Highlands Of Madagascar Research Articles

South-East Asian ovalocytosis among the population of the Highlands of Madagascar: a vestige of the island's settlement

In the Madagascar Highlands, 0·76% of children from 168 random primary schools, and 19 of 150 families from 3 villages, had oval-shaped erythrocytes. Most harboured the deletion in the band 3 gene characteristic of South-East Asian ovalocytosis. This genetic trait supports the Indonesian origin of the Madagascar settlement.

Deforestation in the Madagascar Highlands – Established `truth' and scientific uncertainty

The Madagascar central highlands with their red soils and erosion gullies are often held up as a frightening example of the consequences of deforestation. They are also used as a model of how the entire island will look if so-called `forest unfriendly activities' of local people continue. This insight is based on a narrative that describes the highlands as totally forested by the time of human arrival and gradually deforested as a response to human activities. This paper questions the deforestation narrative of Madagascar and points at alternative explanations for present day land cover. By the use of alternative sources of information the paper presents a counter-narrative that sees the treeless rolling planes of Madagascar as a result of several abiotic and biotic changes and not as the work of one single agent. The paper points at the political nature of the deforestation narrative as an explanation for its hegemonic position. On a theoretical level the paper makes an attempt to investigate the epistemological implication of a social constructivist approach to environmental discourses in Third World settings.

The political ecology of biodiversity conservation on the Malagasy Highlands

During the past two decades biodiversity conservation has become the major focus of international interventions by donor organizations and nongovernment organizations (NGOs) working in Madagascar. Despite these efforts, biodiversity losses continue unabated and the conservation activities undertaken to date appear ineffective. This paper summarizes the findings of a research study conducted on the central highlands of Madagascar that sought to examine possible reasons for the continued erosion of biodiversity in the country. A political ecology research framework is used to identify different social actor groups vying for access to natural resources and the extent to which their actions influence biodiversity on the island. The application of this framework reveals three major findings: (1) the existence of tensions between groups of social actors vying for access to and use of natural resources; (2) the characteristics and role of power differs greatly by actor group and it is the maldistribution of power, and its manifestations of use, that have detrimental effects on regional biodiversity; and (3) there are numerous aspects of the `conventional wisdom' that have permeated the thinking of staff from extralocal actor groups (especially NGOs, donors, and government agencies) and are largely false. Such `wisdom,' perpetrated by discourse and media, lacks understanding of local livelihood patters and local needs. However, endowed with political power and financial influence, this wisdom dominates local knowledge and ultimately leads to the conception and realization of conservation projects that fail to achieve their stated objectives. The behaviors of all of the actors in this situation perpetuate the status quo, including current patterns of biodiversity elimination.

Impact of the malaria control campaign (1993-1998) in the highlands of Madagascar: parasitological and entomological data.

Malaria transmission in the central highlands of Madagascar was interrupted in the 1960s by a national control program that used DDT indoor spraying and mass treatment with chloroquine. At the end of the 1980s in this region, epidemic malaria reappeared. Italian health authorities provided technical assistance to the National Malaria Control Program since the beginning of the resurgence of malaria in the central highlands. Yearly residual house spraying performed for 5 years (1993-1998) and the availability of antimalarial drugs reduced malaria transmission to very low levels, with improvement in parasitologic and entomologic indexes. A significant reduction of malaria prevalence was observed in the villages located at altitudes of 1,000-1,500 m, corresponding to the stratum of unstable malaria that was the main target of the antivector interventions. A significant reduction of malaria prevalence was also observed in the villages located at altitudes of 900-1,000 m, where malaria transmission is stable. The main vector Anopheles funestus was dramatically reduced in abundance and distribution in the sprayed areas.

Time and the ancestors: landscape survey in the Andrantsay region of Madagascar

The landscapes of the central highlands of Madagascar are inhabited by the spirits of the dead as well as by the living. The ancestors are a forceful presence in the everyday world, and the archaeology of the central highlands is intimately entwined with them. This is made manifest both in the on-the-ground experiences encountered during fieldwork, and in archaeological narratives, such as the one presented here. Tombs are a traditional focus of archaeological research, and those that dot the hills of the central highlands are part of a network of beliefs and practices which engage with the landscape as a whole and through which social identity is constructed and maintained. In the central highlands, and indeed elsewhere in Madagascar, there is an intimate relationship between peoples’ understandings of their social and physical location in the world and their understanding of their relationship to the dead.

Lutte contre le paludisme dans le moyen-ouest de Madagascar : comparaison de l'efficacité de la lambda-cyhalothrine et du DDT en, aspersions intra-domiciliaires. II - Étude parasitologique et clinique

For malaria vector control in Madagascar, 10 WP (lambda-cyhalothrin 10% wettable powder) was compared with DDT 75% WP for house-spraying, from November 1997 to September 1998. This study was implemented at the fringe of the malaria epidemic zone, in villages on western slopes of the central highlands, outside the area covered for the past five years by routine DDT house-spraying (OPID). Four types of treatment were compared in different areas: 1) DDT 2 g ai/m2 and 2) lambda-cyhalothrin 30 mg ai/m2 in previously unsprayed villages, 3) no intervention (control); 4) yearly DDT spraying (OPID fifth cycle). To investigate the malariological impact of spraying, cross-sectional surveys of the village populations were performed in each study area at intervals of two months, before and after spraying. In the newly sprayed areas, from December to June, malaria indices decreased by 62% in the ICON area and 44% in the DDT area, whereas in the unsprayed village malaria increased by 32% during the same season. There was a similar decrease in the number of gametocyte carriers in the newly sprayed areas. Active malaria case detection among febrile individuals was performed fortnightly in each village outside the OPID area. Results showed decreased malaria incidence from February (two months post-spraying) in the sprayed villages, despite the rainy season, whereas in the unsprayed area the decline occurred only after the main transmission season. This study demonstrated that, parasitologically as well as entomologically, house-spraying with residual insecticide (DDT or ICON) was an effective method for controlling malaria on the western fringes of the Madagascar highlands epidemic zone. Both products were effective, but ICON had slightly better impact than DDT, i.e. more reduction of malaria indices and of vector longevity, less irritancy of mosquitoes. For best results in this area of transition between stable and unstable malaria, we recommend earlier annual spraying (as soon as November) and extension of the OPID barrier towards western and northern slopes of the Plateau.

Morbidity of schistosomiasis mansoni in the highlands of Madagascar and comparison of current sonographical classification systems

To study the morbidity of schistosomiasis mansoni in the highlands of Madagascar, a cross-sectional study examined the extent to which liver fibrosis occurred in a rural community. The Managil and the Cairo classification systems were used. A second purpose was to investigate the effect of the measurements of 2 different branches of the portal vein (either segmental or sub-segmental branches) on the resulting staging of morbidity using the Cairo classification system. In a rice farmer village, 656 inhabitants (95% of the total population) were parasitologically examined; 561 patients underwent sonographic work-up based on the Managil scoring system, and in 307 randomized patients the outer to outer diameters of both the segmental and the sub-segmental branches of the portal vein were measured and scored by the Cairo classification system. Overall prevalence of schistosomiasis mansoni in the study area in 1994 was 68·3%. Upon sonographic examination and scoring by the Managil system 23·4% of the population showed liver changes (Managil degree I/II/III, 20%/2·5%/0·9%). Measuring the subsegmental branches only and scoring by the Cairo classification, 19% of the study population were found to have liver changes, none with severe fibrosis. By contrast, 82% were found to have liver changes (Cairo degree 1/2/3, 70%/11%/2%) when the segmental branches were measured. The diameters of the subsegmental branches were about two-thirds of those of the segmental branches. Both the Cairo- and the Managil-examination protocols have pitfalls. Using the Cairo classification, a considerable systematic error in classifying morbidity is created by measuring different branches of the portal vein.

Late Quaternary environmental changes from a pollen and diatom record in the southern tropics (Lake Tritrivakely, Madagascar)

A pollen and diatom study of a 40 m-sedimentary sequence from Lake Tritrivakely in the central highlands of Madagascar provides evidence for climatic and environmental changes during the late Pleistocene and the Holocene. The major features of the modern regional climate and vegetation distribution and the hydrological characteristics of the lake basin are first summarized. Lithological units and specific methods used to establish the pollen and the diatom biozones are then presented. The chronology of the upper 13 m of the sequence (0–41 kyr) is supported by AMS radiocarbon dates. It has not been possible to acquire absolute ages for the lower 27 m of the core. However, a tentative age model based on comparisons between the terrestrial pollen and the Vostok δD isotopic temperature records suggests an age of about 150 kyr for the core base. The vegetation primarily responded to variations in temperature. Six major cold/warm cycles are identified. During the cold phases, Ericaceous bush surrounded the site. Assuming that the age model is correct, cold phases (prior to 143, 115–110, 94–88, 75–69, and 22–17 kyr) coincide with periods of high local summer insolation. The warm phases are of different amplitude. They show first the development of a wooded grassland under a dry climate, which may coincide with periods of low summer insolation (around 125, 100, 83, 60 and 10–5 kyr), and then the growth of mid-latitude trees under wetter conditions. The diatom record primarily documents changes in the precipitation minus evaporation balance ( P− E) in the lake catchment, which appear consistent with orbitally-induced changes in P and seasonal contrasts as suggested by our pollen-based age model and the 14C-dated section of the record. An overall trend from a relatively deep, nutrient-rich lake to a shallow, oligotrophic peat-marsh, is partly attributed to the basin infilling. However, wet–dry oscillations are superimposed over this trend. The two first lake high stands (before 143 and around 115 kyr if the age model is correct) and low stands (around 125 and 105 kyr) match pollen-inferred cold and warming phases, respectively. Later, periods of low P− E also fit cold phases, and occur apparently every 12–10 kyr. Summer rain during phases of high summer insolation was not heavy enough to compensate for the large evaporation–evapotranspiration losses during warm summers and dry winters, especially during the cold Last Glacial Maximum, which was drier than today.

Malaria in the highlands of Madagascar after five years of indoor house spraying of DDT

The central region of Madagascar is a vast area of highlands (altitude 700–2000 m). Malaria transmission has re-established itself here since the last epidemic of 1985–1990 and has caused the deaths of 40 000 persons according to the Minister of Health. To combat the main malaria vector in the region, Anopheles funestus, annual programmes of indoor house spraying of DDT were carried out between December 1993 and January 1998 in most rural areas at altitude 1000–1500 m. A parasitological and serological study was then conducted in the highland schools to evaluate the impact of the programme and set up a database on the region. Using a cluster-sampling method 2 independent selections were conducted (one of 130 sites, the other of 40 sites). During the study, 13 462 schoolchildren were examined, 71% living in sprayed villages. Parasite prevalence among schoolchildren declined as altitude increases, from 11% at 700–900 m to 0 ·4 % at >1500 m. Below 1500 m, the impact of the spraying on the prevalence of the parasite was very clear (an average decrease of from 20% to 2 ·7 % below 1000 m and of from 4 ·5 % without spraying to 0 ·8 % at 1000–1500 m). Geographical analysis of the data showed that the marginal regions remained the most affected by malaria (especially outside spraying zones), and persistence of ‘pockets of transmission’ at 1000–1500 m, essentially in areas where spraying has never been used. In 9 schools, anti- Plasmodium antibodies were sought by indirect immunofluorescence on thick smears of parasitized red blood cells. The seroprevalence ranged from 22% to 63%, which suggests that the parasite is still circulating in the region. Even though our data show that vector control continues to be very successful in the Madagascan highlands, rapid reinfection could occur and must be monitored following spraying. To this end, the Minister for Health, with the support of the Italian Co-operation, has placed the region under epidemiological surveillance since 1997. An alert system for the timely detection of the sources of epidemics and the targeting of the antsivectoral campaign is also in operation. Our study suggests that this strategy should be reinforced by the spraying of DDT in the marginal zones in order to consolidate the results obtained at higher altitudes.

Seroepidemiology of human plague in the Madagascar highlands.

We conducted a seroepidemiological survey of human plague in the general population using random sampling in the area of Ambositra, the main focus of plague in the central highlands of Madagascar (520 confirmed and presumptive cases notified during the past 10 years). Sera were tested using an ELISA IgG F1 assay. Considering the internal validity of the assay and the sampling method, the overall corrected prevalence of F1 antibodies was 0.6% (95% CI: 0.2%-1.8%). Being nearly 0 up to the age of 40, the corrected prevalence increased markedly after 45 years to 6.2%. Six of 20 individuals who declared to have been treated for clinical suspicion of bubonic plague in the past had F1 antibodies. The seroprevalence did not differ according to gender except in individuals > 60, where antibodies were significantly more frequent in males. This study suggests that the number of clinically suspected cases of plague provided by the surveillance network was plausible, despite some true cases being missed and a significant number of false positives. We also confirm that Yersinia pestis infections may occur without marked clinical manifestations and patients may recover without treatment, in accordance with old observations of pestis minor.

Seismotectonics of Madagascar

Instrumental seismic activity in Madagascar over a 16-years period, as well as the largest earthquakes experienced over the last century, are studied using seismic signal recordings and the archives of the Antananarivo Geophysical Observatory (Institut et Observatoire de Géophysique d'Antananarivo). Earthquakes occur in most of the island. The known focal mechanisms indicate a general E–W extension, consistent with the stretching currently observed in the Mozambique channel and on the eastern edge of the African continent. Seismic areas cover the main Pliocene–Quaternary tectonic formations. Other events are located near Pan-African tectonics structures, preferentially reactivated within the Precambrian basement complex. A regional origin of stresses cannot explain completely the spatial distribution of all the earthquakes in the central highlands of Madagascar. An other possible stress origin can be a thermal bulge on which the main Pliocene Quaternary tectonic and volcanic structures of this region are situated. The current tectonic phase of extension seems to correspond to an early stage of reactivation. Madagascar could be part of the E–W extension zone associated with diffuse seismicity affecting East Africa south of the Equator.

Reversibility of Schistosoma mansoni-associated morbidity after yearly mass praziquantel therapy: ultrasonographic assessment

A parasitological, clinical and ultrasonographic longitudinal study was undertaken in 1993 in a focus hyperendemic for Schistosoma mansoni infection in the central highlands of Madagascar. All the inhabitants were systematically treated with praziquantel. A complete examination and treatment were repeated each year. Among the 289 villagers who underwent the complete 3 years' follow up, 65·9% excreted eggs at the initial survey and the mean egg count of infected individuals was 202 eggs/g. In 1996, the prevalence of infection was 19·3% with a mean egg count of 27 eggs/g and, among inhabitants aged >44 years, only one was found to be infected. The proportion of individuals complaining of bloody stool decreased from 24·9% in 1993 to 8·4% in 1996. Compared to the initial clinical examination, the age-adjusted prevalence of splenomegaly was significantly lower in 1996, but remained high: 62% in the 10–14 years age group and 59% in individuals aged >24 years. Ultrasonographic examination after 3 years of praziquantel therapy showed a marked decrease of the overall prevalence of schistosomal hepatic fibrosis, from 28% in 1993 to 10·3% in 1996. This improvement had already been achieved during the second year of follow-up for most subjects. Usually, the reversal of morbidity affected individuals classified as stage 1 at the beginning of the study. Stage 3 was not observed in the last 2 surveys. One patient's ascites disappeared during the follow-up, associated with a significant reversal of periportal fibrosis. Our results indicate that repeated praziquantel therapy can lead to improvement of liver morbidity and the prevention of the development of schistosomal hepatic fibrosis, even in an old-established hyperendemic focus.

A 40,000-yr Pollen and Diatom Record from Lake Tritrivakely, Madagascar, in the Southern Tropics

Links between southern and northern hemisphere climates during the Late Quaternary are poorly known, partly due to the scarcity of continuous climatic records in the southern tropics. Pollen and diatom evidence from Lake Tritrivakely (19°47′S) provides information on vegetational and hydrological changes in the central highlands of Madagascar over the past 40,000 yr. Most of the record reflects natural environmental variability since humans arrived on the island ca. 2000 yr B.P. During glacial times, the migration of mountain plants toward lower altitudes is consistent with a temperature decrease and with reduced atmospheric CO2levels. In the lake, a positive mean annual hydrologic balance, from 38,000 to 36,000 and from 17,500 to 9800 cal yr B.P., coincided with periods of decreasing summer insolation and preceded by several millennia lake rises in the northern tropics. A negative hydrologic budget during periods of maximum seasonal contrast in solar radiation is partly attributed to high summer evaporation rate. The last glacial maximum was cool and dry. The deglacial warming occurred in two steps. The first step, accompanied by an increase in wetness, occurred abruptly at ca. 17,000 cal yr B.P., about two millennia earlier than in the northern hemisphere. It is abundantly documented in southern terrestrial data. The second step, at 15,000 cal yr B.P., was in phase with the first major temperature change in the northern hemisphere.

Age-related prevalence of Plasmodium falciparum circumsporozoite antibody in an hyperendemic area of Madagascar, and its relationship with parasite prevalence.

An epidemiologic survey on malaria endemicity levels, including entomologic, parasitologic, and immunologic indicators, was carried out in a village of the Madagascar highlands (Analaroa) at the end of the 1990-1991 rainy season. The results indicate that malaria is hyperendemic and Anopheles funestus is the main vector in the area. The prevalence of parasitemia decreased with age from a maximum level of about 60% in children less than five years of age to a minimum of about 16% among those more than 29 years of age. The prevalence of Plasmodium falciparum circumsporozoite antibodies (Ab-Cs) increased with age from a minimum level of about 10% in children less than five years of age to a maximum of 71.7% among those more than 29 years of age. An inverse correlation was observed between P. falciparum prevalence and levels of Ab-Cs and parasite prevalence. The study confirmed that prevalence and Ab-Cs levels are reliable indicators of malaria endemicity in hyperendemic areas. Schoolchildren between five and 14 years of age are considered the most practical and susceptible group for this kind of epidemiologic study.

Lack of correlation between HLA class II alleles and immune responses to Pf155/ring-infected erythrocyte surface antigen (RESA) from Plasmodium falciparum in Madagascar.

To investigate the relationships between predominant HLA class II alleles and immune responses to the Plasmodium falciparum ring-infected erythrocyte surface antigen (Pf155/RESA), 50 individuals from the highlands of Madagascar were followed-up from 1988 to 1991. The T cell reactivity and antibody responses to synthetic peptides (EENV)4, (EENVEHDA)4, and (DDEHVEEPTVA)3, representing major T and B epitopes of Pf155/RESA antigen, were assessed with an average of five determinations per individual over the four-year follow-up period. The T cell reactivity was investigated by lymphocyte proliferation and assays for interferon-gamma and interleukin-2 release. Antipeptide antibodies were measured using the Falcon assay screening test-enzyme-linked immunosorbent assay. The cumulative prevalence rates of cellular (range for the three peptides = 64-68%) and antibody responders (range = 70-74%) were similar for each peptide. The HLA class II typing was performed using polymerase chain reaction-restriction fragment length polymorphisms. The prevalent alleles or groups of alleles (frequency > 20%) were similar in responders and nonresponders, both for cellular and antibody responses to each peptide. These were HLA-DR 5 group and HLA-DQA1 *0601, *0101-0102-0104, HLA-DQB1 *0301, and HLA-DPB1 *0101-2601 alleles. Allelic distribution was similar in individuals presenting with (74%) or without (26%) a malaria attack during a 20-week follow-up conducted when malaria was hyperendemic (P > 0.05, by Fisher's exact test). Despite repeated immunologic measures that better identify the responders, no relationship was found between HLA class II alleles and the cellular or antibody responses to Pf155/RESA epitopes.(ABSTRACT TRUNCATED AT 250 WORDS)

Presentation of the Plasmodium falciparum antigen Pf155/RESA to human T cells. Variations in responsiveness induced by antigen presenting cells from different but MHC class II identical donors

The antibody response in humans naturally primed to a malaria vaccine candidate antigen (Pf155/RESA) is genetically regulated. Here, the impact of antigen presenting cells (APC) on the control of in vitro T-cell responses induced by Pf155/RESA or synthetic peptides corresponding to its major Pf155/RESA epitopes was studied. T cells and APC were from the peripheral blood of monozygotic or dizygotic twins and their age matched siblings, all living in the central highlands of Madagascar. When induced to proliferate (thymidine incorporation) in vitro by antigenic peptides, the T-cell responses varied less within the twin pairs than between them and their siblings or the entire group, implying that they were genetically regulated. Occasional MHC class II associations of some of the responses were weak and did not reflect underlying MHC class II restrictions. When T cells and APC from different but MHC class II identical donors were incubated in various combinations, antigen charged APC from homologous donors induced in vitro T-cell proliferation which differed from that induced by the T-cell donors' own APC. Pretreatment of the APC with either paraformaldehyde or anti-class II antibodies inhibited or abolished this antigen dependent T-cell proliferation. The results suggest that the observed differences in T-cell responses induced by APC from different donors reflect differences at the level of these cells. Whether they reflect differences in the proteases involved in antigen processing, in the costimulatory signals provided by the APC to the T cells or in the secretion of other regulatory factors remains to be elucidated.

Humoral and Cellular Immune Responses to the Circumsporozoite Protein of Plasmodium vivax in Madagascar

Plasmodium vivax malaria is prevalent during the rainy season in the central highlands of Madagascar. In April 1991, we investigated the cellular and antibody immune responses of 53 inhabitants of Manarintsoa, a village in this area, to four antigens corresponding to B and T cell epitopes of the P. vivax circumsporozoite (CS) protein. Cellular responses were assessed by lymphocyte proliferation assay as well as by detection of interferon-gamma and interleukin-2 production in vitro. Cell culture was performed with two overlapping synthetic peptides (CSVTCGVGVRVRSRVNA [amino acids 311-326]) and VRVRSRVNAANKKPED [amino acids 319-334]) from the vicinity of the highly conserved region II of the CS protein. In at least one of the three assays, cells from seven subjects showed a positive response to CSVTCGVGVRVRSRVNA, while cells form 14 subjects responded to VRVRSRVNAANKKPED. Antibodies directed against the two recombinant antigens, NS1(81)V20 and rPvCS-2, both of which contain the entire central repeat region of the P. vivax CS protein, plus regions I and II in the case of rPvCS-2, were measured by the Falcon assay screening test-enzyme-linked immunosorbent assay. Eight and nine subjects had antibodies to NS1(81)V20 and rPvCS-2, respectively. The presence of antibody responses to both recombinant antigens was related (P = 0.02, by Fisher's exact test), but was not related to the presence of a cellular response to peptides from vicinity of region II (P > 0.1, by Fisher's exact test).(ABSTRACT TRUNCATED AT 250 WORDS)

Human immune responses to the Plasmodium falciparum ring-infected erythrocyte surface antigen (Pf155/RESA) after a decrease in malaria transmission in Madagascar.

A resurgence of falciparum malaria occurred in the central highlands of Madagascar in the 1980s and was responsible for an outbreak in 1986-1987. Since 1989, transmission has decreased dramatically. In April 1991, we investigated the humoral and cellular immune responses of 53 inhabitants of the village of Manarintsoa to six synthetic peptides that reproduced the major B and/or T cell epitopes of the Pfl 55/ring-infected erythrocyte surface antigen (RESA) of Plasmodium falciparum. The presence of RESA peptide-reactive T cells was assessed by lymphocyte proliferation assay as well as by detection of in vitro production of interferon-gamma and interleukin-2. The mean values of these cellular responses were low, and the results obtained in these three tests showed no correlation. Twenty-seven subjects presented with anti-RESA antibodies as detected by modified immunofluorescent assay, but the mean levels of anti-peptide antibodies were low. When compared with data obtained in January 1988 from the same subjects with three of the six peptides, the present data demonstrated a decrease in the response to these peptides in terms of both proliferative response and mean antibody titers. The mean values of anti-RESA antibodies remained unchanged. The fact that cellular and humoral responses to the major Pfl 55/RESA epitopes decreased but did not disappear probably reflects both the remainder of the acquired immunity resulting from the 1986-1987 malaria outbreak, and its conservation by the very low level of transmission since 1989.

Is Immunity to malaria really short-lived?

Protection against Plasmodium falciparum malaria is usually considered to be the cumulative product o f repeated exposure to parasites, and thus a function of age, in endemic areas. The recent outbreak of malaria in the central highlands of Madagascar gave Philippe Deloron and Claire Chougnet the opportunity to compare the incidence of malaria in children and young adults exposed to malaria for the first time, with that in older adults who spent their childhood in the study area before malaria control was introduced. Protection, as well as immune responses to two major P. falciparum antigens, was not related to age. Individuals older than 40 years were more protected than were younger adults. This increased protection was probably due to immunological memory.

Archaeology in Eastern Africa: An Overview of Current Chronological Issues

Even at this still early stage in the development of the chronostratigraphic framework in eastern Africa a number of important advances have been reported. As more attention is paid to the different responses of food producers to the variety of resources provided by the range of available environments then, and only then, will we be in a position to understand the diachronic processes which result in settlement aggregation and urban development.In the Lake Nyanza region at the hub of the Sudanic and Guinea–Congolian regional vegetation centres, early dates for iron working are not yet convincing enough to demonstrate independent invention of iron working, but the region is almost certainly the most important diffusion source of the technique to the eastern and southern sectors of the sub-continent.Currently available data from the Maasai–Somali region show clearly the early adoption of food production techniques and a capacity to absorb iron technology without necessarily abandoning pastoral production. This did not, however, mean a lack of development based on agriculture as the towns of the Somali coast with their advanced craft production clearly show. However, it is interesting that the urban development seems closely linked to the juxtaposition of the valuable agricultural resources provided by the Shabelle river running close to the coast and the marine resources of the littoral.The Zanzibar–Inhambane floral mosaic provides a context for the spread southwards of the early farming communities and for the development of the coastal towns. Particularly important here appears to have been the combination of surface and arboreal forms of agriculture with the exploitation of marine resources. Links eastwards with the specialized floral communities of the Comoro archipelago and Madagascar were also fully established. The highlands of Madagascar experienced the expansion from the eleventh century a.d. onwards of a settlement system increasingly focused upon hydraulic agriculture which culminated in the powerful Merina kingdom and ultimately the present day capital of Antananarivo.On the continent relatively little penetration into the Zambezianmiombowoodland communities was achieved by the coastal urban dwellers. In the woodlands of the vast highlands of the interior different developmental trajectories of settlement systems occurred. Here food production cannot be shown to have become established earlier than the late first millennium b.c. But by the mid first millennium a.d. significant settlement hierarchies based on mixed cropping and cattle keeping were established on the Zimbabwe plateau and the margins of the Kalahari. These together with the incorporation of the opportunities presented by inter-regional exchange and the exotic trade goods penetrating from the coast ultimately gave rise to the powerful state formations of the Mapungubwe and Zimbabwe traditions.Together these developments show a remarkable degree of regional articulation and it remains true that an adequate understanding of the processes giving rise to urbanism in any part of eastern Africa cannot be understood in isolation.