Shell Middens Research Articles

From Shell Midden to Midden-Mound: The Geoarchaeology of Mound Key, an Anthropogenic Island in Southwest Florida, USA.

Mound Key was once the capital of the Calusa Kingdom, a large Pre-Hispanic polity that controlled much of southern Florida. Mound Key, like other archaeological sites along the southwest Gulf Coast, is a large expanse of shell and other anthropogenic sediments. The challenges that these sites pose are largely due to the size and areal extent of the deposits, some of which begin up to a meter below and exceed nine meters above modern sea levels. Additionally, the complex depositional sequences at these sites present difficulties in determining their chronology. Here, we examine the development of Mound Key as an anthropogenic island through systematic coring of the deposits, excavations, and intensive radiocarbon dating. The resulting data, which include the reversals of radiocarbon dates from cores and dates from mound-top features, lend insight into the temporality of site formation. We use these insights to discuss the nature and scale of human activities that worked to form this large island in the context of its dynamic, environmental setting. We present the case that deposits within Mound Key’s central area accumulated through complex processes that represent a diversity of human action including midden accumulation and the redeposition of older sediments as mound fill.

Distinguishing offshore bird hunting from beach scavenging in archaeological contexts: The value of modern beach surveys

Determining whether seabirds recovered from coastal shell middens were obtained via active hunting or scavenging of beached carcasses is a challenge for archaeologists. Traditional methods have included analyzing skeletal part frequencies, abundance, age profiles, and contextual evidence. The assumption has been made, based on limited biological data, that an assemblage of carcasses scavenged from the beach will have more wing elements, and fewer legs and heads. Few studies, however, have embraced modern beaching data to verify this assumption and assess the potential faunal resources available for scavenging. We analyze the skeletal part representation of modern beached birds observed by the Coastal Observation and Seabird Survey Team (COASST), comparing the COASST dataset to two idealized hypotheses used by archaeologists: the human scavenging hypothesis (wings only are recovered, while heads and legs are absent) and the human hunting hypothesis (all body parts are found in equal proportions). Finally, we apply these results to analysis of the bird remains from the Minard site (45-GH-15), a late Holocene coastal site in Grays Harbor, Washington. We find that contemporary beached bird data are closer to replicating the human hunting hypothesis as compared to the human scavenging hypothesis, as >75% of the 19,599 carcasses in the COASST dataset had a combination of head, wings and legs. This result, and the similarity in taxonomic distribution between our contemporary beached bird data and Minard assemblage, suggests that indigenous peoples may have used scavenging as a viable means of resource acquisition in the past. Use of contemporaneous beached bird data may provide zooarchaeology with a statistically defensible baseline of information on the phenology, abundance and condition of bird carcasses.

Paleo environment in the Noguchi shell midden during the Holocene transgression maximum, Aomori Prefecture, northern Japan

Paleo environment in the Noguchi shell midden during the Holocene transgression maximum, Aomori Prefecture, northern Japan

Reconciling oxygen isotope sclerochronology with interpretations of millennia of seasonal shellfish collection on the Pacific Northwest Coast

The results of high-resolution stable oxygen isotope (δ18O) sclerochronology of 139 butter clam (Saxidomus gigantea) shells from nine archaeological shell midden sites on the central coast of British Columbia, Canada demonstrate clear patterns of multi-season harvest at most sites and specific seasonal harvest at particular site locations. Although the results are based on small numbers of shells relative to an unknown number of seasonal clam harvesting events over the centuries or millennia of site occupation, they can be interpreted with confidence based on regional inter-site consistency, fit with a broader range of ancillary archaeological evidence, and the relative stability of seasonal patterns through iterative increases in sample size. On the basis of these results we outline a pragmatic approach to the selection of shells for seasonality analysis and for the interpretation of seasonal harvest patterns. This approach depends neither on the indeterminate reliability of probabilistic sampling of midden deposits and constituents or the prohibitively costly and still potentially unreliable use of large shell samples drawn from deposits distributed across sites.

Looking for the traces of the last hunter–gatherers: Geophysical survey in the Mesolithic shell middens of the Sado valley (southern Portugal)

Abstract This paper presents the results of an application of geophysical surveying methods to Mesolithic sites in Portugal. Magnetic gradiometry maps have been produced at the sites of Vale de Romeiras and Pocas de Sao Bento. At the latter, three excavation seasons have provided information to test the results of the survey. It is concluded that these techniques can be considered efficient tools for the delimitation of the sites and for the identification of Mesolithic and Neolithic structures.

14C and Marine Reservoir Effect in Archaeological Samples from the Northeast Gulf of California

Previous studies of live-collected pre-weapons testing mollusk shells in the northern Gulf of California have demonstrated that the local radiocarbon reservoir effect (ΔR) is large and highly variable. To test the validity of this observation for paired charcoal and shell samples from archaeological contexts, we dated samples from four shell midden locations and six midden layers from the eastern shoreline of the Gulf of California near Puerto Penasco, Sonora, Mexico. Dated charcoal samples were small twigs or brushy plants used to cook shellfish and the shells dated within each midden showed signs of burning. Ages range from approximately 5700 to 1900 cal BP. The offset between the modeled marine 14 C age for calibrated ages of the middens and measured shell 14 C age (ΔR) averaged 425 yr with a standard deviation of 115 yr. DOI: 10.2458/azu_rc.57.18319

Identifying Specialized 19th Century Fishing Camps on California's Northern Channel Islands: Applying AMS Radiocarbon Dating to Historical Sites

California’s Northern Channel Islands have long been an epicenter of specialized fishing economies dating from 13,000 yr ago to the mid-19th century. With thousands of well-preserved shell middens, some dominated by single shellfish species and little to no material culture, it can be difficult to distinguish between specialized prehistoric and historical deposits at some site types. Beginning at least in the Early Holocene and continuing into the Historic period, California mussels, turban snails, and abalone were targeted for specialized collection and processing by Native Americans and later Chinese and Euro-American fishers. Here, we demonstrate how selective accelerator mass spectrometry (AMS) radiocarbon dating of specialized abalone processing camps can help distinguish between prehistoric and historical sites. While unconventional, our case study demonstrates the utility of 14 C dating at sites less than 300 yr old. DOI: 10.2458/azu_rc.57.18397

Could occupation duration be related to the diversity of faunal remains in Mesolithic shell middens along the European Atlantic seaboard?

In order to define the Mesolithic populations that came to settle on the European Atlantic seaboard, it is essential to assess the duration and continuity of human occupation. Although our knowledge of the geographical distribution of these populations is largely truncated by archaeological bias, the presence of more than 240 Mesolithic shell middens along the European Atlantic coast enables us to establish an overview of the published data on the dwelling patterns (seasonal, permanent, recurrent or limited in time) of these groups of maritime hunter-gatherers. A review of the literature shows that information on human occupations is lacking for many known Mesolithic shell middens. Moreover, the data leading to the identification of settlement patterns prove to be very heterogeneous from one site to another. Marine molluscs play a special role among the artefacts used to define the attraction of human populations for the coast, and help us to understand settlement. Although molluscs represent a major component of Mesolithic shell middens, it has taken several decades for the analysis of these faunal remains to be considered from the outset in excavations. One of the questions addressed here is whether the diversity of faunal remains in these shell middens can be related to the duration of human occupations on the coast. This question is based on the idea that people staying in the same settlement for several seasons can explore the surrounding area in greater detail. In this paper we observe that the diversity of faunal remains, including seashells, is not only dependent on the way of life of coastal people and settlement duration. It also affects other factors, such as the accessibility of more or less diversified environments and their resources in the vicinity of settlements and the degree of adaptability of humans exploiting similar resources.

Importance of small-scale paleo-oceanographic conditions to interpret changes in size of California mussel (Mytilus californianus). Late Holocene, Santa Cruz island, California

Abstract Several authors have highlighted the difficulties in disentangling human from natural effects when studying shellfish foraging strategies and archeo-malacogical records. In coastal settings, one of the reasons for this difficulty is the lack of understanding of small-scale oceanographic conditions and their influence on coastal and marine species used by people in the past. The present study evaluated the influence of environmental conditions on shellfish harvesting during the Late Holocene around Santa Cruz Island (Southern California) considering small-scale nearshore oceanographic variability around the Island and its effect on M. californianus (main shellfish species exploited through prehistory along the coast of California). Fluctuations in size and abundance of M. californianus shells through time and local past sea surface temperature (SST) values from oxygen isotope analysis of mussel shells are correlated throughout the stratigraphy of two shell midden sites (2200–500 cal B.P.) nearby coasts with different paleo-oceanographic characteristics. Despite the fact that human harvesting seems to have been the main force shaping length and availability of M. californianus , no clear evidence of human pressure was found on archaeological mussel assemblages. Results also show that fluctuations in regional SST records (Santa Barbara Basin marine core) do not represent local SST variations (oxygen isotope analyses on mussel shells), especially if the archaeological site is located on a coast exposed to strong and consistent upwelling activity. By re-evaluating human impact on mussel beds during the Late Holocene around Santa Cruz Island, the research presented here demonstrates the importance of considering small-scale nearshore oceanography for understanding the archaeological evidence of prehistoric human foraging on important intertidal shellfish species.

A Revised Radiocarbon Chronology of the Aceramic Shell Midden of Ra’s Al-Hamra 6 (Muscat, Sultanate of Oman): Implication For Occupational Sequence, Marine Reservoir Age, and Human Mobility

AbstractRa’s al-Hamra 6 (RH-6) is one of the earliest stratified archaeological sites along the eastern littoral of the Arabian Peninsula. This shell midden was radiocarbon dated to the 6th–5th millennium cal BC, but the majority of the dates were obtained before the advent of accelerator mass spectrometry (AMS) 14C dating and suffer from large uncertainties. In addition, most of these dates were obtained on marine and mangrove shells and required correction for local variations from the global average marine 14C reservoir age (MRA). This proved difficult because no consensus value exists for this period in the area. Recent excavations at RH-6 offered the opportunity to redate this important site in order to precisely determine its occupation history and later use as a graveyard, and establish the marine reservoir effect for this time period. Thirty-eight samples of charcoal, shells, and human bone apatite were selected for 14C dating. Bayesian modeling of the 14C dates suggests that the formation of the shell midden spanned ~1 millennium, between the mid-6th and the mid-5th millennium cal BC. Positive and consistent ΔR values were calculated throughout the entire sequence, ranging from 99±27 to 207±43 14C yr. At the beginning of the 4th millennium cal BC, RH-6 was used as a graveyard, as suggested by the 14C dating of a shell in strict association with an individual buried at the surface of the site. 14C dating of human bone apatite allowed us to calculate that 89% of this individual’s diet derived from marine resources. This finding confirms previous observations showing the overwhelming presence of marine and mangrove-dwelling species in the faunal and charcoal assemblage, and implies a low mobility, or mobility restricted to the coast for this population during the 4th millennium cal BC.

Prey selection, size, and breakage differences in Turbo undulatus opercula found within Pacific Gull (Larus pacificus) middens compared to Aboriginal middens and natural beach deposits, southeast Australia

Qualitative discrimination criteria are employed commonly to distinguish cultural shell middens from natural shell deposits. Quantitative discrimination criteria remain less developed beyond an assumption that natural shell beds tend to contain a wider range of shell sizes compared to cultural shell middens. This study further tests this assumption and provides the first comparative quantitative analysis of shell sizes from cultural middens, bird middens, and beach shell beds. Size distributions of opercula of the marine gastropod Turbo undulatus within two modern Pacific Gull (Larus pacificus) middens are compared with two Aboriginal middens (early and late Holocene) and two modern beach deposits from southeast Australia. Results reveal statistically significant differences between bird middens and other types of shell deposits, and that opercula size distributions are useful to distinguish Aboriginal middens from bird middens but not from beach deposits. Supplementary qualitative analysis of taphonomic alteration of opercula reveal similar opercula breakage patterns in human and bird middens, and further support previously recognised criteria to distinguished beach deposits (water rolling and bioerosion) and human middens (burning). Although Pacific Gulls are geographically restricted to southern Australia, the known capacity of gulls (Larus spp.) in other coastal contexts around the world to accumulate shell deposits indicates the broader methodological relevance of our study.

Billfish foraging along the northern coast of Chile during the Middle Holocene (7400–5900 cal BP)

Early maritime adaptation is well known from eastern South America; however, evidence for navigation and large marine fish exploitation, indicative of skilled foraging techniques, remains scarce. In the Atacama Desert, north central Chile, coastal archaeological sites show that during the Middle Holocene people relied mostly on marine resources. However, evidence from one of the shell middens (Zapatero – 7400–5900cal BP) indicates that not only were shallow waters and marine mammals exploited, but that a dedicated fishery for large pelagic fish (mainly billfish) existed; with indications suggesting that large swordfish, weighing up to 300kg, were being caught and brought back complete to the settlement. Although the peculiar topographic and oceanographic features of the area probably allowed pelagic fish to come well inshore, this is still the oldest evidence worldwide for such a fishery. These results provide new insights into maritime adaptation and seafaring along the Chilean coast from as early as the sixth millennium BCE.

El Collado shell midden and the exploitation patterns of littoral resources during the Mesolithic in the Eastern Iberian Peninsula

The 1980s excavations at the El Collado, a large open-air Mesolithic site on the Eastern coast of Spain, revealed a sequence of human occupations consisting of a large shell midden and 14 human burials dated to the Mesolithic period. Human palaeodietary reconstructions based on bone collagen δ13C and δ15N isotope ratios, identified a variable contribution of marine proteins, ranging from fully terrestrial diets to a maximum input of 25%. Most subsequent research on Mediterranean coastal and dietary adaptations refers to the site's funerary record and palaeodietary study, but the composition of its shell midden has remained unstudied.This work reports the first systematic study of a representative sample of mollusc and a comparatively small assemblage of fish bone recovered from the different stratigraphic horizons of El Collado site. Results indicate a mixed marine–terrestrial mollusc composition of the shell midden. The edible land snail Sphincterochila candidissima and the marine bivalve Cerastoderma glaucum are the best represented species throughout the archaeological sequence. The pattern of intertidal resource exploitation is clearly dominated by bivalves (C. glaucum, Glycimeris violacescens, and Ruditapes decussatus) and gastropods (Cerithium vulgatum, Hexaplex trunculus) inhabiting mud and sand flats in coastal lagoon environments. The presence of rocky shore intertidal species is minimal, mostly related to the manufacture of pierced shell ornaments made from Columbella rustica. On the other hand, fish bone assemblages are overwhelmingly dominated by the gilthead sea bream (Sparus aurata) a coastal species frequent in brackish water coastal lagoons and estuaries, both on rocky and sandy grounds. The biometric analysis of C. glaucum records differences on shell size amongst the different layers, suggesting variations of marine productivity throughout the archaeological sequence.

The timing of postglacial coastal adaptations in Eastern Iberia: A Bayesian chronological model for the El Collado shell midden (Oliva, Valencia, Spain)

El Collado is an open air site containing evidence of 14 burials and a shell midden archaeological deposit with different phases of human occupation dated to the Early Holocene. Previous studies have produced 14 radiocarbon dates using bone collagen samples from human burials. However, no attempt has been made to date the stratigraphic sequence to address the formation time of the shell midden and establish a chrono-stratigraphic framework for the study of bio-archaeological and cultural assemblages. We critically evaluate the available radiocarbon record of the site and present 6 new AMS radiocarbon dates on ungulate bones with anthropic fractures and bivalves from the three major stratigraphic horizons. Then, we integrate our results with previously published AMS radiocarbon determinations on human samples into a chronological Bayesian model, constraining the radiocarbon distributions with prior information about the samples stratigraphic provenance. Finally, we correlate the results of the chronological model with the regional archaeological sequence and the available data on coastal evolution at regional and local scales. The vertical distribution of the radiocarbon dates makes evident some stratigraphic disturbance caused by the repeated presence of pit graves, agricultural activities and low stratigraphic control of the excavation process. However, once the problematic determinations are identified as outliers, the remaining radiocarbon dates grouped according the major stratigraphic divisions, and the calibrated distributions modelled-constrained, the resulting Bayesian phase model reveal high agreement index (Ac = 103). The shell midden formation took place from the bottom of the sequence, spanning 1022–1965 calibrated years (CI 95.4%). The occupations documented at Phase 1 (Level IV) are dated to the Early Holocene (9828–9551 cal BP). The Phase 2 (Level II), encompass radiocarbon evidence of both burial and occupational activities dated to 9437 to 8477 cal BP, spanning 779–985 calibrated years (CI 95.4%). Finally, the Phase 3 (Level I) records Late Mesolithic occupations deposited between 8509 and 8391 cal BP for the start boundary and 8499–8060 cal BP for the end boundary (CI:95.4%). There are no individual radiocarbon dates during or postdating the chronological span of the 8.2 ka climatic event (8300–8140 cal BP).

Shellfish exploitation in Stone Age Arctic Norway: procurement patterns and household activities

ABSTRACTArchaeological evidence for coastal shellfish exploitation in Arctic Norway mostly dates from 5300 to 1500 BC, encompassing the terminal Mesolithic, the Late Stone Age, and the beginning of the Early Metal Period. Shell middens are generally associated with semi-subterranean houses, but may also occur as free-standing deposits. The primary shellfish taxa exploited were Littorina littorea, Arctica islandica and Mytilus edulis. Most of the species utilized were available in the littoral zone and were easily procured in the course of daily household activities, but sub-littoral A. islandica would have required greater investment in procurement technology and time, as well as labor cooperation. Shellfish were an occasional dietary supplement and could also serve as fishing bait. Spatial and temporal variation in shellfish assemblages reflects local marine ecology, sea-level changes, household dynamics, and perhaps field processing. Interpretation of the existing data is constrained by inadequate midden sampling and documentation methods.

Contemporary Indigenous Relationships to Archaeological Features: Agency, Affect, and the Social Significance of Rock Art

One of the major challenges facing archaeologists and cultural heritage/resource managers is how to better approach and understand concepts of significance and value of archaeological features (e.g. stone tools, rock art sites, pottery, stone arrangements, shell middens) in contemporary Indigenous settings. In this article, I focus on exploring contemporary Indigenous engagement and interaction with one specific type of archaeological feature—rock art—to develop a better understanding of how significance and value are attributed to sites and motifs. Through the lens of affectual, relational, and cultural understandings of rock art in northern Australia and the American Southwest I examine the complex nature of encounters and responses (e.g. verbal and gestural) to sites and motifs to illustrate how concerns with significance go beyond the archaeological realm. Using Gell's (1992 The Technology of Enchantment and the Enchantment of Technology. In Anthropology, Art, and Aesthetics, edited by Jeremy Coote and Anthony Shelton, pp. 40–67. Clarendon Press, Oxford) ideas of agency and affect as a starting point, and then unpacking the responses that are elicited from visiting or viewing rock art, I focus on investigating the structure of the relationship between people and the contemporary social significance of archaeological features. A key outcome from this research is the realization that contemporary engagement with rock art is diverse and can be shaped by a variety of factors including ontological and epistemological understandings, emotional experiences (e.g. fear), social memory, and the health and well-being of individuals and communities.

Three Paleocoastal Lithic Sites on Santa Cruz Island, California

Paleocoastal sites, dated between ∼13,000 and 8000 cal yr BP, are relatively abundant on California's Northern Channel Islands but rare on the largest island of Santa Cruz. Here we describe three lithic sites containing chipped-stone crescents and, in one case, a stemmed Channel Island Barbed point fragment. Elsewhere on the islands, datable sites containing crescents are either terminal Pleistocene or early Holocene in age. Two of the lithic sites produced no datable organic remains, but one contains a small patch of intact shell midden dated to ∼8650 cal yr BP. These three sites add to the growing number of Paleocoastal sites on Santa Cruz and the other Northern Channel Islands.

An Unusual Late Aboriginal Assemblage from the Wilson Site (41SS186), San Saba County, Central Texas

The late aboriginal component in the Wilson Site in San Saba County is unusual in that most of the assemblage is consistent with that of Classic Toyah, but the diagnostic projectile point is an unnotched triangular arrow point instead of the typical Perdiz point. The absence of Perdiz points suggests that this component is associated with non-Toyah people and possibly dates to after 1700. Archaeological testing by the Llano Uplift Archeological Society (LUAS) to find supporting evidence for a historic date identified an Austin phase shell midden and a “Late Component” composed of triangular arrow points, end scrapers, a beveled biface and bone-tempered sherds, but no items of European manufacture. Complicating matters, the luminescence dating on a ceramic sample opens the possibility that the Late Component predates the currently accepted beginning of the Toyah phase.

Vegetation history and use of plant resources from the incipient to early Jomon periods at the Torihama shell midden site, Fukui Prefecture, central Japan

Vegetation history and use of plant resources from the incipient to early Jomon periods at the Torihama shell midden site, Fukui Prefecture, central Japan

Análisis morfométrico en Mytilus edulis : una herramienta para el análisis de los sitios concheros y la arqueología de la Costa Patagónica

Fragmentation is one of the most important taphonomic processes in archaeomalacological assemblages. The quantity of fragmented malacological material introduces biases that affect the interpretation of formation processes, shell midden preservation and human gathering strategies. In this paper we analyze the process of Mytilus edulis fragmentation at the PE 133 site, thereby presenting a methodology for studying archaeomalacological assemblages. The results show that no relationship could be established between shell size and the degree of fragmentation. This demonstrates that fragmentation depends on the mollusk characteristics. Therefore, shell thickness, shape and chemical composition, are more important that the size of the individual, the type of sediment, or site formation history. Finally, fragmentation introduces bias into shell length analysis, so that morphometric equations are a useful tool that permits us to to study fragmented individuals, resulting in a decrease in fragmentation bias.