 5800– 4300 cal yr BP and 2100–730 cal yr BP, during storm events favored by episodes of rapid sea-level rise in the eastern coast of US. In contrast, pollen grains export was reduced during 12,170–10,150 cal yr BP and 4200– 2200 cal yr BP, during episodes of intense continental dryness and slow sea level rise episodes or lowstands in the eastern coast of US. The near absence of reworked pollen grains in core JPC32 contrasts with the high quantity of reworked material in nearby but deeper located marine sites, suggesting that the JPC32 record was not affected by the Deep Western Boundary Current (DWBC) since the end of the Younger Dryas and should be considered a key site for studying past climate changes in the western North Atlantic. © 2015 Elsevier B.V. All rights reserved. 1. Introduction Directly linking continental and marine records is one of the best approaches for understanding atmosphere–ocean–land connections and their impact/role in global climate variability (e.g. Heusser and Shackleton, 1979; Naughton et al., 2009; Sanchez-Goni et al., 2012). In recent decades several efforts have been undertaken to understand vegetation response to past climate variability detected in the eastern (e.g. Hooghiemstra et al., 1992; Sanchez Goñi et al., 2000, 2012, 2013; Roucoux et al., 2001, 2005; Tzedakis et al., 2004; Desprat et al., 2007, 2009; Naughton et al., 2007a, 2007b, 2009; Margari et al., 2010; Bouimetarhan et al., 2012) and western North Atlantic regions (e.g. Balsam and Heusser, 1976; de Vernal et al., 1993; McCarthy and Gostlin, 2000; Mudie and McCarthy, 2006) by directly linking both terrestrial (pollen) and marine (e.g. planktonic foraminif- era, dinoflagellates, coccolithophores) proxies from the same sediment. However, before assessing land–sea linkages during past climate changes of a given region and time period, it is important to understand the present and past pollen signals of each area (e.g. Heusser, 1983; McCarthy and Mudie, 1998; Naughton et al., 2007a). This ensures that both present and past pollen signals in a given marine record generally reflect similar trends in the vegetational patterns and/or changes of the neighboring landmasses (e.g. Heusser, 1983; Naughton et al., 2007a). Thus the main source area of pollen grains included in marine sediments is recognized and the processes/mechanisms behind the transfer of pollen grains from the continent to the sea are distinguished Marine Geology 367 (2015) 118–129 ⁎ Corresponding author at: Portuguese Institute for the Ocean and Atmosphere (IPMA), Av. Brasília 6, 1449-006 Lisboa, Portugal. E-mail address: filipa.naughton@ipma.pt (F. Naughton). http://dx.doi.org/10.1016/j.margeo.2015.06.003 0025-3227/© 2015 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Marine Geology journal homepage: www.elsevier.com/locate/margeo