Marine shells were recovered both by hand collection on site and wet sieving of soil samples. Shells were not collected on site during the 1989 excavation with very few exceptions, therefore there are only four 1989 contexts listed in table 1. All shells from hand collections have been identified to genus or species level. Small quantities of shell was recovered from most soil samples but generally this has only been quantified by weight (see archive). Shell from the samples in Area 3 has been dealt with in more detail as part of the analysis of the midden deposits in that part of the site. Apical fragments were identified and the concentration of shell fragments by weight was calculated.
Identifications of shells from hand collections are presented in table 1. The frequencies of the gastropod molluscs (Patella, Littorina, Nucella and Buccinum) are minimum numbers of individuals based on apex counts. The frequencies of the bivalve molluscs (Mytilus, Ostrea, Cerastoderma and Venerupis) are numbers of hinges and therefore must be divided by two to give the minimum number of individuals.
The recording of Mytilus fragments indicates a collection of shell dominated by fragments of Mytilus which as a result of fragmentation is represented by very few countable shells.
Identifications of shells from the area 3 soil samples are presented in table 2. Frequencies of shells were estimated in the way described above although only six samples contained any shells that could be counted. No attempt was made either to systematically identify the nonapical shell fragments or to measure the degree of shell fragmentation.
Shell assemblages from hand collections are inevitably biased and therefore should not be interpreted in detail. The species list is short, only nine taxa, which appear to be present more or less consistently throughout the site. In most contexts, the condition of the shells was poor with considerable fragmentation. This is the cause of the apparent rarity of Mytilus shells (Mussel) and may explain the very low numbers of Patella (Limpet) collected.
Overall, the collections are dominated by species from the intertidal zone of rocky shores and this matches the nature of the present shore at Scalloway. Only Cerastoderma edule (Cockle) and Venerupis pullastra (Carpet shell) are sand burrowing species and these appear to be genuinely rare on the site.
The assemblages from soil samples in area 3 are not affected by the types of bias seen in hand collections. However bias is still introduced by differential fragmentation and this clearly favours the survival and identification of Littorina littorea (Periwinkle). This species dominates the counts of individual shells but is not common in the shell fragments.
Two parallel stratigraphic sequences are represented by these sample columns, 401-406 and 421-434; the contexts are listed in stratigraphic order. Shell is only common in block 9.2 of the first column (402-406) and almost all the counted shells come from these contexts. Shell fragment size was largest also in these block 9.2 contexts with pieces of Ostrea (Oyster) shell up to 5 cm across. The fragment size in the second sample column (421-434) was smaller than any context in the first with almost all fragments smaller than 1 cm and no apparent difference between block 9.1 and 9.2 contexts.
Hand collections of marine shells indicate the exploitation of common species that inhabit rocky shores. Analysis of the shell from the area 3 samples has revealed variation in the degree of shell fragmentation both within and between columns. In terms of midden accumulation processes, this evidence suggests that block 9.2 in the first column consists of relatively undisturbed dumps of refuse with disturbed or slowly accumulated deposits of block 9.1 overlying them. Both block 9.1 and 9.2 in the second column are either heavily disturbed or slow accumulations. Its position, downslope from column one and away from the settlement, suggests that column two was on the periphery of the midden and may have received sediment derived by erosion from primary dumps upslope.
Table 1 Marine shell from hand collections
context block Patella Littorina Nucella Mytilus Ostrea Cerastoderma Venerupis Others
spp. littorea lapillus spp. edulis edule pullastra
6 7.1 2
8 7.1 34
28 7.3 1
150 12.4 3 6
201 1 3 1
202 1 1 2 6 1
203 1 2 5 1 2
204 1 4
206 3.1 1
207 7.1 13 4 4 1 Mytilus
215 3.2 3
235 2.1 1 58 18
236 1 32 3 2 1
237 2.1 3 1 10 5 L.
littoralis
x 1
239 6.1 6
245 1 90
258 7.2 1
260 6.1 38 1 1 23
263 7. 1 12 7 1
265 2.1 3 2 4 2
266 2.1 46 1
268 7.1 7 5 16
275 7.1 2
279 7.1 1 10 8 L.
littoralis
x 1
285 7.1 2 3
287 7.1 1 6 1 2
294 7.1 1 1 1
297 7.1 1
301 10.1 48 2
404 9.3 6 1
404/5 9.3 19 2 1
405 9.3 19 1
433 9.3 2 1 1
507 7.1 10 2 2
510 7.1 1 5 1 Mytilus
517 4.3 1 12 2 9
531 2.1 1 2 1
533 6.1 53 1
539 7.2 1 6 2
540 7.2 1 Mytilus
543 7.2 1 4 6 Mytilus
549 7.3 1
563 7.2 2
570 2.1 2 5
571 2.1 1 1
577 7.2 Buccinum
undatum
586 5.5 3
593 6.2 8
595 6.2 3
611 2.3 2 10 2
613 7.2 15
635 7.3 1 4 Buccinum
undatum
639 6.2 1
648 6.2 1 2 Mytilus
654 7.4 1 9 1
665 5.4 1 33 1
669 6.4 1
679 7.3 7
747 7.5 6
Table 2 Marine shell from sieved area 3 samples
context block sample Total Littorina Littorina Nucella Others
volume weight littoralis littorea lapillus
(g/10ltrs)
401 9.1 559 0.6 2 Venerupis
pullastra
403 9.2 163 7.2
402 9.2 310 37.2 1
404 9.3 415 132.7 77 4 Ostrea
edulis
405 9.3 60 51.7 1 10 1
406 9.3 10 119.1 7
421 9.2 48 1.2
425 9.3 28 3.9
427 9.3 56 4.2
423 9.2 14 0
422 9.3 246 2.4
424 9.3 16 0
433 9.3 50 8.7 Cerastoderma
edule
434 9.3 22 1.4