11,000 Years of Human-Clam Relationships on Quadra Island, Salish Sea, British Columbia

by Ginevra Mae Toniello

B.A., Simon Fraser University, 2015

Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of Master of Archaeology

in the Department of Archaeology Faculty of Environment

Copyright in this work rests with the author. Please ensure that any reproduction or re-use is done in accordance with the relevant national copyright legislation. Approval

Name: Ginevra Mae Toniello

Degree: Master of Arts

Title: 11,000 Years of Human-Clam Relationships on Quadra Island, Salish Sea, British Columbia

Examining Committee: Chair: Dr. Mark Collard Professor Dana Lepofsky Senior Supervisor Professor

Kirsten Rowell Supervisor Director of Research Leadership University of Colorado

Torrey Rick External Examiner Director Program in Human Ecology & Archaeology Smithsonian Institution

Date Defended/Approved: May 26, 2017

ii Abstract

The historical ecological approach provides unique insights into the relationship between humans and clams throughout the Holocene. Combing archaeological and palaeo-fossil records provides a time depth of clam history both with and in the absence of intensive human predation. These results show that butter clam (Saxidomus gigantea) growth was naturally improving from the early-to-mid Holocene and that humans took advantage of the expanding clam resources. Clam garden construction around 2,000 BP promoted the sustainability of clams, and despite increased harvesting pressure there is no evidence for resource depression. Since European contact, decline of traditional management practices and increases in industrial activities have resulted in reduced clam growth rates. Growth rates of living clams reflect the stunted growth of post-glacial early Holocene clams, making them the slowest growing clams in the past ~10,000 years. Deeper-time baselines more accurately represent clam population variability throughout time and are useful for modern coastal resource management.

Keywords: historical ecology; clam gardens; archaeology; palaeoecology; traditional resource management

iii

Acknowledgements

This research could not have been completed without the dedication of and collaboration from my supervisory committee, to whom I am extremely grateful; my senior supervisor Dr. Dana Leposfky, and supervisor Dr. Kirsten Rowell. Thank you also to the Northern Coast Salish and Southern Kwakwaka’wakw First Nations on whose territory this research took place; and to Laich-kwil-tach and We Wai Kai Nations for the dedicated field support. Christine Roberts and Louie Wilson – thank you for bestowing your knowledge and hard work on this project. This project was completed with financial support from the Tula Foundation and Hakai Institute, and Social Sciences and Humanities Research Council of Canada Grants.

Thank you to the many hardworking collaborators and volunteers in the field, laboratory, and elsewhere: Nicole Smith, Christina Neudorf, Amy Groesbeck, Gavia Lertzman-Lepofsky, Travis Crowell, Susan Kidwell, Joanne McSporran, Daryl Fedje, Quentin Mackie, Cal Abbott, Keith Holmes, Helen Gurney-Smith, Louise Williams, Kate Lansley, Mark Wunsch, Alessandria Testani, Katie Neal, Madeleine Lamer, and Kim Figura. Special thanks to members of the Clam Garden Network for beneficial collaborations and knowledge-sharing; to Torben Rick for your recommendations and encouragement; and to my fellow Archaeology graduate students, especially Nyra Chalmer, Julia Jackley, Chelsey Armstrong, Misha Puckett, and Antonia Rodrigues, for offering advice, support, and inspiration. I’m eternally grateful for my family, my partner Taylor Grant, and my parents Shelly and Dino Toniello, for your continued support in all of my endeavours.

Table of Contents

Approval ...... ii Abstract ...... iii Acknowledgements ...... iv Table of Contents ...... v List of Tables ...... vi List of Figures ...... vii

1. Introduction ...... 1

2. Study Area ...... 5

3. Material and Methods ...... 7 3.1 Butter Clams (Saxidomus gigantea Deshayes) ...... 7 3.2 How Does Butter Clam Growth Vary Through Time and What are the Environmental and Cultural Attributes Affecting Growth? ...... 7 3.2.1 Field Sampling ...... 8 3.2.2 Defining and Predicting the Affects of Environmental and Cultural Attributes on Butter Clam Growth ...... 9 3.2.3 Butter Clam Sample Selection and Sclerochronology ...... 12 3.2.4. Statistical Analyses ...... 13 3.3 How Does the Area of Clam Habitat Change with the Construction of Clam Gardens, and How Does this Affect Clam Abundance? ...... 13 3.3.1 What is the Total Area of Clam Habitat in Clam Gardens Today? ...... 14 3.3.2 How Much Clam Habitat was Provided by Building Clam Gardens? ...... 14

4. Results ...... 16 4.1 The Effects of Environmental and Cultural Attributes on Butter Clam Growth ...... 16 4.2 Variation in Age, Size, and Growth Rates of Butter Clams ...... 17 4.3 Variation in Butter Clams Shells by Temporal Context: A Proxy for Environmental and Cultural Influence and Change ...... 21 4.4 Shifted Baselines of Butter Clam Maturation age ...... 25 4.5 Clam Garden Area and Habitat Increase ...... 26

5. Discussion ...... 28 5.1 Relationships Between the Environmental History, Human History and Butter Clam Growth in Northern Quadra Island ...... 28 5.2. Historical Ecology of Butter Clams on N. Quadra Island ...... 32

References ...... 34

List of Tables

Table 1. Predicted effects of attributes on butter clam growth...... 10 Table 2. Environmental and cultural attributes of each butter clam temporal context. .... 11 Table 3. Attributes for assessing the percentage increase in clam habitat created by building a clam garden...... 15

List of Figures

Figure 1. Study area, including Kanish Bay and Waiatt Bay, northern Quadra Island, BC. Clam garden sites (blue dots), large village midden sites (yellow diamonds), and sampling sites (red stars) within the study area...... 4 Figure 2. Examples of clam gardens constructed on bedrock (left), and on soft sediment (right)...... 6 Figure 3. Von Bertalanffy modelled growth curves for butter clams dated to 11,500- 11,000 BP and 10,900-9,500 BP, collected at sites EbSh-5 and EbSh-36...... 17 Figure 4. Von Bertalanffy modelled growth curves for butter clams from seven temporal contexts ...... 19 Figure 5. Butter clam size, age, and growth rates (juvenile = 0-1, 0-2, 0-5 years; and mature = 5-9 years) for each temporal context...... 20 Figure 6. Human settlement, intertidal environment, and butter clam growth histories of northern Quadra Island...... 24 Figure 7. Butter clams from 11,500-11,000 BP (left) and 10,900-9,500 BP (right) showing the difference in shell thickness and shape...... 24 Figure 8. Butter clam size at maturation of each temporal context ...... 26 Figure 9. Clam garden categories, as outlined in Table 3, showing relative percentages of both the total number of clam gardens and the total area of clam garden habitat...... 27

vii

1. Introduction

Many peoples worldwide have developed intimate knowledge of and relationships with particularly valued species. These relationships developed over millennia of observations and interactions with these species in their natural ecosystems (Anderson 2005; Anderson et al. 2011; Deur and Turner 2005; Minnis and Elisens 2000; Turner 2014, 2005). Such long-term relationships are often expressed in oral traditions, in language, in named locations, and in the archaeological and palaeoecological records. Over time, some particularly important species become “cultural keystones” (Garibaldi and Turner 2004) because their cultural saliency is so great that an alteration to their relationship with humans would fundamentally change a person’s identity.

Historical ecology provides a framework for tracking these human-species relationships through time (Armstrong and Veteo 2015; Armstrong et al 2017; Balée 2006; Crumley 1994). By incorporating multiple kinds of evidence and understanding how they fluctuate over the long term, we can begin to understand the ecological and social contexts of a landscape and the various consequences of human actions. Taking into account the deeper history of this relationship is requisite for understanding and better managing human-ecological relationships today (Braje et al. 2012; Erlandson and Rick 2010; McKechnie and Moss 2016; Rick et al. 2016).

Tracking these long-term relationships requires temporally-grounded records that can provide insights into both the ecological and cultural sides of this relationship. Archaeological faunal records on the one hand can provide detailed information about the ecology and the effects of human predation on those resources during times of human harvesting (e.g., Braje et al. 2012; Cannon and Burchell 2009; Erlandson et al. 2008, 2009). On the other hand, some palaeoecological records can provide insights into species ecology in the absence of significant human predation. Taken together, these two records can provide powerful historical ecological insights into the continuum of impacts on natural resources and the human-species relationships. In many landscapes, however, because of the intensity of long-term human- environmental relationships, it is difficult to find ecological baselines that are not influenced by people to some extent, either in the distant past or in the industrial present.

On the Northwest Coast of North America, clams are valued cultural species (Lepofsky

1 and Caldwell 2013; Lepofsky et al. 2015; Moss 1993) whose widespread importance is reflected in origin stories, rituals, language, and in the kilometers of deep and ancient shell middens that line the coastline (Lepofsky et al. 2015). Firm archaeological evidence of clam consumption in this area dates back as early as ~9,000 Cal. BP (Ackerman et al. 1985). However, ground-water dissolution at the base of most archaeological sites almost certainly obscures earlier evidence of clam harvesting. By ~5000 BP, due to changing sea levels and the increasingly large and settled populations who created shell platforms as the foundation of their settlements, shell middens became widespread in the region (Martindale et al 2009). Detailed archaeological and ethnographic research indicates that bivalves, especially butter clams (Saxidomus gigantea), littlenecks (Leukoma staminea), and mussels (Mytilus californianus, M. edulis, M. trossulus, M. galloprovincialis) were eaten both seasonally and year-round, both fresh and dried. These bivalve species were reliable, abundant, easily harvested near people’s settlements (Burchell et al 2013; Moss 1993), and could be managed to increase abundance (Deur et al. 2015; Groesbeck et al. 2014; Jackley et al. 2016; Lepofsky and Caldwell 2013; Lepofsky et al. 2015).

In addition to the regional archaeological record, the palaeoecological record of fossil and subfossil bivalve molluscs can provide long-term information on species abundance and richness in the absence of considerable human intervention (Kidwell 2015, 2002, 2001; Kvenvolden et al 1979), and is important in establishing ecological baseline data (Rowell et al. 2008). Such buried bivalve subfossil assemblages in tidal and intertidal contexts have been noted on the Northwest Coast (Fedje et al. 2011; Moss et al. 1990), but to our knowledge the potential data from these contexts have not been fully harnessed. Combining palaeo- and archaeological faunal records provides a window into the social and ecological contexts of bivalve molluscs throughout the Holocene (e.g., Jackson et al. 2001; Kidwell 2015; Rick et al. 2016).

In this paper, we investigate the historical ecology of butter clams on northern Quadra Island, Salish Sea, British Columbia (BC) (Figure 1) over the last 11,000 years and their response to human harvest and management. To do this, we use both the palaeoecological and archaeological records to evaluate the effects of various environmental and cultural conditions, including human management of clam populations. Our study sites in Kanish and Waiatt Bays have a long history of human-clam relationships, which is evident in the archaeological shell middens and the plethora of clam garden features (Figure 1). Clam gardens are a form of traditional Indigenous clam management that involves the enhancement and expansion of

2 shellfish habitats, and thus clam abundance, through the construction of rock-faced sand terraces in the lower intertidal zone. Such features have been identified throughout the Northwest Coast and date to ~2,000 years ago (Deur et al. 2015; Neudorf et al. 2017; Smith et al. in prep). Traditional ecological knowledge of the Northwest Coast demonstrates that in addition to clam gardens, clams were managed in a variety of ways, including transplanting, tilling, non-human predator deterrence, and habitat enhancement (Caldwell et al. 2012; Deur et al. 2015; Ellis and Wilson 198; Lepofsky et al. 2015; Marlor 2009; Parks Canada 2011; Turner 2005; Woods and Woods 2005).

We investigate the long-term relationship of humans and clams by evaluating butter clam growth rates and size-at-age, and model these attributes in different temporal, ecological, and cultural contexts. We also measure the number and area of clam gardens, as well as the extent of clam habitat created by building clam gardens. Understanding the environmental context of clams prior to extensive human harvesting, when people began to harvest clams in earnest, and when they began to construct clam gardens, allows us to explore the effects of human harvesting and mariculture on clam populations through time, as well as the effects of clam availability on human settlement in Kanish and Waiatt Bays. Furthermore, we can compare these data to the marine landscape of the modern industrialized era, which has not been managed by Indigenous peoples for decades, and is at risk from on-going industrial and housing developments, and the effects of climate change.

Our results show that conditions for clam productivity improve throughout the early to mid Holocene and humans took full advantage of the large clams that resulted from these enhanced growing conditions. This harvesting pressure is reflected in a reduced number of clams in the intertidal subfossil record and a burgeoning archaeological record of harvested clams. Despite this increased pressure, there is no evidence that people were forced to harvest smaller individuals through time. Rather, we hypothesize that the harvesting pressure was ameliorated by the huge increase in clam habitat resulting from the creation of clam gardens. It is not until the introduction of industrial activities and the dramatic depopulation of Indigenous people in the region that we see clam populations suffering. Collectively, our data highlight the value of an historical ecological approach to studying cultural keystone species and the potential for using these data to understand modern ecological processes.

4 2. Study Area

Kanish and Waiatt Bays, located on the northern end of Quadra Island, are within the traditional territories of the southern Kwakwaka’wakw (Laich-kwil-tach) and northern Coast Salish First Nations. The area is relatively remote, with access largely limited to boat travel. Much of the region is provincial park or leased government land. The seasonal and year-round population consists mainly of newcomers to the region and is limited, with less than 15 homes bordering the bays. Both bays are seasonally popular tourist destinations for boaters and hikers, and both local First Nations and visitors harvest clams from some of the beaches. Kanish Bay was heavily impacted by logging in the early 20th century, and logging continues in some upper watersheds today. Evidence of sediment runoff as a result of the logging is observable when digging in some of the beaches. There have also been local impacts from fish farming. With the exception of these few visible and traceable impacts, the beaches are relatively undisturbed by industrial activity.

Kanish and Waiatt Bays were ideal locations for long-term settlement of Indigenous populations throughout most of the Holocene. A short 1.5 km trail between the two bays, which is still used by hikers, provided easy access between bays and bridged the ecosystems, forming a single cultural landscape. Both bays offer numerous sheltered coves, protection from the elements, access to freshwater, and an abundance and variety of terrestrial and marine flora and fauna, including shellfish. In most weather conditions, people could also travel easily among the settlements within each bay via canoe.

Archaeological investigations indicate that the bays were settled at least by 9,000 BP (Toniello et al. 2017), when sea level stands were ~2 meters higher than today (Fedje et al. 2017). Sea level was relatively stable between 9,000-5,000 BP. Over the next 3,000 years, sea level dropped 2 meters, reaching modern levels by ~2,000 BP (Crowell 2017; Fedje et al. 2017). Our testing of archaeological sites suggests that these initial human settlements were relatively small, and probably had low to moderate impact on the landscape in the form of settlement construction and predation on local plants and animals. By as early as 5,000 years ago, people in Kanish and Waiatt Bays were aggregating in large, permanent settlements, evidenced today by enormous archaeological shell midden sites (Toniello et al. 2017). Radiocarbon dates from the base and caps of these sites suggest that by ~2,500 BP, most large villages were settled and continued to be occupied until the late 1700s when European diseases severely decreased

5 Indigenous populations (Harris 1994, Taylor 2009, Toniello et al. 2017). The number and extent of shell middens in the bays illustrates that nearly every portion of foreshore that could be transformed into flat, usable land, was used for settlements.

The importance of clams to the pre-contact Indigenous people of Kanish and Waiatt Bays is reflected in the region’s shell middens and also the expansive record of clam gardens. The density of clam garden features within Kanish and Waiatt Bays are among the highest documented on the Northwest Coast. Our dating of clam gardens through a variety of novel methods indicate that clam gardens were constructed at least as early as 2,000 BP (Lepofsky et al. 2015; Neudorf et al. 2017; Smith et al. in prep). Clam garden walls and adjoining beach terraces are abundant within this study area, present in almost every conceivable nook, and constructed on both soft and hard substrates (Figure 1, 2).

Figure 2. Examples of clam gardens constructed on bedrock (left), and on soft sediment (right).

6 3. Material and Methods

3.1 Butter Clams (Saxidomus gigantea Deshayes)

Butter clams can be found along the Pacific Northwest Coast from Alaska to California in a mixed intertidal substrate of gravel, sand, and shell (Quayle and Bourne 1972). This species lives up to 30cm below the beach surface, and range from the mid-to-lower intertidal to up to 30m below tidal datum (Quayle and Bourne 1972). Today, butter clams can live upwards of 20 years (Harbo 1997), and mature at a mean size of 38mm, which is reached by age 4 or 5 on average in southern areas, and age 8 to 9 on average in northern ranges (Quayle and Bourne 1972). Butter clams are a cosmopolitan species on the west coast of North America, are abundant in archeology sites along the B.C. coast, and are important in traditional diets (Cannon and Burchell 2009; Deur et al. 2015; Lepofsky et al. 2015; Williams 2006). This species is also abundant in the intertidal fossil and subfossil deposits along the coast. The earliest known butter clam in British Columbia, which is from Lax-Kw'alaams, BC, dates to ~15,000 years BP (Hetherington and Reid 2003). In Kanish Bay, butter clams are present in intertidal subfossil deposits throughout the Holocene, as early as 11,500 BP. The widespread temporal distribution of this species within the study area is likely due to the availability of suitable environmental conditions throughout most of the Holocene, including stabilizing sea levels (Fedje et al.2017), gravel sand beaches (Orford et al. 2002), and warm sea surface temperature (SST) (Keinast and McKay 2001).

3.2 How Does Butter Clam Growth Vary Through Time and What are the Environmental and Cultural Attributes Affecting Growth?

There are many factors that affect the abundance of clams available for human foraging. The first order factor is habitat availability, or how much habitat is available to clams. More available clam habitat results in more clams. The second order is quality of habitat which can influence density of clams, growth of clams, and survivorship of clams. We assess how several environmental and cultural factors, including SST, beach slope, intertidal substrate, clam garden construction, and degree of human interaction, may alter the overall productivity of clam habitats (Table 1). We collected butter clam shells from paleo-beach deposits below clam gardens, from

7 within clam gardens, and from archaeological shell middens within Kanish Bay. Each of these sampling universes is affected by different environmental and cultural attributes throughout the Holocene (Table 2). Sclerochronological analyses of butter clam thin-sections provides information about relative growth rates, sizes, and demographics of butter clam populations within different environmental, cultural, and management contexts. Assessing the butter clam subfossil assemblages from middens (harvested clams) and from in situ beach contexts provides insights into the interaction of people and clams, and the effects of building clam gardens on clam growth and abundance.

3.2.1 Field Sampling

S. gigagntea have robust shells that appear to preserve well in the intertidal palaeo- and archeological records, which make them strong candidates for sclerochronology analyses. These specimens can provide deep-time ecological information across all depositional contexts, including archaeological midden deposits (Burchell et al. 2013; Cannon and Burchell 2009, 2016; Hallmann et al. 2013). Butter clams were collected from three site types in Kanish Bay: 1) palaeo-beach deposits below clam gardens (N=3 sites); 2) clam garden deposits (including live- collected and recently dead specimens) (N=2 sites); and 3) archaeological shell midden deposits (N=2 sites) (Figure 1 and Table 2). Paleo-beach and clam garden sampling sites were selected for two reasons. Firstly, because they are situated at relatively higher elevation above chart datum and thus allowed for extended excavation time during the low tide window; and secondly, because they have a soft-sediment substrate that allowed us to collect specimens from both pre- (paleo beach) and post- clam garden contexts.

Evidence from the taphonomy, position, and articulation of both valves from the intertidal subfossil clams from paleo-beach and clam garden deposits indicates that these specimens died of natural causes in situ and since remained relatively undisturbed within the sediment column. To retrieve these samples, excavation units were placed in clam garden terraces, in beach sediments seaward of the clam garden walls, and in clam garden rock walls themselves. When possible, we aimed to reach the basal clay deposited after the retreat of the Cordilleran glaciers 13,500 BP (Blais-Stevens et al. 2001; Ryder et al. 1991). This was not always possible due to the constraints of the tide and groundwater flooding of the excavation units. In each excavation unit, all shells in good condition (full valves, limited breakage) were collected for laboratory analysis. This resulted in some sampling bias towards larger or more robust shells and against the smaller, more fragile shells that are more susceptible to taphonomic processes.

8 Clams that were harvested, and thus did not die in situ, include live-collected specimens and midden-deposited specimens. Living butter clams are relatively rare in Kanish Bay beaches, so we collected every living specimen that was encountered from within five shallow shovel tests along a transect (7m, 16m, 25m, 28m, and 32m perpendicular from the clam garden wall) within the clam garden at EbSh-13. Clam specimens from the two midden locations (EbSh-14, EbSh-13) come from dated column samples excavated in 2013 (Puckett et al. 2014). Since midden specimens represent clams that were harvested selectively by Indigenous people, they do not necessarily represent the full natural range of variation in clam sizes and ages.

3.2.2 Defining and Predicting the Affects of Environmental and Cultural Attributes on Butter Clam Growth

There are several changing ecological conditions that are affected by abiotic environmental and cultural attributes that we can assess through the preserved record of the Holocene and that may have affected butter clam growth (Table 1). We recorded the various conditions of the environmental and cultural attributes throughout the Holocene to determine if any of these conditions influenced clam growth (Table 2). The environmental abiotic attributes that we measure in this study include substrate, beach slope, and sea surface temperature. Cultural attributes include the presence of modified clam habitat (clam gardens), and the degree of human interaction.

The environmental attribute of substrate was evaluated during intertidal excavations, which exposed intertidal sediment stratigraphy. Pre- clam garden beach slope was assessed by analyzing the geomorphology of the beach and surrounding landform, whereas contemporary clam garden beaches are relatively flat. Sea surface temperature data were provided by Keinast and McKay (2001).

The cultural attribute of the presence or absence of clam garden beach modification was evaluated based on the temporal and stratigraphic association of clam garden rock walls and clam garden terrace sediments with butter clam samples. The degree of human interaction with these habitats was determined by temporal and geographical proximity to archaeological settlements. Assessment of these attributes allowed us to make predictions about shell size and growth rates unique to the environmental and cultural context of each temporal context.

9 Table 1. Predicted effects of attributes on butter clam growth.

1. While not all human interactions are beneficial (e.g., over-harvesting, trampling, displacing sediment, leaving open holes, etc.), many will have a positive or at least a neutral effect on the health of clam populations (Lepofsky et al. 2015).

Table 2. Environmental and cultural attributes of each butter clam temporal context.

1. Since shells from midden were selected from the beach by ancient harvesters, they will tend to be biased towards larger specimens at any given age than the other (non-midden) samples. In the case of the living specimens harvested by Toniello, there are no sampling biases as all encountered specimens were collected. However, because they were harvested, live-collected specimens did not reach maximum age or size.

2. Human interaction is inferred from number, size, and proximity of ancient settlements to the harvested beach. The closer a large settlement and shell midden, the greater the influence of tilling and harvesting on intertidal ecology.

11 3.2.3 Butter Clam Sample Selection and Sclerochronology

We selected a sub-sample of 125 butter clam shells in good condition for thin-section preparation and sclerochronological analysis (Table 2). Shells were determined to be in good condition if both valves were present with limited wear, and if the umbo was preserved. As with the in-field sampling methods, this presents some bias towards larger specimens, as larger, more robust shells are more likely to be preserved in full than smaller, more delicate shells. Keeping the selection criteria biased toward larger clams across all sampling contexts (middens and in situ beach assemblages) should maintain consistency and reduce skewed results.

Intertidal subfossil clams were selected for analysis from sites that were known to have adequate temporal representation. The temporal context of intertidal clam specimens was determined in the field by analyzing the depositional context, and in the laboratory by radiocarbon dating select samples from each depositional context. Dated samples were processed by John Southon at W.M. Keck Carbon Cycle AMS Laboratory at UC Irvine using Accelerator Mass Spectrometry (AMS) Radiocarbon dating techniques. Butter clam shells in good condition from each temporal context were sampled at random using a number randomizer.

Thin-sections of butter clam shells were prepared for sclerochronological analysis by cutting along the axis of maximum growth of each butter clam. The preparation and processing of shell thin-sections follows standard laboratory methods (see Schöne et al. 2005). For clams that died naturally, differences in age and size at death are indicative of how well the clam was able to adapt to the ecological and environmental conditions available. Growth rates, as determined by annual growth increments preserved in the shell, were measured to determine growth differences from year to year, in the juvenile stages (years 0-5), and after sexual maturity (years 5-9). Although size is the main factor determining when a specimen matures (Quayle and Bourne), we use age as a proxy for size due to ease of measuring growth annually.

During the first three years of growth, clams grow their fastest, which provides higher resolution in detecting measurable differences in growth during these years. In addition, this early life history stage is the most vulnerable stage of animals - how fast the clam achieves adequate size to ensure survival from predation and other fecundity- related attributes. We chose to define the juvenile growth as 0-5 years because 95% of the specimens we collected are mature by age 5, which fits with modern measurements of maturity occurring between 4-5

12 years (Quayle and Bourne 1972). Analysis of the average growth up to the year of maturity (years 0-5) was analyzed to account for the cumulative effects of annual growth, and to ensure we reduce the impact of growth advantages by accounting for specimens that experience compensatory growth, or “catch up” in size in later years (Munroe et al. 2016). Mature annual growth lines are important in determining how well the clam is growing after sexual reproduction.

3.2.4. Statistical Analyses

We produced von Bertanlanffy (1938) growth curves to model the growth rate of clams in each temporal context. L-infinity values were produced to accompany the growth curve model and assist in assessing differences between butter clams from different temporal contexts. L- infinity values are informed by known parameters (size at age, size at death, age at death), represent an average age of clam longevity, and are a numeric representation of general clam health. These outputs allowed us to generate hypotheses about 1) the juvenile growth rate; 2) growth rates after sexual maturity; 3) the size at age; and 4) the size and age at death. We tested for difference in these parameters between the three different clam living conditions (prior to human settlement, during active clam gardening, and post-European colonization). ANOVA and Welch paired with Tukey’s HSD and Games-Howell post-hoc statistical tests were performed to determine the extent and significance of the variation in growth rates, size, age, and maturity of butter clams between each temporal context. We used multiple regression analyses to explore which of the various environmental and cultural attributes we identified may be contributing to these differences.

3.3 How Does the Area of Clam Habitat Change with the Construction of Clam Gardens, and How Does this Affect Clam Abundance?

To determine if clam availability and abundance shifted throughout time as a result of changes to the area of available intertidal clam habitat, we measured the total surface area of clam habitat within clam gardens today. We then calculated the overall increase in intertidal clam habitat provided by the clam gardens by estimating the area of pre- clam garden habitat. These measurements help us to assess the availability of intertidal clam habitat, and therefore clam abundance, prior to- and during- large-scale human management of clam resources. We

13 recognize that butter clam habitat extends into sub-tidal areas; however, because these areas are not readily accessible to human harvesters or to us as researchers, we only focus our estimates on intertidal clam habitat.

3.3.1 What is the Total Area of Clam Habitat in Clam Gardens Today?

Over 3,000 high-resolution drone photographs were taken during the 2016 field season survey, and were stitched together to create a high-resolution map of the study area with up to 20cm accuracy (Holmes 2016). The western extent of Kanish Bay was not photographed by drone due to poor weather conditions and is supplemented by Google Earth imagery in lieu of higher-resolution drone images. Clam gardens are clearly visible on both the drone- photographed and the Google Earth maps. Clam habitat within each clam garden terrace was delineated using the polygon tool on GIS computer software. We then calculated the area of each polygon giving us a total of the productive clam habitat in clam gardens today. The area of clam habitat available in clam gardens today is a proxy measure of the maximum clam habitat within clam gardens available to people living in Kanish and Waiatt Bays in the past. Our estimates of the areal extent of clam habitat were based on previous surveys (Groesbeck et al. 2014; Harper 2007; Harper and Morris 2004) and our own observations of clam beaches in the study area. At each clam garden location, the area of clam habitat was under-estimated in order to ensure a conservative overall estimate of the productive habitat within clam gardens today.

3.3.2 How Much Clam Habitat was Provided by Building Clam Gardens?

Estimating the area of clam habitat in Kanish and Waiatt Bay prior to the construction of clam gardens required understanding the underlying geomorphology of each clam beach. During the 2013-2016 field seasons, we characterized the underlying substrates by excavating over 100 shovel tests in 14 clam garden terraces and 3 non-walled intertidal sand flats abutting the clam gardens, and 15 trenches through 8 clam garden rock walls.

To quantify the increase in clam habitat provided by clam gardens, four researchers independently evaluated each clam garden beach in the study area based on the pre- clam garden geology at each location. A list of categories was developed to guide the assessment of prior geomorphic beach conditions (Table 3). The pre- clam garden geology was partially informed by the results of our excavation units, and also inferred from surrounding slope and sedimentation processes. The goal of this assessment was to determine whether or not there

14 was viable clam habitat prior to clam garden construction, as well as to estimate the extent of such habitat. Based on these criteria, each researcher independently estimated the percentage increase of area of clam habitat based on their assessment of prior habitat conditions. The percentage-categories (Table 3) provide us with a minimum and maximum value of the estimated increase in clam habitat surface area for each clam garden location. In some cases, single clam gardens were separated into multiple categories. For instance, most Category A clam gardens have smaller Category D “extensions” on either end.

Table 3. Attributes for assessing the percentage increase in clam habitat created by building a clam garden.

15 4. Results

4.1 The Effects of Environmental and Cultural Attributes on Butter Clam Growth

The environmental and cultural attributes outlined in Table 2 account for only a small percent of the total variation in butter clam growth among the samples. In the early developmental years of the butter clam juvenile growth stage (years 0-5), site (p=0.252), substrate (p=0.007), SST(p=0.010), and slope (p=0.212) attributes together account for 26.1% of the variation in butter clam growth rate. In the years after butter clam sexual maturity (years 5-9), only site (p=0.165) and substrate (p=0.076), and the additional attribute of human influence (p=0.020), contribute to the variation in butter clam growth. Together, these three attributes account for 25% of variability in butter clam growth rate during these later years.

The variable of site (i.e., beach), which includes countless attributes that fluctuate throughout time, has a significant influence on the growth of clams throughout their lives. Each of the five sampling sites (EbSh-5, EbSh-13, EbSh-14, EbSh-36, EbSh-77) provided different growing conditions for clams (Table 1, Table 2) that differentially affected clam growth throughout time. Site-specific effects are especially prevalent in the two early Holocene temporal contexts (11,500-11,000 BP and 10,900-9,500 BP), which both include butter clam samples from sites EbSh-5 and EbSh-36 (Figure 3).

Figure 3. Von Bertalanffy modelled growth curves for butter clams dated to 11,500-11,000 BP and 10,900-9,500 BP, collected at sites EbSh-5 and EbSh-36.

4.2 Variation in Age, Size, and Growth Rates of Butter Clams

Differences between butter clams from different temporal contexts are visible in the variations in the von Bertanlanffy modeled growth curves (Figure 4), as well as in the size, age, and growth rates (Figure 5). Modeled growth curves allow us to see patterns between butter clam growth from different temporal contexts, and to further explore which environmental and/or cultural attributes may explain these differences. The only significant differences in butter clam size at death are found in intertidal subfossil samples from two temporal contexts: the 11,500- 11,000 BP samples, which are smaller in size than both the 10,900-9,500 BP and 4,200-2,900 BP samples (ANOVA p=0.000, Tukey HSD p=0.000-0.022); and the 4,200-2,900 BP samples, which are significantly larger in size than butter clams from any other intertidal or midden contexts (ANOVA p=0.000, Tukey HSD p=0.000-0.022) (Figure 5a). Differences in age at death are visible in the 4,200-2,900 BP intertidal subfossil samples, which are significantly older at death than both the early historic 250-100 BP clam garden beach subfossil samples as well as the live-collected samples (Welch p=0.000, Games-Howell p=0.018-0.020). The two midden samples, which are human-harvested between 2,800-2,300 BP and 500-200 BP, are younger at

17 death than all samples pre-dating them (>2,800 BP) (Welch p=0.000, Games-Howell p=0.000- 0.046); and (Figure 5b).

We detected few differences in year-to-year growth and cumulative growth rates between butter clams from different temporal contexts. In the first year of growth, the only significant difference is visible in the 250-100 BP clam garden beach subfossil samples, which grew faster than all other butter clam samples (Welch p=0.000, Games-Howell p=0.000-0.002) (Figure 5c). By the second year of growth (year 1-2), there are no differences among the samples, suggesting that the other clams “caught up” through compensatory growth (Munroe et al. 2016). The legacy of increased growth in the first year of these samples, however, can be detected in the cumulative growth rate from years 0-2, which remains significantly greater than all other temporal contexts except modern (Welch p=0.000, Games-Howell p=0.000-0.0031) (Figure 5d). The absence of detectable statistical differences in clam growth at other ontogenic stages likely reflects the fact that yearly differences are small and our archaeological and intertidal butter clam depositional contexts are temporally broad, which can increase the variability in our sampling.

When the cumulative growth rates for the juvenile stage of growth (years 0-5) in butter clams are examined, we see differences among the butter clams from different temporal contexts that are less detectable when examining the small incremental growth per year. Differences in these pre-mature clam growth rates is visible in the 11,500-11,000 BP subfossil clams, which grow slower than all other temporal contexts examined, except the live-collected specimens (Welch p=0.000, Games-Howell p=0.000-0.003) (Figure 5e). Additional differences are visible in the cumulative growth rates after sexual maturity (years 5-9) between faster- growing butter clam midden samples (2,800-2,300 BP and 500-200 BP) and slower-growing 11,500-11,000 BP subfossil butter clams (Welch p=0.004, Games-Howell p=0.036-0.048) (Figure 5f).

Figure 4. Von Bertalanffy modelled growth curves for butter clams from seven temporal contexts

Figure 5. Butter clam size, age, and growth rates (juvenile = 0-1, 0-2, 0-5 years; and mature = 5-9 years) for each temporal context.

Dark grey represents subfossil shells from intertidal deposits, light grey represents shells from archaeological midden contexts and live-harvested samples.

20 4.3 Variation in Butter Clams Shells by Temporal Context: A Proxy for Environmental and Cultural Influence and Change

Beginning in the early Holocene (11,500 BP) and continuing through to the late middle Holocene (2,900 BP), there appears to be a trend toward increasingly favourable growing conditions for butter clams (Table 1, Table 2). Counts of shell annuli and measurements of growth increments indicate that butter clams grew faster, and lived to be significantly older and larger as time passed. This is indicated by our measurements of significant increased specimen size at death (ANOVA p=0.000, Tukey HSD p=0.000-0.022) (Figure 5a); as well as trends towards older age at death (Figure 5b), increased juvenile growth rate (years 0-5) (Figure 5e), increased mature growth rate (years 5-9) (Figure 5f) and increased L-infinity value (Figure 6) from 11,500 BP to 2,900 BP. The 11,500-11,000 BP intertidal subfossil samples represent some of the first butter clams to settle within this study area after the retreat of the glaciers. These thin and gracile subfossil shells (Figure 7), which have slower growth rates and are smaller at any given age than all other temporal contexts (Figure 4), likely reflect the sub-optimal conditions in which they grew, including cold SST (Figure 6), poorly drained clay substrate, and relatively steeper beach slope (Table 2).

The 10,900-9,500 BP samples are relatively bigger in size, and trend towards longer- lived and having increased juvenile (years 0-5) and mature (years 5-9) growth rates than the previous period (Figure 5e, 5f). These shells are larger in size at any given age than the 11,500- 11,000 samples (Figure 4), and are thick and robust (Figure 7) as a result of their slow growth during the mature stage (i.e., narrow growth rings, represented in the flat growth curve) (Figure 4). Their atypical growth pattern suggests that despite improvements in substrate, these specimens also grew under stressful conditions, likely including the warmer-than-ideal SST during this time (Figure 6), and nonlethal effects of non-human predation (Nakaoka 2000; Smee and Weissburg 2006).

Mid-Holocene butter clams (4,200-2,900 BP) are bigger in size at death (Figure 5a), and trend toward being older at death (Figure3b) than butter clams from all other temporal contexts before and after this period. The juvenile (years 0-5) and mature (years 5-9) growth rates also trend toward faster-growing in this period (Figure 5e, 5f), which is visible in the relatively larger size at age than the previous temporal contexts (Figure 4). The large size and long-lived age of the mid-Holocene clams (4,200-2,900 BP) is likely related to the improved growing conditions, which include preferred SST and substrate (Figure 6), and decreased beach slope (Table 2).

21 Despite an increase in human populations, indicated by the accumulation of large shell middens beginning around ~5,000 BP, the abundance of large, old clams from the 4,200-2,900 BP intertidal subfossil assemblage at site EbSh-77 indicates that this specific population, at least, was not under significant predatory pressure by humans.

Butter clam samples from both midden contexts (2,800-2,300 BP and 500-200 BP) tend to grow similarly and no significant difference in juvenile (years 0-5) or mature (years 5-9) growth rates was detected when compared to the earlier mid-Holocene samples (4,200-2,900 BP) (Figure 5 e, f). The calculated growth curves for these butter clams also follow the mid- Holocene samples (Figure 4), likely due to the beneficial ecological conditions that were established in the mid-Holocene and that remained unchanged (Figure 6). The increase in human influence during this time seems to have no influence on the overall growth rate of butter clams. The multiple-regression analysis of ecological and cultural attributes (above) does show that human interaction influences butter clam growth rate after sexual maturation (years 5-9), though we conclude that this is largely being driven by the affect of human harvesting biases that show up in the midden samples. That is, humans will tend to harvest clams of larger size (i.e., the mature specimens). Our lack of detecting a difference in growth rates for the juvenile years of clam growth suggests that harvest levels where not high enough to affect the life history of these clams. In other words, human harvest was not high enough to affect the gene pool resulting in selection for individuals that grow fast enough and mature early enough to contribute to the next generation. Nor did humans have enough influence on the physiology to trigger a fast growth through thinning the population causing a release from density dependent growth.

Midden samples of butter clams that were harvested from within a clam garden (500-200 BP) show no difference in size at death (Figure 5a), age at death (Figure3b), nor growth rates (Figure 5c-f) from the midden samples that were harvested from a non-walled beach (2,800- 2,300 BP). The lack of change in growth rate suggests that while the change in slope and substrate provided by the construction of clam gardens renders a more suitable habitat, they were not enough to have a significant effect on butter clam growth rate. In addition, the lack of differences in age at death and size at death between the two midden samples suggests that the resources were not being harvested to the extent that people were forced to take smaller and smaller clams. As we discuss further below, we hypothesize that this lack of resource depression was largely due to increases in clam habitat as a result of clam garden construction.

22 The 250-100 BP intertidal subfossil specimens, which lived and died within clam garden beaches, are slightly smaller in size and younger at death (Figure 5a, b), and trend toward a decreased size at age after maturity (Figure 4) compared to the earlier intertidal subfossil butter clams from the mid-Holocene. Although not significant, the overall decrease in growth of these samples is also visible in the decline of the L-infinity value (Figure 6). Most environmental and cultural attributes that we assess in this research remain unchanged during the late Holocene, with the exception of the reduction in human interaction with clam habitats after European contact. We conclude that the trend we observe toward smaller size and younger age at death is a result of this decline in human interaction, including lack of management of clams through tilling, rock removal, predator removal, and harvesting.

The butter clams living within clam gardens today continue to show a trend toward decreased size at death (Figure 5a), age at death (Figure 5b), and growth rate (Figure 5f, Figure 4), which is also visible in the decreasing L-infinity value (Figure 6). Our findings of slightly younger age and smaller size at death for this sample group is likely due to the live-collected samples being harvested before natural death. However, we feel this should only slightly bias our results because we selected the largest individuals found for these analyses. Although there are no significant differences in cumulative growth rates (years 0-5 and 5-9) between the live- collected clams and any other temporal context, the modeled growth curve (Figure 4) shows that the size at any given age of these live-collected butter clams is most similar to the early Holocene samples. We attribute their smaller size at age to the absence of human intervention at this time, as well as an increase in silt deposition as a result of industrial logging (Table 2, Figure 6).

Figure 6. Human settlement, intertidal environment, and butter clam growth histories of northern Quadra Island.

Regional sea level (RSL) data provided by Fedje et al. (2017) and Crowell (2017). Sea surface temperature (SST) data for the Northeast Pacific provided by Kienast and McKay (2001), indicating temperature outside (red) and within (green) preferred range for butter clams.

Figure 7. Butter clams from 11,500-11,000 BP (left) and 10,900-9,500 BP (right) showing the difference in shell thickness and shape.

24 4.4 Shifted Baselines of Butter Clam Maturation age

Our data allow us to evaluate changes in butter clam maturation size and age throughout the Holocene and indicates that modern butter clams are maturing at smaller sizes than those represented in our intertidal and archaeological subfossil assemblages. In particular, live-collected butter clams from Quadra Island mature at the same size as modern clams collected from elsewhere in the Salish Sea, BC (Quadra mean size = 38.79mm, SD 9.1mm; Salish Sea mean size = ~38mm [Quayle and Bourne 1972], T-test p=0.743). That these living butter clam samples fall within the modern regional average indicates both that our methods for assessing maturation are sound and that our data are somewhat representative of the regional pattern.

We note, however, that most of the intertidal and archaeological subfossil assemblages collected for this research are consistently larger than our live-collected samples or modern butter clams from elsewhere in Salish Sea. That is, each of the subfossil assemblages, except for midden samples from 500-200 BP, are significantly larger than the contemporary maturation data from the Salish Sea (Quayle and Bourne 1972; t-test, p ≤ 0.047). This in turn suggests that the modern measurements of butter clams reported in the literature reflect a shifted baseline and that contemporary fisheries data may not always be an accurate representation of the growth patterns of past clam populations (Figure 8). Establishing a deeper-time baseline for fisheries data that more accurately represents population variability through time can be useful for modern coastal resource management.

Figure 8. Butter clam size at maturation of each temporal context

Dotted lines indicate the mean size at maturation as indicated by contemporary fisheries data (~38mm, red) (Quayle and Bourne 1972), and the subfossil clams collected for this research from 11,500 BP to 100 BP (44mm, green).

4.5 Clam Garden Area and Habitat Increase

Our calculations of of clam garden area and analysis of clam habitat increase indicates that the building of clam gardens had a significant effect on the availability and productivity of clams. The creation of clam gardens reduced human harvest pressure on the more limited natural clam beaches in the bays. The increased productivity of clams in clam gardens and increased habitat for clams resulted in a lack of resource depression observed between the butter clam midden samples dating pre- and post- clam garden construction. We calculate a total of 15.6 hectares (ha) of clam habitat surface area represented by clam gardens within Kanish and Waiatt Bays today, which is a conservative estimate for the maximum area of clam gardens last used by local populations. Based on the categories created to estimate clam habitat increase (Table 3), clam gardens increased the surface area of clam habitat between 4.4-8.3 ha, or by 28-53% from non-altered pre- clam garden clam habitats.

26 Categories A-C clam gardens have relatively even distribution in percentage of total number of clam gardens, but vary in the relative percentage of total area they represent (Figure 9). Category D clam gardens, which were constructed on bedrock and/or boulder outcrops, are the most numerous, representing 40% of the total number of clam gardens. Bedrock clam gardens are numerous, and account for 75-100% increase in clam habitat at each bedrock clam garden location, yet only represent 2.1 ha (13.5%) of the total clam garden surface area due to their small size. Despite an overall small contribution to the surface area of clam habitat, bedrock clam gardens represent an intensification of bivalve-resources. The intensification of clam harvesting, as depicted in the construction of bedrock clam gardens ~1,500 years BP (Neudorf et al. 2017), occurs during a time of village site expansion (Toniello et al. 2017). This suggests that human population increase likely prompted the need for additional clam resources.

Figure 9. Clam garden categories, as outlined in Table 3, showing relative percentages of both the total number of clam gardens and the total area of clam garden habitat.

This analysis included 192 measurements on 176 clam gardens. Clam garden category A represent an estimated 0-25% increase, category B represent an estimated 25-50%, category C represent an estimated 50-75%, and category D represent an estimated 75-100% increase in clam habitat surface area.

27 5. Discussion

5.1 Relationships Between the Environmental History, Human History and Butter Clam Growth in Northern Quadra Island

With the retreat of the Cordilleran ice sheets ~13,500 BP, coastal areas provided increased habitat for many species – including humans and bivalves. Our earliest intertidal butter clam subfossils dating to 11,500-11,000 BP burrowed into sands mixed with the glacial- marine clays below, on steep beach slopes, and in relatively cold SSTs (Table 2, Figure 6). These stressful conditions likely contributed to their small size and age at death (Figure 5a, 5b, 6), their slow growth rates (Figure 4, 5e, 5f, 6), and the gracility of their shells (Figure 7). We do not know the nature, if any, of human-clam interactions during this earliest Holocene period. Currently, evidence for early human occupation of northern Quadra is limited to inferences based on the location of high-elevation, shell-free archaeological sites. These sites are associated with marine deposits in association with higher sea level dated to 13,000 to 12,500 BP (Fedje et al. 2017). Based on the early Holocene archaeological record of other locations on Quadra Island (Fedje et al. 2017), we suspect that humans were minimally visiting Kanish and Waiatt Bays by at least 10,000 years ago, and probably earlier. We recognize that there will always be a bias against the preservation of clam shells in older archaeological deposits due to the acidic soils of the Pacific Norwest, so if early sites are confirmed, it would be difficult to explore the role of clams in the human diet.

Bivalve environmental conditions began improving ~11,000 BP. These changes include a transition to gravel-sand substrate due to paraglacial deposits and hydrodynamic erosion (Orford et al. 2002), an increase in SSTs, and stabilizing sea levels (Fedje et al. 2017) (Table 2, Figure 6). These factors likely contributed to an increased phytoplankton food supply (Eppley 1972) and more stable substrates for bivalve settlement (Roegner 2000). These improved conditions are reflected in our data and observations of relatively larger size, faster growth rates, and older age at death of the 10,900-9,500 BP intertidal subfossils compared to the earlier temporal context (Figure 4 & 7). The robustness of the 10,900-9,500 BP shells (Figure 7), however, suggests that these specimens were also experiencing some form of stress that we hypothesize is due to either the rapid fluctuation in SSTs, or the warmer-than-ideal SSTs during this time (Figure 6). The abundance of early Holocene subfossil shells within the intertidal sediment column, and the absence of securely-dated archaeological sites from this time period,

28 suggests that these clam populations were not under significant predatory pressure by humans during this time.

There is a temporal gap in our clam samples from the period 9,500-4,200 BP. Sea level and SSTs remain stable during this time. We suggest that taphonomic factors are responsible for the lack of intertidal subfossil shells representative of this ~5,000-year period. In particular, we suspect that the persistent wave action caused by relatively stable sea level during this time (Crowell 2017; Fedje et al. 2017), coupled with a possible period of reduced sedimentation (e.g., Aitken and Bell 1998), resulted in increased sediment erosion and the displacement and deterioration of intertidal subfossil shells. We hypothesize that before and after this time, the steady decline in sea level (Fedje et al. 2017) resulted in less wave action at any given location, and that the regular influx of sediment caused dead clams to be buried more quickly making them more likely to be preserved in the intertidal sediment column (Claassen 1998).

By the late mid-Holocene, environmental conditions affecting butter clam growth continued to improve. These improvements include the further accumulation of gravel-sand substrates and stabilized SSTs (Table 2, Figure 6). These beneficial environmental conditions are visible in the trend of increased age at death, size at death, and growth rate of the 4,200- 2,900 BP intertidal subfossil butter clams as compared to those from the early Holocene (Figure 5a, b, e, f & 6). These apparently healthier, longer-lived clams were living under less stressful conditions than those preceding them. The large size of these clams were certainly appealing to human harvesters, who began to harvest clams in earnest by ~5,000 BP, as indicated by the preservation of shells within the middens (Puckett et al. 2014; Toniello et al. 2017). The abundance of long-lived clams in the intertidal subfossil record, however, suggests that this specific clam assemblage dating to 4,200-2,900 BP and originating from one particular beach (EbSh-77) 0.8km from the nearest village, was not under heavy predatory pressure by humans. We suspect that the absence of evidence for heavy human predation on this sampling population reflects the fact that humans were focusing their harvesting pressure on beaches immediately adjacent to village sites at this time.

After ~3,000 BP large village settlements increase in number in Kanish and Waiatt Bays, filling all inhabitable coastal landforms, and reflecting an increase in human populations (Toniello et al. 2017). As a result, clam harvesting intensified (M. Puckett and T. Crowell unpublished data; Toniello et al. 2017). Consequently, clam shells are not as prevalent in the intertidal palaeo-record, and instead are deposited in archaeological shell middens. The growth

29 rates of midden butter clam shells indicate that favourable butter clam growing conditions established by the mid-Holocene continued until the early historic period. SSTs remained within the preferred temperature range for butter clams, and at least some clam beaches offered sand- gravel and shell hash substrates (Table 2, Figure 6). These preferred conditions are reflected in the relatively steep growth curve of the 2,800-2,300 BP shell midden samples, which follows the same growth-rate as the 4,200-2,900 BP intertidal subfossil samples (Figure 4). The similarity in size and growth rates of the 2,800-2,300 BP midden samples and the intertidal subfossil shells immediately predating them suggests that humans were taking full advantage of this enhanced clam resource.

By ~2,000 years ago, human relationships with clams intensified through the construction of clam gardens (Neudorf et al. 2017; Smith et al. in prep). The construction of clam gardens was an on-going inter-generational process (Lepofsky et al. 2015; Neudorf et al. 2017), which ultimately resulted in an estimated 27-54% increase in the area of clam habitat in Kanish and Waiatt Bays. A significant portion of this increase in habitat (an estimated 10.3- 13.5%) occurred in areas that did not have clam habitat prior to human modification, such as on bedrock and boulder outcrops (Table 3 and Figure 9D). Most of the clam gardens within Kanish and Waiatt Bays were constructed on pre-existing clam habitat. Building clam gardens in these locations not only expanded the area of viable clam habitat, but also increased accessibility of clams to human harvesters by decreasing beach slope, increasing the amount of time the beach is exposed during low tide, and increasing the proximity of clams to village settlements (Deur et al. 2015). The act of investing time and energy into building harvestable clam habitat and creating clam habitats where none existed before is a sign of clam resource intensification and demonstrates the importance of this resource to growing human populations.

We found no significant differences between growth rates in clams living in pre-clam garden beaches (2,800-2,300 BP midden samples) and those from clam garden beaches (500- 200 BP midden samples). We expect that long-term harvesting of clams would show some effects of predatory pressure, such as a gradual decline in harvested clam size over time (cf. Broughton 2002, Butler 2001, Erlandson et al. 2008). Instead, our samples suggest that there is variability in growth, and that the size and growth rates of harvested clams appear to be relatively stable between these two contexts. Increased available clam habitat due to clam garden construction, combined with other management techniques (e.g., tilling, removal of non- human predators, altering substrate, rock removal), could be responsible for this stability during a time of increased harvest and growing numbers of human settlements. The size and

30 abundance of mature clams within the midden samples (79% of midden samples were sexually mature and over 50mm when harvested) (Figure 5a & b) further suggests that there was a culturally prescribed set of harvesting restrictions or preferences about clam size.

The suitability of clam habitats within the study area began to decline ~250 years ago, coinciding with the decrease in Indigenous populations after European contact. This is reflected in the trend of overall decline in butter clam growth rate in the 250-100 BP clam garden beach subfossil samples and the live-collected butter clam samples compared to those pre-dating them (Figure 4). This decline occurred despite the relative stability of intertidal environmental conditions since the mid-Holocene (Figure 6). We hypothesize that the lack of human management of clam resources contributed to the degradation of habitat and trend toward decreased growth of these recent samples (Figure 4, 6). In general, the decimation of Indigenous populations ~250 years ago not only resulted in far fewer clams being harvested, and thus a return of intertidal subfossil clam assemblages, but also a decline in clam habitat maintenance that was part and parcel of daily human-clam interactions.

The increase of industrial logging activities in the past ~100 years (Taylor 2009) resulted in the deposition of land-based silts that further stunted the growth of butter clam specimens living within clam garden beaches today. The relatively slower growth of these live-collected butter clams resulted in these specimens being smaller at age than any other temporal group in the past ~9,500 years. The growth curve of the live-collected specimens is most similar to those of the early Holocene samples (11,500-11,000 and 10,900-9,500 BP), which also grew in silt- rich substrates (Figure 4, 6). Our results suggest that without regular maintenance of clam beaches by humans, butter clams are especially vulnerable to the negative effects of contemporary industrial activities such as logging. In fact, the shallow slope of the clam garden beaches and the lip of the clam garden wall may be helping to trap the influx of fine silts from recent logging events. These results demonstrate that clam garden beaches today could be worse habitat for butter clams than the pre-industrial and human-managed beaches that existed within Kanish and Waiatt Bays in the past.

Previous findings by Groesbeck et al. (2014) show that butter clam productivity today, as viewed through density and biomass measurements, is heightened within clam garden habitats. Our analyses show that butter clam productivity within clam gardens today, as viewed through growth rate measurements, is less than clams living in tended clam gardens (500-200 BP midden), tended un-walled beaches (2,800-2,300 BP midden), and un-tended late mid-

31 Holocene (4,200-2,900 BP intertidal) beaches in the past. The contemporary butter clam populations are stressed and are not representative of the clams living within clam gardens in the past. We suggest, based on our analyses, that clam garden habitats that are traditionally managed and maintained are likely to produce clams that far exceed the increased productivity percentages reported by Groesbeck et al. (2014).

5.2. Historical Ecology of Butter Clams on N. Quadra Island

Our exploration of the historical ecology of humans and clams on Quadra Island illustrates the long-term intertwined histories of these two species. Clams were an important part of of the settlement history and subsistence practices in Kanish and Waiatt Bays, as indicated by the abundance of clam garden features and expansive shell middens. The intertidal subfossil assemblages extending back to the early Holocene provides the baseline data from which to assess how clam growth changes throughout time as the human-clam relationship intensifies and changes. We are able to recreate this history by using evidence from archaeological shell middens as well as from the intertidal subfossil record. Collectively, these sources of data reflect both Indigenous clam management and use, as well as the palaeo-ecological record of Holocene butter clams. Holocene-scale analyses allow us to see patterns in butter clam growth that has been difficult to detect with a narrower temporal focus (i.e., Groesbeck et al. 2014; Quayle and Bourne 1972). This long-term perspective also allows us to identify that modern estimates of butter clam maturation size reflect a shifted baseline. Taken together, our study emphasizes the value of an historical ecological approach that 1) fully integrates humans into ecological studies, and 2) takes into account the long-term histories in order to fully understand the role of humans within ecosystems.

Our data demonstrate how the relationship between humans and clams fluctuates over time. In the pre-contact era, from the middle Holocene onward, humans directly, and in many cases intentionally, manipulated the intertidal landscape through intensive clam harvesting and the creation of clam habitat. The construction of clam gardens by humans not only provided a desirable habitat for clams, but also expanded the available area in which clams live, ultimately increasing clam abundance and combatting resource depression. In turn, the increased availability of clams positively affected human populations by providing a stable food source that allowed human settlements to expand and thrive within Kanish and Waiatt Bays. The ability to

32 increase available habitat by constructing clam gardens, coupled with a conservation ethic embedded within the worldview of coastal First Nations (Deur et al. 2015; Turner 2005, 2014), resulted in stable clam resources during times of intense harvesting pressure. It is not until the post-European era that we see negative human impacts on clam populations from the lack of Indigenous traditional management practices coupled with industrial development.

The present-day intertidal landscape of Kanish and Waiatt Bays retains the legacy of millennia of Indigenous traditional resource management practices, but also reflects the absence of Indigenous people and practices since European contact. Today, many coastal First Nations have observed that clam beaches are unhealthy and less productive because of a lack of traditional management (Marlor 2009; Parks Canada 2011). Intertidal environments are at continued risk of habitat disruption from on-going industrial activities (logging, fishing), increased tourism, pollution, and climate change. Concerns about marine environment degradation and the loss of coastal resources have emerged in recent years, and efforts to manage, restore, and conserve coastal resources and improve biodiversity and food security have developed (Berkes 2015). Management strategies that incorporate local and traditional ecological knowledge (e.g., Augustine and Dearden 2014) as well as archaeological and palaeoecological data (e.g., Jackson et al. 2001; Rick et al. 2016) are key to the long-term sustainability and resilience of ecosystems (Berkes 2015; Berkes and Turner 2006; Berkes et al 2000; Jackley et al. 2016; Smith 2009;). Examining butter clams from a variety of ecological and cultural contexts throughout the Holocene provides useful data for the management of marine clam resources along the Pacific Northwest Coast, and beyond.

33 References

Ackerman, R.E., Reid, K.C., Gallison, J.D., & Roe, M.E. (1985). Archaeology of Heceta Island: A survey of 16 timber harvest units in the Tongass National Forest, Southeastern Alaska. Unpublished manuscript. (Rep. No. 3, 144). Pullman, WA: Center for Northwest Anthropology, Washington State University.

Aitken, Alec E., and Trevor J. Bell (1998). Holocene Glacimarine Sedimentation and Macrofossil Palaeoecology in the Canadian High Arctic: Environmental Controls. Marine Geology 145(3-4):151-171.

Anderson, Eugene N., Deborah M. Pearsall, Eugene Hunn, and Nancy Turner (Eds). (2011). Ethnobiology. Wiley-Blackwell, Hoboken, NJ.

Anderson, M. Kat. (2005). Tending the Wild: Native American Knowledge and the Management of California's Natural Resources. University of California Press, Berkley and Los Angeles, CA.

Armstrong, Chelsey Geralda., and James R. Veteto (2015). Historical Ecology and Ethnobiology: Applied Research for Environmental Conservation and Social Justice. Ethnobiology Letters 6:5-7.

Armstrong, Chelsey Geralda., Anna C. Shoemaker, Iain McKechnie, Anneli Ekblom, Péter Szabó, Paul J. Lane, Alex C. McAlvay, Oliver J. Boles, Sarah Walshaw, Nik Petek, Kevin S. Gibbons, Erendira Quintana Morales, Eugene N. Anderson, Aleksandra Ibragimow, Grzegorz Podruczny, Jana C. Vamosi, Tony Marks-Block15, Joyce K. LeCompte, Sākihitowin AwaÃsis, Carly Nabess, Paul Sinclair, Carole L. Crumley (2017). Anthropological Contributions to Historical Ecology: 50 Questions, Infinite Prospects. PloS one 12(2): e0171883.

Augustine, Skye, Philip Dearden (2014). Changing Paradigms in Marine and Coastal Conservation: A Case Study of Clam Gardens in the Southern Gulf Islands, Canada. The Canadian Geographer 58(3): 305–314.

Balée, William (2006). The Research Program of Historical Ecology. Annual Review of Anthropology 35:75-98.

Berkes, Fikret (2015). Coasts for People: Interdisciplinary Approaches to Coastal and Marine Resource Management. Routledge, New York, NY.

Berkes, Fikret, and Nancy J. Turner (2006). Knowledge, Learning and the Evolution of Conservation Practice for Social-Ecological System Resilience. Human Ecology 34(4):479.

Berkes, Fikret, Johan Colding, and Carl Folke (2000). Rediscovery of Traditional Ecological Knowledge as Adaptive Management. Ecological Applications 10(5):1251-1262.

Bernard, Frank R. (1983). Physiology and the Mariculture of Some Northeastern Pacific Bivalve Molluscs. Canadian Special Publication of Fisheries and Aquatic Sciences No. 63. Dept. of Fisheries and Oceans, Ottawa, ON.

Blais-Stevens, Andrée, B.D. Bornhold, A.E.S. Kemp, J.M. Dean, A.A. Vaan (2001). Overview of Late Quaternary Stratigraphy in Saanich Inlet, British Columbia: Results of Ocean Drilling Program Leg 169S. Marine Geology 174:3-26.

Braje, Todd J., Torben C. Rick, and Jon M. Erlandson (2012). A Trans-Holocene Historical Ecological Record of Shellfish Harvesting on California’s Northern Channel Islands. Quaternary International 264:109-120.

Broughton Jack M. (2002). Prey Spatial Structure and Behavior Affect Archaeological Tests of Optimal Foraging Models: Examples From the Emeryville Shellmound Vertebrate Fauna. World Archaeology 34:60–83.

Burchell, Meghan, Aubrey Cannon, Nadine Hallmann, Henry P. Schwarcz, and Bernd R. Schöne (2013). Inter-Site Variability in the Season of Shellfish Collection on the Central Coast of British Columbia. Journal of Archaeological Science 40(1):626-636.

Butler, Virginia L. (2001). Changing Fish Use on Mangaia, Southern Cook Islands: Resource Depression and the Prey Choice Model. International Journal of Osteoarchaeology 11:88–100.

Butman, Cheryl Ann, Judith P. Grassle, and Christine M. Webb. (1988). Substrate Choices Made by Marine Larvae Settling in Still Water and in a Flume Flow. Nature 333: 771-773.

Caldwell, Megan E., Dana Lepofsky, Georgia Combes, Michelle Washington, John R. Welch, and John R. Harper (2012). A Bird’s Eye View of Northern Coast Salish Intertidal Resource Management Features, Southern British Columbia, Canada. The Journal of Island and Coastal Archaeology 7: 219-233.

Cannon, Aubrey, and Meghan Burchell. (2016). Reconciling Oxygen Isotope Sclerochronology with Interpretations of Millennia of Seasonal Shellfish Collection on the Pacific Northwest Coast. Quaternary International 427(A):184-191.

Cannon, Aubrey, and Meghan Burchell. (2009). Clam Growth-Stage Profiles as a Measure of Harvest Intensity and Resource Management on the Central Coast of British Columbia. Journal of Archaeological Science 36: 1050-1060.

Claassen, Cheryl. (1998). Shells. Cambridge University Press, Cambridge, UK.

Crowell, Travis. (2017). Settlement and Sea-Level Histories: Refining Holocene Sea-Levels in Kanish and Waiatt Bays, Quadra Island. In progress Master’s thesis, Department of Archaeology, Simon Fraser University, Burnaby, BC.

Crumley, Carole. L. (editor). (1994). Historical Ecology: Cultural Knowledge and Changing Landscapes. School of American Research Press, Santa Fe, NM.

Dame, Richard F. (2012). Ecology of Marine Bivalves: An Ecosystem Approach. CRC press, Boca Raton, FL. de Lumley, H. (1969). A Paleolithic Camp at Nice. Scientific American 220: 42–50.

35 Deur, Douglas, and Nancy J. Turner (editors). (2005). Keeping it Living: Traditions of Plant Use and Cultivation on the Northwest Coast of North America. University of Washington Press, Seattle, WA.

Deur, Douglas, Adam Dick, Kim Recalma-Clutesi, and Nancy J. Turner. (2015). Kwakwaka’wakw “Clam Gardens”: Motive and Agency in Traditional Northwest Coast Mariculture. Human Ecology 43(2):201-212.

Ellis, David W., and Solomon Wilson (1981). The Knowledge and Usage of Marine Invertebrates by the Skidegate Haida People of the Queen Charlotte Islands. Monograph Series No. 1. The Queen Charlotte Islands Museum Society, Skidegate, British Columbia.

EPA (U.S. Environmental Protection Agency) (2005). Environmental Monitoring and Assessment Program (EMAP) Database Benthic Grab Information, the Benthic Replicate Abundance data. Encyclopedia of Life webpage, accessed September 2016

Eppley Richard W. (1972). Temperature and Phytoplankton Growth in the Sea. Fish Bull 70(4):1063–1085.

Erlandson, Jon M., and Torben C. Rick. (2010). Archaeology Meets Marine Ecology: The Antiquity of Maritime Cultures and Human Impacts on Marine Fisheries and Ecosystems. Annual Review of Marine Science 2:231-251.

Erlandson, Jon M., Torben C. Rick, and Todd J. Braje (2009). Fishing up the Food Web?: 12,000 Years of Maritime Subsistence and Adaptive Adjustments on California's Channel Islands. Pacific Science 63(4):711-724.

Erlandson, Jon M., Torben C. Rick, Todd J. Braje, Alexis Steinberg, and Rene L. Vellanoweth (2008). Human Impacts on Ancient Shellfish: a 10,000 Year Record from San Miguel Island, California. Journal of Archaeological Science 35(8):2144-2152.

Fedje, Daryl, Jenny Cohen, Rob Commisso, and Nicole Smith. (2011). 2010/11 Gwaii Haanas Archaeology Programme. Report on file with the British Columbia Archaeology Branch, Victoria, BC.

Fedje, Daryl, Duncan McLaren, Tom James, and Quentin Mackie. (2017 in prep). A Revised Sea Level History for Quadra Island on the West Coast of Canada.

Garibaldi, Ann, and Nancy J. Turner (2004). Cultural Keystone Species: Implications for Ecological Conservation and Restoration. Ecology and Society 9(3).

Goong, Stuart Alan, and Kenneth K. Chew. (2001). Growth of Butter Clams, Saxidomus Giganteus Deshayes, on Selected Beaches in the State of Washington. Journal of Shellfish Research 20(1):143-147.

Groesbeck, Amy, Kirsten Rowell, Dana Lepofsky, and Anne K. Salomon. (2014). Ancient Clam Gardens Increased Production: Adaptive Strategies from the Past Can Inform Food Security Today. PloS one 9(3).

Green, Mark A., George G. Waldbusser, Lane Hubazc, Eric Cathcart, and Joshua Hall. (2013).

36 Carbonate Mineral Saturation State as the Recruitment Cue for Settling Bivalves in Marine Muds. Estuaries and Coasts 36(1):18-27.

Hallmann, N., Burchell, M., Schöne, BR, Brewster, N., Martindale, A. (2013). Holocene Climate and Shellfish Collection at the Dundas Island Group, Northern British Columbia, Canada: A Bivalve Sclerochronological Approach. Palaeogeography, Palaeoclimatology, Palaeoecology 373(1):163-172.

Harbo, Rick M. (1996). Shells & Shellfish of the Pacific Northwest: A Field Guide. Harbour Publishing, Madiera Park, BC.

Harper, John R. (2007). Clam Garden Field Inventory. Contract Report by Coastal and Ocean Resources Inc., Sidney, British Columbia to the Hamtla Treaty Society. Report on file, Hamtla Treaty Society, Campbell River, British Columbia.

Harper, John R., and Mary C. Morris. (2004). Quadra Island Clam Garden Survey. Contract Report by Coastal & Ocean Resources Inc. Sidney, British Columbia to the Hamatla Treaty Society. Report on file, Hamatla Treaty Society, Campbell River, British Columbia.

Harris, Cole (1994) Voices of Disaster: Smallpox around the Strait of Georgia in 1782. Ethnohistory 41(4):591-626.

Hetherington, Renée, and Robert GB Reid. (2003). Malacological Insights into the Marine Ecology and Changing Climate of the Late Pleistocene Early Holocene Queen Charlotte Islands Archipelago, Eastern Canada, and Implications for Early Peoples. Canadian Journal of Zoology 81(4)):626-661.

Holmes, Keith. (2016). The Great Walls of Quadra: Mapping Quadra Island’s Monumental Clam Gardens, Which Line Over One-third of the Coastline. In: Hakai Blog, webpage. Accessed April 2017.

Jackley, Julia, Lindsay Gardner, Audrey F. Djunaedi, and Anne Salomon (2016). Ancient Clam Gardens, Traditional Management Portfolios, and the Resilience of Couple human- Ocean Systems. Ecology and Society 21(4).

Jackson, Jeremy B. C., Michael X. Kirby, Wolfgang H. Berger, Karen A.Bjorndal, Louis W. Botsford, Bruce J. Bourque, Roger H. Bradbury, Richard Cooke, JonErlandson, James A. Estes, Terence P. Hughes, Susan Kidwell, Carina B. Lange, Hunter S.Lenihan, John M. Pandolfi, Charles H. Peterson, Robert S. Steneck, Mia J. Tegner andRobert R. Warner. (2001). Historical Overfishing and the Recent Collapse of Coastal Ecosystems. Science 293(5530):629-637.

Kidwell, Susan M. (2015). Biology in the Anthropocene: Challenges and Insights From Young Fossil Records. Proceedings of the National Academy of Sciences 112(16):4922-4929.

Kidwell, Susan M. (2002). Time-averaged Molluscan Death Assemblages: Palimpsests of Richness, Snapshots of Abundance. Geology 30(9):803-806.

Kidwell, Susan M. (2001). Preservation of Species Abundance in Marine Death Assemblages. Science 294(5544):1091-1094.

Kienast, Stephanie S., and Jennifer L. McKay. (2001). Sea Surface Temperatures in the Subarctic Northeast Pacific Reflect Millennial-Scale Climate Oscillations During the Last 16kyrs. Geophysical Research Letters 28(8):1563-1566.

Kvenvolden, Keith A., David J. Blunt, and H. Edward Clifton. (1979). Amino-acid Racemizarion in Quaternary Shell Deposits at Willapa Bay, Washington. Geochimica et Cosmochimica Acta 43(9):1505-1520.

Lepofsky, Dana, and Megan E. Caldwell. (2013). Indigenous Marine Resource Management on the Northwest Coast of North America. Ecological Processes 2:1–12.

Lepofsky, Dana, Nicole F. Smith, Nathan Cardinal, John Harper, Mary Morris, Gitla (Elroy White), Randy Bouchard, Dorothy I.D. Kennedy, Anne K. Salomon, Michelle Puckett, Kirsten Rowell, and Eric McLay (2015). Ancient Shellfish Mariculture on the Northwest Coast of North America. American Antiquity 80(2):236-259.

Marlor, Chantelle. (2009). Ways of Knowing: Epistemology, Ontology, and Community Among Ecologists, Biologists and First Nations Clam Diggers." Ph.D. Dissertation, Department of Sociology, Rutgers, New Jersey.

Martindale, Andrew, Bryn Letham, Duncan McLaren, David Archer, Meghan Burchell, and Bernd R. Schöne. (2009). Mapping of Subsurface Shell Midden Components Through Percussion Coring: Examples From the Dundas Islands. Journal of Archaeological Science 36(7):1565-1575.

McKechnie, Iain, and Madonna L. Moss.(2016). Meta-Analysis in Zooarchaeology Expands Perspectives on Indigenous Fisheries of the Northwest Coast of North America. Journal of Archaeological Science: Reports 8:470-485.

Minnis, Paul E., and Wayne J. Elisens, (eds.). (2000). Biodiversity and Native America. University of Oklahoma Press, Norman, OK.

Moss, Madonna L. (1993). Shellfish, Gender, and Status on the Northwest Coast of North America: Reconciling Archaeological, Ethnographic, and Ethnohistorical Records of the Tlingit. American Anthropologist 93(3):631-652.

Moss, Madonna L., Jon M. Erlandson, and Robert Stuckenrath. (1990). Wood Stake Weirs and Salmon Fishing on the Northwest Coast: Evidence from Southeast Alaska. Canadian Journal of Archaeology/Journal Canadien d'Archéologie 14:143-158.

Munroe, Daphne, S Borsetti, K Ashton-Alcox, D Bushek. (2016). Early Post-Settlement Growth in Wild Eastern Oyster (Crassostrea virginica Gemlin 1791) Populations. Estuaries and Coasts 40:880–888.

Nakaoka, Masahiro. (2000). Nonlethal Effects of Predators on Prey Populations: Predator- mediated Change in Bivalve Growth. Ecology 81(4):1031-1045.

Neudorf, Christina M., Nicole Smith, Dana Lepofsky, Ginevra Toniello, Olav B. Lian. (2017). Between a Rock and a Soft Place: Using Optical Ages to Date Ancient Clam Gardens on the Pacific Northwest. PLoS ONE 12(2):e0171775.

Orford, Julian D., Donald L. Forbes, and Simon C. Jennings. (2002). Organisational Controls, Typologies and Time Scales of Paraglacial Gravel-Dominated Coastal Systems. Geomorphology 48(1): 51-85.

Parks Canada. (2011). Gulf Islands National Park Reserve— Hulq’umi’num Treaty Group Shellfish Traditional Knowledge Research Project Report. Manuscipt on file, Parks Canada, Sidney, British Columbia.

Puckett, Michelle, Dana Lepofsky, Terry Clark, Sue Formosa, Megan Caldwell, and Louise Williams. (2014). Quadra Island Clam Garden Study: 2010-2013 Fieldwork in Kanish and Waiatt Bays HCA Inspection Permit No. 2010-0251. Unpublished report on file with British Columbia Archaeology Branch, Victoria, BC.

Quayle, D.B., and N. Bourne. (1972). The Clam Fisheries of British Columbia. Fisheries Research Board of Canada, Ottawa, Canada.

Rick, Torben C., Leslie A. Reeder-Myers, Courtney A. Hofman, Denise Breitburg, Rowan Lockwood, Gregory Henkes, Lisa Kellogg, Darrin Lowery, Mark W. Luckenbach, Roger Mann, Matthew B. Ogburn, Melissa Southworth, John Wah, James Wesson, and Anson H. Hines. (2016). Millennial-Scale Austainability of the Chesapeake Bay Native American Oyster Fishery. Proceedings of the National Academy of Sciences, 201600019.

Roegner G. Curtis. (2000). Transport of Molluscan Larvae through a Shallow Estuary. Journal of Plankton Research 22(9):1779–1800.

Rowell, Kirsten, Karl W. Flessa, David L. Dettman, Martha J. Román, Leah R. Gerber, and Lloyd T. Findley. (2008). Diverting the Colorado River Leads to a Dramatic Life History Shift in an Endangered Marine Fish. Biological Conservation 141(4):1138-1148.

Ryder, June, Robert Fulton, and John Clague. (1991). The Cordilleran Ice Sheet and the Glacial Geomorphology of Southern and Central British Colombia. Géographie Physique et Quaternaire 45(3):365-377.

Schöne, Bernd R., Elena Dunca, Jens Fiebig, Miriam Pfeiffer. (2005). Mutvei’s Solution: An Ideal Agent for Resolving Microgrowth Structures of Biogenic Carbonates. Palaeogeography, Palaeoclimatology, Palaeoecology 228:149– 166.

Smee, Delbert L., and Marc J. Weissburg. (2006). Hard Clams (Mercenaria mercenaria) Evaluate Predation Risk Using Chemical Signals from Predators and Injured Conspecifics. Journal of Chemical Ecology 32(3):605-619.

Smith, Bruce D. (2009). Resource Resilience, Human Niche Construction, and the Long-term Sustainability of Pre-Columbian Subsistence Economies in the Mississippi River Valley Corridor. Journal of Ethnobiology 29(2):167-183.

Smith, Nicole F., Dana Lepofsky, Ginevra Toniello, Christina M. Neudorf, Louie Wilson, Christine Roberts (in prep.) Dating Clam Gardens on the Northwest Coast of North America.

Taylor, Jeanette. (2009). The Quadra Story: A History of Quadra Island. Ed. Pamela Robertson.

39 Harbour Publishing Co. Ltd., Madeira Park, BC.

Toniello, Ginevra, Travis Crowell, Dana Lepofsky, Bryn Letham, Nicole F. Smith, Christina Neudorf, Christine Roberts, Louie Wilson. (2017). Quadra Island Clam Garden and Archaeological Settlement Project 2015-2016 Fieldwork in Kanish and Waiatt Bays. Unpublished report on file with British Columbia Archaeology Branch, Victoria, BC.

Turner Nancy J., and Douglas Deur. (2005). Keeping It Living Traditions of Plant Use and Cultivation on the Northwest Coast of North America. Vancouver: University of British Columbia Press.

Turner, Nancy J. (2014). Ancient Pathways, Ancestral Knowledge: Ethnobotany and Ecological Wisdom of Indigenous Peoples of Northwestern North America. Vol. 74. McGill-Queen's University Press, Montreal, QC.

Turner, Nancy J. (2005). The Earth's Blanket: Traditional Teachings for Sustainable Living. D & M Publishers Inc., Vancouver, BC.

Von Bertalanffy, Ludwig. (1938). A Quantitative Theory of Organic Growth (Inquiries on Growth Laws. II). Human Biology 10(2):181-213.

Williams, Judith. (2006). Clam Gardens: Aboriginal Mariculture on Canada's West Coast (Vol. 15). Transmontanus.

Woods, D. J., and D. Woods (producers). (2005). Ancient Sea Gardens. Mystery of the Pacific Northwest. DVD. Directed by Aaron Szimanski. Aquaculture Pictures, Toronto, Ontario.