Section 4: Review of Results by Sample Unit

Unlike excavation levels beneath it in Test Unit 4, the excavation of level 6 was relatively uncomplicated. I excavated from the cutbank edge south to the nearest east-west 1m grid line, resulting in an excavation level more than 0.1 m3 in volume. By computing the volume from field notes and maps, I estimate the level's volume to be 0.12 m3, and its volume control constant to be K_V = 0.83.

Of the 98 bird specimens weighing a total of 39.4 g, I identified 56 specimens totaling 31.1 g to taxon and anatomical element. At least 15 individual birds, representing at least eight species in four families, were present in the sample (Table 4.3) . Most abundant were anatids (ducks, geese, and swans), with the remains of at least 2 individuals of the genus Anas (marsh ducks), and at least 3 individuals of other, unidentified genera.

Members of the auk family (Alcidae) were also well-represented, with the remains of at least 3 individuals--one each of the genera Cepphus (guillemots) and Uria (murres, probably Uria aalge, the common murre), and one individual of an unidentified genus. I also identified the remains of three individuals of the gull family (Laridae) --one sea gull (Larus sp.) and two kittiwakes (Rissa sp.)--and one individual of the loon family (Gavia sp.). 24 bones lacking articular or other easily recognizable surfaces were of similar size to small anatid or alcid bones in the comparative collection, and could not be identified, along with 42 specimens of undifferentiated (usually crushed and burned) bone.

I recovered and analyzed 599 fish bones from TU4, level 6, weighing 372.8 g ; of these I identified 324 specimens, weighing a total of 268.3 g and representing at least 14 individual fish (Table 4.4) . Gadus macrocephalus were the most abundant fish in this sample, with the remains of at least seven individuals represented by more than 250 specimens. Next most abundant were rockfish (Scorpaenidae , probably Sebastes sp. ), with two individuals represented by 27 specimens. Flatfish (Pleuronectidae) and salmonids (Salmonidae) were also each represented by the remains of at least one individual of each family.

281 specimens comprised the mammalian faunal sample in TU4, Level 6. I identified 127 specimens, representing six individuals in 6 families (Table 4.5) . The mammal remains from TU4, Level 6 were particularly poorly-preserved. Many of these showed evidence of having been burned, crushed, or pulverized; even large intact specimens were badly eroded. Phocids and Rangifer (caribou or reindeer) were the most abundant taxa by bone weight and number of individual specimens. Phocids were represented by 12 specimens connoting at least one juvenile individual. Odobenus rosmarus was represented by ten specimens--six rib fragments and four pieces of ivory, all of which show evidence of shaping for use as tools. I identified a wide variety of other mammal remains: a small canid's (probably Vulpes sp. ) teeth, a single Rangifer molar, and a burned Mustelidae (weasel or mink) humerus. A long bone fragment and paired mandibles represents a single rodent, probably a lemming (Synaptomys sp. or Dicrostonyx sp.). This rodent may have been intrusive, a strong possibility given the turbated nature of the soil.

Several shell fragments indicated the presence of mollusks in the TU4, Level 6, sample. Unlike other samples with significant amounts of shellfish remains, the most abundant species was not Mytilus edulis (blue mussel) but a variety of clam, probably Macoma . This bivalve was represented by 53 large fragments weighing a total of 37.3 g, more than twice the total weight of the highly fragmentary Mytilus sample of 17 g (Table 4.6) . Eight partial Macoma hinge fragments also suggested the presence of at least four individual clams. Mytilus remains may have been underrepresented due to excavation and screening methods, but the Macoma remains from this sample still outweighed all other non-Mytilus shellfish remains from other analyzed samples.

Test Unit 4, Level 14 (Sample 4/14)

Although difficult to excavate because of its hard, concrete-like consistency, permafrost helped preserve archaeological materials in this unit. The frozen matrix minimized root penetration, rodent burrowing, and microorganism activity, so strata were reasonably intact, and archaeological materials were well preserved. I located several concentrations of fish and bird bone that probably denoted discrete dumping locations. Excavation Level 14 contained a large proportion of the midden material above Feature 1, in Stratum II-9. Moderately sticky, silty sandy loams with high organic content characterized the Stratum II soils of Level 14. Several substrata intersected the excavation level, each of which may have represented unique depositional processes. For the purposes of my analyses, however, I assumed that the midden materials in substratum II-9 represented a single activity locus.

By any measure (MNI, NISP, or bone weight), fish remains--especially Pacific cod (Gadus macrocephalus) --overwhelmed the other fauna in the TU4, Level 14 sample (Table 4.8) . As in other samples, Pacific cod was the most abundant fish, with at least 19 individuals represented by more than 618 specimens. 19 atlas vertebrae indicated a minimum of 19 cod; other elements (10 frontal bones; 13 left exoccipitals, 17 basioccipitals) supported the high Gadus macrocephalus MNI based on cranial elements. A few other fish were represented in the sample as well--twelve vertebrae indicated at least one salmonid, and eight vertebrae indicated a single scorpaenid. 550 elements weighing 113 g had the consistency of other, identified Gadus bones, but were crushed and not easily differentiated as to element. Six specimens possibly do not belong to Gadus , but were also undifferentiated bone fragments.

I recovered and analyzed a wide variety of mammal bones from TU4, Level 14 -- approximately 170 specimens representing at least nine individuals in six taxa, and weighing a total of 378 g (Table 4.9) . Dog remains--probably Canis familiaris , the domestic dog--figure prominently in the identified sample, with nine identified specimens. A small phocid is also present, with four elements representing a single individual. I also noted 31 bones--ribs, and fragments of carpals and tarsals--of the appropriate size to belong either to a large dog or small phocid, so these taxa may be underrepresented in Table 4.9. I also identified 2 long bone fragments of a porous consistency similar to sea mammal bones in the comparative collection. Other sea mammal remains were present in this sample--at least one walrus and one sea lion. Two large fragments of porous bone signified the remains of a single cetacean. Four tiny left mandibles indicated the presence--probably intrusive--of lemmings of the genera Synaptomys or Discrostonyx .

The excavation crew recovered a few invertebrate remains from TU4, Level 14, during excavation. Most were badly crushed shell fragments or pieces of periostracum, the leathery "skin" covering mussels shells; altogether they weighed 40 g (Table 4.10) . Five Mytilus hinge fragments--three right and two left--indicated the presence of at least three individual mussels, weighing a total of 38.9 g. I also noted a single fragment of Macoma shell.

Test Unit 4, Level 18 (Sample 4/18)

Owing to the refrigerative properties of permafrost, Level 18 was largely unimpacted by bioturbation, decay, and soil formation processes, and it did not appear to intersect dense regions of refuse, as with Level 14. Substrata soils vary widely, from sticky anthropic sandy silt loams to thin sheets of sands, perhaps eolian in origin. I have interpreted the variety of soil types here as evidence of turbation during or immediately after deposition, perhaps as a result of episodes of trash dumping, interspersed with episodes of eolian sand deposition.

Excavation of Level 18 was constrained by the large stones of Feature 1, which could not be removed without risking damage to the excavation walls. The excavation crew worked from the cut bank face south, in essence digging into the level sideways. We also left a large cantilevered pedestal to support the weight of Feature 1. These two conditions produced an excavation level considerably smaller than the ideal 0.1 m3; my calculations placed the volume of Level 18 at around 0.05 m3, with a Volume Control Constant of 2.0.

Of the 54 total bird specimens from the TU4, Level 18 sample, I identified 35 specimens (Table 4.11) . Alcids seemed particularly abundant in the Level 18 sample, represented by two Cepphus individuals, one Cerorhinca individual, one Uria individual (probably aalge ), and two unidentified Alcidae. I noted the remains of at least four anatid individuals--one each Anas, Bucephala, and Melanitta , and an unidentified small anatid. In addition, I noted eight specimens with similarities to anatids in the comparative collection. Larids were poorly represented, with only a single individual of Larus indicated by three elements; two other elements identified to Lariidae suggested the presence of a second larid.

Analysis of bulk sample materials recovered in the laboratory using 1 8" screens revealed the presence of a large number of Clupea bones in the 4/18 sample. I noted 149 Clupea harengus vertebrae and one basioccipital; together, these bones weighed a mere 1.1 g and represented three individuals. Since bulk samples varied wildly in volume and represent a tiny fraction of the total excavated volume, I could not extrapolate the overall abundance of Clupea in TU4, Level 18. However, excavators will have missed virtually all of the tiny Clupea remains, whatever their abundance, and herring played some role in the economy of Nash Harbor people.

Test Unit 4, Level 20 (Sample 4/20)

Excavation in Level 20 proceeded more smoothly than in Level 18, although Feature 1 again forced us to excavate this unit from the cut bank into the dune. I estimated the volume of Level 20 to be 0.07 m3. Dividing the ideal excavation unit volume of 0.1 m3 by the actual volume excavated yielded a Volume Control Constant of 1.43.

Of the 22 bird bones from the TU4, Level 20 sample, I identified 14 to taxon (Table 4.14) . Cliff-dwelling birds dominated the identified sample, with at least one individual each of murrelets (Brachyramphus sp.), Cepphus columba, Uria sp. (probably aalge ), and one other alcid. I also noted a single cormorant and one kittiwake individual. Anatids were entirely absent from the sample, but a single grebe (Aechmorphus occidentalis) and one loon were present.

Test Unit 4, Level 24 (Sample 4/24)

The crew experienced no particular difficulty in excavating Level 24. Permafrost proved a problem in the southern margin of the unit, but the loose sandy matrix meant that the cold soil thawed quickly and was easily excavated even when partially thawed. The estimated volume of TU4, Level 24, is about 0.06 m3, with a Volume Control Constant of 1.6.

I recovered a small amount of faunal material from Level 24, less than 25 g of material in all (Table 4.17) . Most of the material was badly eroded or broken; I speculated that this degradation may have happened before or during deposition. Thick mats of vegetal matter found in these strata imply that permafrost had effectively refrigerated this level, halting biological decay of the faunal materials shortly after deposition.

I could not identify the three bird bones from the TU4, Level 24 sample. A single bird ulna had some similarities to Larus bones in the comparative collection, but was too deteriorated to identify positively. Two gadid bones and eight unidentified but possible gadid bones comprised the total fish sample. Two large mammal bones--a vertebra possibly belonging to a sea lion or fur seal and a fragment of porous bone probably belonging to a cetacean--suggested the presence two large sea mammals. A few crushed barnacle shells and fragments of Mytilus periostracum, together weighing only a total of 5 g, comprised the total invertebrate remains from Level 24.

Test Unit 4, Level 28 (Sample 4/28)

As with TU4, Level 24, the excavation of Level 28 proceeded smoothly. Permafrost complicated digging, but the sandy soil thawed quickly and could be dug when partially frozen. I estimate the volume of TU4, Level 28 to be 0.08 m3, and calculate a Volume Control Constant of K_V = 1.25.

The excavation crew recovered few faunal remains from Level 28 (Table 4.18) , an unsurprising result, considering the large amount of sterile beach sands comprising the excavated soils. Two walrus bones--rib and long bone fragments--weighed 222 g and comprised the vast majority of faunal material by weight. I could not distinguish these remains as a separate individual from the Odobenus bones found in Level 20, however. Another unidentified mammal was indicated by 16 small, badly eroded mammal bones. I analyzed five fish bones as well, two of which I identified as salmonid and scorpaenid. A few fragments of Mytilus periostracum indicated the presence of shellfish. The total lack of both bird bones and gadid bones set the TU4, Level 28 sample apart from all other analyzed faunal samples.

Test Unit 7, Level 7 (Sample 7/7)

Bioturbation appeared to have played an important role in the history of the soils in TU7. Root penetration mixed soils in the upper 50 cm, and rodent activity disturbed apparent buried surfaces throughout the unit. A distinct band of dense shell and bone midden ran through the unit, designated Stratum II-4 on the stratigraphic profile (See Figure 2.5) . The majority of faunal remains--especially fish and shellfish--were derived from the midden stratum.

Unlike Test Unit 4, Test Unit 7 lacked frozen soils and large features of flat stones; the excavation of this unit proceeded more easily than in Test Unit 4. Because of the erosion along the sea cliff, the unit was irregular in cross-section. I computed the total excavated volume of Level 7 at 0.08 m3, with K_V = 1.25.

I identified a large number and wide variety of bird remains--almost all from pelagic and cliff-dwelling birds--from the TU7 Level 7 sample (Table 4.19). Cormorants (Phalacrocorax sp.) were the most abundant birds by any measure, with 14 specimens weighing 36 g signifying the presence of at least two individuals. Other pelagic birds included at least one guillemot and one murre, and another alcid unidentified to genus. Gulls were also represented, with one Larus individual indicated by five specimens and one Rissa individual indicated by four specimens. Four specimens indicated the presence of at least one small loon, and two specimens were identfied as an anatid of unknown genus. Undifferentiated small long bone fragments, similar in size to alcid, small anatid, or small larid bones, remained unidentified.

I analyzed 399 fish bones and fragments weighing a total of 276 g from the TU7, Level 7 sample; of these, I identified 235 specimens (Table 4.20). Although Gadus remains comprised a majority of fish specimens identified, other fish made up a significant amount of the piscine fauna. One hundred thirty-four elements signified at least three Gadus individuals. At least three rockfish (Scorpaenidae, probably Sebastes sp.) and four individuals of a second variety of gadid (probably tomcod, Microgadus proximus ) also contributed to the analyzed sample.

Identification of the mammalian fauna in the TU7 proved more difficult than in other samples. Of the 64 bones and bone fragments, I identified only 11 specimens (Table 4.21). About 50 bone fragments, weighing 40 grams, were burned, and badly eroded. Many of the remaining bones were also worn and damaged, possibly by immediate post-depositional weathering. The remains of phocids and Rangifer were the most abundant, by weight and number of individual specimens, but I identified only one individual of each group. I also noted a portion of a small rodent cranium.

The TU7 Level 7 sample was most notable for its abundant invertebrate remains, especially Mytilus shells (Table 4.22) . Weighing 196 g, mussel shells outweighed fish remains in the analyzed sample. As discussed in Section 3, the loss of a large amount of shell due to 1 4" screens used during excavation meant that this amount actually underrepresents the total number of shellfish remains. I noted 41 individual Mytilus hinges or hinge fragments, but did not determine their sides; I estimated the presence of at least 20 mussels. I found the remains of a few other invertebrate species--one complete marine snail (family and genus unknown), and eleven fragments of chalky clam shell, probably belonging to Macoma . A single fragment of acorn barnacle shell discovered during analysis proved enigmatic; it may have arrived in the midden as a hitchhiker on mussel shell collected by Nash Harbor residents.

Table of Contents

Section 3: Faunal Analysis: Goals & Methods

Section 5: Distribution of Faunal Materials