Surveying and Excavating Ancient Samoan Agriculture

We’ve been back in Samoa for a few weeks now surveying in rock walls, mounds and other likely agricultural features, some of which are identified on lidar map by Dr Seth Quintus, as part of our Marsden funded project investigating agricultural variation and socio-political complexity in the past.

Survey involves transects across the landscape, recording characteristics of features to add to the lidar and GPS based GIS that we are building. We use these data to detect and explain patterns of agricultural infrastructure over time.

Alex Queenin recording a stone wall. Many of these features, like this one, have had the ancient stones removed to make new walls.

We have also begun excavating features recorded last year. Our primary goal is to recover datable material, so we can begin to build a chronology of construction in the Falefa Valley. We will add more features and different types of features as we continue this year and next.

A rough 3D model of a walled walkway stretching across several hundred metres of the Falefa Valley.

3D model of the walled walkway after excavation of layer 1, feature rocks and associated sediment.

3D model after excavation of layer 2, a deposit associated with feature construction with some feature rocks embedded within it.

3D model of the base of excavation showing continuation of layer 3, a more orange sediment that had charcoal flecking at the top, predating wall construction and with abundant natural eroding basalt boulders and cobbles.

We’re getting great dating material from these excavations, so I’m confident are analyses and results will have much to contribute to Samoan archaeology and studies of agriculture and social complexity.

Hiding out from the rain under a banana leaf.

What explains the differences in the ancient socio-politics of Samoa and Tonga?

Why did Tonga form a single paramountcy (later a tripartite one) while Samoa became several (presumably competing) chiefdoms? We are pretty sure the colonising/early demography of Tonga and Samoa was different (Rieth and Cochrane 2012; Cochrane et al. 2013, 2016; Cochrane and Rieth 2016; Cochrane 2018), so is this early difference key? Does this continue with later demographic patterns in the two archipelagos? Are these differences related to variation in agriculture and other subsistence strategies between Tonga and Samoa (Quintus and Cochrane 2017, 2018)?

After surveying a gigantic mound in Saoluafata (Pule, Aumua Matiu, Ethan). February 2019

How can we begin to tackle these questions through archaeology?

  1. Our Marsden funded research is a start. With this project we are investigating the development of agriculture in the Falefa Valley, Samoa, focusing on the timing of field development and drainage ditch systems, environmental variation, human-induced changes and the rise of corporate or monumental structures.
  2. Agricultural terraces are a big question–ie, there are lots of terraces on the lower slopes of mountains bounding Falefa and it is not clear if these were agricultural of for some other purpose. And of course the use of features can change.
  3. Identifying variation in socio-politics is part of the problem as well. Having Just read Quintus’ and Clark’s (2019) pigeon mound article make me think of…mounds. What could a survey of mounds, description and classification of these structures in Falefa tell us? What evolutionary and transmission processes explain their variation? Can they tell us about changing socio-politics, group formation in Samoa?

Rieth, T. M., & Cochrane, E. E. (2012). Archaeological Monitoring and Data Recovery in Support of the Federal Emergency Management Agency Permanent Housing Construction Program, Tutuila Island, American Samoa. Retrieved from International Archaeological Research Institute, Inc. Honolulu, Hawaii: 

Cochrane, E. E., Rieth, T. M., & Dickinson, W. R. (2013). Plainware ceramics from Sāmoa: Insights into ceramic chronology, cultural transmission, and selection among colonizing populations. Journal of Anthropological Archaeology, 32(4), 499-510.

Cochrane, E. E., Kane, H., Fletcher, C., Horrocks, M., Mills, J., Barbee, M., . . . Tautunu, M. M. (2016). Lack of suitable coastal plains likely influenced Lapita (~2800 cal. BP) settlement of Sāmoa: Evidence from south-eastern ‘Upolu. The Holocene, 26(1), 126-135.

Cochrane, E. E., & Rieth, T. M. (2016). Sāmoan artefact provenance reveals limited artefact transfer within and beyond the archipelago. Archaeology in Oceania, 51(2), 150-157.

Quintus, S. J., & Cochrane, E. E. (2017). Pre-Contact Samoan Cultivation Practices in Regional and Theoretical Perspective. The Journal of Island and Coastal Archaeology, 1-27.

Cochrane, E. E. (2018). The Evolution of Migration: the Case of Lapita in the Southwest Pacific. Journal of Archaeological Method and Theory, 25(2), 520-558.

Quintus, S., & Cochrane, E. E. (2018). The Prevelance and Importance of Niche in Agricultural Development in Polynesia. Journal of Anthropologial Archaeology, 51, 173-186. 

 Quintus, S., & Clark, J. T. (2019). Ritualizing Hierarchy: Power Strategies and Pigeon Catching in Ancient Samoa. Journal of Archaeological Research, 0(0), 000-000.

Paleoenvironmental Coring of Samoan Lakes

In what may be the world’s tardiest research update…

Our fabulous core team. You really got to like the people you work with to spend 10 hours with then on an 8 m2 platform in the middle of a lake.

Last September, Matthew Prebble, Paul Augustinus, Jennifer Hayman and I transported a Uwitech coring system to Samoa to extract sediment cores from Lake Lanoto in eastern ‘Upolu. Our goal is to generate a Holocene and earlier environmental record as part of our Marsden study on the development of agriculture and social complexity in Samoa.

We extracted about 11 m of core and this should get us back past the Holocene, but we haven’t done any dating yet. The cores were split with half to Paul’s lab at the University of Auckalnd for ITRAX scanning and half to colleagues at ETF Zurich were a host of analyses, including 210Pb and 14C dating, and human, pig, taro biomarkers. We should be able to report on preliminary results soon.


Jen, Paul, and Mat taking 1 cm samples from our surface water core.

Yes, I’m happy we got these suckers out of the lake bed and into the ute.

This was a complicated undertaking and owes much to the support of the Centre for Samoan Studies at the National University of Samoa. Our work should also fit into the accelerating research linking detailed paleoclimate studies and human activities in the ancient Pacific. Check out related work by David Sear, Melinda Allen, Janelle Stevenson

Nice work, if you can get it.


A week of rain, survey and coring in Falefa, Samoa

Nick Porch examining a D-section core. Bugs in there will tell us about ancient ecosystems. Photo by Ron Lloren.

Our third week in the field was taken up with a variety of projects: working in the crater lakes to extract paleoenvironmental cores, continuing to map drainage ditches in Falevao and more D-section coring in Falevao. Plus we had several rain days.

We’ve now mapped an impressive network of ditches in Falevao and these mostly seem to connect to the river and likely served to channel water off the plantations during river flooding—we’ll do an elevational and hydrological analysis to test this idea.

Crossing the river to get to a coring site. Photo by Ron Lloren.

PhD student Ron Lloren from the Swiss ETF (Institute of Technology) is here to collect water samples for Hydrogen isotope analysis. The proportion of H isotopes in various water bodies can tell us about periods of aridity over millenia. We’ll correlate these data with our work on the field systems, flora, and insect fauna to generate as complete a picture of changing ecosystems as we can.

Ron takes rain water samples every day to get a baseline Hydrogen isotope signature.

Falefa survey area. The blue feature is a large walled walkway. The red features are single rock walls.

We’re now in our last week or so of fieldwork and back to Falefa mapping landscape features, including massive walled walkways that predate living memory and perhaps represent early divisions of the landscape. Excavations next year will help us determine this.

What one does on Sunday. Photo by Ron Lloren

More on the Land and Agriculture Marsden Project

We’ve recorded mostly walls (yellow) in Falefa near the coast and ditches (pink) in Falevao inland. There are different agricultural strategies depending on local environment.

Our second week of field work had us training another set of NUS Centre for Samoan Studies students, boys this time. We moved inland a bit to survey the lands of Falevao village (some of our Falevao survey area is the banner image), primarily noting and correcting drainage ditches identified in lidar.

We recorded a “fortified site” or at least a set of ridgeline features with platforms, terraces, and “umi ki”. It’s not on any of the Green and Davidson maps, so there’s still some discovery left out there.

Dr Matt Prebble getting ready to face this river section.

We’ve also been facing river sections and taking charcoal and sediment samples. There is a consistent burn layer at about 1.2 m deep across the area with massive charcoal chunks suggesting a big burn of forest. We’ve also found coconut and candlenut endocarp and seeds in the burn layer.

There is a similar burn layer in nearby by Fagaloa Bay that we recorded last year with breadfruit charcoal, and which dates to about 1000 BP.

Taking a break with Dr Brian Alofaituli, student Loto, assistant Pena, and me.

An interesting question for Falefa: why would populations burn off a presumably productive forest? Perhaps (raised-field) taro agriculture became more advantageous? We’ll return to this area in the upcoming weeks for D-section coring and more sampling. Dating should be interesting as well.

Speaking of D-sections: we hiked into a crater yesterday (on maps, Mt. Savaii) to take cores for pollen, micro- and macro-fossils, invertebrates and other goodies to begin reconstructing the paleoenvironment. It was a successful day with an easy 4 m core. We’ll go back to get down to 5 m and do some more collecting.

The D-section extracts sediment in intact 50 cm sections for later analysis.

For the next two weeks we’ll be making more GIS based maps and coring several locations, all to ultimately figure out the processes that explain agricultural and socio-political change in ancient Samoa.

Marsden Research on Ancient Samoan Agriculture and Political Complexity

We’ve kicked off field work for our research on the development of Samoan agriculture and political complexity in the Falefa Valley, ‘Upolu, working with the villages of Falefa and Falevao. Last week Associate Investigator Matt Prebble and I had National University of Samoa Students working with us as part of their HSA380 Field Archaeology course. The upcoming week we will have a new batch of students on the course. University and Centre for Samoan Studies staff, lead by Matiu Matavai Tautunu and Dionne Fonoti, have also helped behind the scenes and in the field.

Samoan archaeology student Vinimua captures the discussion of something important over there.

Our first goals is to map the extent of agricultural features such as rock walls, ditches, and terraces through a combination of lidar-based feature extraction and ground-truthing. We are working with the feature extraction data produced by Associate Investigator Seth Quintus at the University of Hawaii. With our field data, Seth should be able to tweak his extraction procedures to get better results. One obvious issue we are encountering is the use of rocks from old walls to build new structures. The old wall are visible in the field as exposed base-courses, but don’t show up very well in the lidar.

Samoan archaeology student Rachel records a rock wall.

We use a combination of Bad Elf GNSS Surveyor units, bluetoothed to iPads with ArcGIS Collector to collect field data. This is a really slick system and allows us to upload data every day to our ArcGIS enterprise cloud where Seth (currently in Hawaii) can check on our work, offer advice and ask for corrections. The Bad Elfs are working surprisingly well. I was nervous about their accuracy in heavy canopy, but they are consistently getting sub-5 metre accuracy and down to a metre or less with clear sky. Trimble has got a serious competitor here.

Some of the features recorded thus far include walls, starmounds, and stratigraphic relationships.

Our field data after just five days at work is pretty extensive. Although the valley is huge, I think that after a couple more month long sessions over the next year we will have generated an impressive amount of spatial data on agricultural infrastructure.

It’s hot surveying in the Samoan bush.

In a little over a week’s time we will be working with the D-section corer in the inland lake-caldera. For more check out the Centre for Samoan Studies, Research Gate, Google Scholar and Twitter.

Recording Ideas for Marsden Falefa Project

Seth Quintus, Mat Prebble and I have been planning upcoming field work associated with our Marsden project in association with the National University of Samoa, Centre for Samoan Studies (

One direction we will likely travel is down the path of full digital recording with iPADs/tablets and various GPS technology. This article in Internet Archaeology is informative, to a degree, but also bogged down in “reflexive”-speak. However, access to various levels of recording in real-time has great promise.





Land and Agriculture in Samoa

Here’s the 100 word summary of my recent grant submisstion:

“Despite being the archetype of socially stratified and hierarchical ancient cultures, we do not know how the Polynesian chiefdoms arose from less complex societies. The untested orthodox theory proposes that innovations in land tenure and agriculture lead to chiefly control of populations. Our international team will test this theory through new drone-based mapping and innovative soil analyses that track changes in land boundaries and agricultural productivity. Our research will discover the origins of the Polynesian chiefdoms with significant results for explanations of Polynesia prehistory, the functioning of stratified societies, and the global rise of persistent, institutionalized inequality.”

Falefa viewed from Lemafa pass.

This is very much a LKFS type of project where we will try to determine the chronology of agriculture and different types of land use in Falefa Valley, ‘Upolu, Samoa. My grant collaborators are Seth Quintus and Matthew Prebble.