Seasonal climate change across the Roman Warm Period/Vandal Minimum transition using isotope sclerochronology in archaeological shells and otoliths, southwest Florida, USA

Abstract

Archaeological evidence suggests that southwest Florida experienced variably warmer and wetter climate during the Roman Warm Period (RWP; 300 BC–550 AD) relative to the Vandal Minimum (VM; 550–800 AD). This hypothesis was tested by reconstructing seasonal-scale climate conditions for the latter part of the RWP (1–550 AD) by using high-resolution oxygen isotope ratios (δ¹⁸O) of archaeological shells (Mercenaria campechiensis) and otoliths (Ariopsis felis). Eight shells radiocarbon-dated to 150–550 AD recorded that the RWP summers at 150–250 AD were insignificantly different from today and the RWP summers at 250–350 AD and 450–550 AD were drier relative to today. Eight otoliths indicate that the winters were variable during the RWP, colder than today at 150–200 AD and 250–300 AD, similar to today at 200–250 AD, 300–350 AD and 450–500 AD, and warmer than today at 500–550 AD. The climate reconstructions agree with archaeological observations and are partially coherent with the history of sea-level change, with a drying and cooling trend at the 95% confidence level across the RWP/VM transition. The climate transition is not only consistent with falling sea level, but also coherent with reduced solar radiation. Reduced solar radiation may have triggered a change in atmospheric circulation patterns that precipitated the observed climate transition.

Introduction

The pre-European Calusa people in southwest Florida left behind abundant shell middens/mounds, artifacts, and other cultural remains (Marquardt, 2004). Archaeological evidence from these deposits suggests that the Calusa were impacted by abrupt climate change and sea-level fluctuation during two climate episodes in the first millennium: the Roman Warm Period (RWP; 300 BC–550 AD) and the Vandal Minimum (VM; 550–800 AD) (Marquardt and Walker, 2001, Marquardt and Walker, 2012). Both settlement (e.g., high-elevation versus low-elevation shoreline locations) and subsistence patterns (e.g., fish, shellfish, migratory-duck availability) were conditioned by variably warm/cool temperatures, high/low water levels, wet/dry precipitation levels, as well as storm intensity/frequency. An earlier study reconstructed seasonal climate change during the VM using oxygen isotope proxy data (δ¹⁸O) preserved in archaeological Mercenaria campechiensis shells and Ariopsis felis otoliths from discrete chronostratigraphic layers within these Calusa middens and mounds (Wang et al., 2011). Following the earlier VM study, the primary intent of this study is to reconstruct climate change in the RWP with oxygen isotope ratios of archaeological shell–otolith pairs and provide isotope evidence to test the archaeological findings across the RWP and VM climate transition. Archaeological shell–otolith pairs are useful study proxies for paleoclimate reconstructions and human–climate relationships. The shell–otolith pairs provide high-resolution climate information on changes in summer wet season conditions and winter sea surface temperature (SST) (Walker and Surge, 2006). Because archaeological shells and otoliths were deposited during the RWP by the Calusa of southwest Florida, the climate proxies also reflect aspects of human behavior in the late Holocene.

Oxygen isotope ratios of mollusc shells and fish otoliths have been widely used in high-resolution temperature or precipitation reconstructions and provide climate information on seasonal timescales (Jones et al., 1989, 1990; Jones and Quitmyer, 1996; Ivany et al., 2000; Wurster and Patterson, 2001; Surge and Walker, 2005, 2006; Wang et al., 2011, 2012). Moreover, there is increasing paleoclimate evidence that tropical/subtropical climate is more variable and dynamic than previously thought (Winter et al., 2000; Haug et al., 2001; Hodell et al., 2001; Black et al., 2007; Richey et al., 2009). The model simulation conducted by Barnett et al. (1992) identified the climate variability within the latitudes of 0°–30° as the primary contributor to global climate variability at multidecadal to centennial timescales. Adjacent to the study area, the Gulf of Mexico experienced a larger magnitude of cooling than the mean magnitude of northern hemisphere cooling during the Little Ice Age (Richey et al., 2009). Therefore, subtropical southwest Florida (26–27 °N) is sensitive to climate change as are other low-latitude regions, and is appropriate for studying multidecadal to centennial climate oscillations, such as the VM and RWP.

This study presents a multi-proxy record of δ¹⁸O values that records the variability of summer precipitation and winter temperature during the RWP in southwest Florida. It further compares the RWP climate reconstruction with archaeological evidence to test whether they are in agreement. Additionally, the climate records of RWP and VM are integrated with the history of sea level and solar activity to gain insights into the climate mechanism(s) driving late Holocene climate change in southwest Florida.