Climatic and Paleoenvironmental Changes in Central Africa over the Last 300,000 Years: Evidence from the Rhizophora/Podocarpus Pollen Ratio in a Marine Core from the Gulf of Guinea ()
1. Introduction
Marine sedimentary archives are key recorders of Quaternary climatic and environmental variations because of their temporal continuity and their ability to integrate continental and oceanic signals at large spatial scales [1]-[3]. In Central Africa, marine cores recovered from the Gulf of Guinea, therefore, provide a particularly suitable framework for analyzing how tropical continental systems respond to major global climatic oscillations.
Continental pollen grains exported to the ocean by large rivers, especially the Congo River, represent robust tracers of vegetation dynamics and regional climatic conditions. In contrast to continental archives, which are often influenced by local effects related to basin morphology or hydrological conditions, deep-sea marine archives integrate a regional pollen signal at the scale of vast drainage basins [4] [5]. This integrative characteristic gives marine records particular value for reconstructing Central African paleoenvironmental changes over long-time scales.
Numerous palynological studies conducted in the Gulf of Guinea have demonstrated that marine pollen assemblages reliably reflect the major vegetation units of the Congo Basin, provided that fluvial transport processes and marine hydro sedimentary dynamics are taken into account [6] [7]. More recent studies have further refined these interpretations by highlighting the key role of coastal hydrodynamism and marine circulation in pollen dispersal, redistribution, and preservation on the Cameroonian continental shelf [8], as well as the sensitivity of marine pollen assemblages to climatic variations during the Late Pleistocène and the Holocene [9] [10].
Marine core KW23, recovered from the Congolese margin of the Gulf of Guinea, provides an exceptional record covering approximately the last 300,000 years. Previous sedimentological and isotopic studies have revealed a succession of warm and cold climatic phases that are well correlated with marine isotope stages [11] [12]. However, palynological investigations of this sequence have remained partial and were previously restricted to the uppermost levels.
This study aims to extend the palynological analysis of the KW23 marine core to reconstruct the climatic variations in Central Africa over the last 300,000 years. It is based on a rigorous assessment of the main pollen groups and indicator taxa, particularly those associated with mangroves and Afro-montane formations. Previous work has highlighted the importance of Rhizophora during high sea levels and the extension of low-altitude Podocarpus during cooler periods in equatorial and southern Africa [13]-[15]. The relative abundances of Rhizophora and Podocarpus, as well as the variations in their ratio, will serve as a climatic proxy to distinguish the warm and humid phases from the cooler periods. These two taxa, with contrasting ecological requirements, constitute excellent climatic bioindicators during glacial and interglacial periods: Rhizophora is typical of coastal mangroves in hot and humid climates, while Podocarpus characterizes montane forests associated with cooler conditions.
2. Materials and Methods
2.1. Geographical and Environmental Setting
2.1.1. Location and Geological Context
The study site is located in the Gulf of Guinea, on the Congolese margin of the equatorial Atlantic Ocean. Marine core KW23, recovered during the WALDA cruise aboard the R/V Jean-Charcot, was collected at a water depth of approximately 2300 - 2330 m, between coordinates 3˚12′S-9˚18′E and 3˚46′05″S-9˚17′05″E (Figure 1). The site lies about 250 km west of the Congolese coastline and 300 - 425 km offshore from the Congo River estuary.
This area belongs to a sedimentary basin formed during Mesozoic rifting associated with the opening of the South Atlantic Ocean [16] [17]. Its stratigraphic
Figure 1. Location of core KW23 and vegetation. a: in Giresse et Barusseau (1989); b: in Aubreville et al. (1958).
succession includes pre-Albian continental deposits overlain by Upper Cretaceous and Cenozoic marine sediments. The distal position of core KW23, which is strongly connected to the Congo Basin (≈3.7 × 106 km2), favors the recording of an integrated regional sedimentary and pollen signal dominated by continental terrigenous inputs [7] [18]-[20].
2.1.2. Terrigenous Inputs and Marine Dynamics
The Congo River represents the main source of terrigenous inputs to the Gulf of Guinea, with an average discharge of approximately 40,900 m3∙s−1 and an annual suspended matter flux of 31 to 45 million tons, including sandy, pelitic, and organic fractions [21]-[23]. The fine fraction, dominated by kaolinite, smectites, and illite, results from the weathering of rocks and tropical ferrallitic soils within the Congo Basin [24], whereas the organic fraction notably includes pollen and spores produced by continental vegetation.
The dispersion and deposition of these materials are controlled by a complex system of currents, including the Guinea Current, the South Equatorial Current, the South Equatorial Countercurrent, the South Atlantic Current, and the Congo River outflow. Together, these circulation systems ensure water mass mixing and the transfer of sediments toward the deep basin [25] [26].
2.1.3. Atmospheric Circulation, Climate, and Vegetation
The climate of Central Africa is dominated by the seasonal latitudinal migration of the Intertropical Convergence Zone (ITCZ), which controls the alternation between a humid season characterized by oceanic monsoon influence and a dry season driven by continental trade winds [27]-[29]. The Congo Basin experiences a warm and humid equatorial climate, with annual rainfall ranging between 1500 and 2000 mm, favoring the development of dense and highly diversified vegetation [30] [31].
Lowland areas are dominated by Guineo-Congolian evergreen tropical rainforest, which transitions toward the basin margins into semi-deciduous forests, forest–savanna mosaics, and wooded savannas. Coastal zones host mangrove ecosystems dominated by Rhizophora [32] [33], while higher-altitude regions support montane forests where Podocarpus is a common component. These vegetation formations constitute the main sources of the pollen signal recorded in marine sediments from the Gulf of Guinea.
2.2. Material
Marine core KW23 was recovered using a Kullenberg coring system. Visual lithological observations reveal locally stratified layers, 1 to 2 cm thick, showing marked color variations, particularly olive to very dark olive lutite levels [11]. The sedimentary units, with thicknesses ranging from 10 to 250 cm, generally display gradual contacts, although sharp or bioturbated contacts are locally observed. Estimated sedimentation rates range between 3 and 8.5 cm∙ka−1.
A standard analytical protocol was applied, including measurements of water content, CaCO₃ concentrations, and organic carbon and nitrogen contents, as well as detailed grain-size analyses at selected levels. Binocular microscope observations of the sandy fractions (>40 µm) allowed both quantitative and qualitative analyses of the main sedimentary components, including fecal pellets, foraminifera, diatoms, sponge spicules, quartz grains, and pyrite concretions.
In total, 58 samples, regularly distributed along the sequence and covering marine isotope stages 1 to 9, were analyzed.
2.3. Methods for Pollen Diagram Construction and Ecological Significance of Taxa
In long Quaternary marine sequences, the frequent overrepresentation of spores, as well as certain taxa such as Rhizophora and Podocarpus, requires a specific methodological approach for the construction and interpretation of pollen diagrams [6]. The data were processed using two distinct datasets. The first dataset includes all identified pollen taxa and allows the analysis of bioindicator taxa (Rhizophora, Podocarpus, and spores), whereas the second dataset excludes spores in order to improve the readability of relative variations in the other major taxa, following practices commonly applied in regional marine palynological studies [6] [10].
Pollen diagrams were constructed using percentage values calculated from each dataset. The ecological significance of the taxa is based on their modern ecology and on regional reference frameworks derived from previous palynological studies conducted in Central Africa, particularly for the interpretation of forest, savanna, littoral, and montane vegetation formations [34] [35]. Within this methodological framework, changes in the relative abundances of Rhizophora and Podocarpus are summarized by calculating the Rhizophora/Podocarpus pollen ratio, which is used as a quantitative indicator for analyzing paleoenvironmental and climatic variations throughout the sequence, in continuity with approaches developed in recent studies on the Cameroonian continental shelf [10].
3. Results
The main results of the geological study (lithology, mineralogy, and micropaleontology) of the biological and mineral components (Figure 2) are based on the work of Bonifay and Giresse (Bonifay, 1987; Bonifay and Giresse, 1992).
3.1. Biogenic Components
3.1.1. Foraminifera
Foraminifera constitute the main components of the calcareous tests in core KW23, with a general increase in carbonate content during warm isotope stages.
The species Melonis barleeanum (benthic) and Globorotalia menardii (planktonic) were used for oxygen isotope analyses and for biostratigraphic reconstruction. Benthic foraminifera tests are generally more resistant to dissolution processes than planktonic tests; a decrease in the value of P/B (planktonic foraminifera compared to benthics) is therefore linked to dissolution [36]. For each sample,
Figure 2. Lithostratigraphic section of core KW23 (Adapted from Extracts from the figures in Bonifay and Giresse, 1992).
15 to 20 specimens of Melonis barleeanum were selected for the determination of δ1⁸O values. The resulting isotopic curve allowed the sequence to be correlated with Marine Isotope Stages (MIS) 1 to 9 according to the chronology of [2], as revised by [3], with most stages displaying secondary fluctuations corresponding to substages. The top of the core is dated by radiocarbon (14C) at approximately 5000 years BP, with the uppermost part of marine isotope stage 1 being poorly represented, whereas the base of the core, located at 16.40 m depth, corresponds to an age of about 303,000 years.
According to the ecostratigraphic zonation of Ericson and Wollin (1968), based on the relative abundance of Globorotalia menardii, its presence individualizes the warm biozones (Z, X and V), while its disappearance delimits the “cold” hypothermal biozones (Y and W). The top of biozone “X” is located at 460 cm; the disappearance of G. menardii at the top of isotopic stage 5 may be related to selective dissolution. The hypothermal biozone W, located between 820 and 970 cm, represents only a tiny fraction of MIS 6. Strong variations in the abundance of G. menardii within biozone V indicate significant climatic fluctuations.
3.1.2. Calcareous Tests
The P/B index generally follows the evolution of the carbonate curves. The cumulative curve of the carbonate particle flux (Figure 2) shows several marked decreases, mainly associated with warm Marine Isotopic Stages (MIS 5.0, 5.4, and 7), while smaller increases correspond to certain cold stages, particularly MIS 5.2 and 2. Lithological analysis reveals stratified sedimentary levels 1 to 2 cm thick, with varied colors. The chronostratigraphic framework of core KW23 allows estimating the periodicity of lamination formation between 116 and 330 years. Stages 3 and 5c are characterized by high sedimentation rates, whereas lower values indicate a greater distance from detrital input. These zones, distributed randomly throughout the stratigraphy, are mainly associated with biozones “V” and “Y”.
3.1.3. Fecal Pellets and Siliceous Skeletons (Figure 2)
Fecal pellets and siliceous skeletons represent important biogenic components of core KW23. Their abundance reflects variations in oceanic primary productivity, with high concentrations generally associated with cold periods. Diatoms show maximum abundances during marine isotope stage 2, at the top of stage 3, as well as in the lower part of the sequence, whereas fecal pellets are particularly abundant during stages 3 and 5.4, at the base of stage 6, and at the top of stage 7. Sponge spicules, in turn, display increased abundances during marine isotope stages 2, 3, 7, and 8.
3.1.4. Organic Carbon and Nitrogen (Figure 2)
Profiles of total organic carbon and organic nitrogen display a closely correlated evolution throughout the sequence, indicating a predominantly marine origin of the organic matter. Carbon content, ranging between 2% and 3%, reflects high levels of organic matter. The C/N ratio, which is higher in the lower part of the core, highlights differences in the preservation state of organic matter within the sequence.
3.2. Mineral Components
3.2.1. Quartz
In the marine sediments of core KW23, the abundance and variations of coarse quartz (>50 µm) are directly related to terrigenous inputs and long-distance transport processes. Morphological observations indicate grains of multiple origins, including aeolian inputs attributed in particular to the Saharan and Chadian regions, as well as aquatic inputs linked to continental drainage. The highest quartz concentrations are mainly recorded during the cold and dry phases of the sequence (Figure 2).
3.2.2. Clay Minerals
Clay minerals are dominated by kaolinite, with proportions ranging between 40 and 75%, followed by smectites (15% - 30%) and illite (up to 10%). Kaolinite content decreases with increasing distance from the mouth of the Congo River. This clay fraction reflects substantial continental inputs, mainly derived from pedogenesis developed on adjacent land areas, with proportions varying along the sequence (Figure 2).
3.3. Pollen Identification, Abundance, and Taxonomic Diversity
The initial study carried out on 18 levels of the upper part of the nucleus (800 cm) and covering the last 135,000 years (Bengo and Maley, 1991), identified 152 types of pollen. Spores account for nearly half of the grains counted, followed by the main tropical families (Euphorbiaceae, Caesalpiniaceae, Rubiaceae), bioindicator taxa (Rhizophora and Podocarpus), taxa characteristic of open formations (Gramineae, Cyperaceae) and temperate taxa. For this follow-up study, the species richness remained similar, but a total of 17,262 pollen grains (53.5%) and 14,991 spores (46.5%) were recorded. The main taxa identified are Podocarpus (4246 grains; 24.60%), Rhizophora (3633; 21.05%), Gramineae (2511; 14.54%), Cyperaceae (2080; 12.05%), temperate taxa (1081; 6.26%), other taxa (2995; 17.35%), and indeterminate pollens (716: 4.15%). Detailed values for each level are presented in Table 1, corresponding to dataset 2.
Table 1. Counting table and percentages of the main taxa and plant groups.
3.4. Variations in Pollen Climatic Bioindicators during the Late Pleistocene
The pollen diagram of climatic bioindicators, constructed from dataset 1, highlights significant variations throughout the Late Pleistocene (Figure 3). Spectra derived from this dataset show a strong dominance of spores, whose proportions locally reach nearly 80%. Diagrams constructed from dataset 2, in which spores are excluded, allow clearer visualization of the relative variations of Podocarpus and Rhizophora.
These two taxa exhibit marked fluctuations, often opposite or alternating, along the sequence. These variations are summarized by changes in the Rhizophora/Podocarpus pollen ratio (Figure 3), which shows values lower than 1 during phases dominated by Podocarpus and distinctly higher than 1 during phases characterized by Rhizophora dominance.
Figure 3. Pollen diagram (1): ecological bioindicators.
3.5. Variations in Vegetation Formations Inferred from Major Pollen Taxa
The spectrum of the Rhizophora/Podocarpus pollen ratio and of open savanna-type formations, represented by the taxa Gramineae and Cyperaceae, is shown in Figure 4. It highlights the contrast between savanna formations and forest formations, the latter being grouped under the category of “other taxa” in dataset 2,
Figure 4. Pollen diagram (2): main taxa.
with the exception of temperate taxa.
This representation makes it possible to identify the different vegetation types present during successive climatic phases, as defined by the alternation of bioindicator taxa. Analysis of the transition zones corresponding to a value of 1 in the Rhizophora/Podocarpus ratio further allows the identification of the boundaries of the nine marine isotope stages as well as those of interstadials over the last 300,000 years.
3.6. Correspondence between the Rhizophora/Podocarpus Pollen Ratio and Established Climatic Models
Figure 5 illustrates the correlation between the pollen spectrum (Rhizophora/Podocarpus ratio), the δ1⁸O isotopic curve of the benthic foraminifer Melonis barleeanum, Milankovitch orbital cycles, and data from the Vostok ice core. This analysis highlights convergent trends among these indicators. Variations in the pollen ratio coincide with the main oscillations in isotopic and orbital records,
Figure 5. Comparison of the Rhizophora/Podocarpus ratio curve with standard climatic curves.
demonstrating a strong coherence between the Rhizophora/Podocarpus ratio and the oxygen-18 isotopic curve, particularly at stages 1, 2, 3, 4, 5a, 5b, 5e, the beginning of stage 7, as well as stages 8 and 9, over the past 300,000 years.
4. Discussion
4.1. Taxonomic Diversity and Representativeness of the Pollen Signal
Pollen assemblages derived from the complete analysis of core KW23 exhibit a high level of taxonomic diversity, comparable to that observed in earlier preliminary studies conducted on the upper part of the core [6] [37]. This diversity reflects the multiplicity of vegetation formations within the vast Congo Basin and confirms the ability of deep-sea marine archives from the Gulf of Guinea to integrate a continental pollen signal that is representative at the regional scale. Apart from spores, pollen from Podocarpus, Rhizophora, and Gramineae constitute the dominant taxa, highlighting the combined importance of montane, littoral, and open vegetation formations in the composition of the marine pollen signal.
The strong agreement observed between pollen profiles recorded at different levels of the core and the present-day and past distribution of continental flora underscores the major role of terrigenous inputs from the Congo River and its tributaries in transferring pollen to the marine domain. By draining a basin covering more than 3.7 million km2, the Congo River acts as a powerful integrator of signals originating from diverse vegetation units, thereby limiting the influence of localized sources and enhancing the regional representativeness of the pollen assemblages recorded at site KW23.
Gramineae (Poaceae) constitute the main pollen contribution from herbaceous vegetation and indicate the presence of open environments such as savannas and grasslands within the Congo Basin. Their mean proportions remain relatively stable through time, except during certain phases marked by climatic cooling and a strong abundance of Podocarpus, which induces a relative dilution effect.
This relative stability can be explained by the extensive spatial coverage of savannas, which account for nearly one third of the surface area of the Congo Basin, and by their continuous contribution to the pollen flux exported to the ocean. Consequently, in deep-sea marine archives, Gramineae maxima do not systematically oppose forest phases, in contrast to what is frequently observed in small continental drainage basins.
This contrast is particularly clear when marine records are compared with those derived from small lacustrine sequences, such as Lake Barombi Mbo in Cameroon, where variations in Gramineae appear more strongly contrasted and directly opposed to phases of expansion of dense humid forest [38]. In such lacustrine contexts, forest expansion is often marked by the dominance of pioneer taxa (Alchornea, Elaeis) and forest taxa such as Pycnanthus, Uapaca, and Aucoumea klaineana, as well as by various families characteristic of Guineo-Congolian forests (Caesalpiniaceae, Sapotaceae, Sterculiaceae, Meliaceae, Sapindaceae) [39].
These observations fit within the broader debate on the Quaternary history of the African tropical rainforest. Many authors have suggested that during arid periods, the forest persisted in the form of forest refugia, while peripheral areas were invaded by savanna formations of the Sudanian–Zambezian type [40]. The results obtained from core KW23 suggest that, at the regional scale integrated by the Congo Basin, pollen signals record both the persistence of forest formations and the long-term importance of open environments, without a systematic opposition between the two. This pattern reflects a complex landscape mosaic modulated by Quaternary climatic fluctuations.
4.2. Dominance of Rhizophora
High proportions of Rhizophora pollen have been observed in recent dredging samples [8] and during marine isotope stage 1 of the Late Holocene in marine cores C61 [9] and CF [10] from the Cameroonian continental shelf. Comparable abundances have also been reported in several marine sequences from the Gulf of Guinea, notably off Gabon (GIK16867) [7], Congo (GeoB1008) [6] [14], and Angola (GeoB1016) [15], indicating a regionally consistent distribution of this taxon during recent warm and humid phases.
Rhizophora pollen is mainly recorded in coastal and marine sediments, either in offshore settings or in estuarine environments such as the Niger Delta [41] [42] and the estuaries of the Congolese coast, particularly around Pointe-Noire [43]. It is rarely found in continental archives, as pollen grains are predominantly exported to the marine domain by fluvial networks, a pattern confirmed by surface studies conducted in the Sanaga River basin [44].
Along core KW23, Rhizophora-rich levels reflect periods of shoreline stabilization and mangrove expansion under warm and humid conditions, characterized by mean temperatures exceeding 22˚C [33]. The concomitant absence of Podocarpus during these phases contrasts with conditions prevailing at the end of the mid-Holocene, which are marked by higher proportions of this taxon and are interpreted as cooler climatic conditions favoring a downslope expansion of montane vegetation [10].
The progressive decline of Podocarpus during the Late Holocene is a well-documented phenomenon in Central Africa, notably on the Batéké Plateau [45], in the Pointe-Noire region and at Sinnda [46], as well as in Cameroon at Lake Barombi Mbo [39] [47], Lake Bambili [48]-[50], Lake Njupi [51], Lake Mbalang [52], Lake Mboandong [53], and Lake Ossa [54].
4.3. Dominance of Podocarpus
In absolute terms, the taxon Podocarpus dominates large portions of core KW23. Throughout the sequence, its abundance peaks consistently oppose those of Rhizophora, which are strongly developed during warm and humid phases, suggesting that Podocarpus is associated with cooler climatic conditions. This taxon is indeed characteristic of cool and humid montane environments, generally located above 1,000 m a.s.l., where frequent fog and persistent cloud cover provide substantial moisture inputs in the form of occult precipitation, typical of cloud forests [55].
Podocarpus represents a major component of Afromontane forests, notably in the Kivu-Ruwenzori massifs, along the Cameroon Mountain chain that marks the southern margin of the Congo Basin, and on the southern flank of the Chaillu Massif near the Atlantic coast [56]. Its presence has also been documented in low-altitude continental archives, such as the Bois de Bilanko site, located at approximately 600 m a.s.l. and about 40 km north of Brazzaville. Pollen analyses at this site show that prior to the onset of the Holocene, montane forest vegetation dominated by Podocarpus (exceeding 50%), associated with Olea hochstetteri and Ilex mitis, was widely established across the Batéké Plateaus [34].
The convergence of these continental and marine data indicates that during cold periods of the Late Quaternary, an Afromontane vegetation dominated by Podocarpus expanded downslope over large sectors of the western Congo Basin and adjacent regions, producing a clearly identifiable pollen signal in deep-sea marine archives such as core KW23.
4.4. Bioindicators and Climatic Zones
The alternation and successive variations in the abundances of Podocarpus and Rhizophora along core KW23 reflect major climatic changes over the last 300,000 years. The Rhizophora/Podocarpus pollen ratio curve highlights climatic transition zones that closely correspond to the boundaries of marine isotope stages, as defined from the benthic foraminifer Melonis barleeanum, as well as to standard Milankovitch orbital forcing curves [57] and isotopic and temperature records derived from the Vostok ice core [58] [59].
Changes in the Rhizophora/Podocarpus ratio make it possible to distinguish nine climatic zones corresponding to the first nine marine isotope stages. Warm climatic zones, characterized by Rhizophora dominance and successive grouped data points with ratio values greater than 1, coincide with odd-numbered isotope stages (MIS 1, 3, 5, 7, and 9). In contrast, cold climatic zones, dominated by Podocarpus and marked by ratio values lower than 1, correspond to even-numbered isotope stages (MIS 2, 4, 6, and 8). Marine isotope stage 5 is distinguished by several internal fluctuations, consistent with the existence of classically recognized interstadials (5a to 5e; or 5.0 to 5.5), and represents a major interglacial period situated between the Riss and Würm glaciations.
Comparable patterns of alternation between Podocarpus and Rhizophora have been identified in numerous marine cores from the Gulf of Guinea and the tropical Atlantic off Africa, notably offshore Liberia [60], Côte d’Ivoire [5] [13], Ghana [61], Niger [62], Cameroon [9], Gabon [7], Congo [14], and Angola [15]. This regional convergence confirms the relevance of the Rhizophora/Podocarpus ratio as an indicator of major Quaternary climatic oscillations in Central Africa.
5. Conclusions
This study aimed to: 1) reconstruct paleoenvironmental and climatic variations in Central Africa over the last 300,000 years based on a marine pollen record, 2) assess the relevance of the Rhizophora/Podocarpus pollen ratio as a climatic proxy, and 3) compare regional vegetation signals with the main global climatic forcings.
The palynological analysis of marine core KW23, based on 58 samples and more than 32,000 counted grains, reveals a strong dominance of spores (46.5%) and pollen (53.5%), indicating an abundant and well-preserved continental signal. Among the main pollen taxa, Podocarpus accounts for 24.60%, Rhizophora for 21.05%, Gramineae for 14.54%, Cyperaceae for 12.05%, temperate taxa for 6.26%, while other forest taxa together represent 17.35%. This distribution confirms the regional representativeness of the pollen signal, which is largely dominated by fluvial inputs from the Congo Basin.
The contrasted evolution of the two major bioindicators shows that cold climatic phases are characterized by high proportions of Podocarpus, an Afromontane taxon, whereas warm and humid phases are marked by a strong increase in Rhizophora, indicating the expansion of littoral mangroves. The Rhizophora/Podocarpus pollen ratio thus makes it possible to identify nine climatic zones corresponding to marine isotope stages MIS 1 to MIS 9, in excellent agreement with the foraminifera benthic isotopic (δ1⁸O) record, Milankovitch orbital cycles, and data derived from Antarctic ice cores.
These results demonstrate that vegetation formations of the Congo Basin have responded sensitively, repeatedly, and in a quantitatively measurable way to past climatic variations. The strong relative contributions of mangroves, Afromontane forests, and open vegetation formations highlight the vulnerability of Central African tropical ecosystems to changes in temperature, humidity, and sea level. In this sense, the present study provides an essential paleoenvironmental reference framework for assessing the potential impacts of ongoing and future climate change on one of the world’s major reservoirs of biodiversity and carbon.
Acknowledgements
For their material and scientific contributions that made it possible to write this article, sincere thanks are given to:
1) Jean Maley, Honorary Research Director of the IRD, Associate Researcher at the University of Montpellier II (France);
2) Pierre Giresse, Professor at the University of Perpignan (France).