OU Logo
Oklahoma Archeological Survey
HomeResearchEducationResource ManagementActivitiesStaffOklahoma's Past
Spiro Mounds site shell engravingOklahoma's Past

Pushmataha County

Click on Oklahoma map to return to county listings

The Kiamichi Fish Weir Site

In 1994, a landowner reported a "weird rock wall" eroding from the base of a 22 ft. tall terrace of the Kiamichi River. The landowners had found artifacts related to the very late Late Archaic or very early Woodland period (see time periods at the end of this page) from the upper 6 ft. of the terrace above the river as it eroded the bank, but the rock wall was at the base of the terrace directly on a gravel bar of the river channel. Though it resembled a foundation for an old cabin, it was clearly too old to relate to a historic occupation, and would have been rather damp in any case. They called the Oklahoma Archeological Survey for help in understanding what was coming out of their river bank.

The setting of the site is not unique in the Ouachita Mountains An outcrop of shale and a trapped gravel bed constrains the flow of the river during normal periods and forms a nice pool of water upstream. This geological feature prevents downcutting by the river and forces a broadening of the stream during flood episodes, thus cutting back the terrace. There were two severe rises along the Kiamichi River between 1989 and 1994, and each of these cut back the terrace edge. After one flood, the stumps of cypress trees were found preserved where they had grown into the gravel bed and underlying shale outcrop, in the same circumstances as the rock wall, and about 25 ft. north of the wall. Cypress do not grow in this part of the river valley under modern conditions.

The rock wall was composed of large, roughly rectangular rocks, probably from a ridge outcrop about one-quarter of a mile to the south. The rocks were stacked in a linear fashion without mortar. We straightened the nearly 6 ft. tall terrace remnant overlaying the rocks, and revealed a thick layer of blue (organic stained) sand over the rocks. As archeologists removed the sand at the west end of the alignment, they exposed what seemed to be the end of another rock wall extending to the south and forming a wide V with the exposed wall. At the apparent apex, two large and several smaller rocks lay parallel to one another and seemed to form an opening or chute between the ends of the walls. Further excavation over the next week revealed the second wall bearing off to the south, but unlike the first wall, this one was more disturbed rather than being stacked. After studying the stream and the gravel bar, archeologists concluded the south wall extended across the stream flow to divert water, and fish, through the chute or funnel. Water flowing against this wall dispersed it over time, while the first wall was parallel to the stream flow and was therefore not disturbed. Archeologists realized they had identified an archeological feature unique in Oklahoma prehistory, a fish weir (trap), before it was destroyed by natural processes.

Archeologists wanted to learn if the weir was associated with the archeological materials found in the upper part of the terrace. In the process of excavation, they recovered several samples of wood, one from beneath one of the largest undisturbed rocks of the feature, and one from within the arms of the weir. Charred hickory and pecan hulls were recovered from the culture bearing deposits in the terrace. All these materials were radiocarbon dated. The wood from beneath the large rock was identified as oak and dated to 3150+60 B.P. (about 1200 B.C.). The wood from inside the arms of the weir dated to 2990+80 B.P. (about1100 B.C.), but the charred nut hulls dated much later, 2380+80 B.P. (about 400 B.C.) showing the use of the weir was far earlier than the better known occupation of the site. Interestingly, the cypress trees found after the 1992 floods dated to 2990+60 B.P., or virtually contemporary with the dates from the weir.

Drawing of rock fish weir feature.

What have we learned about Oklahoma prehistory by finding this one feature? 1) Fishing was more important to food gathering than we might have known. We knew about bone fish hooks from sites such as 34LT11 and net weights from sites along the Mountain Fork and Glover Rivers located above natural fish traps, and we suspected that some of the crush black walnut and hickory husks may have been used as a fish poison, but the weir required an investment of labor to construct and maintain. 2) The age of the weir relates it firmly to the Late Archaic Period. 3) The radiocarbon dates from the cypress and the pine informs us about the vegetation and the climate. Cypress does not grow in the area under modern conditions and likes wetter and warmer conditions with warmer winters. Pine was previously thought to have become common about 2,000 years ago, but the date from the pine shows its presence over 1,000 years earlier. Again, we learned of the climate and vegetation from this piece of burned wood. 4) We now know that the age of the component in the upper part of the terrace is just before and at the beginning of pottery manufacture and use in eastern Oklahoma, and we suspect that use of the site by native Americans continued over a long time span on a seasonal basis.

Prehistoric Sites in Pushmataha County Identified to Time Period

Paleo = ?-8,000 BP / Archaic = 8,000-2,000 BP / Woodland = 2,000-1,000 BP / Village 1000-500 BP
BP (before present)


graphic decoration

Oklahoma Archeological Survey 111 E. Chesapeake Norman OK 73019-5111 (405)325-7211 Contact Webmaster: archsurvey@ou.edu

Home | Research | Education | Resource Management | Activities | Staff | Oklahoma's Past