The Study of the Past in Mozambique
Photo: Zimbabwe of Manyikeni, Mozambique, 1978: João Morais standing (foreground, left), Paul Sinclair (background right) and Ricardo T. Duarte (sitting). © J. Morais.
Note: This work is a review of scientific outcomes, in commemoration of the 'golden anniversary' (1975-2026) of the Department of Archaeology and Anthropology, Eduardo Mondlane University [Work in Progress]
Contents
The Source Material
Unfolding Landscapes
The Dawn of Humankind
Modern Humans
New Cultural Frontiers (late foraging communities.)
From Foragers to Farmers
Coastal Trading Towns
References
Africa is firmly established as the cradle of humanity, but understanding our evolution requires more than just analysing fossils and tools: a comprehensive approach must integrate environmental, cultural, and social data to explain how early humans adapted, dispersed, and behaved.
Modern, interdisciplinary, archaeology now uses precise dating (potassium-argon, uranum series, OSL), ancient DNA, LiDAR, alongside geological, botanical and anthropological data. The goal has shifted from simply excavating remains to explaining how complex landscapes shaped the trajectory of early human history.
The source material
1.1. Historical references
Portuguese explorers and settlers left a vast body of primary documentation for Southeastern Africa from the late 15th century onward. Through the eyewitness testimonies of government officials, missionaries, merchants, and navigators, these records offer invaluable insights into the social, cultural, and political traditions of local communities, covering the coast and hinterland between the Cape of Good Hope and Mombasa, including present-day Mozambique, Zimbabwe, and Tanzania. The documentary sources include several volumes of "Documentos sobre os Portugueses em Moçambique e na Africa Central" covering the period 1497 to 1560, the roughly contemporary narratives of the "Historia Trágico-Maritima" and the complementary volumes "Records of South Eastern Africa" and "The Portuguese in South Africa" (Theal 1964 and 1969). The first contacts of the Portuguese in the Cape Province of South Africa (1486) and in the Natal coast (1497-1499) reveal the presence of several hunter-gatherer and pastoral communities of Khoe-Kwadi speakers, and provide details on their appearance, housing, clothing, diet, implements and ornaments, as well the type of cattle and sheep they raised. Supplementary reference to the coastal to hinterland trade network describe a wider geographic distribution of foraging communities by 1512, at least as far as the Zimbabwean plateau, which may also have included BaTwa foragers in areas North of that region. Historical Portuguese records have furthermore been used to analyse historical climatology for decadal-to-seasonal-scale reconstructions over the past 500-year climate variability and societal responses to historical extreme events.1
Competition between the newcomers and the influential Arab mercantile establishments on the coast at Sofala and Mozambique Island illustrate the fact that pre-Portuguese interchange was already consolidated among most of the seaside populations particularly North of Sofala, transactions which are documented by early arab writers at least from the 9th to the 15th centuries (e.g. Al-Masʿūdī, Al-Idrīsī, Ibn Battūṭa). From late 15th century several Portuguese records provide first-hand information (e.g. ‘the fortress of the King of Menomotapa [...] made of stone [...] without mortar’, as described by António Fernandes in 1511) while reporting on trading activities along the south-eastern coastal areas and, particularly, from Sofala to the Zimbabwean plateau. Such descriptions highlight how farmers and their rulers operated, along with information on production, distribution, technological systems and crafts, including housing, settlement patterns, mining procedures, trading routes as well as political competition, namely between the Mwene Mutapa and Rozvi states struggling for regional leadership. Furthermore, the sources also contribute in interpreting long-lasting historical processes among major present-day ethnic groups. This is the case for the connotation of a Karanga (Shona) minor entity among the Thonga in the Inhambane region. At least one document dated from 1501 establishes dissimilarities between the "Mocaranga" and the "Botonga", which propounds the former as an earlier and specialized community still surviving in the contemporary Guambe chieftainship in southern Mozambique. Continental trade included copper in exchange for cloth, which was obtained at the coast in exchange for ivory. Beyond elephant hunting, regular livelihood included agriculture mainly performed by women, cattle-raising, and fishing with spears in dug-out canoes. Among some of the most distinct features for the Thonga was the manufacture of fabrics and containers out of tree bark and the utilization of bows and arrows which may suggest an ancient cultural presence of Khoe-Kwadi speakers among this group. Further South in the Maputo Bay and inland region the early accounts cover the period 1552-1554 and provide details concerning trade (mainly ivory beads and cloth), housing and settlement patterns, valuable to the the study of the evolution of chieftainships such as the Inhaca, "Zembe" (Tembe), "Rumo" (Fumo), "Domanhica" (Manhica) and Lebombo.
1.2. Early archaeological and ethnological work
First-hand Portuguese documentation reflects emerging geopolitical interests as well as a sense of ‘Renaissance curiosity’ that the early travellers had towards new civilisations being ‘discovered’ to the world. Beyond their written accounts several ethnographic collections would have reached Portugal and were offered by the crown to the ‘cabinet of curiosities’ assembled at the Vatican by the influential naturalist Michele Mercati (1541-1593). The existence of the relics, only disclosed in 1717, may in turn explain the enthusiastic announcement four years later (1721) by the Bishop of Mozambique to the new Royal Academy of Portuguese History in Lisbon the discovery of pictographs in his diocese, the first known announcement of African rock art to the world.
This extemporaneous interest will only be revived much later in the 19th century as part of the European "scramble for Africa" when ethnographic accounts regain momentum. However, contrary to what one would expect, the Portuguese colonial take-over did not prioritise cultural heritage management in the same way as the neighbouring British colonies, regardless of the establishment in Mozambique in 1943 of a somewhat ineffective ‘Commission on Monuments and Historical Relics’. The reason for this attitude lies in the modest role of archaeology and museology within Portugal itself as the metropolitan power. It is also rooted in the lack of scientific training and formal institutions, and the perceived absence of spectacular Mozambican monuments, such as Great Zimbabwe, or early human fossils comparable to the 1924 Taung find in South Africa. Until 1930, research was limited to surveys of border regions carried out by foreign scholars, who focused specifically on rock paintings and associated assemblages such as the ones at Chifumbaze shelter in Tete and the Niamara stone wall in Manica2. Following Salazar’s nationalistic reforms and influenced by the 1st Congress of Colonial Anthropology (Porto, 1934), the newly formed 'Mozambican Ethnographical and Anthropological Mission' conducted six fieldwork campaigns between 1936 and 1957 3 which occasionally included collaboration with visiting archaeologists from South Africa 4. More importantly, during the two decades leading up to Mozambican independence in 1975, public interest stimulated a surge in fieldwork initiatives by a growing number of scholars, laying the groundwork for subsequent research 5.
1.3. A New Vision of the Past
To assist with scientific rescue work at the Massingir Dam site the first Archaeological Unit is established in 1973 at the Department of Earth Sciences, Instituto de Investigação Científica de Moçambique (IICM) 6. Following independence in 1975 African Prehistory is introduced into the university syllabus 7 and new history schoolbooks, regular reporting on research outcomes is provided to the media and the public, namely with the printing of a postal stamp collection displaying major archaeological sites. Considering the absence of museology dealing with the human past - beyond the Natural History Museum founded in 1913 in Maputo and the decaying National Museum of Ethnography in Nampula founded in 1956 – the new National Museums and Antiquities Services is created in Nampula and archaeological surveys initiated in the province 8. The Archaeological Unit assisted with specific capacity-building tasks in support of the new ‘Agentes de Cultura’ (district culture delegates).
'Archaeological Sites of Mozambique' Postal Stamp Collection, Moçambique Postal Office 1981 (Image © J.Morais)
In 1976 the African Studies Centre (CEA) is founded at the IICM and the Archaeological Unit placed under its administration, taking over the premisses of the extinct Earth Sciences department and associated laboratories 9. The Unit’s emerging archaeological ‘vision’ aimed to contribute to the new nation’s requirements for: (i) a knowledge-based historical identity; (ii) evidence-based education; and (iii) national cultural awareness. In support of these goals, multiple initiatives were launched to define a pioneering archaeological survey programme, with the primary objective of documenting Mozambique's extensive precolonial heritage.10. Among other initiatives, important work was initiated at Matola and products of the emerging investigation made available through extensive public media and scientific reporting 11.
In 1977 on the grounds of non-complementary research agendas and methodologies, the governance of the Archaeological Unit is transferred to IICM and soon thereafter instituted under the Faculty of Arts of Eduardo Mondlane University (UEM) as the new Archaeology Department 12. During two seasons (1976-1977) research work is conducted at the stone enclosure at Manyikeni as new archaeologists join in from abroad 13. During 1978 a major Swedish research grant (the first of UEM's external funding for subsequent decades) is allocated for the ‘Archaeological Survey of Mozambique’ project. The initial Swedish partners are the Swedish National Heritage Board and, from 1980, Uppsala University. Among other sites (e.g. Matola, University Campus, Zitundo) fieldwork is conducted at Manyikeni 14 where an Open-Air Museum was inaugurated under increasing war risk, from landmines to enemy (i.e. Rhodesian) helicopters flying for surveillance over the site and the museum building burned down during civil war in the late 80’s. A dedicated effort is made to continuing providing information to schoolbooks and media reporting. Over this period extensive surveys are also implemented in the Save-Vilanculos bay region and Niassa 15. Considering the growing need of interdisciplinary research and in the absence of a ‘home’ at UEM, the Department of Archaeology and Anthropology (the present DAA) is established in 1979 and ‘Trabalhos de Arqueologia e Antropologia’, the new departmental publication, is launched 16 . During two seasons in 1978 and 1979 archaeological fieldwork is carried out in the northern coastal regions at Murrupula, Mossuril, Ilha de Moçambique and Ibo 17.
Manyikeni Museum Brochure, October 1979 (Image © J.Morais)
During the period 1980-1985 the Swedish National Heritage Board assigns to DAA, for limited time-periods, osteologist Leif Jonsson, marine archaeologist Per-Inge Lindqvist and archaeologists N. and G. Nydolf. Late in 1983 Ricardo T. Duarte returns to Maputo, joining the DAA to conduct research in northern Mozambique. During this period, when survey work more than doubled previously known sites, new fieldwork is implemented in the south at Daimane, Ponta Dundo, Zitundo, University Campus, among others. To complement T. Cruz e Silva’s groundwork at Matola, the site is re-excavated by J. Morais, L. Jonsson and P-I. Lindqvist, aiming at enlarging collections and reassess dates. Following earlier work in Northern Mozambique (T.C. Silva and P. Sinclair), L. Adamowicz conducted archaeological surveys at 92 sites, namely at Chakota, Riane, Muse and Namolepiwa rock-shelters, the Nakwaho site providing the first occurrence of grindstones associated with early farming. In 1981 the first computer is installed (an IBM PC 5150) and a thematic library regularly updated. Notwithstanding considerable progress, particularly later into the 80’s with increasing civil war and unrest (1977-1992), the fieldwork component of the 'Archaeological Survey Programme' gradually comes to a halt. The head of department was João Morais until leaving Mozambique in 1884, being replaced by Ricardo T. Duarte.
1.4. Research work in the last few decades
From 1980, in the challenging context of the civil war (1977-1992), Mozambican archaeology underwent a transition from broad regional surveys (the ‘Archaeological Survey Programme’) to more specialized site excavations closer to urban centres away from the conflict zones. This period also provided the context for critical review of archaeological theories and methods 18. Despite security constraints, research into the origin of the early peopling of Mozambique continued in southern, central and northern regions. Notable southern sites include Matola, University Campus, Chongoene, Xai-Xai, Bilene, Siaia, Caimane, Zitundo, Inhaca, Tembe and Nhachengue, the central sites of Manyikeni and Chibuene, and the northern sites of Xakota, Nakwaho, Riane, Muse, Muhekane, Namikopo, and Namolepiwa, to name a few.
Following the end of the civil war, Mozambique has experienced a steady, three-decade buildup of human and scientific capacity in archaeology. This progress began in the 1990s with the introduction of a BA in Archaeology, which has included a Cultural Heritage Management component since 2011. A landmark development occurred in 1994, when the government enacted Law No. 2/94, establishing legal protections for cultural heritage and requiring developers to fund rescue archaeology for projects affecting archaeological sites. This legislation significantly transformed the sector by creating a commercial market for private archaeological companies and requiring a high level of academic competence from civil servants working within the National Directorate for Cultural Heritage and associated national archive (ARPAC). While UEM remains the primary institution providing an undergraduate degree in archaeology, the Department of Archaeology and Anthropology (DAA) has established an active research node on Mozambique Island—the Centre of Archaeology, Research, and Resources of Mozambique Island (CAIRIM)—which is notable for its work in maritime archaeology. Additionally, a dedicated Museum of Archaeology, opened in 2017, is curated by the DAA.
Postgraduate archaeological training (MA and PhD) has grown significantly, marked by a steady increase in alumni—including João Morais (1987), Ricardo T. Duarte (1993), Paula Meneses (1999), Solange Macamo (2005), Hilário Madiquida (2015), Albino Jopela (2017), Mussa Raja (2020), Celso Simbine (2021), Cézar Mahumane (2024), Omar Madime (2024), and Décio Muianga (2025). This progress is rooted in long-standing cooperation with Swedish aid (Sida) and Uppsala University (notably P. Sinclair and A. Ekblom). The network has now expanded to include partnerships with UNESCO (underwater heritage at Mozambique Island), NORAD, and the German Archaeological Institute, alongside research support from the universities of Algarve, Bergen, Calgary, and Hamburg, under the guidance of N. Bicho, T. Saetersdal, J. Mercader, and N. Babucic. These international partnerships remain essential for training the next generation of Mozambican archaeologists and anthropologists.
2. Unfolding Landscapes
Understanding modern-day Mozambique offers a crucial foundation for interpreting its deep history, from the origins of humanity to the development of societies along the Eastern and Southern African coasts. Although shifting climates and human activity have constantly reshaped the landscape, its core features—ranging from coastal mangroves and sand dunes to interior savannas, woodlands, and western highlands—have provided a consistent, enduring setting throughout the prehistoric timeline.
Examining this natural history allow us to identify the plants and animals that sustained human populations from early foraging communities through to historical societies. While modern isotope analyses of paleontological bone and tooth have proven vital by offering direct evidence of diet and mobility, the archaeological record is often skewed. Because bone preserves significantly better than plant remains in open-air sites—a process accelerated by acidic soils, high temperatures, and excessive rainfall—traditional interpretations have often overemphasized 'man-the-hunter' at the expense of 'man-the-gatherer.' Consequently, the best-preserved, stratified assemblages are found in caves and rock-shelters, which unfortunately represent only a minority of archaeological sites in Mozambique.
2.1. Geography
Topographically, Mozambique is highly diverse. Situated in Southeast Africa, the country is defined by a 2,700 km coastline—one of the longest on the continent— and a tropical savanna climate characterized by distinct wet and dry seasons. The landscape is characterized by extensive littoral lowlands, wetlands, and mangrove swamps that dominate roughly 45% of the country, primarily in the south and along the coast. Moving inland, the terrain rises into plateaus and hills (200–500m), followed by elevated meso-plateaus (500–1000m) in the north and centre. Western areas along the border are characterized by highlands (>1000m), with mountainous regions including the highest point, Monte Binga (2 436m). The drainage is shaped by major rivers, including the Zambezi, Limpopo, Rovuma, Save, Lúrio, Pungwe and Búzi which drain into the Indian Ocean across vast, flood-prone alluvial plains. Due to this geography, two-thirds of the population resides in coastal provinces, where livelihoods have traditionally been centred on marine and riverine resources.
Influenced by monsoons, latitude, and elevation, the climate is predominantly tropical to subtropical, alternating between a hot, rainy season (November to April/May) and a cooler, dry season (May to October). Regionally, conditions vary: the North and central highlands are warmer and wetter, with rainfall exceeding 1 000mm due to the Inter-tropical Convergence Zone (ITCZ). In contrast, the South and interior are warmer, drier, and experience more significant drought conditions. While low-lying coastal areas stay hot and humid, the inland western plateaus and northern highlands provide more temperate, milder weather. Overall, precipitation decreases as you move south or inland. In terms of agricultural resources, the northern and central regions have fertile soils for cassava, cashew and maize, while the south presently faces erratic rainfall and recurrent droughts.
Vegetation is divided into two primary ecological zones, shaped by climate, soil, topography, and anthropogenic factors such as fire and deforestation: (i) the Zambezian Domain, featuring the dry forests and grasslands of the Zambezi basin, characterized by Miombo woodland, Mopane savanna woodland, and Acacia tree savanna; and (ii) the Indian Ocean Coastal Belt, a subtropical region stretching south toward South Africa. High population density along the coast is driving significant ecological transformation, converting native mangroves and dry forests into a mosaic of woodland and savanna. The Zambezian Domain now occupies approximately 75% of the region, having gradually replaced what was once the Indian Ocean Coastal Belt due to human pressure.
MAP: Mozambique Vegetation based on Wild and Fernandes 1967, and Werner 1979 (in Morais 1988:20)
2.2. The Human-Environmental footprint
Modern vegetation provides a baseline for reconstructing the ancient environments that shaped human society. Key considerations include past paleoclimatic variations, extensive landscape changes, and associated vegetation shifts, especially regarding freshwater availability. As climate change shifted ecosystems, shorelines, and water sources over million years, these paleo-landscapes became central to the co-evolution of communities and wildlife. Tracking these environmental shifts across geological epochs reveals human ingenuity, as hominins constantly had to adapt to fluctuating waterways, coastlines, unsteady habitats and uncertain livelihoods.
By analysing past hydroclimates, vegetation cycles and faunal diversification, we gain a clearer understanding of how deep-time landscape transformation—particularly during periods of water scarcity—impacted ancient ecosystems, food security, and settlement patterns. Analysing these shifts within the framework of geological epochs allow for a better understanding of their impact on ecosystems, human societies, and food resources. Ultimately, settlement patterns highlight how humans time and again redefine their existence in response to significant changes, such as deep-time emergence and submergence of coastal plains.
Research methods involve analysing natural archives—such as tree rings, speleothems, pollen, lake/marine sediments, and ice cores—utilizing rigorous radiometric, chronostratigraphic, and numerical dating tools. While these methods are robust, substantial uncertainties remain regarding the precise causes and consequences of these changes in deep geological time.
Key epochs in human evolution include: (i) the Pliocene (5.3–2.6 Myr 19) marking the emergence of early hominins like Australopithecus; (ii) the Pleistocene (2.6 Myr to 11,700 years ago,), witnessing the evolution of the Homo genus, from H. habilis/erectus to H. sapiens; and (iii) the Holocene (11,700 ya–to present), when H. sapiens transitioned from foraging to agriculture and urbanization. Transitions between these stages were heavily influenced by abrupt climate shifts, driving significant environmental and biological adaptations, and ultimately lifestyles.
Grappling with deficient terminology: biased by stone tool technology and stratigraphical considerations —ignoring the broader role of livelihoods and lifestyle—African archaeologists use Early, Middle and Late Stone Ages (abbreviated to ESA, MSA and LSA) when referring to prehistoric foraging communities. For subsequent and more recent periods, shifting the focus from tool technology (‘Iron Age’) to a lifestyle, the author promoted from 1980 the use of Early and Late Farming Communities (EFC, LFC) which is now accepted terminology.
Global climate of the last 5 million years has seen major transformations and dramatic shifts. The Pliocene epoch (5.3–2.6 Myr) was overall warmer than modern times, with significantly higher sea levels. However, this period was interrupted by brief, sharp cooling events (‘glaciations’) at 4.9, 4.0, 3.6, and 3.3 Ma, leading to temporary sea-level falls. Be as it may, coastal forests along the Indian Ocean could have provided a refugium for early hominins in periods of variable climate with strong seasonality. From this coastal refuge hominins would disperse inland via vegetated humid corridors during stable climate periods.20
During the mid-Pliocene warm period (approximately 3.3–3.0 Myr), the region was generally warmer and characterized by savanna-woodland mosaics, with sea levels peaking 15–30 meters higher than today.21. While Australopithecus afarensis thrived for nearly a million years (c. 3.9 to 2.9 Myr) within the tectonically active landscapes of East Africa, Southern African Australopithecus africanus (c. 3.5 to 2.0 Myr) 22 adapted to a similar mosaic savanna-woodland and riverine forests, increasingly utilizing cave-associated environments as climates changed, where their fossils are now found. But more on this later.
Relative climatic stability declined between 2.7 and 2.5 Myr, coinciding with the intensification of Northern Hemisphere glaciation and the Pliocene–Pleistocene transition. This shift initiated an era of progressive aridification and high-amplitude oscillations across Southern Africa, driven by rigorous glacial-interglacial cycles. As glaciations intensified from approximately 2.6 Myr, forests contracted into grass-dominated savannas, triggering a critical shift in faunal communities and food resources, while sea levels dropped as much as 125 meters below present markers at the Limpopo River shelf.
During periods of increased aridity in the interior, human populations likely increased their reliance on the coastline, utilizing shellfish and marine intertidal resources. Simultaneously, the exposed—and now submerged—continental shelf, formerly a vast grassy plain, would have provided vital foraging and hunting opportunities to supplement the gathering of grasses, fruits, tubers, and roots. The period between 2 Myr and 1.25 Myr marks a critical evolutionary juncture characterised by the coexistence of at least three distinct hominin genera living in proximity: Australopithecus sediba, Paranthropus robustus, and early Homo erectus. Ultimately, these open-land environmental fluctuations likely favoured the superior behavioural and technological adaptability of Homo over the highly specialized dietary niche of Paranthropus, which may explain its gradual extinction.23
The evolution of glacial-interglacial cycles between 1.25 Myr and 600 000 ya, the so-called ‘Middle-Pleistocene Transition’, represents a global climate shift from milder and more frequent 40 000-year cycles to longer and stronger 100 000-year glacial cycles. In southern Africa, the glacial-interglacial variability ranged from cooler and arid conditions with expanded grasslands (glacial periods) to relatively warmer and wetter conditions with expanded coastal forest and woodlands (interglacial periods) which profoundly impacted faunal communities and influenced the development and dispersal of early humans particularly in areas of open grassland, lakes and rivers. During this period local sea-levels fluctuated from c. 120–130-meter lowstands, exposing extensive areas of the continental shelf, to highstands (during interglacial periods) about 15-20 meters above current levels.
The subsequent Middle Pleistocene epoch (c. 774 000 to 129 000 ya) marks a decisive evolutionary phase characterized by the transition from Early to Middle Stone Age foraging strategies and the emergence of archaic Homo sapiens lineages in the period around 700 to 400ka24. This evolutionary process was likely driven by dramatic glacial-interglacial cycles, where phases of intense aridity pressured populations to adapt to rapidly changing landscapes through significant technological innovation.
The Late Pleistocene (c. 129,000–11,700 ya) represents the dramatic transition from the warm Last Interglacial to the stable, modern Holocene epoch. This era was characterised by extreme climatic shifts. During the peak of the Last Interglacial (c. 130 000–115 000 ya), the climate was warmer than today, with sea levels 4 to 9 meters higher. In contrast, during the Last Glacial Maximum (c. 26 000–20 000 ya) the climate became cold and dry, with temperatures 5–6°C lower than at present, causing sea levels to plummet 120–130 meters below modern levels. This massive drop exposed vast tracts of the continental shelf, creating new coastal foraging landscapes while severely reducing resources in the arid hinterland. Although subsequent melting restored warmer conditions and allowed humans to inhabit the Mozambican coastline, this warming trend was briefly interrupted by the ‘Younger Dryas’ event (c. 12,900–11,700 ya). This final cold, dry snap functioned as the ultimate transition into the stable, warm climate of the Holocene.
A recent marine pollen record from southern Mozambique (Delagoa Bight) covering the past c. 16 600 years generally correlates with established regional paleoclimate scenarios 25. The data indicates that prior to 15 000 ya, cooler conditions prevailed, characterized by a prevalence of shrub vegetation and minimal forest or woodland cover. From c. 13 000 ya onwards, a gradual increase of savanna taxa suggests climate warming, aligning with the end of the last ice age and the inception of the Holocene period.
Woodland expansion occurred across Southern Africa between 11,200 and 9,200 ya, driven by increased moisture that supported coastal forest spread during a period of higher sea levels. Though this moist climate persisted into the mid-Holocene (5 000-6 000 ya), records indicate a significant shift toward arid conditions in Southern Africa from c. 3,500 ya, causing a decline in coastal forests.
A similar trend is also noted from 5 000 ya with the desiccation of the Sahara/ Sahel, the fragmentation of the West-Central African rainforest and savanna expansion in the Congo basin. These environmental shifts likely provided multi-directional migration corridors through forest and savanna, facilitating the southward movement of small groups of Bantu-speaking farmers into southern and eastern Africa. 26 Nevertheless, the water-for-crops requirements of these early farmers made them more vulnerable to drought than the indigenous foraging communities, who experienced population growth in the Eastern Cape region from 5,000 to 2,000 ya.
2.3. Winds of Change
Over the past two millennia, southern Africa has experienced significant, spatially heterogeneous climate variability, driven by alternating wet and dry conditions. High-resolution proxy data, such as cave speleothems and shell middens, indicate that a "Medieval Warm Period" (c. 900–1300 CE, ‘Current Era’) brought warmer, wetter conditions—with peak temperatures potentially 3°C higher than present—followed by the "Little Ice Age" (c. 1300–1800 CE). The latter period featured maximum cooling around 1500–1700 CE, including a notable ~1°C drop in mean annual temperature which, together with increased farming and forest cleaning from around 1000 CE triggered the decline in woody cover27.
Reconstructions of climate variability in Eastern and Southern Africa are consistent: derived from multi-proxy datasets including lake sediments, speleothems, tree rings, and ice cores (Kilimanjaro and Mount Kenya)—reveal a highly dynamic environmental history. High-resolution sediment cores from Lake Nhaucati near the Chibuene site in southern Mozambique have provided high-resolution data covering the last 1 600 years indicating a significant vegetation shift from a savanna-forest mosaic towards more open woodland.28
The arrival of farming communities in Mozambique around 2 000 ya was facilitated by a warm, wet climate (c. 40–400 CE) that favoured settlement within a mosaic of coastal forests, riverine woodlands and savannas and sea-levels like present-day; these conditions become increasingly variable after 800 CE. Subsequently, a shift to cooler and drier conditions—peaking between 1360 and 1560 CE—drove widespread ecosystem changes that significantly impacted major political centres like Mapungubwe (c. 1075-1300 CE), and Great Zimbabwe (c. 1250-1450 CE). Their demise was likely driven by a drought-induced resource depletion, further compounded by socio-political stressors, including shifting trade routes and intensified geopolitical competition. Such pressures would have also impacted the Manyikeni Zimbabwe settlement near the Vilankulos coast, which was occupied between c.1200 and 1700 CE.
A dramatic reduction in coastal forest cover occurred between 1400–1600 CE, which aligns with severe drought conditions associated with the 'Little Ice Age'. This suggests that climate shifts, rather than simply anthropogenic factors, were the primary driving force of ecological change during this period. Beyond archaeological, linguistic and genetic archives, multidisciplinary evidence—including rock art, oral traditions, and Portuguese records from 1505 CE—details human-environmental interactions, providing information on how climate extremes (such as droughts, floods and disease outbreaks) triggered famine, population displacement, and migration.
3. The Dawn of Humankind: the earliest southern Africans
Who were Mozambique’s earliest inhabitants? How did they lived and evolved, which tools did they used or manufactured? Learning from the tools they left us, they were hominins who succeeded through gradual evolutionary adaptations. By developing the ability to walk and run upright, they freed their hands to create tools rather than relying on using natural ones. Combined with increased intelligence and a diverse diet, these adaptations allowed early Homo to survive harsh environmental changes and outcompete other species. Although no fossils have been discovered in the country yet, the abundance of early stone tools makes Mozambique a crucial, high-priority frontier for paleoanthropological research.
Situated at the southern terminus of the Great Rift Valley, Mozambique likely served as a vital migratory corridor between East African tropical landscapes and temperate southern regions. While locations such as the Niassa Rift shoulders, Urema Rift, Cheringoma Plateau, Gorongosa Massif, and the karstic caves of the Buzi and Buchane rivers hold immense potential, finding well-preserved fossil sites is challenging. 29 Like South Africa, many of these fossil-bearing caves are severely eroded. Furthermore, systematic research is hampered by dense vegetation, intense chemical weathering, and a lack of volcanic layers for precise chronometric dating—unlike the northern rift valleys. Consequently, until more Mozambican data emerges, our understanding remains largely dependent on comparative evidence from South and East Africa.
Although this potential was recognized decades ago, the Mozambican Civil War (1977–1992) prevented planned collaboration between the International Louis Leakey Memorial Institute in Nairobi and the Department of Archaeology and Anthropology in Maputo. In 1978, at Richard Leakey’s invitation, the author visited the Koobi Fora sedimentary outcrops at Lake Turkana. It was there that Leakey had discovered two landmark Homo erectus fossils (KNR-ER 3733 and 2598) in 1974 and 1975. Dated to 1.63 and 1.85 Ma respectively, the latter remains one of the oldest H. erectusspecimens, surpassed only by the South African Drimolen ‘Simon’ find (DNH 134), dated to approximately 2.04 Ma30.
The fossil record in South Africa over 2 million years is complex. Evidence is often fragmentary and taxonomic classification lively debated. The coexistence of Australopithecus sediba, Paranthropus robustus, and early Homo erectus between 2 and 1.25 Myr, against a backdrop of high faunal diversity, highlights a period of significant environmental instability and territorial partitioning in a mosaic of habitats explored by three distinct genera. Under these conditions, the superior resourcefulness of Homo likely outperformed in time the specialized (vegetarian) dietary strategy of Paranthropus and the inability of Australopithecus to adapt to the increasingly loss of woodland habitats being replaced by expanding savannas.
Simplified phylogeny of hominins, in John Hawks, “Secrets within the teeth of the first Homo fossils” https://www.johnhawks.net/p/homo-habilis-edj-morphology-australopithecus
The archaeological presence of multiple hominin species at this early date makes it difficult the attribution of specific stone tool traditions. Fossil records and archaeological data are often difficult to fully match. While it is expected that communities assigned to Homo would have acquired a higher skill of toolkit production, there is evidence that the South African Paranthropus robustus might also have been able to manufacture stone tools.31
Because organic materials like bone, wood, and skin seldom survive, stone tools are the primary record of early human history. African prehistory is traditionally divided into three main eras based on tool technology, stratigraphy, and age: the Early, Middle, and Late Stone Ages (ESA, MSA, LSA). Narrowly focused on lithic technology over lifestyle, such periodisation defines broad geographical and temporal "technocomplexes" and specific regional "industries"—assemblages with shared production and design, often named after a type-site. Key Southern African ESA examples include the Oldowan (c. 2–1.5 Myr), the Acheulean (c. 1.5 Myr – 300 ka), and the transitional Fauresmith (c. 600–200 ka), illustrating overlapping, evolving traditions. 32
Oldowan stone tools from Bed 1, Olduvai Gorge, Tanzania, flake tools (1-4), polyhedron (5) small chopper tools (6 & 7), large chopper tools (8 & 9), discoids (10-12), in R.R.Inskeep 1978:41
Although mostly undated, Mozambican Early Stone Age sites reflect well the typology of southern African sites where pioneering hunter-gatherers evolved over 2 million years ago. The archaeological record shows a lengthy technological transition from simple Oldowan pebble tools to more complex Acheulean handaxes and picks. The subsequent emergence of smaller tools —such as large blades and convergent points— indicate the transition to the Middle Stone Age (MSA), s period likely associated with the appearance and non-linear evolution of archaic Homo sapiens between 700-400 ka across the African continent 33.
A recent study of over 650 Stone Age sites in Mozambique identifies clear settlement patterns based on elevation: 93 Early Stone Age sites are located primarily at low elevations, 252 Middle Stone Age (MSA) sites at higher elevations, and 197 Late Stone Age (LSA) sites at intermediate heights 34. While climate, shelter, and natural resources likely shaped these choices, the concentration of ESA sites along the coast and riverine floodplains suggests a specific, highly mobile, habitat preference. Unlike South Africa’s elevated inland caves, river valleys and hilltops, it is expected that (as noted above 2.2) the increasing aridity and the shift from woodlands to savannas—combined with a 125-meter drop in sea levels 2.6 million years ago—may have forced communities to rely seasonally on resources now submerged on the continental shelf.
3.1. Life in the Deep Past ... (TO BE CONTINUED)
1 Nash, D.J. and M.J. Hannaford “Historical climatology in Africa: a state of the art”, Pages Magazine 28 (2), (2020): 42-43.
2 Carl Wiese in 1907, with pottery described in the 1970’s by David Phillipson as the *Chifumbaze complex’; and H. A. Wieschoff in 1930, as part of the 9th Frobenius Institute expedition to Niamara; There is however an earlier report in 1915 by E.J. Wayland regarding an odd stone tool from Búzi river sent in 1913 to J. Leite de Vasconcelos (director of the Ethnographic Museum officially opened in 1906 in Lisbon).
3 Prominent members included J.R. Santos Júnior and A. Mendes Correia; notably, Jorge and Margot Dias conducted seminal work among the Makonde in the late 50’s.
4 van Riet Lowe, L.H. Wells and the French scholar H. Breuil.
5 Major names include M. Bettencourt Dias, Lereno Barradas, Simões Alberto, Rosa de Oliveira, A. Rita-Ferreira, R. Dickenson, G. Liesegang, G. Smolla, M. Korfmann, R. Derricourt, Soares de Castro, Amaro Monteiro, C. Ervedosa and M. Ramos, among others.
6 The Unit was established by Prof. G. Soares de Carvalho and included Teresa Cruz e Silva, Ricardo and Maria da Luz T. Duarte, João Morais, all involved with the Massingir project; and briefly J. Senna Martinez (b. 1948 - d. 2022) before leaving the country in 1976. The work was published in Memórias IICT,10-B, 1974. Professor Revil Mason provided valuable guidance from Witwatersrand University where R.T. Duarte and J. Morais carried out comparative studies of archaeological collections during a two-month scholarship.
7 In parallel with their work at the Archaeological Unit, João Morais and Ricardo T. Duarte were appointed junior lecturers of African Prehistory at the History Department, Faculty of Arts, where Teresa Cruz e Silva also lectures from 1976.
8 Ricardo and Maria da Luz T. Duarte pioneered this initiative from 1976, rescuing the National Museum of Ethnography in Nampula from critical plundering of its assets. The Museum was originally created by A.J. Soares de Castro in 1956 following seminal work by Jorge and Margot Dias among the Makonde.
9 Aquino de Bragança was appointed director of the CEA and the Archaeological Unit integrated as an additional research body with João Morais as head of unit. The present department of Archaeology and Anthropology inherited the premisses of the extinct Earth Sciences Department.
10 for a detailed description of the undertaking see J. Morais, The Early Farming Communities of Southern Mozambique, 1988: 48-58.
11 c.f. J. Morais, Prehistoric Research in Moçambique: the earlier research, the present projects, plans and proposals, 1976, which includes reports from T. Cruz e Silva, R. Duarte and J. Senna Martinez.
12 with agreement from director Aquino de Bragança and rector Fernando Ganhão the governance of the unit is placed under IICM’s director Prof. Pedro Alcântara and soon thereafter transferred to the Faculty of Arts (director Manuel Araújo) as a new Archaeology Department, a transition process managed by J. Morais.
13 The Manyikeni site is excavated by J. Morais and P. Garlake. The new co-workers included Michael (b. 1949 - d. 1999) and June Stephen starting from 1976 to 1979 (before leaving for Swaziland in 1979), Paul Sinclair (b. 1949 - d. 2023) starting in 1977, previously working in Zimbabwe, who acted as principal investigator until leaving in 1980 for Uppsala University, and Leonard Adamowicz (b.1945 - d. 2020) starting in 1980, from Poland.
14 by J. Morais and P. Sinclair
15 by T. Cruz e Silva, P. Sinclair, A. Loforte and, for shorter periods, J. Morais at Chibuene, Hola-Hola, Bazaruto, Ponta Dundo and Marrape sites while concurrently M. and J. Stephen conducted survey work at Maua cave in Niassa.
16 Ana Loforte and Christian Geffray (from France) join DAA. During the academic year 1979-1980 J. Morais initiated doctoral work at the University of Oxford, and Ana Loforte replaced him during the period as head of department.
17 Work implemented by T. Cruz e Silva and P. Sinclair
18 See J. Morais “Mozambican Archaeology: Past and Present.” African Archaeological Review 2 (1984)
19 throughout this document I will use Ma (mega-annum) for absolute dates and Myr (million years) for time durations
20 Joordens Josephine C.A., C. S. Feibel, H.B. Vonhof, et.al, “Relevance of the eastern African coastal forest for early hominin biogeography”, Journal of Human Evolution, 131, (2019: 176-202. And, Carvalho, S., V. Aldeias, Carvalho, S., V. Aldeias, Z. Alemseged, et.al. “Gorongosa National Park and the biogeography of human origins in the Mio-Pliocene”, Proceedings of the European Society for the Study of Human Evolution, 6 (2017): 36.
21 Hearty, P. J., A. Rovere, M.R. Sandstrom, et.al. “Pliocene‐Pleistocene stratigraphy and sea‐level estimates, Republic of South Africa with implications for a 400 ppmv CO 2 world”. Paleoceanography and Paleoclimatology, 35, (2020): 1-23.
22 Recent dates from Madupe, P.P., et al. “Results from an Australopithecus africanus dental enamel fragment confirm the potential of palaeoproteomics for South African Plio-Pleistocene fossil sites”. South African Journal of Science, 121(1/2), (2025); and Thackeray, F. A “biochronological date of 3.6 million years for “Little Foot” (StW 573, Australopithecus prometheus from Sterkfontein, South Africa)”. Evolutionary Anthropology: Issues, News, and Reviews, 33(6), (2024).
23 C.f. evidence from the Drimolen palaeocave in South Africa, c.f. Herries, A. I. R., J. M. Martin, A.B. Leece, et. al.) “Contemporaneity of Australopithecus, Paranthropus, and early Homo erectus in South Africa.” Science, 368 (6486), (2020).
24 Stringer C. “The origin and evolution of Homo sapiens.” Philos Trans R Soc Lond B Biol Sci. 371(1698), (2016)
25 Neumann, F.H., J. Finch, A. Hahn, et.al. “Vegetation and climate dynamics in a 16,600-year marine sequence offshore Mozambique in Delagoa Bight, south-eastern Africa”. Quaternary International, Vol. 747 issue 109956, (2025).
26 Koile E., S.J. Greenhill, D.E. Blasi, et. al, “Phylogeographic analysis of the Bantu language expansion supports a rainforest route”, Proc. Natl. Acad. Sci. U.S.A. 119 (32), (2022).
27 Norström, Elin, G. Norén, R.H. Smittenberg, et.al. ”Leaf wax δD inferring variable medieval hydroclimate and early initiation of Little Ice Age (LIA) dryness in southern Mozambique”, Global and Planetary Change 170 (2018): 221–233
28 Ekblom, A. “Forest–savanna dynamics in the coastal lowland of southern Mozambique since c. AD 1400”, The Holocene, 18(8), (2008):1247–1257.
29 Ongoing work, e.g., Fourvel, J-B, A. Beaudet, M. Zanolli, et al. “The HOMME project – Human Origins in Mozambique and Malawi Environments: looking for our origin in the Mozambican karst”. Karstologia Mémoires, 5, no. 25 (2022): 219–222.
30 Lepre C.J. Dennis V. Kent, “Chronostratigraphy of KNM-ER 3733 and other Area 104 hominins from Koobi Fora,” Journal of Human Evolution, 86, (2015): 99-111; Hammond A.S., S.S. Mavuso, M. Biernat et.al. “New hominin remains and revised context from the earliest Homo erectus locality in East Turkana, Kenya.” Nat Commun.,12(1), (2021):1939; and Herries Andy I.R. et al., “Contemporaneity of Australopithecus, Paranthropus, and early Homo erectus in South Africa”. Science, 368 (6486), (2020).
31 Dusseldorp G., M. Lombard, S. Wurz S. “Pleistocene Homo and the updated Stone Age sequence of South Africa”. S Afr J Sci. 109 (5/6), (2013): 1-7.
32 Lombard, M., J. Bradfield, M.V.Caruana et. al, “The Southern African Stone Age Sequence Updated (II)”, South African Archaeological Bulletin 77 (217), (2022):179
33 Stringer C. (2016), op.cit.
34 Bicho, N., C. Gonçalves, H. Madiquida et. al, “Mozambique Stone Age Landscape: A Decade of Archaeological Research”, Preprint/ under review, Journal of Paleolithic Archaeology, posted 12 March 2026.