Research Article: 3500 years of shellfish mariculture on the Northwest Coast of North America

Date Published: February 27, 2019

Publisher: Public Library of Science

Author(s): Nicole F. Smith, Dana Lepofsky, Ginevra Toniello, Keith Holmes, Louie Wilson, Christina M. Neudorf, Christine Roberts, Michael D. Petraglia.


Ancient systems of mariculture were foundations of social-ecological systems of many coastal Indigenous Peoples. However, since such systems either do not leave tangible remains in the archaeological record, and/or are hard to date, we know little about their development and use. Clam gardens, traditional mariculture features located within the intertidal zone along the Northwest Coast of North America, are composed of a rock wall positioned at the low tide mark and a flattened terrace on the landward side of the wall. Because these features are largely composed of rock and sediment, and have complex formation histories, they can be difficult to age. On northern Quadra Island, British Columbia, we identify three variations in clam garden form, constructed in different geomorphological settings, each of which require different sampling approaches to obtain ages on construction and ongoing use. To age the clam gardens, we consider radiocarbon dating of invertebrates that inhabit beach deposits (both pre- and post-garden construction), and the relationship of the gardens and clam samples to the local sea level history and taphonomic processes. Within our study area, we find clam gardens have been in use for 3500 years, likely corresponding to other social and ecological changes of the time. These data allow us to formulate guidelines on samples most suitable to constrain the age of initial and on-going wall construction and use of clam gardens, which can be extrapolated to dating other ancient mariculture features in other regions. Such dating programs are the foundation for understanding the long-term development of traditional marine management practices and how they are situated in broader social-ecological systems.

Partial Text

Indigenous and local peoples around the world have developed a suite of management techniques to maintain and enhance culturally important resources [1–3]. However, documenting such techniques in the archaeological and paleoecological records, and thus the larger social-ecological contexts in which they are embedded, can be elusive [4, 5]. Clam gardens, an ancient mariculture technique used by Indigenous People of the Northwest Coast of North America, provide a unique opportunity to study the tangible aspects of ancient marine management systems. These rock-walled intertidal terraces, in combination with a variety of cultivation techniques, enhanced clam productivity and abundance through a variety of mechanisms [6–9]. In the past, as today, these features were linked to the governance, livelihoods, and identity of coastal First Nations from Alaska to Washington State [6, 8, 10, 11]. The persistence of these rock features in the archaeological record of the region allows us to track the development of this form of management through time and space.

Quadra Island is located off the northeast coast of Vancouver Island in southwest British Columbia, Canada in the traditional territories of the Laich-Kwil-Tach and northern Coast Salish peoples. Our study area at the north end of Quadra Island (Fig 1) focuses on Kanish and Waiatt Bays, two large water bodies separated by a narrow 750 m stretch of land traversed easily by foot. Recent refinement of the relative sea level (RSL) history [13, 14] indicates that sea level at the north end of Quadra Island fell rapidly following the last glacial maximum from a height of over 195 m to within approximately 2 to 4 m of modern levels by 12,300 years ago [13]. This drop was followed by a minor transgression before RSL fell gradually from about 3 m above modern around 10,500 years ago, down to about 1 m above modern by 4000 years. By 2000 years ago, sea level was within 35 cm of modern day levels and then gradually declined to current levels [13, 14]. Tracking the progression of RSL change is vital for locating shoreline proximal sites and for understanding clam garden development through time.

Of the nine clam gardens sampled and dated in Kanish and Waiatt Bays, seven sites represent Form 1 clam gardens on soft sediment beaches (EbSh-5, EbSh-13, EbSh-58, EbSh-36, KB14-05, EbSh-77, EbSh-23), while two represent Form 3 gardens on previously steep and eroding shorelines (WB02, WB08). Due to unclear stratigraphy and high water tables we were not able to collect suitable dating samples from the Form 2 gardens. Two Form 1 study sites contained multiple walls and terraces (EbSh-5, EbSh-58 Fig 3D and 3E). In total, we analyzed sixty-one radiocarbon samples from the nine sites (S1 Table, S2 Table). We divided these in two groups: samples which are unsuitable for determining when clam gardens were built and/or used, and those which are suitable. All samples have allowed us to establish sampling guidelines for dating clam gardens (S1 File) in areas of falling sea levels. We discuss each group of unsuitable and suitable samples below followed by a discussion of how the relative sea level history and tidal heights of the samples and walls can also be used to ascertain clam garden ages. All radiocarbon dates discussed in the text are the median calibrated (cal BP) dates.

Our dating program allows us to generalize about the best contexts to age clam gardens under a variety of geomorphological and ecological settings (see S1 File for sampling guidelines) and demonstrates that meaningful age determinations can be established for clam garden terraces, despite being primarily inorganic constructions. Clam gardens are complex living landscapes subject to continuous abiotic, biotic, and cultural processes. As such, there is value in exploring multiple sampling contexts to understand the chronology of clam garden construction and use, and in considering the role of local sea level histories in formulating those chronologies. Our proposed methodology works in places where sea level is dropping, such as the Quadra Island study area, but elsewhere the details of sampling may need refinement to suit the local sea-level history.




Leave a Reply

Your email address will not be published.