Northern Headwaters Region on Minnesota Map
Institute for Minnesota Archaeology logo From Site to Story logo

Northern Headwaters Region








The Ogema Geshik Point Site         
Site Number(s):   21IC12  
County:   Itasca, MN  
City Township:   Unorganized Territory  
Image Archive:    (big thumbnails)     (medium)     (small)       

Shovel testing in dense cover
Shovel testing in dense cover.

Ogema Geshik Point was occupied by as many as four different Woodland cultural groups and has supplied the earliest known evidence for wild rice use in the Headwaters area.


The site is located on a peninsula that juts out into the Bowstring River approximately one mile northwest of Bowstring Lake in central Itasca County. Along the base of the peninsula is a large ridge about two meters high that runs parallel to the river.


Local residents say that this point was often used by Indian people in recent times as a ricing camp -- a place from which to harvest the plentiful stands of wild rice that grow on the river.

Modern Wild Rice
Modern wild rice grains.

Preliminary Investigation

First recorded in 1979, the site was revisited in 1990. At that time the only information on it was that it contained "mounds and a possible village." It was determined that there were 56 pit features of varying sizes. In 1993 the site was inadvertently disturbed during construction activities. This accidental disturbance led to an intensive archaeological survey and the licensing of excavation to define more fully the extent and nature of the site. The study was carried out by the Leech Lake Heritage Sites Program.

Research Design

Proper archaeological method requires archaeologists to ask questions that will provide an important base for future archaeological research. Questions that the survey and excavations at Ogema Geshik Point were designed to answer included:

  • How large was the site?
  • Which cultural groups occupied it?
  • Were the pits on the ridge of cultural origin, and if so, could their function be determined?
  • Did the people that lived at the site in precontact times use wild rice in the same way the local residents do today?

Excavation and Data Collection

The first priority was to identify exactly how big the site was. This was accomplished through shovel testing. Tests were excavated at every ten meters along the peninsula and parallel to the ridge. A total of 78 shovel tests were made, and of these 52 contained precontact artifacts.


Excavation unit
Excavation units 5 and 6.

Four of the pit features were then chosen for study. Three were selected from the peninsula area, while the last pit was located at the top of the ridge. Each pit was bisected by two adjoining l x 1 meter excavation test units, for a total of eight units. During excavation, more than 2,200 artifacts were recovered.


The soil from each unit was sifted through a screen, leaving only the artifacts such as pottery fragments, flakes of stone from tool making, projectile points, scrapers, and cutting tools, along with fish and animal remains.


Cross section of units 7 and 8
Cross section drawing of units 7 and 8.

Excavation units 7 and 8
Excavation units 7 and 8.

Once the test unit was deep enough, a profile of the pit could easily be seen in the wall of the unit.


Unfortunately, 1993 was a year of high water in the Northern Headwaters area, and four of the excavation units were flooded before reaching the bottom. However, excavation was completed in the remaining four. They revealed two straight-walled, flat-bottomed, stratified pits, measuring betrween 35 and 50 centimeters deep at the time they were created. Soil samples were taken from each pit feature, so that plant remains too small to be seen during excavation could be analyzed later.

Water in excavation unit
Water in excavation unit.

Analyzing the Data

Analysis of the material from the shovel tests showed evidence associated with at least five different cultural groups. They were Brainerd, Laurel, Blackduck, Sandy Lake, and Oneota, all dating from the Woodland period in the Northern Headwaters.



Sandy lake ceramics
Sandy Lake ceramics.

Each cultural group made its own very distinctive type of pottery. The location of pottery vertically within each unit provides the archaeologist with a clue as to which groups may have occupied the site first and which came later. For example, Brainerd ceramics were located near the bottom of each unit, indicating that the people who made this type of pottery, with a distinctive net-impressed exterior, were the earliest inhabitants of the site. Blackduck ceramics were often found closer to the top of the unit, indicating that they were deposited at the site later.


Blackduck ceramic rim
Blackduck ceramic rim.


Analysis of the soil samples showed that blueberry and sunflower were a source of food for the inhabitants of the site. However, surprisingly, no wild rice appeared. Since wild rice did not show up in the analyzed soil samples, the archaeologists turned to phytolith analysis. Phytoliths are silica bodies created by plants as a result of photosynthesis. Generally, each plant species produces its own type and shape of these silica bodies, and by looking at them under a microscope, scientists can determine which plants were present.

Therefore researchers began to look for plant remains that had been preserved by cooking. The best place to find these is on ceramic sherds with a black, charred "crust" on the interior, which may contain the remains of food cooked in the vessel. This crust is often used to learn the age of the sherd through radiocarbon dating, and it can also be used for phytolith analysis.

Wild Rice Phytolith
Phytolith showing wild rice.
Brainerd rim with crust
Brainerd rim with residue.

A Brainerd pot rim was found that contained a large amount of this crust. A sample was examined for phytoliths, while another portion was radiocarbon dated. The results were quite unexpected. Not only were wild rice phytoliths present in large numbers on the sherd, but the radiocarbon date was very early, approximately 170 CE (1,830 years ago). This is the earliest dated evidence for wild rice use in Minnesota.

Conclusions and Interpretation

With analysis complete, what were the findings regarding the four research questions?

  • How big is the site? Shovel testing revealed that the site was approximately 56,700 square meters, or 170,000 square feet in extent.

  • Which cultural groups occupied it? Shovel testing also revealed artifacts from at least five cultures: Brainerd, Laurel, Blackduck, Sandy Lake, and Oneota. Some or all of them may have occupied the site over the last 3,000 years.

  • Were the pits on the ridge of cultural origin, and if so, could their function be determined? Excavation revealed that the pits were not naturally formed, but were dug by occupants of the site. These pits were straight-walled, flat bottomed, and stratified, but they contained only a few artifacts. Their function is unclear. Since the soil samples analyzed from two pits did not contain any wild rice, their use for wild rice threshing (a common historic practice) is unlikely. They may have been used for storage or had some other unknown function. Since no charcoal was located in the pit features, they could not be dated, and therefore it is unclear whether they are of precontact or historic origin.

  • Did the people who lived on Ogema Geshik Point in precontact times use wild rice in the same way the local residents do today? We know from phytolith analysis and radiocarbon dating of the crust on potsherds that the people who made Brainerd pottery almost 2,000 years ago were collecting, processing, and cooking wild rice.


Sources Stories Credits Search Contents Links
Northern Headwaters Twin Cities Metro Area Red Wing Locality

From Site to Story web address
© 1999 The Institute for Minnesota Archaeology
Email us:
Updated 27 Jun 1999