Introduction

Archaeological finds of the Iberomaurusian are well-known from many caves and rockshelters across Morocco. Until now, however, knowledge of its dating has been limited to relatively few sites and, although the Iberomaurusian is generally assumed to fall within the time period >20,000–10,000 bpFootnote 1 (Lubell 2001), very few long stratigraphic sequences have been published. A further assumption is that the Iberomaurusian represents a fairly stable cultural adaptation with only relatively minor typological and technological changes throughout the whole of its presumed timespan of over 10,000 years.

The question of the dating of the Iberomaurusian in Morocco is also becoming increasingly important to anthropologists because of its bearing on broader issues concerning the dispersal of modern humans into NW Africa. In particular, it has been suggested that the Iberomaurusian microlithic bladelet tradition marks a major break with preceding technologies (Close and Wendorf 1990, 43) and, since it is known to be associated with biologically distinctive populations of skeletally robust Mechta-Afalou human types (Arambourg et al. 1934; Ferembach et al. 1962; Ferembach 1985; Hachi 1996; Hadjouis 2002), the question arises whether the introduction of this technology coincided with an appearance of new human types in NW Africa or whether it emerged and diffused independently. This of course remains an unresolved and highly controversial issue (for discussion see Irish 2000) but it is clear that for the debate to be substantially advanced there is now a requirement for further developments in the refinement and precision of the chronological evidence. In this paper we consider the existing chronological framework for Morocco and present new AMS dates on the Iberomaurusian which we hope will inform this process and enable further progress to be made.

Background

It is generally accepted that the Iberomaurusian represents the earliest Upper Palaeolithic technology in the Maghreb (modern Morocco, Algeria and Tunisia). The cultural term was introduced by Pallary (1909) to describe an industry from Abri Mouillah (western Algeria) characterised by “une profusion de très petites lames à dos retouché et à pointe très aiguë”. The same type of industry has been referred to under a variety of different names, most notably ‘Oranian’ (Gobert and Vaufrey 1932), ‘Mouillian’ (Goetz 1945–6 cited in Roche 1963) and ‘Eastern Iberomaurusian’ or ‘Eastern Oranian’ (McBurney 1967; Close 1986).

Geographically, the Iberomaurusian is known from localities stretching over a very wide area from Morocco (10° W) in the west as far east as Cyrenaica (22° E), with some apparent gaps in between such as in western Libya (Lubell 2001). The southerly extent of its distribution is not well understood but is known to reach as far as approximately 33° N along the Atlantic coast of Morocco or possibly even further south (Grébénart 1975, 1997; Fig. 1). There is a recognisable focus in the location of sites in the present coastal belt (Brahimi 1970) although other authors have noted that the distribution does include sites further inland (Lubell et al. 1984; Close and Wendorf 1990).

Fig. 1
figure 1

General location of Iberomaurusian sites in Morocco and the Maghreb including those mentioned in the text. 1 El Khenzira, 2 Contrebandiers, 3 Harhoura II (Zouhra), 4 Ghar Cahal, 5 Hattab II, 6 Kehf el Hammar, 7 Ifri el-Baroud, 8 Ifri n’Ammar, 9 Taforalt, 10 La Mouillah, 11 Rachgoun, 12 Columnata, 13 Afalou, 14 Tamar Hat, 15 Taza, 16 Kef oum Touiza

Full size image

In terms of the lithic inventories, Camps noted that the Iberomaurusian is a bladelet industry with a bias towards microliths ‘L’industrie lithique ibéromaurusienne est essentiellement une industrie lamellaire de tendance microlithique’ (Camps 1974: 62). Estimated frequencies of backed pieces could reach as high as between 40–80% of the retouched tool assemblages (Camps 1974: 63). The high proportion of backed points manufactured using the microburin technique is also a feature on which various authors have commented (Camps 1974: 76; Barton et al. 2005).

Existing Problems with Chronology

Despite its availability as a method since the 1950s very little systematic application of radiocarbon dating has been employed in relation to Iberomaurusian sites in Morocco. A rare exception is the work by Abbé Roche at Contrebandiers Cave (Témara) and at Grotte des Pigeons (Taforalt) where a series of conventional 14C dates was obtained for Iberomaurusian layers (Roche 1976). Until now, his dates of 21,900 ± 400 bp (Gif-2587) and 21,100 ± 400 bp (Gif-2586) for Taforalt have stood as the oldest for Morocco and are broadly comparable to the lowermost Iberomaurusian layer at Tamar Hat, Algeria which produced an age of 20,600 ± 500 bp (MC-822; Saxon et al. 1974).

For a number of reasons the above dates and others from sites on the Atlantic coast (Debénath et al. 1986) have proved problematic and are now urgently in need of revision. To begin with, many of the original dates were produced not on single charcoals or bone but on bulked material and from the documentation it is not always easy to attribute them to a precise archaeological level or stratigraphic unit. A typical case in point concerns radiocarbon dated bone from Contrebandiers Cave said to come from pre-Iberomaurusian (Aterian) levels but which produced anomalously young ages with central values of between 12,500–12,320 bp (Roche 1976). On the basis of the dates obtained it was concluded that the samples must have come from pits of Iberomaurusian age that had intruded into the Aterian levels (Roche 1976: 161). However considerable doubts remain over the integrity of the samples and the dates themselves since they were based on bulked material. A similar difficulty arises over the interpretation of the oldest Iberomaurusian dates from Taforalt. Here the samples were based on carbon extracted from burnt sediments where the source of error including the possibility of admixture of charred materials of different ages is particularly high. Moreover, some of the levels sampled for 14C assay by Roche can now be demonstrated to pre-date the Iberomaurusian at this cave.

Some of the limitations alluded to above can now be overcome through the application of better screening and pre-treatment methods and the ability to measure smaller quantities of charred materials using Accelerator Mass Spectrometry (AMS). As a consequence, in 2001 we began a systematic programme of resampling at new as well as previously known sites in northern and eastern Morocco which seemed to offer the best potential for dating work. AMS dating was undertaken on single identified charcoals from layers only where there was a secure Iberomaurusian association. Initially, it was also hoped to obtain accelerator dates on bone and shell from the same stratigraphic units. The aim of this work was partly methodological, to provide comparative samples for testing apparent discrepancies recently noticed between radiocarbon dates on charcoal and those on bone and shell (Jöris et al. 2003). However, in the event, the choice of material was restricted to charcoal because of the poor preservation of protein in the bone samples. Similarly, much of the shell recovered was of thin-walled terrestrial molluscan species. Although not in itself an obstacle to dating (Preece 1991), we were aware of problems concerning the uptake of ‘old’ inorganic carbon from geological sources, when the snail was alive, and the effects of post-depositional diagenesis (Goodfriend 1987; Goodfriend and Stipp 1983), not to mention the common habit of some these forms to burrow.

Within these constraints we have obtained highly consistent results on charcoal and, as a by-product for palaeoenvironmental studies, have often been able to identify the types of wood from the preserved fragments (Ward 2007). The dating results with species identifications are shown in Table 1.

Table 1 Iberomaurusian: AMS determinations on single charcoals from Ghar Cahal (GC), Kehf el Hammar (KEH) and Taforalt Sector 8 (TAF)
Full size table

Sites and AMS Radiocarbon Dating Results

Ghar Cahal

This cave site is located in the extreme north of the Tingitane Peninsula less than 2 km from the present coast (Fig. 1). Situated at GPS reference 5°25.517′ W, 35°52.607′ N, the cave lies at an altitude of 285 m and was originally investigated by Tarradell (Tarradell 1954, 1955). Further work took place in the 1980s by the former Mission Préhistorique et Paléontologique Française au Maroc (MPPFM) and has not yet been fully published (Ouchaou 1998–9). In 2001 we made a stratigraphic study of the cave deposits as part of a wider survey of sites in the Talembote and Sebta/Ceuta areas. Fifteen individual sedimentological units were recognised and described (Bouzouggar et al. 2006). Although previous researchers at Ghar Cahal had identified mainly Neolithic occupation, we were able to pinpoint two layers beneath the Neolithic sequence attributable to the Iberomaurusian. Unlike the overlying deposits, the finds from layers 10 and 12 are aceramic and the lithic technology is dominated by blades and backed bladelets (Bouzouggar et al. 2008). Also present in the large fauna in the same layers was Barbary sheep (Ammotragus lervia), including one heavily cut-marked astragalus in layer 12. In addition to these finds, an enigmatic bone was recovered from layer 10 (Fig. 2). The piece is a small splinter of bone (21 × 13 mm) with scrape marks in a chevron pattern on its outer surface and overlain by a series of parallel deeply incised grooves. The marks are not the result of butchery and none of the incisions are deep enough to have weakened the bone sufficiently to allow it to have been snapped. On present evidence it would appear to be a deliberate engraving on bone and as such would be the first piece of artwork recovered in an Iberomaurusian context from this cave.

Fig. 2
figure 2

Engraved bone from Ghar Cahal, layer 10. Scale in millimetre

Full size image

Fine-mesh sieving of layer 10 has recovered a rich assemblage of small mammals, comprising bats, insectivores, rodents, a lagomorph and a small carnivore. The rodents and insectivores are taken to have been brought to the cave by owls or other birds of prey, but the bats may have been living in the cave. The small mammal assemblage is dominated by the Algerian mouse Mus spretus (minimum number of individuals = 59), which accounts for just over three-quarters of the individuals. Today, the Algerian mouse is limited to the Mediterranean climatic zone along the North African littoral from Morocco to Libya, where it is found in grassland, shrub and woodland margins. In North Africa it is commonly found together with the eastern orchard dormouse Eliomys quercinus, which is also found in the small mammal fauna from layer 10. The abundance of Algerian dormouse together with the scarcity of arid-adapted rodents, such as gerbils (Gerbillus sp., MNI = 4), suggests a mosaic of habitats and a relatively humid climate typical of the Holocene (see Denys et al. 1996).

The dated layers at this cave bracket layer 11, which shows greatly reduced anthropogenic influence (and even this, mostly in the form of fine charcoal fragments, is possibly intrusive from above) and has a generally finer limestone clast content than layers 12 and 10. It is of interest, however, that this about 50 cm thick sub-sequence (layers 12, 11 and 10) does not contain sharp internal boundaries and no significant unconformity is apparent, suggesting that a more or less continuous sediment record is present, plausibly spanning the Pleistocene/Holocene transition, including the Younger Dryas.

Hattab II

Hattab II Cave (GPS 5°4′ 60″ W. 35°26′26″ N) is located in the Talembote region on a tributary of the River Oued Laou about 12 km from the Mediterranean coast (Fig. 1). The site is one of two caves named Hattab but only Hattab II contains evidence of Iberomaurusian activity. The site is situated about 30 m above the river and at a point where the narrow gorge opens out into a wider floodplain. No previous excavations had taken place there before our work in 2002 and 2003. The sediments could be divided into nine stratigraphic units and a human burial was discovered in layer 8 resting on the floor of the cave. The grave and its contents are described in detail elsewhere (Barton et al. 2008). The salient points of interest are that the burial is of an individual, possibly male, with an estimated age of between 25 and 30 years at the time of death. The skull shows an absence of the central upper incisors characteristic of those reported in other Iberomaurusian burials (but not exclusively, see Humphrey and Bocage this volume). The skeleton was discovered in anatomical position with no signs of disturbance. The accompanying non-lithic grave goods included bone sagaies, a marine gastropod and a gazelle horn.

A small but homogeneous Iberomaurusian lithic assemblage consisting of straight- and curve-backed bladelets was recovered from the same layer as the burial. There was no evidence of the use of the microburin technique. Efforts to obtain a direct AMS radiocarbon date on bone samples of a rib and the left and right clavicles of the skeleton failed due to poor preservation of bone collagen. However, the presence of a burnt lithic artefact immediately adjacent to the human burial provided a Thermoluminescence (TL) age determination of (K0311) 8.9 ± 1.1 ka years. This date, though unexpectedly young, is securely within the same archaeological layer as the burial. No charcoal dates have yet been obtained from this site.

Kehf el Hammar

Kehf el Hammar (GPS 5°10′56″ W, 35°20′37″ N) is a cave occupying a small lateral valley of the River Oued Laou, not far from Hattab II Cave. It lies some 10 km inland from the present Mediterranean coast (Fig. 1) and at an altitude of 97 m. Previous test excavations undertaken by the Mission Maroco-Espagnole (MME) in 1992 (Slaoui 1995) confirmed the presence of multiple Iberomaurusian occupation layers, very rich in artefactual and organic remains, and extending over a depth of more than 2 m. Fragmentary human skeletal remains of adults and children were also reported from the uppermost layers (Slaoui 1995). In 2001 and 2002, we undertook small scale excavations to obtain dating and palaeoenvironmental samples from sections left by the earlier excavations (Barton et al. 2005). The cave is extremely dry and conditions are exceptionally favourable for the preservation of bone, charcoal and phytoliths. The dating results on individual charcoals are presented in Table 1.

Iberomaurusian artefacts are found in successive layers from 1 down to the base of 6. No major technological or typological variation is visible throughout this sequence, but the sample is still relatively small. Layer 1 is heavily bioturbated but otherwise shares with layer 2 an abundance of backed bladelets and the presence of microburins. Layer 3 also has backed bladelets with the addition of notched bladelets. This layer is separated from a lower series of gritty carbonate and silty deposits by a complex angular unconformity, which could represent a significant interval of time. Beneath this in layers 4 to 6 the industry again comprises high frequencies of straight- and curve-backed bladelets made using the microburin technique. A potential change in archaeological sequence occurs in layer 7 where retouched blades outnumber backed bladelets.

Some information on the palaeoenvironmental context of the Iberomaurusian is available from charcoal, microfaunal and sediments evidence (Barton et al. 2005). The combination of charred wood species recovered in layers 5 and 6 (Leguminosae, juniper, deciduous oak and pine) is normally found today in supra-Mediterranean bioclimatic zones (Blondel and Aronson 1999). Such a grouping for this latitude would usually be expected to occur at altitudes of 1200 m and above (Blondel and Aronson 1999, Fig. 4.1), thus implying much cooler, semiarid conditions than are experienced in the area of the site today (Aura Tortosa et al. 2002). This interpretation appears to be supported by the relative abundance in the microfauna of gerbils (Gerbillus sp.) and jirds (Meriones sp.). The sediments also reveal markedly less lamination (wash input) and fewer biogenic carbonates in these levels. After layer 5, there is a minor unconformity, followed by increased signs of wetness in layer 4 before a major angular unconformity and a possible return to cooler drier conditions as indicated by occurrences of cedar (Cedrus) and fir (Abies) in layer 3. There would thus appear to be at least one or more climatic oscillations present in the cave sequence.

Grotte des Pigeons, Taforalt

Grotte des Pigeons is located in north-east Morocco (34°48′38″ N, 2°24′30″ W), close to the village of Taforalt in the Beni Snassen mountain range (Fig. 1). It is situated at an altitude of about 720 m and lies approximately 40 km inland from the Mediterranean coast. The cave was first reported in 1908 and became the subject of major excavations in 1944–1947, 1950–1955, and 1969–1977 (Roche 1953, 1963, 1967, 1969, 1976), with further investigations taking place during the 1980s (Raynal 1979–80; Courty et al. 1989).

The cave contains both Iberomaurusian and Middle Palaeolithic occupation evidence. Only the Iberomaurusian levels have been published in detail by Roche (1963). Based on his excavations, he was able to estimate that the archaeological deposits spanned a combined depth of over 10 m, with the Iberomaurusian layers occupying the top 2–3 m of the sequence. Since 2003 when our work at the cave began, we have re-sampled most of the sequence from standing profiles left by Roche and other previous excavators. As a result we have been able to obtain fresh archaeological and palaeoenvironmental evidence and now have well over 50 dates for the Iberomaurusian and underlying Middle Palaeolithic layers, based on a variety of independent techniques (AMS radiocarbon, OSL, TL and U-series dating). Several preliminary accounts of this work have been published or are in press (Bouzouggar et al. 2006, 2007; Barton et al. 2007; Bouzouggar et al. 2008).

The Iberomaurusian layers are most thickly developed on the south side of the cave where they form a series of dark grey deposits containing ash, charcoal and fire-cracked rock. Also contained in these multiple midden layers are large quantities of burnt shells of terrestrial land snail (including Helix) together with cut-marked and modified bone and vast amounts of lithic debitage and retouched tools. This ‘Grey Series’ sequence also contains evidence of human burials (Ferembach et al. 1962), mainly from near the back of the cave. Underlying the grey sediments are ‘Yellow Series’ deposits dominated by natural processes that reveal a fine-grained and very well stratified (laminar and lenticular bedding) sequence with evidence of individual and sometimes well separated Iberomaurusian occupation horizons.

Dating evidence for the Iberomaurusian based on single identified charcoals is presented in Table 1. The dates come from a continuous sequence in Sector 8 on the south side of the cave and demonstrate high integrity. Amongst other things, they reveal that, whereas the ‘Grey Series’ deposits accumulated very rapidly over a period of no more than 1.5 k radiocarbon years (beginning just below OxA-13478), the lower ‘Yellow Series’ cave earth sediments built up more gradually and perhaps with some erosive gaps. Archaeologically, the grey layers contain high percentages of backed bladelets with a full spectrum of straight, pointed and curved outlines and with evidence for frequent use of the microburin technique (Roche 1963; Barton et al. 2007). The artefacts in the ‘Yellow Series’ are in the course of study but indicate no great changes. Only in the lower levels (from about OxA-16273 17,515 ± 75 bp) are some differences detectable in the greater representation of flakes and marginally retouched blades and fewer tools made on bladelets.

Much of the analysis of the palaeoenvironmental data is in progress, however we can report on some trends already visible in the charcoal evidence, notably the fluctuating relationship between species such as Cedrus atlantica (cedar) and pines and evergreen oaks (Ward 2007). The former is a reasonably good indicator of ‘montane’ influences, since it only grows at higher altitudes today in Morocco. Evidence for cedar can be found near the base of the Iberomaurusian sequence in the ‘Yellow Series’ but higher up it disappears and there is a noticeable increase in woodland characterized by Aleppo pine (Pinus halepensis) and evergreen oaks (Quercus evergreen). Virtually absent throughout the ‘Grey Series’ cedar reappears once again at the very top of these sediments in layer G89, equivalent to OxA-13480 (Table 1). This later occurrence could be linked with environmental cooling and/or drying very close to the onset of the Younger Dryas. Another phase of relatively open and dry conditions is marked by the presence of Barbary ground squirrel (Atlantoxerus getulus) in the basal ‘Grey Series’. These animals shelter in burrows and prefer rocky habitats with isolated trees (Kingdon 1997). Their present distribution is restricted to the extreme west of Algeria, with a northern limit of around 33° N (Kowalski and Rzebik-Kowalska 1991), which implies slightly increased aridity in these layers.

Before closing this section, mention should be made of the site of Ifri el-Baroud (34°46′ N, 3°19′ E) in eastern Morocco (Fig. 1) where published radiocarbon dates on charcoal suggest an ensemble of Iberomaurusian deposits spanning the period between about 9.7 and 17 ka bp (Table 2) with an age of 16,777 ± 83 bp on charcoal for the oldest Iberomaurusian level (Eiwanger 2001; Görsdorf and Eiwanger 1998). Unfortunately no information is given on charcoal species or on the relative stratigraphic positions of each of the samples which are merely presented in chronological order. Nevertheless they seem to provide a very coherent series of dates and it is interesting to note that the lowest grey deposits of what is described as an escargotière show a remarkably similar date of inception (Bln-4748 12574 ± 65 bp) to the grey deposits at Taforalt. The most recent Iberomaurusian is attributed an age of 9677 ± 60 bp (Moser 2003, 100) but no explanation is provided for an even younger date from the site (Bln-4872). The same team has excavated a long sequence (3–4 m) of Iberomaursian deposits in the cave of Ifri n’Ammar (Fig. 1), also in the eastern Rif (Mikdad et al. 2004; Moser 2003). Here a list of calibrated and uncalibrated dates have been published but interpretation is hampered by an absence or incompleteness of information on the material dated and which samples were obtained by accelerator and conventional radiocarbon methods (Moser 2003, 101). There is unfortunately no information either on the calibration curve employed. The reported radiocarbon determinations range in age from the youngest at (Erl-4394) 10,022 ± 80 bp (11,653 ± 427 Cal bp) to (Utc-6180) 16,411–16,150 Cal bp.

Table 2 AMS determinations on charcoal from Ifri el-Baroud (after Görsdorf and Eiwanger 1998)
Full size table

An interesting observation at Ifri n’Ammar, which seems to corroborate our preliminary findings at Taforalt, is the gradual change in the retouched components from marginally retouched blades (so-called ouchtata blades) near the base of the sequence (enlèvement 28, as yet undated) to the appearance of backed points in enlèvements 25–24 and above which date from 16,411–16,150 Cal bp. The latter are made on bladelets and comprise 50% of the tool assemblage (Moser 2003, 100). The microburin technique is also in evidence from the same levels. The only other change noted is the occurrence of small geometric microliths (mainly crescents) which first appear in sediments dating to around (Erl-4399) 11,009 ± 144 bp (enlèvement 11). We have yet to analyse in detail the layers of equivalent age at Taforalt but it is noteworthy that similar geometric forms have been reported by Roche (1963) from the grey series sediments.

Discussion and Conclusions

The re-assessment of the dating evidence for Iberomaurusian sites in northern and eastern Morocco, though still relatively restricted, does not refute the model of population continuity in the Late Pleistocene in this region but at the same time neither does it rule out the possibility of some gaps in settlement, notably during the Younger Dryas interval. From the dating evidence presented here it can be deduced that a sustained phase of occupation began about 18–17 ka bp and continued with an intensification of cave use (at Taforalt and Ifri el-Baroud) from around 13 ka bp. In addition, though so far based on only one of our sites (Ghar Cahal), there is possible break in settlement from around 11–10 ka bp.

The possibility that internal typological and technological developments occurred within the Iberomaurusian is a question that cannot be addressed at present until further detailed analyses of individual assemblages are complete. According to preliminary results we would note only that the use of the microburin technique (a common by-product of microlith manufacture) occurs in some of the oldest Iberomaurusian levels at Taforalt and throughout the sequence at this site. The appearance and longevity of the microburin technique is a feature that was first recognised by Camps (1974, 76). Various questions arise about the evolution of the Iberomaurusian and whether the potential break(s) in settlement coincided with any modifications in the technology. While no obvious pattern is detectable in changes of microlith size through time in the studied samples, we would expect variations to emerge in the relative frequencies of different microlith shapes and also in the range and composition of microlith types. For example we would not be surprised if there was a greater representation of small crescents in some of the later assemblages. The only other potentially relevant characteristic is that, in the earliest layer so far recorded with Iberomaurusian artefacts at Taforalt in Sector 8, we have observed that blades, some with marginal retouch, are more common than bladelet forms.

In terms of the overall chronological framework, if we admit for the time being only the AMS determinations on single charcoals, it is fairly clear that the earliest Iberomaurusian is unlikely to date much before about 18 ka bp, at least in presently known cave sites in Morocco. This is an interesting observation because such a dating would suggest an appearance after a period of very sharp cooling in the global climatic record referred to as the Last Glacial Maximum (LGM).

By correcting the AMS determinations using the IntCal04 calibration curve (Reimer et al. 2004) we can also gain a clearer understanding of how the ages of these Iberomaurusian occurrences compare with climatic events recorded in the Alboran Sea cores (especially with MD95-2043) and more generally with Greenland ice δ18O core records (Figs. 3 and 4). Using this approach we can begin to form an idea of human presence in this region with reference to Dansgaard–Oeschger warming and cooling events as well as so-called Heinrich Events when cold polar waters penetrated into the Mediterranean (Cacho et al. 1999). Of particular significance are phases marked by severe decreases in sea surface temperatures (probably reflecting Heinrich Events) in the west Mediterranean (Figs. 3 and 4) which appear to be linked to increased aridity on the adjacent landmasses (Moreno et al. 2004). Viewed in this context it is noteworthy that the first appearance of Iberomaurusian follows a prolonged period of aridity (HE2 and Greenland Stadial 4). On the other hand, the climatic oscillations associated with Heinrich Event 1 do not seem to have unduly affected local populations, since occupation continued at least intermittently during this phase and was accompanied by no significant changes in subsistence or technology. This is particularly well illustrated at Kehf el Hammar where the palaeoenvironmental evidence points to a period of rapid climatic oscillation with humans present throughout and even during phases of increased aridity.

Fig. 3
figure 3

Iberomaurusian AMS dates for Ghar Cahal, Kehf el Hammar and Taforalt Sector 8 in relation to ice core and marine record data. The Cal bp AMS determinations are corrected using the IntCal04 calibration curve (Reimer et al. 2004) and the OxCal 4.0 calibration software (Bronk Ramsey 2001). The AMS dates are plotted against variations in δ18O composition of ice in the GISP2 (Greenland Ice Sheet Project 2) record; the relative abundance (%) of N. pachyderma in core MD95-2043 from the Alboran Sea, Western Mediterranean and; variations in alkenone-derived sea-surface temperatures (SST) in marine core MD95-2043. Also indicated are positions of Heinrich Events (HE1 and HE2) and the Younger Dryas cooling event (YD) according to Cacho et al. 1999

Full size image
Fig. 4
figure 4

Iberomaurusian AMS dates for Ifri el-Baroud in relation to ice core and marine record data. The rest of the caption is the same as in Fig. 3

Full size image

Later in the development of the Iberomaurusian, the more or less contemporaneous accumulation of massive midden deposits in caves such as Taforalt and Ifri el-Baroud is an interesting phenomenon that needs to be explained. Indeed, if this is a pattern reproduced at a regional scale, as we suspect, then it might signal significant changes in subsistence strategies at around 13 ka bp. Clearly, much more work now needs to be done to determine the strength of correlation between environmental change (and more importantly the rapidity or variability of change) and human behaviour. For example, when and why was human culture able to buffer against the effects of environmental fluctuation and is this reflected in changes in the uses of caves ? Finally, the idea of a break in occupational sequences corresponding to the last cooling of the Pleistocene (Younger Dryas) is something which needs to be explored further. Of relevance in this respect is whether the technologies that followed into the Holocene (at sites like Ghar Cahal) show any differences with those of the Latest Pleistocene.

In concluding this brief review it can be inferred on present evidence that microlithic bladelet industries of Iberomaurusian type made a fairly sudden appearance in this part of Africa soon after the LGM and not quite as early as previously asserted by Roche. However, it remains to be seen whether the technology originated in the Maghreb or outside this region, and whether its abrupt appearance can be linked to wider patterns of demic diffusion across areas north of the Sahara and/or in response to rapid climatic change (in this case to a rise in humidity following the LGM). We believe that in order to investigate this question more fully similar studies to the one outlined here will need to be conducted in adjacent areas of the Maghreb and in the Saharan south of Morocco.