Mesolithic to Neolithic transitions: new results from shell-middens in the western Algarve, Portugal. (Research).

Stiner, Mary C. ; Bicho, Nuno F. ; Lindly, John 等

Questions

The Algarve shell-midden sites span roughly 9500 to 6500 years before the present and represent the Mesolithic and early Neolithic cultural periods (Table 1). All of the Algarve middens to be discussed contained fire-cracked rock, and some of them have ceramics in the upper layers. Traditionally, archaeological sites in Portugal are classified as Neolithic if ceramic sherds are present, following the guidelines applied elsewhere in Europe. This zoo-archaeological study of the shell middens in the western Algarve region of Portugal questions the nature of the Mesolithic-Neolithic economic transition. Does the appearance of ceramics correspond to the adoption of agriculture or livestock herding? Or did hunting and gathering continue in this area, with the minor addition of utilitarian ceramic vessels to local tool kits? Shellfish exploitation was one aspect of Mesolithic and early Neolithic subsistence in the Algarve, but the perennial importance of marine resources here suggests that the malacofaunas can register important shifts in subsistence organization, regardless of whether shellfish were primary or "fall-back" sources of animal protein, fats, and other nutrients. The nutritional rewards of some shellfish are considerably greater than those of many terrestrial wild animals, such as hares, the latter having been heavily exploited at some inland sites in Portugal during the later Palaeolithic (e.g., Hockett & Bicho 1999; see also Hockett & Haws 2001).

The sites

Several shell middens were identified by the authors during surveys of the western Algarve, and those tested between 1997 and 1999 yielded the collections used in this paper. All of the sites lie west of Faro (Figure 1), most within a few hundred meters of the modern Atlantic shore, often perched on limestone coastal cliffs 10 to 60 m above sea level with access to rocky collection areas below (Figure 2). Rocha das Gaivotas (Mesolithic) and Vale Santo I (early Neolithic) occur in sand dunes and near good flint sources; the former lies on the modern coastline and the latter 1 km inland. At Barranco das Quebradas II and I another coherent but somewhat older faunal series occurs in collapsed rockshelters along a narrow valley, a few hundred meters from the Atlantic shore and adjacent to a fresh water spring. Ribeira de Alcantarilha, an early Neolithic occupation, is located on a high clay terrace of a river valley, now 5 km inland, although it must have been on an estuary at the time of occupation. In contrast to the other shellmidden sites, Ribeira de Alcantarilha is associated with shorelines dominated by soft-sediments and the mollusc species in this midden vary accordingly.

[FIGURES 1-2 OMITTED]

The shellfish

Virtually all the faunal remains in the Algarve middens are from marine molluscs. The sediment chemistry of these sites would seem suitable for bone preservation, as the shells are in good condition, leaving us with the conclusion that bones were seldom deposited. The shellfish were definitely collected by humans (see below) and nearly all of them (>99%) were consumed as food; shell ornaments are present but rare. The genera commonly exploited (Table 2) have similar adult soft tissue weight ranges, if scaled to human dietary need, and most of the genera inhabit the rocky intertidal zone: mainly mussels (Mytilus), limpets (various Patella species and, rarely, Siphonaria), turbans (Monodonta), and a large drill known as Thais haemastoma. At Ribeira de Alcantarilha, by contrast, carpet venus clams (Ruditapes decussatus) were particularly important, supplemented by other soft-substrate species such as cockles (Acanthocardia tuberculatum and Cerastoderma edule), scallops (Pecten maximus), bittersweets (Glycymeris sp.), striped venus clams (Chamelea gallina), and razor clams (Solen sp. or Ensis sp.). Thais occurs on a wide range of substrates but is much more common on rocky shores.

These shellfish represent significant sources of fat and carbohydrates in autumn, winter, and particularly spring (Fernandez-Reirez et al. 1996; Hove 1992). Well-fed mussels, for example, may convert 4-5% of the algal protein they consume into lipids, 33-64% to carbohydrates, and 44-47% to body protein prior to the spawning season (Fidalgo et al. 1994). Dry tissue of sand-dwelling Ruditapes decussatus may contain 7-8% lipids, 57-68% protein, and 6-21% glycogens prior to spawning (Beninger and Lucas 1984). Comparable information is not available for Thais haemastoma, but the closely related subtropical species, T. cattifera, can yield as much as 56% protein dry weight; 6.5% crude fat, 12% ash, and 24% carbohydrates prior to spawning (Udoh et al. 1995). The nutritional yields of these common shellfish potentially exceeds that obtainable from common terrestrial game such as rabbits and deer in late winter and spring.

Formation of the middens

The taphonomy of the shells was studied to determine how the middens were formed and the molluscs were processed. Shell damage analysis emphasized three criteria (Stiner 1994, 1999): burning by fire, indicated by shell discoloration, condition of break edges, specifically sharp versus wave-abraded surfaces; and fragmentation calculated as the completeness index, MNI/NISP, or the minimum number of individual animals divided by the number of identified shell fragments. The assemblages from Ribeira de Alcantarilha had suffered extensive fragmentation at the time of excavation, barring application of the fragmentation index, and limiting burning observations to minimum frequency estimates.

The consistent lack of wave-induced wear on the shells from the rocky shore sites as well as the altitude of the sites above modern sea level, refutes the possibility that the middens were accumulated by wave action. Sharp break edges indicate that the molluscs were collected while alive, and nearly all of the breakage reflects human processing activities. Burning damage frequencies generally are low but vary significantly by taxonomic group: it is fairly common on mussels (10-13% of total NISP) and Thais (2-20%), but very rare on limpets and turbans, if present at all. The soft-substrate malacofauna of Ribeira de Alcantarilha displays little in the way of burning damage, apart from the Thais shells. Shell completeness (MNI/NISP) also varies by taxonomic group. The index of shell completeness is consistently high for turbans (0.50-1.00 are complete) and limpets (0.40-.63 complete), although the former possess a thick shell and the latter a thin shell. Thick-shelled Thais is always fragmented to some degree but not as extensively as the shells of other taxa (0.13-.26 complete). Mussels are most fragmented of all (0.05-.21 complete), even in situations where equally thin limpet shells are not.

Much of the fragmentation of mussel shells occurred prehistorically from a combination of heating and trampling. The differences in burning frequencies among the mollusc genera are significant and partly linked to fragmentation. Ethnographic accounts reveal common cooking techniques for bivalves, usually baking on a bed of wood ash and older shell litter (reviewed in Stiner 1994:186-193). Heating weakens the mineral structure of shells, making them more brittle under any form of mechanical pressure (Stiner et al. 1995).

Distinct processing treatments are apparent for the turbans and limpets, which were not extensively heated, and bivalves and Thais, frequently prepared with the aid of fire. Humans removed turbans from their shells with minimal damage, punching out the spire or splitting a section of the body whorl to break the vacuum between the animal and its shell. Heating mussels, presumably to open them, was a widespread practice in the Mediterranean area, and the damage patterns observed in the Portuguese cases closely resemble those from Paleolithic sites (Stiner 1994, 1999).

The gathering of large drills and the damage patterns that resulted from humans processing them are most closely linked to mussel exploitation. Thais haemastoma is a large gastropod with a strong columellar muscle and heat may have made prying them from their shells easier. However, this does not explain the much lower frequencies of burning damage on similarly structured turban shells. It seems likely that, because Thais were gathered as they were encountered on mussel beds, these large drills were put on the same hot coals as mussels during processing.

Harvesting patterns

Two aspects of variation in shellfish exploitation--rank order dominance in the genera harvested (Table 3) and molluscan predator-prey ratios--can be examined. The edible rock-dwelling shellfish differ in diet and the extent of colony formation: turbans are herbivorous gastropods and live scattered in loose patches; mussels are large colonial bivalves that feed by filtering suspended particles from the water; limpets are semi-solitary grazers, scattered in loose patches; and medium to large predatory drills (Thais haemastoma) feed on bivalves among other prey (Tornaritis 1987). There is much repetition in the genera collected from hard substrates. A possible trend is apparent in the combined faunal series (Figures 3 and 4): turbans are most abundant, sometimes in combination with mussels, in the lower cuts and decrease with time (15 time-ordered calculable ratio values, [r.sub.s] = 0.75, 0.80, p =.001); limpet frequencies vary essentially randomly. The proportion of drills appears relatively stable, which is significant for the exploitation of shellfish as a whole.

[FIGURES 3-4 OMITTED]

Inter-site differences could explain some of the variation in shellfish species representation, but the rocky shore sites share similar settings. It is difficult to exclude the possibility that natural, stochastic variation in intertidal community structure that cause the relative live abundances of turbans, mussels and limpets to cycle, is responsible for the trends in species exploited. Such variation occurs naturally over very short time-intervals of a decade or less (on various marine provinces, compare Buss 1986; Dayton 1971; Dethier 1984; Hartnoll & Hawkins 1985; Lubchenko & Menge 1978; Petraitis 1995). While the archaeofaunal accumulations certainly represent fine-scale palimpsests of multiple exploitation events that would average away much of the "noise" caused by short-interval variation, it remains difficult to separate human-caused trends in the relative frequencies of mussels, turbans, and limpets from those arising from local marine community dynamics.

Possible indications of human predation intensity

Humans' ranking of shellfish resources depends upon the search and processing costs in relation to nutrition gained. Rock-dwelling genera present the advantages of being exposed and easily found. Although mussels, limpets, turbans, and drills coexist within the intertidal band of rocky Mediterranean shores, only mussels tend to form tightly packed colonies (Chelazzi et al. 1994; Lintas and Seed 1994; Svane and Setyobudiandi 1996). Dense resource patches make harvesting easier and, for this reason, may cause humans to rank mussels higher than other, more scattered shellfish such as limpets. Differences in the detachment and processing costs of these molluscs are also considerations. Opening bivalves is made easy by gentle heating, although they must be carried to a hearth in order to do so. The processing costs for turbans and drills are heightened by their thick shell, which prehistoric humans nearly always punched, split or cracked to extract the animal. Limpets do not require heat to extract, as the shell aperture is large and cannot be closed, but prying them from rock surfaces takes concentration and time, even with the benefit of stealth or a strong spatula. For these reasons, mussels appear to have been especially highly ranked by prehistoric foragers in the Algarve and elsewhere on the Mediterranean Sea. Because of their potentially large size and high visibility, the same may be suggested for Thais when encountered on mussel beds.

A second issue, independent of the ranked values of food molluscs is their relative productivity and its effect on species abundance on rocky shores under variable levels of exploitation. Turbans, limpets, mussels, and Thais differ in the rates of individual development in the wild, properties that translate to significant differences in population growth and turnover rates (sensu Stiner 2001; Stiner et al. 2000). Mussels are extremely productive, often maturing within 1-2 years of age, turbans in about 1-3 years, so their populations replenish rapidly. Many limpet species, and especially large drills, take longer to mature on average (usually 2-5+ years), making population replacement slower (cf. Ardizzone et al. 1996; Bombace 1995; Branch 1981; Gardner et al. 1993; Hockey & Bosman 1986; Kendall 1987; Koch & Wolff 1996; McLachlin & Lombard 1981; Moreno et al. 1984; Oliva & Castilla 1986; Paine et al. 1985; Pombo & Escofet 1996; Richardson et al. 1990; Southward & Southward 1978; Sukhotin & Kulakowski 1992). The latter pair of genera are therefore more vulnerable to over-harvesting by humans and most likely to signal this effect prehistorically in the form of body size diminution (e.g., Branch 1975; Dye et al. 1994) and/ or shifts in the abundance of high- and low-turnover types. As a relatively large-bodied bivalve predator, Thais falls at the least resilient end of the population turnover continuum and is the most sensitive to population decline under intensified predation by humans.

The question of harvesting pressure can be examined here by comparing the relative frequencies of the drills to the molluscs upon which they normally prey; extensive shell fragmentation in the Algarve middens has so far prohibited diminution analysis. Table 4 presents the ratios of predatory Thais to herbivorous and suspension-feeding molluscs in the time-ordered rocky shore sites. While MNI counts are not perfectly equivalent to mollusc biomass, MNI is permissible for this comparison because the molluscs eaten by humans in the Algarve have similar mean adult body masses, except for the carnivorous drill, which can be somewhat larger. The question is whether the ratio of Thais is constant or otherwise time-independent, or instead diminishes with time. If human exploitation intensified, as might be expected to result from increased human population density with a major transition to farming in the area, productive intertidal mollusk species are unlikely to be affected much. However, Thais would be adversely affected on account of its nutritional desirability and large body size in combination with a substantially lower potential for population recovery (see also Davis 1987).

In fact the ratio of predator to its prey is independent of time ([r.sub.s] = 0.17, multiple squared r = 0.0345, N = 15, p = 0.51). The ratio is much lower at Ribeira de Alcantarilha, where the mollusc assemblages are dominated by soft-substrate species, particularly Ruditapes decussatus, but this difference is probably explained by Thais' preference for hard or gravel substrates). The close relation between the relative frequencies of Thais and productive molluscs is likely to reflect normal saturation in the balance between this large drill and its prey--the number of adult Thais an intertidal community can support--and secondarily, humans' opportunistic collection of these predatory gastropods in the context of mussel exploitation on rocky shores.

Apparently exploitation was not sufficiently intense to reduce the relative abundance of Thais between ca. 7800 and 4800 years BC in the western Algarve region. The lack of evidence of Thais decline in the context of prehistoric shellfish harvesting is likely to mean that there was no change in human population densities across the Mesolithic to early Neolithic boundary. While the data are preliminary, they assist in refining existing hypotheses that can be tested with additional, well-dated faunal cases. A question for the future is whether this mollusc predator-prey ratio changed in later Neolithic phases in response to harvesting pressure from humans.

Context for the pottery

The existence of ceramics in the two shell-midden series does not seem to be explained by downward mixing of sherds from distinctly Neolithic components, as proposed by Zilhao for other series (2000; see also Carvalho 2002:243). This is because other aspects of the Rocha das Gaivotas-Vale Santo I and Barranco das Quebradas II-I series do not change with time.

It is not yet clear whether the ceramic vessels were made locally. The ubiquity of fire-cracked rock in the shell middens suggests that stone-boiling and related heating techniques were widely used in the study area. Under these circumstances, the benefits of adopting a few ceramic containers may have been considerable, without necessarily altering other aspects of subsistence or technology. Other kinds of evidence, such as patterns of animal exploitation, help to address the possibility of economic reorganization on the scale of a forager-farmer transition in the Algarve of southernmost Portugal.

Concluding discussion

Preliminary findings on the ratios of productive rocky shore mollusks, such as turbans and mussels and more sensitive drill populations, suggest little change in the intensity of marine exploitation by humans between the Mesolithic and early Neolithic periods in the study area. The appearance of ceramic vessels defines the onset of the Neolithic in this region of southwestern Europe but not necessarily that of agrarian economics (Bicho et al. 2000). This is not to say that no subsistence reorganization could have occurred in the earliest Neolithic in the study area. However, it seems unlikely in light of the apparent lack of a substantial demographic pulse at that time, in contrast to the story for other regions where the forager-farmer transition occurred more abruptly (e.g., Carvalho 2002; Dennell 1992; Silva and Soares 1987; Whittle 1996; Zilhao 2001).

It is possible, even probable, that the study area we have chosen is biased to locally persistent forager populations of the region. Nonetheless, any general loss of habitat to invading farming populations could suppress the food supplies of resident foragers more generally, causing them to intensify their use of those resources that remained accessible to them. Marine shellfish tended to be major sources of animal protein in this region prehistorically. The lack of change in the relative abundance of intertidal molluscs to slow-growing large drills in the archaeofaunal assemblages suggests that there was no less or more predator pressure on marine resources, or by extension, human demographic growth at this time.

At least two hypotheses for the Mesolithic-Neolithic transition in the Algarve have been proposed:maritime pioneer colonization from the north/northwest, marked by the abrupt intrusion of a full-blown farming and herding adaptation (Zilhao 2001); and gradual acculturation of, or at least exchange with, indigenous Mesolithic foragers (Silva and Soares 1987).These are not necessarily conflicting viewpoints, as an extended period of forager-farmer coexistence is considered likely in southern Portugal (Carvalho 2002), essentially a heterogenous patchwork of human populations and economic systems.

The findings presented by this first study of marine shells only sharpen the questions being asked of the Algarve midden sites, underscoring their considerable research potential and the need for additional testing and dating of Algarve coastal and inland sites.

Table 1. Sites, radiocarbon dates, and presence/absence of ceramics
for the marine shell assemblages from the western Algarve, Portugal.

 Cut Date Laboratory
Site name (cm bd) ybp ([dagger]) code

Ribeira de Alcantarilha - 6120 [+ or -] 70 Wk-6672
 - 6160 [+ or -] 60 Wk-6851
Vale Santo I - 6340 [+ or -] 120 Wk-6673
Rocha das Gaivotas - 6890 [+ or -] 70 Wk-6075
Barranco das Quebrades II 0-22 7980 [+ or -] 80 Wk-8940
 60-70 8400 [+ or -] 60 Wk-8951
Barranco das Quebradas I 40-50 8580 [+ or -] 70 Wk-8939
 110-120 8640 [+ or -] 70 Wk-8950

 Calibrated
Site name date BC Ceramics

Ribeira de Alcantarilha 5207-4859 Present in all units
 5227-4999 Present in all units
Vale Santo I 5470-5145 Present in all units
Rocha das Gaivotas 5839-5671 None in units studied
Barranco das Quebrades II 7053-6771 Present on surface only
 7569-7377 None
Barranco das Quebradas I 7707-7543 None
 7731-7587 None

([dagger]) Because all dates were obtained on marine shell, it is
expected that the true ages of these sites are somewhat younger,
with a correction factor of 360 [+ or -] 35 years subtracted from
the initial value, after Soares (1993). Calibration of the
radiocarbon dates (BC, 1 sigma) was carried out using CALIB 25.
Fire-cracked rock is abundant in every assemblage.

Table 2. Marine mollusk species identified in the Algarve shell
middens, along with information on diet, preferred substrate, and
approximate adult shell length.

 Name Common
Family Genus species source name

SCAPHOPODA
Dentaliidae Dentalium sp. - tusk shell
 ([section])

GASTROPODA
Patellidae Patella caerulea * Linne limpet
 Patella vulgata * Linne limpet
 Patella Pennant limpet
 depressa=intermedia *
 Patella aspera * Lamarck limpet
 Patella
 rustica=lusitanica * Gmelin limpet
Siphonariidae Siphonaria sp. * - false limpet
Trochidae Monodonta lineata * da Costa turban
 Gibbula sp. ([dagger]) - small top shell
Muricidae Urosalpinx cinerea Say oyster drill
 ([dagger])
Thaididae Thais haemastoma * Linne dog winkle/drill
Nassariidae Nassarius=Hinia Linne netted dog whelk
 reticulata ([dagger])

BIVALVIA
Mytilidae Mytilus Lamarck mussel
 galloprovincialis *
 Mytilus edulis * Linne mussel
Pectinidae Pecten maximus * Linne giant scallop
Ostreidae Ostrea edulus * Linne flat oyster
Glycymeridae Glycymeris sp. - bittersweet
 ([section])
Cardiidae Acanthocardia=Cardium Linne cockle
 tuberculatum *
 Cerastoderma=Cardium Linne cockle
 edule *
Veneridae Ruditapes=Tapes Linne carpet venus
 decussatus *
 Chamelea=Venus gallina * Linne striped venus
Solenidae Solen marginatus * Montagu razor clam
 Ensis siliqua * Linne razer clam

 Adult
Family Genus species Diet Substrate size (mm)

SCAPHOPODA
Dentaliidae Dentalium sp. C m,s 35-50
 ([section])

GASTROPODA
Patellidae Patella caerulea * H r 30-45
 Patella vulgata * H r 30-45
 Patella H r 30-45
 depressa=intermedia *
 Patella aspera * H r 30-45
 Patella H r 25-30
 rustica=lusitanica *
Siphonariidae Siphonaria sp. * H r 20-30
Trochidae Monodonta lineata * H r 20-35
 Gibbula sp. ([dagger]) H r,s,w 10-15
Muricidae Urosalpinx cinerea C s,m,r 10-15
 ([dagger])
Thaididae Thais haemastoma * C r 60-75
Nassariidae Nassarius=Hinia SC s,m 15-22
 reticulata ([dagger])

BIVALVIA
Mytilidae Mytilus F r 20-100
 galloprovincialis *
 Mytilus edulis * F r 20-100
Pectinidae Pecten maximus * F s 100+
Ostreidae Ostrea edulus * F r,s,m 100+
Glycymeridae Glycymeris sp. F s 30-80
 ([section])
Cardiidae Acanthocardia=Cardium F s,m,g 25-90
 tuberculatum *
 Cerastoderma=Cardium F s,m,g 30-50
 edule *
Veneridae Ruditapes=Tapes F s 40-70
 decussatus *
 Chamelea=Venus gallina * F s 35-40
Solenidae Solen marginatus * F s 80-120
 Ensis siliqua * F s 150-200

Sources: Hayward and Ryland 1995; Saldanha 1995; Tornaritis 1987.
(*) used mainly as a food source; ([section]) ornament; ([dagger])
incidental inclusion, probably arrived attached to shell of a larger,
edible marine mollusk. Mollusk diet codes are (H) herbivore;
(SC) scavenger; (F) filter feeder; (C) carnivore. Substrate codes are
(r) rock and other firm surfaces; (m) mud; (s) sand; (w) weeds;
(g) gravel or coarse sand.

Table 3. Relative frequencies (as a fraction of total MNI=1) of
rock-dwelling marine mollusks consumed as food by humans in
time-ordered midden assemblages.

 (grazer) (filter-feeder)
Site, trench & cut turbans mussels

Vale Santo:
 Surface 0.02 0.71
 1 0.05 0.64
 2 - **
 3 - **

Rocha das Gaivotas:
 Surface 0.02 0.57
 T1, 3 - *
 T1, 6 - **
 T1, 7 - 0.80
 T1, 8 - 0.73
 T1, 9 - 0.59
 T1, 11 0.02 0.55
 T1, 12 - **
 T1, 13 - **
 T3, B-C 5 - **

Barranco das Quebradas II:
 0-22 cm 0.19 0.30
 22-30 cm 0.25 0.28
 30-40 cm 0.32 0.11

Barranco das Quebradas I:
 0-30 cm 0.63 0.07
 30-40 cm 0.62 0.07
 40-45 cm 0.60 0.03
 50-60 cm 0.31 0.32
 60-70 cm 0.29 0.28

 (grazer) (carnivore) total
Site, trench & cut limpets Thais MNI

Vale Santo:
 Surface 0.11 0.15 84
 1 0.13 0.18 39
 2 * * 7
 3 - ** 9

Rocha das Gaivotas:
 Surface 0.19 0.21 209
 T1, 3 - * 3
 T1, 6 * * 10
 T1, 7 0.18 0.02 45
 T1, 8 0.25 0.02 55
 T1, 9 0.37 0.04 133
 T1, 11 0.41 0.02 54
 T1, 12 * - 12
 T1, 13 * - 10
 T3, B-C 5 * * 21

Barranco das Quebradas II:
 0-22 cm 0.38 0.13 177
 22-30 cm 0.29 0.18 187
 30-40 cm 0.48 0.08 179

Barranco das Quebradas I:
 0-30 cm 0.22 0.07 199
 30-40 cm 0.13 0.18 68
 40-45 cm 0.19 0.17 93
 50-60 cm 0.23 0.13 172
 60-70 cm 0.23 0.19 92

Note: A few large Siphonaria specimens (pulminates) are present
among the limpet-shaped shellfish, apparently also exploited as
food. (*) present or (**) dominant, but total MNI is too small for
calculation. Rare oysters have been excluded from consideration.

Table 4. Ratio of Thais haemastoma, a large carnivorous drill, to
the total number of herbivorous and filter-feeding mollusks in rocky
shore sites, arranged in chrono-stratigraphic order.

 Thais herbivore & Predator
Site MNI filter-feeder MNI ratio

Vale Santo I 22 118 0.16
Rocha das Gaivotas 59 498 0.11
Barranco das Quebrades II 72 471 0.13
Barranco das Quebradas I 84 540 0.13

Predator ratio is calculated as Thais MNI, divided by the sum of
Thais MNI and herbivore (grazing limpet and turban) and filter-feeder
(mussel, clam and cockle) MNI. Only the larger assemblages are
considered.

Acknowledgments

This research was supported by grant from the Fundacao para a Ciencia e Tecnologia PRAXIS/PCSH/C/HAR/ 70/96 (to N.F.B.), and the U.S. National Science Foundation (to M.C.S., SBR-9511894).

References

ARDIZZONE, G. D., A. BELLUSCIO, M. F. GRAVINA & A. SOMASCHINL 1996. Colonization and disappearance of Mytilus galloprovincialis Lam. on an artificial habitat in the Mediterranean Sea. Estuarine, Coastal and Shelf Science 43:665-676.

BENINGER, P. G. & A. LUCAS. 1984. Seasonal variations in condition, reproductive activity, and gross biochemical composition of two species of adult clam reared in a common habitat: Tapes decussatus L. (Jeffreys) and lapesphilippinarum (Adams & Reeve). Journal of Experimental Marine Biology and Ecology 79:19-37.

BICHO, N., M. STINER, J. LINDLY & C. R. FERRING. 2000. O Processo de Neolitizacao na costa sudoeste. In Neolitizacao e Megalitismo da Peninsula iberica. Actas do 30 Congresso de Arqueologia Peninsular. Vol. III: 11-20.

BOMBACE, G. 1995. Ambiente, pesca e risorse marine. In La Fauna Italiana. XII Giornata dell'Ambiente, Roma 6 guigno 1994, pp. 29-80. Accademia Nazionale del Lincei, Roma.

BRANCH, G. M. 1975. Notes on the ecology of Patella concolor and Cellana capensis, and the effects of human consumption on limpet populations. Zoologica Africana 10(1):75-85.

BRANCH, G. M. 1981. The biology of limpets: physical factors, energy flow, and ecological interactions. Oceanogz Mar. Blot Ann. Rev. 19:235-380.

Buss, L. W. 1986. Competition and community organization on hard surfaces in the sea. In (J. Diamond and T. J. Case, Eds.) Community Ecology, pp. 517-536. New York: Harper and Row.

CARVALHO, A. F. 2002. Current perspectives on the transition from the Mesolithic to the Neolithic in Portugal. In (Badal, E., Bernaeu, J., and Marti, B. eds.) Neolithic Landscapes of the Mediterranean. Valencia: Saguntum, Papeles del Laboratoria de Arquelogia de Valencia, Universitat de Valencia, pp. 235-250.

CHELAZZI, G., P DELLA SANTINA & G. SANTINI. 1994. Rhythmical timing and spatial scattering of foraging in a homer limpet (Patella rustica). Behavioral Ecology 5(3):288-292.

DAVIS, S. J. M. 1987. The Archaeology of Animals. London: Batsford.

DAYTON, P. K. 1971. Competition, disturbance, and community organization: the provision and subsequent utilization of space in a rocky intertidal community. Ecological Monographs 41:351-389.

DENNELL, R. W. 1992. The origins of crop agriculture in Europe. In (Cowan, C. W., and Watson, P. J. eds.) The Origins of Agriculture: An International Perspective. Washington, D. C.: Smithsonian Institution Press, pp. 71-100.

DETHIER, M. N. 1984. Disturbance and recovery in intertidal pools: maintenance of mosaic patterns. Ecological Monographs 54:99-118.

DYE, A. H., G. M. BRANCH, J. C. CASTILLA & B. A. BENNETT. 1994. Biological options for the management of the exploitation of intertidal and subtidal resources. In (W. R. Siegfried, ed.) Rocky Shores: Exploitation in Chile and South Africa, pp. 131-154. Berlin: Springer-Verlag.

FERNANDO-REIRIZ, M. J., U. LABARTA & J. M. F. BABARRO. 1996. Comparative allometries in growth and chemical composition of mussel (Mytilus galloprovincialis Lmk) cultured in two zones in the Ria Sada (Galicia, nw Spain). Journal of Shellfish Research 15(2):349-353.

FIDALGO, J. P., A. CID, I. LOPEZ-MUNOZ, J. ABALDE & C. HERRERO. 1994. Growth and biochemical profile of juvenile mussels (Mytilus galloprovincialis Lmk) fed on different algal diets. Journal of Shellfish Research 13(1):67-75.

GARDNER, J. P. A., D. O. F. SKIBINSKI & C. D. BAJDIK. 1993. Shell growth and viability differences between the marine mussels Mytilus edulis (L.), Mytilus galloprovincialis (Lmk.), and their hybrids from two sympatric populations in S. W. England. The Biological Bulletin (Woods Hole) 185:405-416.

HARTNOLL, R. G. & S. J. HAWKINS. 1985. Patchiness and fluctuations on moderately exposed rocky shores. Ophelia 24:53-63.

HOCKETT, B. & N. F. BICHO. 2000. The rabbits of Picareiro Cave: small mammal hunting during the late Upper Paleolithic in the Portuguese Estremadura. Journal of Archaeological Science 27:715-723

HOCKETT, B. & J. A. HAWS. 2002. Taphonomic and methodological perspectives of leporid hunting during the Upper Paleolithic of the western Mediterranean Basin. Journal of Archaeological Method and Theory 9(3):269-302.

HOCKEY, P. A. R. & A. L. BOSMAN. 1986. Man as an intertidal predator in Transkei: disturbance, community convergence and management of a natural food resource. Oikos 46:3-14.

HOVE, H. T. 1992. Class-variable profiles and principal component analysis used to study the effects of the seasonal cycle on the fatty acid composition in the tissue of Mytilus edulis. Chemometrics and Intelligent Laboratory Systems 14:271-278.

KENDALL, M. A. 1987. The age and size structure of some northern populations of the trochid gastropod Monodonta lineata. Journal of Molluscan Studies 53:213-222.

KOCH, V. & M. WOLFF. 1996. The mangrove snail Thais kiosquiformis Duclos: a case of life history adaptation to an extreme environment. Journal of Shellfish Research 15(2):421-432.

LINTAS, C. & R. SEED. 1994. Spatial variation in the fauna associated with Mytilus edulis on a wave-exposed rocky shore. Journal of Molluscan Studies 60:165-174.

LUBCHENCO, J. & B. A. MENGE. 1978. Community development and persistence in a low rocky intertidal zone. Ecological Monographs 59:67-94.

MCLACHUN, A. & H. W. LOMBARD. 1981. Growth and production in exploited and unexploited populations of a rocky shore gastropod, Turbo sarmaticus. The Veliger 23(3):221-229.

MORENO, C. A., J. P. SOTHERLAND & H. F. JARA. 1984. Man as predator in the intertidal zone of southern Chile. Oikos 42:155-160.

OLIVA, D. & J. C. CASTILLA. 1986. The effect of human exclusion on the population structure of key-hole limpets Fissurella crassa and F. limbata on the coast of central Chile. Marine Ecology 7(3):201-217.

PAINE, R. T., J. c. CASTILLO &; J. CANCINO. 1985. Perturbation and recovery patterns of starfish-dominated intertidal assemblages in Chile, New Zealand, and Washington state. The American Naturalist 125(5):679-691.

PETRAITIS, F. S. 1990. Direct and indirect effects of predation, herbivory and surface rugosity on mussel recruitment. Oecologia 83:405-413.

POMBO, O. A. & A. ESCOFET. 1996. Effects of exploitation on the limpet Lottia gigantea: a field study in Baja California (Mexico) and California (U.S.A.). Pacific Science 50(4):393-403.

RICHARDSON, C. A., R. SEED & E. NAYLOR. 1990. Use of internal growth bands for measuring individual and population growth rates in Mytilus edulis from offshore production platforms. Marine Ecology Progress Series 66:259-265.

SALDANHA, L. 1995. Fauna Submarina Atlantica (3rd Edition): Portugal Continental, Acores, Madeira. Mem Martins: Publicacoes Europa-America, Lda.

SILVA, C. T. & J. SOARES. 1987. Les communautes du Neolithique ancien dans le Sud du Portugal, Premieres Communautes Paysannes en Mediterrane Occidental. Paris: CNRS, pp. 663-671.

SOARES, A. 1993. The 14C content of marine shells: evidence for variability in coastal upwelling of Portugal during the Holocene, In Isotope Techniques in the Study of Past and Current Environmental Changes in the Hydrosphere and the Atmosphere. pp. 471-485. International Atomic Energy Agency.

SOUTHWARD, A. J. & E. C. SOUTHWARD. 1978. Recolonization of rocky shores in Cornwall after use of toxic dispersants to clean up the Torrey Canyon spill. Journal of Fisheries Research Board of Canada 35:682-706.

STINER, M. C. 1994. Honor Among Thieves: A Zooarchaeological Study of Neandertal Ecology. Princeton: Princeton Univeristy Press.

STINER, M. C. 1999. Trends in Paleolithic mollusk exploitation at Riparo Mochi (Balzi Rossi, Italy): food and ornaments from the Aurignacian through Epigravettian. Antiquity 73:735-754.

STINER, M. C. 2001. Thirty years on the "Broad Spectrum Revolution" and Paleolithic demography." Proceedings of the National Academy of Sciences 98:6993-6996.

STINER, M. C., WEINER, S., BAR-YOSEF, O. & KUHN, S. L. 1995. Differential burning, fragmentation, and preservation of archaeological bone. Journal of Archaeological Science 22:223 [not equal to]-237.

STINER, M. C., N. D. MUNRO & T. A. SUROVELL. 2000. The tortoise and the hare: small game use, the Broad Spectrum Revolution and Paleolithic demography. Current Anthropology 41(1):39-73.

SUKHOTIN, A. A. & E. E. KULAKOWSKI. 1992. Growth and population dynamics in mussels (Mytilus edulis L.) cultured in the White Sea. Aquaculture 101:59-73.

SVANE, I. & I. SETYOBUDIANDI. 1996. Diversity of associated fauna in beds of the blue mussel Mytilus edulis L.: effects of location, patch size, and position within a patch. Ophelia 45(1):39-53.

TORNARITIS, G. 1987. Mediterranean Sea Shells (Cyprus). Nicosia, Cyprus: Proodos.

UDOH, A. P., R. I. EFFIONG & D. O. EDEM. 1995. Nutrient composition of dogwhelk (Thais cattifera), a protein source for humans, Tropical Science 35:64-67.

WHITTLE, A. 1996. Europe in the Neolithic: The Creation of New Worlds. Cambridge: Cambridge University Press.

ZILHAO, J. 2000. From the Mesolithic to the Neolithic in the Iberian Peninsula. In (Price, T. D., ed.) Europe's First Farmers. Cambridge: Cambridge University Press, pp. 144-182.

ZILHAO, J. 2001. Radiocarbon evidence for maritime pioneer colonization at the origins of farming in west Mediterranean Europe. Proceedings of the National Academy of Sciences 98:14180-14185.

Mary C. Stiner, Nuno F. Bicho, John Lindly & Reid Ferring *

* Stiner, Dept. of Anthropology, University of Arizona, Tucson, AZ 85721-0030, USA. mstiner@u.arizona.edu Bicho, Dept. de Historia, Arqueologia e Patrimonio, FCHS, Universidade do Algarve, Campus de Gambelas, 8000 Faro, Portugal. nbicho@ualg.pt Lindly, SWCA, Phoenix Office, 2120 North Central Ave., Suite 130, Phoenix, Arizona 85004, USA. jlindly@swca.com Ferring, Institute of Applied Sciences, University of North Texas, P.O. Box 13078, Denton, TX 76203-3078, USA. ferring@unt.edu

Received 3 January 2002; Revised 5 November 2002; Accepted 9 January 2003.