The excavations were scheduled for the time of lowest water levels, mid-March to mid-April. On a cold and windy March 15, Ktunaxa elder Phyllis Nicholas said a prayer to the creator to bless the excavations (Figure 4). 

A series of seven 1 x 1 meter test units were judgmentally located along the bank of the Columbia River to sample several different parts of the site.  Most of these were located on the highest part of the terrace defined by a small creek (Figure 5, Figure 6).  A datum was established on the highest point of this terrace and a contour map was prepared.  Four initially tests units were excavated labeled XU 1, 2, 3, and 7.  Based on the results of these, a larger block of units was excavated adjacent to XU2 (Figure 7) and a row of three units was excavated near XU7.  Additional 1 x 1 metre units were excavated further away of the bank to assess the extent and density of the cultural materials. 

All of the excavations were dug in 5 cm arbitrary levels.  This was felt to provide the best control on the materials and soil deposition.  During March, frost was encountered immediately below the sod layer.  Once the soils had been exposed to the sun for a short time they tended to thaw but were often water saturated.  Most of the excavation was carried out with short handled spades and trowels.  All of the excavated soils were then screened through a 5 mm mesh screen with water pumped from the Columbia River (Figure 9).  A settling pond was constructed of hay bales to trap the silt to prevent river siltation and to trap the soil for backfilling.  Our excavation system was monitored by the B.C. Ministry of Environment personnel to ensure that siltation did not have adverse environmental impacts. 

All cultural materials were bagged in zip-lock plastic bags with the provenience data written on the bag with indelible felt pens.  The same information was recorded on a 3 x 5 inch card and included in the bag.  Each excavator was assigned a notebook for keeping notes on their unit while they were excavating.  Soil profiles and photographs were prepared for selected units.  All units were excavated until basal river gravel was encountered (usually about 70 cm) or until the water table was reached (commonly at 75 cm).  Where units reached the water table before encountering river gravels, an attempt was made to continue excavating until this became impossible.

Fire broken rock was common on the site.  In layers where large concentrations were encountered, it was weighed with a spring scale on site and discarded.  In layers where small quantities were located it was bagged and taken to the laboratory where it was weighed with a balance scale prior to discarding. 

Because of the water saturated nature of the soils, organic materials such as bone and wood were commonly packed in a plastic bag or other container, usually with the surrounding soil so that the general humidity levels were not drastically altered.  These were then stored in a refrigerator.  Because of the potential for highly perishable items (e.g. basketry, netting ) in such a wet environment, the crew had also be briefed on the recovery and storage of such items by a professional conservator, however no such items were located.  All of the wood pieces were carefully examined for evidence of human use but none was identified.  Some of the wood appeared to have been beaver gnawed.

Although a systematic surface collection was originally proposed for the site, the difficulties of extending the site grid into highly water saturated areas made it unfeasible.  As the riverbank began to become slightly drier, or when it froze overnight sufficiently to support a person a few selected items were collected from the surface.  This included a small number of  bone or antler tools and some finished stone tools, which were used to augment our excavated sample of materials from the site. 

Several meetings were held with the Columbia Lakes and Shuswap Elders to show them significant recovered items and to discuss interpretations.