Anthracological analysis of charcoal from Bara, Southwestern Nigeria: implications for paleoenvironment and firewood history

Abstract

Charcoal samples from different archaeological contexts at the site of Bara, in southwest Nigeria, were subjected to anthracological study. The sites investigated—residential structures, refuse deposits, and ceremonial centers-were from different cultural phases and settlement contexts ranging from ca. 400 BC during the Early Iron Age to the Oyo Imperial period during AD 1570–1835. The total number of charcoal pieces investigated is 2,972, constituting 4,458.2 g. All charcoal pieces that were more than 4 mm in size were included in the study. The anatomical features of each taxon were investigated in detail by the cross-sectional, tangential, and radial faces. The charcoal types were compared with burned wood/charcoal reference materials and a description of taxon characters present in the InsideWood database and IAWA list of microscopic features for hardwood identification. The identified types include species native to sub-Saharan African dry savanna, woodland savanna, and moist forests. This study offers the first comprehensive anthracological database in southwest Nigeria and provides the basis for reconstructing the long-term environmental history and household plant utilization from c. 400 BC to AD 1835.

1 Introduction

The long-term history of human ecology is still spotty in most parts of West Africa. This article contributes to the understanding of the complex relationships between humans and their natural environments in southwest Nigeria through a high-resolution anatomical analysis of wood charcoal from archaeological contexts. The five-year archaeological research in Bara (2018-22), a woodland savanna area that straddles periods between the Early Iron Age (c. 400 BC–40 AD) and the Oyo Imperial period (1570-1835), involved the systematic collection of charcoal samples from residential, ceremonial, and refuse contexts (Figures 1–3). A video of the landscape is also provided in the Supplementary Material section). The excavations yielded 2972 charcoal pieces—hand-picked, obtained from 1/8” and 1/4” screens, and selected from limited flotation procedures as described in detail in the methods section. By identifying the wood taxa of these samples, based on anatomical structure—cross-sectional, tangential, and radial features, we seek to understand the long-term human-environment interactions (e.g., fuel selection), changes in tree compositions, management of the flora landscape (e.g., bush-burning), and the climate-related vegetation shifts over the past two thousand years in Bara.

Figure 1

Figure 2

Figure 3

This is the first comprehensive and chronologically deepest anthracological study in Nigeria and the most robust anthracological dataset in West Africa. Its temporal depth enables the reconstruction of vegetation history over nearly two millennia, capturing shifts associated with climatic fluctuations, agricultural intensification, cultural factors, and political transformations. The diversity of contexts from which the wood charcoal samples originated offers critical insights into household economies, domestic organization, community life, and everyday practices across different subsistence and political economic regimes, from the environmentally low-impact Early Iron Age (c. 400 BC–40 AD) and Later Iron Age (c. 800–1500 AD) occupations to the high-impact Oyo Empire period (c. 1570–1835). By conjoining the anthracological evidence with archaeological contexts, we seek to reconstruct plant use, especially fuel selection across different social domains; and assess the entanglement of ecology, economy, and society over the longue durée. In this sense, the study bridges environmental archaeology with social history, illuminating how people engaged with and transformed their vegetal landscapes. It will also give us insights into past vegetation mosaics as well as the nature of transitions and relationships between the dry savanna and moist forest ecologies. In this regard, the study contributes to the debates about climate variability, human-induced landscape modification, and the resilience of West African environments over time.

Bara and adjacent areas were actively occupied by Later Stone Age communities, possibly as early as 3000 BC, as evidenced by the ubiquity of microliths in archaeological deposits (Amusa 2021; Ogundiran and Ade 2021; Willett 1962). However, none of these LSA sites have well-preserved wood charcoal in secured stratified deposits. The oldest Early Iron Age homesteads in southern Nigeria appeared in Bara c. 400 BC at a site christened BSM6, where a cultural complex called Yangi has been documented. The occupation lasted till c. 40 AD. Thereafter, there was a hiatus until about 800–900 AD when a new cultural complex, Atokuta, appeared. The Atokuta are the makers of the Diogun ceramics that Robert Soper first named. Unlike the Yangi complex, whose members had a stable, long-term residential strategy, the Atokuta lived in scattered hamlets across the landscape and established ceremonial centers demarcated by stelae and potsherd pavements. The arrival of the Oyo-Yoruba c. 1500–1570 seems to have contributed to the end of the Atokuta occupation phase at Bara. The latter's scattered hamlet-type settlement landscape was replaced by the Oyo Yoruba large settlement complexes, including villages and towns. By 1600, if not earlier, Bara had become a part of the sprawling Oyo metropolitan area; in fact, a suburb of the empire's capital city, Oyo-Ile (Figure 1). The use of the landscape intensified at this time to support the metropolitan area's dense human population, large herbivores—especially, horses and cattle—and specialized production, including large-scale iron and textile manufacture (Akogun et al. 2026; Ogundiran et al. 2019). A significant portion of the landscape was therefore used for grazing animals and cultivating food. Contemporary observers noted vast fields of cotton farms during this imperial period (Lander 1830: 94), and a significant amount of wood would have been used for the intensive iron smelting activities in the metropolitan area (Bandama et al. 2025). Several sites of the imperial period have been excavated in Bara, including six residential units and one refuse deposit, and several Shovel Test Pits. This study focuses on the charcoal samples from five residential units and one refuse deposit.

2 Materials and methods

2.1 Methods: recovery, anatomical descriptions and ecological features

Excavation was conducted stratigraphically, based on soil color, texture, and inclusion. All the excavated soil from each stratum was screened with a 1/8-inch mesh. However, in occasional situations when the soil is clayey and wet, we used a ¼-inch mesh screen. A 20 L soil sample was routinely collected from each stratum for flotation. In each context or features (e.g., ash lenses) where wood charcoal and other charred materials were expected to be abundant, such as hearths, pits, and refuse deposits, an additional 20 L of sediments (especially ashes) were subjected to flotation. All the excavated units are either domestic contexts (BSM6, BEC, S600, BAO086, and BAO021) or refuse deposits (S1002). We used a simple gravitational flotation method in which a 50 L basin is filled with water. The 20 L excavated sediment is poured into the basin and then stirred with a stick. The water and the agitation break the clay lumps in the sediment apart, the heavier materials (e.g., soil) sink, and the lighter materials, such as charred seeds and charcoal, rise to the surface. The latter are collected through a 0.5 mm mesh and air-dried in a shaded area. This flotation method yielded important finds, such as the endocarps of oil palm (Elaeis guineensis) and shea butter (Vitellaria paradoxa) nut seeds, but did not significantly increase the number of charcoal fragments larger than 4 mm. In sum, every excavated context, all features, and every stratigraphic phase (except the topsoil) at each of the six sites were sampled. Charcoal was meticulously hand-picked during excavation, obtained from 1/8- and 1/4-inch mesh screens, and collected from flotation.

Taxonomic classifications of the charcoal were proposed using the methodology of Höhn and Neumann (2016) for charcoal identification in species-rich environments. The description of charcoal types follows the IAWA (International Association of Wood Anatomists) nomenclature (Richter et al. 2004). Charcoal fragments with similar anatomical features are grouped into types. The main diagnostic characters of each charcoal type are described in three sections: transverse (TS), tangential longitudinal (TLS), and radial longitudinal (RLS). Descriptions of the anatomical characters of the types are accompanied by scanning electron microscope (SEM) photos and short summaries of the ecological distributions and economic benefits in each identified taxon. All charcoal pieces that are relatively larger in size (>4 mm) are included in the study (Bodin et al. 2024).

Even if we follow the IAWA characters for identification, not all IAWA characters can be evaluated, as some are often difficult to identify in archaeological charcoal because anatomical features can be compromised by preservation conditions and age. Sorting of the charcoal into types was undertaken using an incident-light microscope with magnification objectives of X100, X200, X400, and X500, available at the Material History Laboratory of the Department of History at Northwestern University. Anatomical features of the types were documented using the S-3400N Hitachi Scanning Electron Microscope (SEM) present at the Atomic and Nanoscale Characterization Experimental Center (NUANCE) of the same University and using the Richard M. Dillaman Biological Imaging Facility at the University of North Carolina Wilmington.

The allocation to the Bara sites charcoal types was done by comparing with burned wood/charcoal reference materials of West African woods at the archaeobotany laboratory of Northwestern University, and with images and descriptions of taxon characters present in the InsideWood database and the IAWA list of microscopic features for hardwood identification (IAWA Committee 1989; InsideWood 2004). The identification process was further supported by ethnobotanical, archaeological, archaeobotanical, and vegetation history literature from the specific locality of the Oyo Empire and from the wider west-central and western African regions (Keay, 1953; 1989; Sowunmi, 1979; D’Andrea et al., 2006; Höhn and Neumann, 2016). Botanical nomenclature accepted by the International Plant Names Index (IPNI 2020) is used to refer to the same species when comparing their wood anatomies. The ecological distributions of the species associated with each charcoal type are assigned according to the classification of vegetational zones in Nigeria: i) coastal (mangrove, freshwater swamp), ii) rain belt (rainforest), iii) Guinea savanna (woodland and tall grass savanna, Montane), iv) Sudan savanna (short grass savanna) and v) Sahel savanna (Marginal-savanna) (Jones, 1963; Keay, 1989; Killick, 1959; Neumann et al., 2001).

3 Results

3.1 Taxonomic results

A total of 2,972 charcoal pieces were studied, of which 2,702 were identified at various taxonomic levels. The remaining 169 fragments were not identified primarily due to sample size and poor preservation. The identified taxa were grouped into 8 types, 2 at family levels and 6 at genus levels. The only species level identification is for Elaise guineensis as the archaeobotanical assemblage is only represented by carbonized endocarps. Detailed anatomical descriptions and images of the identified charcoal types and endocarps are presented above and in the supplemental files submitted with this article. Tables 1, 2 summarize the identified taxa, distribution of sites, ubiquity, and the ecological distributions of the identified taxa.

The dominant charcoal types in Bara, in terms of comparable abundance and ubiquity, are Entandrophragma spp. (n = 550), Terminalia spp. (n = 518), and Pericopsis spp. (n = 511) with main concentration at S1002 (Table 1). Entandrophragma spp. is recovered from every site, except BSM6. Terminalia spp. and Pericopsis spp. were ubiquitous in 5 of the six sites. Although in relatively smaller sample size, Sapindaceae, Senegalia spp.,Khaya spp., Moraceae, and Chrysobalanaceae were also recorded in 5 of the six sites. The least representation of the identified taxon belongs to Elaise guineensis with a total sample size of 16 and recovered from two sites: BSM6 and BEC.

3.2 Ecological assessment of the charcoal types

Most of the charcoal types recovered from the site of Bara belong to Sudano-Guinean savanna and woodland vegetations. This transitional zone is an ecotone between the Guinean forest/savanna mosaic to the south and the Sudanian savanna to the north. Tree species of Entandrophragma spp.,Terminalia spp.,Pericopsis spp.,Khaya spp., and Chrysobalanaceae basically grow in this vegetation zone (Table 2). E. utile, E. Candollei, E. Congoense, Lecaniodiscus spp., Blighia sapida, Zanha spp., K. ivorensis, M. excelsa, and M. regia are among the species that require higher humidity during their growing season and could also grow in the moist evergreen rainforest. To the north of the Sudanian savanna is the Sahel savanna, a semi-arid ecozone that forms a transition belt between the desert and the savanna woodlands of tropical Africa. In this ecozone, drought-tolerant trees and scattered thorny shrubs and trees grow. This ecozone is represented by the woody vegetation of Senegalia spp. like S. ataxacantha, S. polyacantha, S. Senegal, and S. hockii. In comparison to other ecozones with high species diversity, the Sahel has lower tree species diversity, which simplifies the correlation between charcoal types and the ecosystem (Höhn and Neumann, 2016).

Milicia excelsa is the only species with a variety of habitat preferences. It has a broad ecological range and can be found in several vegetation zones, including rainforests, gallery forests, and savanna woodlands. However, this species is not commonly chosen for firewood; it is more highly valued for its timber, which is known for its high density, durability, and hardness. Allophylus species also thrive in diverse ecosystems, such as moist woodlands, gallery forests, forest-savanna transition zones, riverine vegetation, and secondary forests.

Almost all the taxa identified as charcoal types are good sources of firewood and charcoal production (Table 2). Among these charcoal types, the species with high-quality fuel wood include Blighia sapida, Senegalia Senegal, T. avicennioides, and Parinari curatellifolia. These four woody species are documented in the vegetation composition of the nearby Oyo-Ile (Sowunmi 1979). Anthracological interpretation is often based on the assumption that taxa with good firewood qualities will continuously show up in the charcoal assemblages of settlement contexts (Höhn and Neumann, 2016). The wood types are also concentrated in the savanna woodland, gallery forests, and the Sahelian zone.

4 Discussion and conclusion

The archaeobotanical evidence (Table 3; Figure 2) shows that during the Early Iron Age (Yangi cultural complex), ca. 400 BC–AD 40, the identified plant types include Elaise guineensis and varieties belonging to Chrysobalanaceae, Moraceae, and Sapindaceae (Table 3). This period is historically documented as environmentally low-impact Early Iron Age with well-documented settlement patterns (Ogundiran and Ade 2021). High-quality charcoal-producing wood types like Blighia sapida and Parinari curatellifolia were present as early as this period. The plant varieties that are continuously present in subsequent cultural contexts, from the Atokuta period to the Oyo Empire period, such as Senegalia spp., Terminalia spp., Pericopsis spp., and Khaya spp. are not documented during the early phase.

The charcoal distribution per context (Table 3; Figure 4), reveals both striking changes and important continuities in the composition of fuelwood used at Bara from the Yangi period (BSM6) through the Atokuta (BEC) and Oyo (S600, S1002, BAO086, BAO021) periods. Taken together, the data suggest a long-term shift from a relatively narrow and distinctive fuel profile in the earliest phase (c. 400 BC – 40 AD, BSM6) to a broader and more mixed woodland resource base in later centuries (Oyo period), while also showing that some preferences remained stable. We have undertaken this comparison fully aware that the excavation unit sizes vary per context and that the sample sizes differ dramatically: the Yangi period is represented by only 18 pieces, compared with 847 for Atokuta and a combined 2,074 for the four Oyo contexts. Nevertheless, the same protocols (described above) were used to collect the samples, and the percentages of the charcoal distribution per context (Figure 2) should reflect broad patterns in changing access to fuel resources, shifts in vegetation, and evolving sociocultural choices in wood selection.

Figure 4

The archaeobotanical assemblage in the Yangi period is dominated by a small set of taxa. Elaeis guineensis (all charred palmendocarp) is especially prominent. The other representative taxa include Sapindaceae at 33.3%, Moraceae and unidentified woods at 25% each, and Chrysobalanaceae at 16.7%. Several taxa that become important later— Entandrophragma spp., Senegalia spp., Terminalia spp., Pericopsis spp., and Khaya spp.—are entirely absent in this earliest assemblage. Even allowing for the very small sample, the Yangi pattern is notable for the strong presence of oil palm and the absence of several hardwood genera. This may reflect a local environment in which palm and a limited set of readily available woodland species were more accessible around residential zones, or it may indicate a cultural preference for easily gathered, familiar fuel sources in this early settlement phase. This low investment in cutting and chopping trees (especially hardwood) for fuel may also reflect the low technology and manpower in the autonomous Early Iron Age homestead (BSM6).

The Atokuta period (BEC) marks a major transformation. From the hard woods, Moraceae falls from 25% to 2.1%, Sapindaceae declines from 33.3% to 9.4%, and Chrysobalanaceae decreases from 16.6% to 4.7%. By contrast, several taxa absent in Yangi become central in Atokuta: Terminalia spp. rises to 18.16%, Pericopsis spp. to 17.56%, Senegalia spp. to 14.7%, and Entandrophragma spp. to 13.38%, while Khaya spp. also appears at 8.84%. This redistribution suggests that by the Atokuta period people were drawing fuel from a more varied and probably more mature woodland environment, or at least from a wider catchment. The decline of palm endocarp as fuel is especially significant. Because palm endocarp originated from oil palm fruits used in cooking and as a source of ointment, among others, its reduced use as fuel may indicate changing resource priorities or reduced abundance in the area, or discontinuity in relying on oil palm for cooking and other domestic uses. However, it is more likely that this drastic reduction in the use of palm kernels in the archaeobotanical assemblage by 1100 AD reflects a sharp decline in the population of palm trees in Bara throughout the first millennium AD. In contrast, the rise of taxa such as Terminalia, Pericopsis, and Entandrophragma points in BEC (1200–1550) to greater reliance on hardwood species, which burn differently and last longer, and may have been preferred for domestic tasks requiring steadier heat. Also, it appears that in the Atokuta phase, the technology and manpower needed to harness these hardwood resources for fuel were present and stable.

The Oyo period continues some Atokuta trends but also introduces new internal variability. Across the four Oyo contexts, Entandrophragma spp. becomes one of the most important fuel sources, ranging from 14.93% and 15.5% in BAO086 and S1002 to 33.8% in S600 and 44.0% in BAO021 respectively. Terminalia spp. remains consistently important in all Oyo contexts, varying within a relatively narrow band of 8.6% to 20.9%, which suggests continuity in its value as a dependable fuelwood. Senegalia spp. is likewise present in every Oyo assemblage between 8.9% and 18.4%. These recurring taxa indicate that by the Oyo period a core package of preferred woods had emerged. Yet there is also clear variability within the period. Pericopsis spp. is only 5.5% in S600 but rises to 23.0% in S1002, while Sapindaceae is absent in BAO021 but present at 13.1%–18.7% in the other Oyo contexts. Such differences likely reflect functional or spatial distinctions between deposits, unequal access to woodlots, or short-term seasonal fluctuations in what was available.

Overall, the data suggest that Bara experienced a long-term shift from an early fuel economy in which oil palm and a few associated taxa played a major role, to a later fuel regime centered on a broader set of hardwood woodland species. The Atokuta period appears to be the key moment of transition, with the sharp decline of Elaeis guineensis and the simultaneous rise of Entandrophragma, Terminalia, Senegalia, and Pericopsis. The Oyo period then consolidates this transformed pattern while showing local variability across contexts. These changes are best understood as the combined result of environmental and cultural processes. On the one hand, they may register changes in vegetation structure, resource pressure, and the location of exploitable woodlands relative to settlement. On the other hand, they point to increasingly patterned choices in fuel selection, implying that people did not simply burn whatever was at hand, but drew on accumulated knowledge about the qualities and uses of different woods. The anthracological asemblages therefore document more than botanical variation, they capture a history of human engagement with landscape, resource management, and daily practice over the longue durée at Bara.

Despite the changes in the sources of cooking wood fuel over time, Bara's inhabitants, from the Yangi (Early Iron Age) phase to the Oyo Empire period, consistently relied on woody plants from a mosaic environment rather than on a single species. Even though the dominant taxa change, the assemblages in all periods show selective use rather than indiscriminate burning. The later prominence of Entandrophragma, Terminalia, Senegalia, and Pericopsis suggests sustained familiarity with woods that were probably valued for particular burning qualities, including density, burn duration, and heat output.

A comparison of the three residential Oyo-period contexts—S600, BAO086, and BAO021—also suggests meaningful differences in household economies and domestic organization. BAO021 is the most distinctive because of its very strong concentration in Entandrophragma spp.. (44.0%), with Terminalia spp. a distant second at 20.0%, and Sapindaceae and Moraceae entirely absent. This narrow fuel profile may indicate a household economy that relied on a more restricted or highly patterned supply network, perhaps reflecting either stronger specialization in procurement or a preference for a particular category of hardwood suited to repetitive domestic tasks. The prominence of Entandrophragma may point to organized acquisition of dense fuelwood capable of producing sustained heat, suggesting a household with relatively regularized cooking or production routines. By contrast, S600 displays a somewhat broader mix, with Entandrophragma spp. still dominant at 33.8% but accompanied by substantial Sapindaceae (18.7%) and Senegalia spp. (17.7%), while Terminalia spp. is lower at 8.6% and Pericopsis spp. only 5.5%. This broader spread may reflect a more diversified household economy in which fuel was gathered from multiple micro-environments or through less centralized provisioning. Although the quantity of charcoal remains identified from BAO086 is lower than the other two residential contexts, the representation is the most balanced of the three assemblages: Terminalia spp. (20.9%), Senegalia spp. (18.4%), Sapindaceae (17.9%), Entandrophragma spp. (14.9%), and Pericopsis spp. (13.9%) all appear in relatively comparable proportions. Such evenness may indicate a household organization that distributed labor across different fuel-gathering strategies or drew on a wider catchment without depending too heavily on any single wood source. In this sense, BAO086 may represent a more flexible domestic economy, while BAO021 suggests stronger selectivity and S600 an intermediate pattern. At this point, these differences do not necessarily imply social hierarchy, but they do suggest that Oyo-period households at Bara were not uniform units: they likely varied in how they organized procurement, managed labor, and matched fuel choices to the practical needs of everyday domestic life. The analysis of other archaeological assemblages from these contexts is ongoing, especially pottery and fauna. A combination of these datasets with those of anthracological data might tell us more about differences in class/rank and strategies of household organization in Bara during the Oyo Imperial period.

In the meantime, the refuse context—S1002 stands out among the four Oyo contexts for its relatively high Terminalia (20.4%) and Pericopsis (23.0%). The charcoal assemblage in S1002 is especially informative because the context captures two distinct histories of use: first as a pottery pit kiln and later as a refuse dump. Its species composition differs in significant ways from the residential Oyo contexts and therefore helps clarify both wood selection for specialized production and the later accumulation of household waste from the cooking areas in the surrounding residences. S1002 contains especially high proportions of Pericopsis spp. (23.0%) and Terminalia spp. (20.4%), followed by Entandrophragma spp. (15.5%) and Sapindaceae (13.1%). Senegalia spp. (8.9%) and Khaya spp. (6.2%) occur in smaller amounts. Compared with the household contexts, S1002 is most distinct in its unusually high Pericopsis content. Pericopsis is only 5.5% in S600, 13.9% in BAO086, and 6.0% in BAO021, but rises to 23.0% in S1002, making it the strongest single indicator that this assemblage reflects more than ordinary household discard. Terminalia also stands out: at 20.4%, it is far higher than in S600 (8.6%) and very close to BAO086 (20.9%) and BAO021 (20.0%). By contrast, Entandrophragma, which dominates BAO021 at 44.0% and is also high in S600 at 33.8%, is much lower in S1002 at 15.5%, but close to BAO086 at 14.9%. Sapindaceae is likewise lower than in S600 (18.7%) and BAO086 (17.9%) but higher than in BAO021, where it is absent. These contrasts suggest that S1002 was not simply a dump filled with charcoal from one or more households. Rather, its profile combines a selective emphasis on woods especially suitable for firing with a broader range of taxa also found in domestic settings.

The elevated proportions of Pericopsis and Terminalia are likely associated with the phase when the site functioned as a pottery pit kiln, suggesting deliberate selection of dense hardwoods capable of generating sustained and controlled heat for pottery firing. Neither S600 nor BAO021 shows this same pairing so strongly: S600 is dominated by Entandrophragma and supplemented by Sapindaceae and Senegalia, while BAO021 is overwhelmingly concentrated in Entandrophragma. BAO086 comes closest to S1002 in its balance, but even then, the proportion of Pericopsis in the S1002 assemblage exceeds the proportion in BAO086 by nearly 10 percentage points. At the same time, S1002 also contains Entandrophragma, Sapindaceae, Senegalia, Khaya, Moraceae, Chrysobalanaceae, and unidentified woods in proportions that overlap with the household assemblages, especially BAO086 and S600. That broader mix is consistent with the later use of the abandoned pit as a refuse dump, when sweepings, hearth cleanings, and other domestic debris were likely deposited there by the surrounding community. The result is an assemblage that records two layers of practice: an earlier phase of specialized fuel selection for ceramic production, and a later phase in which the pit absorbed the by-products of everyday domestic life.

This study provides the first comprehensive anthracological dataset for Nigeria, joining a sparse list of such studies for West Africa, especially in Burkina Faso and Togo (Eichhorn and Robion-Brunner 2017; Höhn and Neumann 2012). It demonstrates the value of charcoal analysis for reconstructing long-term arboreal history and household plant use at Bara from ca. 400 BC to AD 1835. The results show that Bara's inhabitants consistently relied on a mosaic of plant species for household fuel use, but the dominant taxa changed across the three periods in this study. The fuel wood became more diverse as sociopolitical complexity increased. The Yangi phase is marked by a narrow fuel profile, and a limited set of woody taxa, while the Atokuta period records a major transition toward greater taxonomic diversity and increased reliance on hardwoods such as Terminalia, Pericopsis, Senegalia, and Khaya. During the Oyo Imperial period, this broader fuel regime became consolidated, although important variation among residential and refuse contexts points to differences in household organization, procurement strategies, and specialized activities such as pottery firing. Taken together, the evidence suggests that changing charcoal assemblages reflect both environmental shifts and socio-politically and culturally informed choices in fuel selection. In this way, the Bara anthracological record not only illuminates past vegetation dynamics, but also reveals how households engaged, managed, and transformed their vegetal landscape over the longue durée.

References

• 1

  AkogunM. O.SzpakP.OelzeV. M.LeishmanK.HilsdenJ.SpellerC.et al (2026). Multiproxy stable isotope analysis provides insights into diet, animal management, and residential mobility in old Bara, a metropolitan suburb of the oyo empire, West Africa. PLoS One21 (4), e0345981. 10.1371/journal.pone.0345981

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  AmusaD. (2021). Analysis of Lithic Material Finds from Bara, Old Oyo (Oyo Ile), in Kaiama Local Government Area of Kwara State, Nigeria. BA Essay, Department of Archaeology and Anthropology. Ibadan, Nigeria: University of Ibadan.

  • Google Scholar
  • View reference in article

• 3

  BandamaF.OgundiranA.MtetwaE. (2025). Knowledge Circulation or Extraction? Iron Production in the Oyo Empire’s Territorial Expansion. 27th Biennial Meeting of the Society of Africanist Archaeologists, University of Algarve, Faro.

  • Google Scholar
  • View reference in article

• 4

  BodinS. C.NeumannK.HenselE. A.VogelsangR.DemissewS.Casas-GallegoM.et al (2024). Afromontane forests and human impact after the African Humid Period: Wood charcoal from the Sodicho rock shelter, SW Ethiopian highlands. Vegetation History and Archaeobotany. 33 (4), 529–543. 10.1007/s00334-023-00977-3

  • CrossRef
  • Google Scholar

• 5

  CommitteeI. A. W. A. (1989). IAWA List of microscopic features for hardwood identification. IAWA Bull. n.s10, 219–332.

  • Google Scholar
  • View reference in article

• 6

  D’AndreaA. C.LoganA. L.WatsonD. J. (2006). Oil palm and prehistoric subsistence in tropical West Africa. J. Afr. Archaeol.4, 195–222. 10.3213/1612-1651-10072

  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  EichhornB.Robion-BrunnerC. (2017). Wood exploitation in a major pre-colonial West African iron production centre (bassar, Togo). Quat. Int.458, 158–177. 10.1016/j.quaint.2017.08.073

  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  HöhnA.NeumannK. (2012). Shifting cultivation and the development of a cultural landscape during the iron age in the northern sahel of Burkina Faso. Quaternary Int.249, 10.1016/j.quaint.2011.04.012

  • CrossRef
  • Google Scholar
  • View reference in article

• 9

  HöhnA.NeumannK. (2016). The palaeovegetation of janruwa (Nigeria) and its implications for the decline of the nok culture. J. Afr. Archaeol.14 (3), 331–353. 10.3213/2191-5784-10296

  • CrossRef
  • Google Scholar
  • View reference in article

• 10

  InsideWood (2004). Available online at:http://insidewood.lib.ncsu.edu/search (Accessed August 7–24, 2014)

  • Google Scholar
  • View reference in article

• 11

  JonesE. W. (1963). The forest outliers in the Guinea zone of northern Nigeria. J. Ecol.51, 415–434. 10.2307/2257694

  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  KeayR. W. J. (1953). An Outline of Nigerian Vegetation. Lagos: Government Printer.

  • Google Scholar
  • View reference in article

• 13

  KeayR. W. J. (1989). Trees of Nigeria. Oxford: Clarendon.

  • Google Scholar
  • View reference in article

• 14

  KillickH. J. (1959). The ecological relationships of certain plants in the forest and savanna of central Nigeria. J. Ecol.47, 115–127. 10.2307/2257251

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  LanderR. (1830). Records of Captain Clapperton’s Last Expedition to Africa: With the Subsequent Adventures of the Author, vol. 1. London: Henry Colburn and Richard Bentley.

  • Google Scholar
  • View reference in article

• 16

  NeumannK.SchochW.DétienneP.SchweingruberF. H. (2001). Woods of the Sahara and the Sahel: An Anatomical Atlas. Paul Haupt. Bern, Stuttgart: Wien.

  • Google Scholar
  • View reference in article

• 17

  OgundiranA.AdeO. (2021). Mejiro rockshelter: revisiting the later stone age sequence in Oyo-ile, Nigeria.”. West Afr. J. Archaeol.51, 1–15.

  • Google Scholar
  • View reference in article

• 18

  OgundiranA.MangutM.MoyibO. (2019). The archaeological landscape of Bara, near oyo-ile. West African Journal of Archaeology48, 15–32.

  • Google Scholar

• 19

  RichterH. G.GrosserD.HeinzI.GassonP. E. (Eds.). (2004). IAWA List of Microscopic Features for Softwood Identification. IAWA Journal. 25 (1), 1–70.

  • Google Scholar

• 20

  SowunmiM.A. (1979). The Vegetation of Old Oyo - Field Trip Report, Department of Archaeology, University of Ibadan.

  • Google Scholar

• 21

  WillettF. (1962). The microlithic industry from old Oyo, Western Nigeria. Actes IVe Congres Panafr. Prehistoire I'etude Quaternaire111, 261–272.

  • Google Scholar