Final project

C’aira Cassavant Anth 368 Final Project

 What did the prehistoric diet of people on Tutuila (American Samoa) consist of 1,000 years ago?

 Has it changed over time? Or remained relatively stable?

 Both ethnographic and archaeological studies have been conducted to answer this question

 George Turner (1989[1884]:105) stated that 5 food items made life possible on Tutuila:  Bread-fruit  Taro  Yams  Bananas  Cocoa-nuts

 For additional nourishment the people of Tutuila turned to the lagoons and reefs which provided a large supply of fish and shellfish

 And “occasionally all, but especially persons of rank, regaled themselves on pigs, fowls, and turtles” (Turner 1989[1884]:105).

 Watters (1958) similarly emphasized 6 cultivated staple vegetable foods:  Taro  Yams  Ta’amu (Alocasia macrorrhizos)  Coconut  Bread-fruit  Banana

 With taro and bread-fruit being the most popular resources

 Their diet was supplemented with resources from the sea and forest:  Lagoon fish: mullet and mackerel  Pelagic fish: bonito and shark  Shellfish  Wild birds: pigeon, dove, etc.

Taro

Ta’amu (Watters 1958:340)

 Several archaeological studies have been conducted on Tutuila by:

 Green (2002)  Morrison and Addison (2008,2009)  Addison et al. (2010)  Burley and Clark (2003)

 Green’s evidence suggests continuity in subsistence patterns and a heavy reliance on marine resources in both early and late sites  However the faunal assemblages also included rat, bird, turtle, chicken, and pig

 Morrison and Addison also suggest stable marine resource use that spans across multiple islands in American Samoa

 Based on archaeological evidence these researchers all suggest a somewhat consistent/stable diet in Tutuila over the last 1000 years

 However, they also agree that more sites need to be examined in order to fully understand subsistence patterns on Tutuila and what role other resources or practices played on these patterns (i.e. domestication, agricultural practices, climate change and socio/cultural modifications)

(Valentin et al. 2011:474)

 To effectively answer the question of what the diet on Tutuila consisted of over the last 1000 years, the authors of this paper turn to isotopic data

 Using AMS 14C dating and Carbon (C) and Nitrogen (N) stable isotope data measured on bone collagen, they assess the dietary patterns of prehistoric communities on Tutuila

(Valentin et al. 2011:473)

 Collagen is a protein formed mainly from food proteins

 It has the ability to help reconstruct food- consumption patterns:  Using an analysis of the stable Carbon (C) and

Nitrogen (N) isotopic compositions found within bones we can observe dietary intakes of protein

 According to Hedges et al. (2007) the data gathered from an adult bone reveals protein intake for approximately the last decade of life

 The two specific dietary components/signatures examined in this paper are:  δ13C  δ15N

 Because their ratios characterize specific foods it is possible to qualitatively reconstruct this prehistoric diet

(Valentin et al. 2011:474)

 This investigation used the remains from 15 human burials previously excavated by Addison and Asaua (2006) from the Tutuila villages of  Ili’ili (II)

▪ 1 adult burial

 Lauli’i (LL) ▪ 1 child burial ▪ 6 adult burials

 Fatu-ma-Futi (FF DR) ▪ 2 infant burials ▪ 5 adult burials

 The analysis was further divided into three time periods that roughly span ~1000-200 cal B.P.:  Period 1: ~1500-1000 cal B.P. (coincides with the

abandonment of pottery and the end of the “Samoan Dark Ages”)(Valentin 2011:475)

 Period 2: ~1000 B.P.-historic period (corresponds with large scale tool export and the construction of larger settlements)

 Period 3: Historic period (which began after Western contact in A.D. 1722)

(Valentin 2011:475)

II

FF DR

LL

 Each of the 15 sets of remains were analyzed using:  Accelerator Mass Spectrometry

(AMS) Radiocarbon Dating ▪ C14 is an isotope of carbon with 6 protons

and 8 neutrons; making it unstable and radioactive

▪ Once a plant or animal dies it ends its oxygen/carbon relationship with the environment and C14 begins to decay at a known rate

▪ Radiocarbon dating allows you to compare the amount of C14 in a dead plant/animal to the available carbon in the atmosphere to determine when it died

▪ AMS uses kinetic energy to rapidly move ions and measures atomic weight to detect or count the number of Carbon 14 atoms within a sample

(Hirst 2016)

 Carbon 13 is a stable isotope of Carbon with 6 protons and 7 neutrons that does not change over time (unlike the radioactive C14)

 There is a constant ratio in earths atmosphere between C12 and C13 (100 C12 atoms to 1 C13 atom)  Through the process of photosynthesis

carbon from the atmosphere gets absorbed into the surrounding flora

 Eventually these plants get eaten by animals/humans and the carbon signatures get passed onto them

 Therefore, by determining the ratio between C12 and C13 in an animals bones/teeth you can determine the types of plants it ate during its lifetime ▪ The amount of C12 atoms compared to

C13 atoms tells you if the animal came from a region with lots of sun and little rainfall, or a wetland, or forest

 Nitrogen 15 is a rare, stable isotope of Nitrogen with 7 protons and 8 neutrons that also does not change over time

 Like Carbon, Nitrogen is absorbed by plants and can be measured in proteins in the body (i.e. bone collagen)

(Hirst 2016)

 In order to extract protein data from the remains the samples were pre-treated to a gelatin form  First the bones were acid washed (decalcified) in a

2% HCl solution to isolate the bone proteins (or collagen)

 Then the ‘acid insoluble’ collagen was gelatinized by heat in slightly acidic water (pH=3 at 90○C)

 After 4 hours the gelatin is freeze-dried and C and N isotopes measurements can be taken

 Quality assurance criteria was used to verify the preservation of the collagen to ensure it would be representative of the individuals diet  Gelatin yield: > 2%  C content: > 30 weight percent (wt. %)  N content: ≥ 11 wt. %  C:N ratio: between 3.1-3.5

(Valentin et al. 2011:475) (van Klinken 1999:691)

 This study utilized already published isotopic plant and animal data to determine what the prehistoric diet on Tutuila was comprised of

 The data were collected from locations with a similar environment to Tutuila including Fiji, Solomon Islands, Cook Islands, Australia, Marquesas Islands, New Britain, Bahamas, and New Ireland  However Valentin et al. does briefly

note that these resources might not be ideal for comparison because local conditions may affect isotopic signatures

 Using the δ13C and δ15N signatures from the published resources 8 dietary groups were formed  1: terrestrial C3 plants  2: terrestrial C4 plants  3: terrestrial animals  4: freshwater fish  5: coastal products (mollusks, Crustacea,

macro-algae)  6: coral-reef fish  7: non-reef marine fish  8: marine mammals

 From here a computer program was used to help determine the proportions of these food groups found within a given sample

(Valentin et al. 2011:475-476)

 Of the 15 sets of remains:  1 (LL2)had no surviving collagen

▪ Therefore no data could be collected

 1 (FF DR6) had C and N percentages that fell below the quality assurance recommendations ▪ However the gelatin yield and C:N ratios were

within satisfactory ranges so it was used in the overall evaluation

 The 2 infants (FF DR5 and FF DR6) were not used in the diet reconstruction because they were most likely breastfed until their death and would not represent the true diet on Tutuila

(Valentin et al. 2011:478)

 Fatu-ma-Futi:  5 individuals (including both infants) dated

to ~1000 years ago  2 individuals dated to ~500 years ago  δ13C range: -19.4% to -17.2%  δ15N range: 9.1% to 14.3%  %C range: 24% to 42.7%  %N range: 8.4% to 14.8%  C:N ratio average: 3.3

 Lauli’i:  Child burial had no surviving collagen  6 adult individuals dated to ~500 years ago  δ13C range:-18.4% to -17.7%  δ15N range:10.4% to 12.3%  %C range: 38.3% to 44.1%  %N range: 14.5% to 16.5%  C:N ratio average: 3.1

 Ili’ili:  Single individual post dates 260 cal. BP

(Historic Period)  δ13C: -17.5  δ15N: 10.5%  %C: 43%  %N: 14.9%  C:N ratio: 3.4

(Valentin et al. 2011:477-478)

Ili’ili

Lauli’i

Fatu-ma-Futi

Burial # Site Age Calibrated Age Yield (%) %C %N C/N Ratio δ13C (%) δ15N (%)

FF DR 7 Fatu-ma-Futi Adult 1067±31 BP 4.4 34.6 12.4 3.3 -17.5 11.9

FF DR 5 Fatu-ma-Futi 2 yrs old 1065±34 BP 3.2 42.7 14.8 3.4 -18.1 12.8

FF DR 6 Fatu-ma-Futi 2 yrs old 1045±31 BP 3.4 24 8.4 3.4 -17.2 14.3

FF DR 1 Fatu-ma-Futi Adult 1037±31 BP 3.5 33.4 12 3.3 -17.9 10.2

FF DR 8 Fatu-ma-Futi Adult 1007±37 BP 3.6 39.5 14.8 3.1 -17.7 11.1

FF DR 3 Fatu-ma-Futi Adult 516±33 BP 4.3 36.5 12.8 3.3 -19.4 9.1

FF DR 2 Fatu-ma-Futi Adult 475±31 BP 5.0 37 13.2 3.3 -18.1 9.5

LL 4 Lauli’i Adult 499±33 BP 4.4 38.7 14.7 3.1 -18.2 10.6

LL 3 Lauli’i Adult 445±33 BP 5.9 44.1 16.5 3.1 -18.4 11.2

LL 6 Lauli’i Child 7-8 yrs

old

469±32 BP 4.2 39.2 14.6 3.1 -18.1 11.5

LL 1 Lauli’i Adult 441±34 BP 5.9 41.9 15.7 3.1 -17.7 12.3

LL 7 Lauli’i Adult 420±34 BP 3.6 39.6 14.6 3.2 -18.2 11.4

LL 5 Lauli’i Adult 410±32 BP 4.2 38.3 14.5 3.1 -18.4 10.4

LL 2 Lauli’i Child 7-8 yrs

old

No surviving

collagen

- - - - - -

II 1 Ili’ili Adult 244±31 BP 4.4 43 14.9 3.4 -17.3 10.5

(Valentin et al. 2011:478)

 The Fatu-ma-Futi and Lauli’i sites both showed a relatively homogeneous distribution of resources and no statistical difference between sites

 The Ili’ili remains fall into the range of variability provided by the other two sites

 ~1000 years ago:  Carbon stable isotopes ranged from:

-18.1 to -17.2%  Nitrogen stable isotopes ranged from:

10.2 to 14.3%

 By ~500 years ago this changed:  Carbon: -19.4 to -17.7%  Nitrogen: 9.1 to 12.3%

 In this sample, Carbon values were significantly higher in older individuals than younger

 However, the Ili’ili remains – dated to the historic period – had the highest C values of all  A comparison with previous studies

shows that these signatures resemble those found on Fiji

 A majority of the samples show higher δ15N values than δ13C values  This could indicate that a high amount

of terrestrial foods, rich in 15N comprised their diet

 Another comparison indicated that the calculated C values of all the remains resemble diets from sites in Sigatoka Valley, Viti Levu, and Fiji

(Valentin et al. 2011:477-478)

>

 The results of this analysis have provided enough data to attempt a reconstruction of dietary patterns over the last 1000 years

 Based on the results taken from the 12 useable samples their diet consisted of:  100% terrestrial C3 plants:

▪ Including root crops, bananas, and coconut

 Terrestrial animals: ▪ Terrestrial endemic and domesticated animals (i.e. rats,

pigs, dogs, reptiles)

 Freshwater products: ▪ Freshwater animals (i.e. eels)

(Valentin et al. 2011:478)

 According to 13C and 15N signatures this diet was supplemented with either marine resources or C4 plant items  As marine resources and C4 plants have very similar isotopic

signatures it is difficult to determine which is which

 However, the isotopic signatures identified in this analysis suggest that dietary content on Tutuila was more influenced by marine foods/products

 Due to the difficulty of detecting certain low protein foods using a bone collagen protein analysis, two “concentration-dependent” models were developed to determine what marine resources were most important:  Model 1  Model 2

(Valentin et al. 2011:479)

MODEL 1

 This diet represented:  C3-plants

 Coastal products

 Non-reef-marine fish

 Non-reef-marine fish have higher δ15N values

 The results indicate that 9.3% of dietary content was comprised of non- reef fish

MODEL 2

 This diet represented:

 C3-plants

 Coastal products

 Coral-reef-fish

 Coral-reef fish have lower δ15N values

 The consumption of this coral-reef marine food would have lead to lower 15N enrichment

 Results indicate that 17.7% of dietary content was comprised of reef fish

(Valentin et al. 2001:479)

 Both Models 1 & 2 emphasize the importance of C3 plants in this diet

 According to the results of the isotopic signatures the diet on Tutuila was more influenced by marine products with lower δ15N values (i.e. coral-reef fish)

 Two marginal individuals seem to ‘bracket’ the average diet consumed on Tutuila  FF DR3:

▪ Diet characterized by 80% C3-plants

 II1: ▪ Diet characterized by 30% marine products ▪ Low δ15N values

 Based on these results there seems to be a wide diversity of dietary contributions here

Food Sources Tutuila

(N=12)

Marginal

Individual:

FF DR 3

Marginal

Individual:

II 1

Model 1

Terrestrial C3-

Plants

72.0% 82.% 66.2%

Non Marine

Reef Fish

9.3% 2.6% 9.6%

Coastal

Products

18.8% 15.1% 24.2%

Model 2

Terrestrial C3-

Plants

75.4% 83.3% 69.7%

Coral Reef Fish 17.7% 5.0% 18.4%

Coastal

Products

6.9% 11.7% 11.9%

(Valentin et al. 2011:480)

 Valentin et al. believes there is enough potential evidence to comment on changes in diet over time  They state that “statistically significant

carbon isotopic shifts between periods hint at possible changes in diet through time on Tutuila” (Valentin et al. 2011:480)

 Their findings indicate that individuals who lived 1000 years ago display a less terrestrial diet than those who lived 500 years ago ▪ This means individuals who lived 1000 years ago

more heavily relied on marine products ▪ A trend can be seen moving toward lower

trophic levels and a lower reliance on marine resources over time

 Valentin et al. does note that further studies are needed in order to confirm these findings

 The authors also offer other explanations that could account for these significant isotopic shifts including:  Climate change/influences  Marine resource depletion  Change in social or cultural practices  Subsistence needs  Landscape use  Or a combination of these factors

(Valentin et al. 2011:481-482)

 Valentin et al. conclude that the average diet on Tutuila 1,000 years ago consisted of :  Terrestrial proteins (C3 plants/vegetables and animals) supplemented with

coastal reef marine resources  Which indicates an already heavy reliance on terrestrial products by 1000 B.P.

 Over time subsistence strategies began to change and marine resources were utilized less frequently in favor of more terrestrial resources (for whatever reason)

 At first look – and according to previous ethnographic and archaeological investigations- it can seem as though the diet on Tutuila has not changed over the last 1,000 years however evidence presented here challenges that idea

 The authors have created a ‘working hypothesis’ and believe they have detected an evolution in dietary strategies on Tutuila but want more research to be conducted in order to solidify their findings

(Valentin et al. 2011:482)

 Addison, D., and Asaua, T.

2006. 100 New Dates From Tutuila and Manu’a: Additional Data Addressing Chronological Issues in Samoan Prehistory. Journal of Samoan History 2:95-119.

 Hedges, R.E.M., Clement, J.G., Thomas, C.D.L., O’Connell, T.C.

2007. Collagen Turnover in the Adult Femoral Mid-Shaft: Modelled from Anthropogenic Radiocarbon Tracer Measurements. American Journal of Physical Anthropology 133:808-816.

 Hirst, K.

2016 (May. 24) AMS Accelerator Mass Spectrometry Radiocarbon Dating. http://archaeology.about.com/od/amthroughanterms/g/ams_radiocarbon.htm

 Hirst, K.

2016 (May. 24) Stable Isotope Analysis in Archaeology- A Plain English Introduction. http://archaeology.about.com/od/amthroughanterms/g/ams_radiocarbon.htm

 Petchey, F. and Higham T.F.G.

2000. Bone Diagenesis and Radiocarbon Dating of Fish Bone at the Shag River Mouth Site, New Zealand. Journal of Archaeological Science 27:135-50.

 Turner, G.

1989 [1884]. Samoa a Hundred Years Ago and long Before. London Missionary Society, MacMillion and Co, London, United Kingdom, 1989. 3rd reprint by Institute of Pacific Studies, University of the South Pacific, Suva, Fiji.

 Valentin, F., Herrscher, E., Petchey, F., and Addison, D.

2011. An analysis of the last 1000 years human diet on Tutuila (American Samoa) using carbon and nitrogen stable isotope data. American Antiquity Vol. 76, No. 3, pp 473-486.

 van Klinken, G. J.

1999. Bone Collagen Quality Indicators for Paleodietary and Radiocarbon Measurements. Journal of Archaeological Science 26:687-695.

 Watters, R.F.

1958. Cultivation in Old Samoa. Economic Geography 34(4):338-351.