Biomes: Concepts, Characteristics and Terminology

Abstract

A cornerstone of ecology is the mapping and classification of biogeographical and ecological diversity, defining patterns in species distribution and abundance. This Chapter presents the concepts of biomes, biogeographical realms and centres of endemism, and the relationship of Angolan and southern African biogeographic units to the rest of the continent. Key ecological terms are defined and characterised using Angolan examples, within the context of the biomes and transitional mosaics found across Africa. The 16 ecoregions of Angola are mapped and their surface area, altitudinal range, and mean annual rainfall quantified. The ecological characterisation of rain forest, mesic savanna and arid savanna biomes are explained in quantitative and graphic detail.

FormalPara Key Concepts and Questions: This Chapter Explains

• How the concepts of biomes, biogeographic realms and centres of endemism are related.

• How Africa’s biomes are distributed and characterised.

• What linkages Angola’s biomes have with Africa’s centres of plant diversity and endemism.

• What key concepts and terms are used to describe the life forms and ecological characteristics of plants and the communities they develop.

• How ecological characteristics distinguish savannas from forests.

• How plant-available nutrients and plant-available water and disturbance pressures (fire and herbivory) influence the distribution of mesic, arid and mixed savannas and of closed forests.

Context: The Classification of Biogeographical and Ecological Diversity

To be meaningful, a classification has to take into account floristic, ecological and physiognomic features, which are difficult to integrate in a single system and even more difficult to portray on a map.

Adams (1996)

Understanding patterns in the distribution and abundance of species and the processes of ecosystem functioning are cornerstones of ecology. In a country as large and diverse as Angola, developing such knowledge is a challenging objective. It can most easily be achieved by proceeding from the general to the specific, scoping down from global to regional to local patterns and processes. Many patterns in nature are hierarchical, functioning at embedded spatial and temporal scales. In Angola, six levels of biotic and habitat organization can be recognised.

• At the highest conceptual level of pattern are the biogeographic realms. Realms reflect global scales of shared evolutionary relationships (lineages) of floras and faunas. Eight biogeographic realms are recognised across the globe, of which two—the Palearctic and the Afrotropical—are found in Africa. Angola, together with the rest of Sub-Saharan Africa, falls wholly within the Afrotropical Realm.

• Next, and visible from outer space, are the major habitat groups, or ecological formations (biomes). Each biome has unique ecological, structural and functional characteristics. Globally, biomes are recognised through their structure, not their floras and faunas, which differ from one continent to another (Olson et al. 2001). Six biomes are found in Angola. Where two or more biomes meet, transitional mosaics are often found, such as at the broad interface between rain forest and mesic savanna biomes in northern Angola.

• At the continental scale, and within the Afrotropical Realm, nine regional centres of endemism are recognised (White, 1971, 1983). Regional centres of endemism indicate floristic evolutionary linkages. The floristic composition of a centre of endemism is often unrelated to the ecological structure of the vegetation, which might include grasslands, savannas and forests. Four of Africa’s regional centres of endemism are represented in Angola.

• Next are ecoregions, large units of land with distinctive assemblages of species, habitats and processes, that serve as useful units for purposes of mapping related ecosystems at a regional rather than a national scale (Burgess et al., 2004). Land use development and management interventions are often taken at this level. Angola has 16 ecoregions, ranging from the tropical rain forests of Cabinda to the hyper-arid desert of the Namib.

• Ecoregions are assembled from the mapped vegetation types that represent major groupings of plant communities. Barbosa (1970) described 32 vegetation types in Angola and his account still reflects the current state of knowledge of Angola’s vegetation as a whole.

• Finally, and key to characterizing all the above levels of the hierarchy are the species of animals and plants that represent a site’s species richness, constituting, with its ecosystems and genetic diversity, a country’s biodiversity.

In this book, the focus will be on the biomes and ecoregions which represent the great biological and ecological diversity in Angola and which demonstrate the impacts of environmental conditions (climate), resources (radiation, landscapes, soils), and disturbances (fire and herbivory) on the patterns of plant and animal distribution and abundance.

1 Biodiversity, Biogeography and Centres of Endemism

An important first step in developing an ecological understanding of the environment commences at the species level. Knowing the characteristics, and ideally the recognised scientific names of the species of plants and animals that form a given community or ecosystem, is a key learning objective. Species are the fundamental units of biological diversity, collectively referred to as an area’s biodiversity. Biodiversity has been defined by the United Nations Convention on Biological Diversity as: ‘the variability among living organisms and the ecological complexes of which they are part; this includes diversity within species, between species, and of ecosystems’. The collective term biota includes animals, plants and microorganisms.

Analyses of Angola’s species distribution and emergent biodiversity patterns have been limited by the lack of accurate and comprehensive geo-referenced data on the fine-scale distribution of its rich biodiversity. However, a synthesis of Angola’s biodiversity was recently prepared by teams of experts on major taxonomic groups (Huntley et al., 2019), Table 2.1. The synthesis provides outlines of the biodiversity of all vertebrate and several invertebrate groups, relating these to biogeographic patterns across Africa. Synopses of vertebrate species composition and vegetation types are given in the biome and ecoregional accounts of this chapter, while brief notes on behavioural, physiological and ecological adaptations of selected animal groups are included throughout this volume. Key ecological concepts relating to species richness, endemism and community structure of animal groups are included in Part III.

Table 2.1 Species richness and endemism in Angolan plant and animal groups

Biogeographers have studied patterns in the distribution of plant and animal species across Africa for over a century and have proposed many phytogeographic (plant) and zoogeographic (animal) classification systems. Terms such as phytochoria and zoochoria are used by some researchers for the units of biogeographic classification. These have been based on extensive, intuitive field knowledge (e.g. Chapin, 1932; White, 1971, 1983), or advanced statistical analyses of plant and animal distribution records based on specimens held in herbaria and museums (e.g. Linder, 2014; Linder et al., 2012; Rodrigues et al., 2015), or on data from field surveys (Fayolle et al., 2018, Osborne et al. 2018).

As already described, there are successive levels of habitat, floristic and faunistic classification of importance to understanding Angola’s terrestrial ecology. At broad scales, centres of endemism and biomes are most commonly used as basic frameworks. There is general consensus on the key patterns of plant distribution across Africa, based on the centres of endemism described in the work of British botanist Frank White (1971, 1983). White (1971) defined a regional centre of endemism as an area having more than 50% of its species confined to it, and a total of more than 1000 endemic species. Angola has representatives of four of White’s regional centres of endemism:

• Guineo-Congolian Regional Centre of Endemism: mosaics of forests, thickets and tall grass savannas;

• Zambezian Regional Centre of Endemism: arid and mesic woodlands, savannas, grasslands and thickets;

• Karoo-Namib Regional Centre of Endemism: desert, shrublands and arid savannas; and

• Afromontane Archipelago-like Regional Centre of Endemism: forests, savannas and grasslands.

White’s (1971) centres of endemism and transition zones reflect the floristic composition and relationships that form the basis of the UNESCO Vegetation Map of Africa. The production of the map (White, 1983) drew on the experience of over 100 specialists in the flora and vegetation of Africa, and it remains a primary reference on the subject.

Linder (2014) has used vast herbarium databases to further define African floras, with distinct geographic centres, ages of radiation and of speciation rates. Four of these floras are represented in Angola and correspond closely to White’s regional centres of endemism. Linder provides a chronology of the evolution of the floras, relevant to understanding the floristic and vegetational history of Angola. He describes a sequential adding of plant families, genera and species to Africa’s floral diversity:

• The first tropical angiosperm lowland forest flora (Guineo-Congolian) was established by the early Cenozoic, during the Paleocene (66–56 Ma).

• An arid flora (Karoo-Namib) appeared from the Eocene (55–34 Ma).

• A tropical montane flora (Afromontane) evolved with the volcanic and mountain-building episodes of the Neogene (23–2.6 Ma).

• Finally, a savanna flora (Zambezian) was introduced with the development of open grasslands and savannas which replaced earlier floras as the incidence and impact of fires increased in the Miocene (23–5 Ma).

Geological time periods and their ecological importance are described in Chap. 4.

2 Biome and Ecoregion Definition and Characterisation

Whilst biogeographic units such as centres of endemism are of great value in explaining the evolutionary relationships of the major plant and animal communities, a more useful tool towards understanding ecological relationships is the biome concept. Introduced a century ago by the American ecologist Frederick Clements (1916), the biome has become widely used to identify the largest category of habitat and the associated plant and animal life-forms. Biomes are the large scale units that can be identified in remote-sensing data such as satellite images or aerial photographs—forests, savannas, grasslands and deserts. Plants and animals of individual biomes have physiological processes that are adapted to similar macroclimatic, soil, water and disturbance regimes, and have similar appearance, complexity and functional characteristics. Biomes are most easily recognised by their physiognomy (structure of plant life forms) and phenology (seasonal growth, flowering and fruiting patterns). Definitions of key structural and functional attributes are provided in Box 2.1.

Both biomes and centres of endemism reflect the long history of their constituent floras, faunas and life forms, and have been described as ‘theatres of evolution’. Over the long geological history of the planet, the size and distribution of biomes have ebbed and flowed. While some have gone extinct, others survive to this day. The coal-forming swamps of the Carboniferous geological period (359–299 Ma) are long gone, while swamp-forests that date back to the Miocene (23 Ma) still survive in Borneo. Many families of animals and plants of southern hemisphere biomes have their origins in the super-continent Gondwana. The separation of Gondwana to form South America, Africa, India and Australia resulted in floristic and faunistic fragmentation or disjunction, isolation and speciation (known as vicariance events). The ‘Gondwana’ elements of the African biota thus have links going back to the Cretaceous (145–66 Ma). In addition, modern floras include many taxonomic groups that arrived through long-distance trans-oceanic dispersal after the breakup of Gondwana (Pennington et al., 2004). However, much of the fauna and flora we encounter today has evolved more recently, in-situ. Recent global paleoecological models (Allen et al., 2020) suggest that much of the Earth’s land surface has experienced at least one biome change over the past 140,000 years. In summary, the biota and biomes of Africa have been in constant flux over geological and evolutionary time scales.

The key determinants of biome distribution, as will be discussed in Part II, are environmental conditions (climate), resources (solar radiation, landscapes and soils) and disturbance (fire and herbivory). It is important to note that biomes are not classified by the species of plants (flora) or animals (fauna) that are found in them. Biomes are characterised by their structural and functional attributes reflected in their vegetation. Vegetation in turn reflects the habitat and its environmental conditions. Thus the term ‘savanna biome’ is used in Brazil (for cerrado), in Angola (for miombo) and in Australia (for mulga), despite these savannas having no plant species in common. What they have in common, however, are the ecological-evolutionary functional traits. A trait is a morphological, biochemical, physiological, structural and phenological adaption to survive, grow and reproduce. Traits (such as the thick bark of savanna trees) have evolved in response to similar climates, soils and disturbance regimes. The biome concept is defined by the similarity of structure and ecological roles of plants within different environments.

Environmental factors and evolutionary processes select for the best adapted plant life form (grass, shrub, tree) and vegetation physiognomy (the vertical layering and horizontal spacing of vegetation components. Plant life form and physiognomy characterise different appearance of vegetation (grassland, savanna, thicket, woodland, forest) for a given site (Box 2.2). Climate determines whether the dominant life form of forests, woodlands or savannas (trees) are evergreen (retaining leaves for several years in non-seasonal climates such as humid equatorial rain forest) or deciduous (in seasonal climates such as tropical savannas where leaves drop and are replaced every year). Disturbance impacts of fire and herbivory are also factors that influence the nature of biomes, especially in Angola, where fire and large mammal herbivores have contributed to shaping biomes across evolutionary time. Finally, landscape (geomorphological) and edaphic (soil drainage, texture, nutrient status) factors influence biome characteristics at a local to regional level.

While the biome concept is used to classify major units of habitats, such as the tropical savannas and woodlands that cover approximately 50% of Africa, some biomes, such as the montane grasslands and forests of Angola, are represented by very small, scattered and isolated patches in an archipelago of ‘island relicts’, such as the Afromontane forests of Mount Moco and Mount Namba. Biological attributes, not size, thus determine what is recognised as a biome.

Within each biome, distinctive plant and animal communities occupy the different ecological conditions and habitats found across a landscape. Communities can vary in scale from groups of animals or plants in a small wetland pool, a patch of grassland, or an extensive forest. An ecosystem was defined by British ecologist Arthur Tansley (1935) as a community of organisms living in conjunction with non-living components of their environment, interacting and functioning as a system.

At continental and regional scales, ecosystems are too numerous and diverse for either mapping or descriptive purposes at those scales, so a more pragmatic category, the ecoregion, has been used to bring an intermediate level of detail within biomes. An ecoregion is defined as: “A large unit of land or water that contains a distinct assemblage of species, habitats and processes, and whose boundaries attempt to depict the original extent of natural communities before major land use change.” (Dinerstein et al., 1995). Throughout this book, the use of the term ecoregion will follow this definition. Burgess et al. (2004) defined a total of 119 terrestrial ecoregions for Africa and its islands. It is impressive to note that Angola has not only the largest diversity of biomes, but also the second largest representation of ecoregional diversity found in any African country.

3 The Biomes of Africa in Outline

Before examining Angola’s biomes, it is important to consider the distribution and characteristics of Africa’s biomes as a whole to obtain a perspective of the continent’s diversity of landscapes, ecosystems and biodiversity.

Some researchers (Burgess et al., 2004, Mucina & Rutherford, 2006; Olson et al., 2001) follow a strictly structural classification of African biomes. A purely structural approach can mask important ecological relationships (such as the distinction between arid and mesic savannas) while exaggerating others (recognizing edaphic ecosystems such as flooded grasslands as a biome). It is important to note that some traditional European botanists use the term ‘dry tropical forests’ for the savanna biomes. Modern ecologists apply the term tropical savannas for the Acacia, Brachystegia, Baikiaea, Colophospermum, Combretum systems of Africa, and similar biomes in South America and Australia. These ‘dry tropical forests’ lack many key forest physiognomic features and ecological traits such as closed canopies and the absence of a grass understorey. Furthermore, some units of classification are compromises between ecological reality and the constraints of mapping scale. Taking these factors into consideration, in this account seven African biomes, and two transitional mosaics are recognised (Fig. 2.1; Table 2.2).

Fig. 2.1

Biogeographic realms, biomes and transitional Mosaics of Africa, redrawn and simplified after White (1983). Mangroves are not mapped

Table 2.2 Estimated area of African biomes and mosaics

In simple terms, the biomes of Africa can be described, and are mapped (Fig. 2.1), as a succession of broad bands of biomes and linking transitional mosaics. To these may be added the more restricted Mangroves of the tropical coastal estuaries and the Montane Forests and Grasslands of the highlands. Within the Mesic and Arid savannas, locally extensive flooded grasslands and halophytic basins occur. From the Equator, the successive bands and transitional mosaics, illustrated in Fig. 2.1, comprise:

• The Guineo-Congolian Rain Forest Biome of the Congo Basin and West African lowlands;

• To the north and south of the rain forests lie two belts of Forest/Savanna Mosaics—the Guinean and Northern Congolian, and the Southern and Western Congolian Forest/Savanna Transitional Mosaics.

• Beyond these mixed transitional mosaics, to the north are the Sudanian, and to the south the Zambezian Mesic Savanna Biome of southern and eastern Africa. Extensive flooded grasslands are found in some river basins within this biome.

• Next, to the north and south of the mesic savannas are representatives of the Arid Savanna Biome—the Acacia savannas of the Sahel, and the Acacia/Adansonia/Commiphora savannas of southern and eastern Africa. Some internal drainage basins within this biome have large halophytic grasslands and shrublands.

• Beyond the Arid Savannas are representatives of the Desert Biome—the Sahara in the north and the Namib/Karoo in the south.

• On the tropical coasts of Africa, the Mangrove Biome occupies the estuaries of great rivers.

• Along the East African coast, a mosaic of grasslands, savannas and forests form the Indian Ocean Coastal Forest/Savanna Mosaic.

• At the northern and southern tips of the continent, examples of the Mediterranean-Type Biome are found—the Maquis of North Africa and the Cape Fynbos of South Africa.

• Finally, rising above the lowlands of Africa are the highlands and mountains carrying forests and grasslands of the Afromontane Biome.

A Synopsis of the African Biomes and Mosaics

Mediterranean Biome

(Area 858,000 km², 2.9% of Africa). The North African Mediterranean Biome, including maquis and garrigue is 10 times larger than the South African fynbos (779,000 v 79,000 km². However, the fynbos flora, with over 6000 species and with 70% endemics, is vastly richer than that of the maquis at ca. 2000 species and with few endemics. Diagnostic genera in the North African Mediterranean shrublands and forests include Quercus, Juniperus, Olea and Ceratonia; and for the South African sclerophyllous shrubland and heathland the characteristic genera include Protea, Erica and Restio. The key ecological determinants of Mediterranean climate ecosystems are cool wet winters and warm dry summers. Great climatic and physiographic diversity is found in this floristically and ecologically distinctive biome.

Guineo-Congolian Rain Forest Biome

(2,343,000 km², 8.0%). The Guineo-Congolian Rain Forest Biome occurs below 1200 m in West and Central Africa. Rainfall ranges from 1000–2500 mm per annum, with short dry seasons of 1–3 months. The surface areas of the major blocks differ considerably from one to another. The Guinean block covers 559,000 km², most of this comprises fragmentary remnants of forest in highly transformed landscapes. The central Congolian block covers 1,794,000 km². Much of this is continuous forest but with significant levels of transformation. Diagnostic rain forest tree genera include Entandrophragma, Gilbertiodendron, Lophira and Pentaclethra.

Mesic Savanna Biome

(6,177,000 km², 21.2%). The mesic savanna blocks lie to the north (2,885,000 km²) and south (3,292,000 km²) of the equatorial rain forest blocks, dominated respectively by Sudanian and Zambezian floristic elements. These formations are often referred to as ‘dry tropical forests’ by European botanists. Within the mesic savanna biome several extensive floodplain grasslands (Sudd, Zambezian, Lake Chad) are located, fed by the catchments of the mesic savannas.

The Sudanian Mesic savannas of northern Africa are floristically depauperate compared to the Zambezian Mesic savannas of southern Africa. Further, the Sudanian Mesic savannas share few species with the Zambezian Mesic savannas, but tree genera such as Burkea, Isoberlinia, Combretum, Terminalia and grasses such as Andropogon, Hyparrhenia, Loudetia and Pennisetum are common to both. Anogeissus and Boswellia dominate many Sudanian savannas, but do not occur in the Zambezian savannas. The Sudanian Mesic savannas lie at 200–1000 m, receiving from 600–1600 mm rainfall per annum. The transition from Sudanian Mesic savannas to Sahelian Acacia Arid savannas is gradual and poorly defined but in general follows the 650 mm isohyet. Combretum species are important in the transitions between mesic and arid savanna biomes.

The Southern, Central and East African Mesic savannas lie at 1000–1600 m on the leached soils of cooler plateaus, typically receiving 650–1200 mm rainfall per annum. Miombo ecosystems dominate the southern Mesic savannas, which cover 3,000,000 km². Trees of the genera Brachystegia, Burkea, Julbernardia and Isoberlinia characterise miombo. Mesic savanna soils are typically nutrient poor and trees are thornless but protected against herbivores by chemical defences. Grasses are productive of biomass but of low nutritional value with the consequence that mammalian herbivore biomass is low. Following extended dry seasons, natural (and increasingly human-mediated) fires, are regular and widespread. Trees, shrubs and grasses have multiple traits that tolerate fire. These savannas are ecologically characterised as Mesic/Dystrophic Savannas. The Mesic savannas feed large floodplain grasslands on the Nile, Niger, Cubango and Zambezi rivers.

Arid Savanna Biome

(7,376,000 km², 25.3%). Arid Savannas occur in three main blocks. To the north of the Sudanian Mesic savannas, the Sahel Acacia Arid savannas cover 3,108,000 km². Semi-arid thorn scrub and savannas on highlands within or adjoining the Sahara Desert cover an additional 469,000 km². Spinescent woody genera found across the Sahel include Acacia, Balanites, Boscia, Commiphora and Zizyphus—all genera common in the eastern and southern African Arid savannas. Within the arid savanna biome, several large internal drainage basins result in extensive salt pans with halophytic vegetation, such as Etosha and Makgadikgadi.

In East Africa, Acacia-Commiphora Arid savannas cover 1,633,000 km². These include the rich volcanic soils of iconic landscapes such as the Serengeti.

In Southern Africa, Arid savannas cover 2,193,000 km². Of this area, 1,326,000 km² is dominated by Acacia, Adansonia, Commiphora and Sterculia savannas. In addition, Colophospermum dominates or is a conspicuous feature of savannas, woodlands and shrublands that cover 607,000 km² of the region.

Arid savannas occur between sea level and 1200 m, on comparatively nutrient-rich soils of hot valleys and low plateaus, receiving from 250 to 650 mm rainfall per year with dry seasons of up to eight months. The rich soils carry thorny tree species and nutritious grazing. The Arid savannas thus carry a high diversity and biomass of ungulate herbivores whose grazing and browsing activities influence grass and tree demography and dynamics. Fire is less frequent than in Mesic savannas. Ecologically, the Arid savannas are referred to as Arid/Eutrophic Savannas to distinguish them from Mesic/Dystrophic Savannas.

Desert Biome

(8,290,000 km², 28.4%). Deserts form the largest biome in Africa, with the Sahara occupying 7,458,000 km² of North Africa and the Karoo/Namib 580,000 km² of southwestern Africa. Rainfall is erratic and usually less than 50 mm in the Sahara, 150 mm in the Namib and 250 mm per annum in the Karoo. Diagnostic genera for the Namib include Acanthosicyos, Stipagrostis and Welwitschia and with members of the Mesembryanthemaceae and Asteraceae characterizing the Karoo. The Karoo has over 3000 species of succulent plants, and is recognised as a global floristic ‘hotspot’. The geological substrate is highly variable over the vast areas covered by desert, with shallow soils, sands or bare rock exposures.

Mangrove Biome

(71,000 km², 0.2%). Examples of this biome are restricted to the mouths, deltas and floodplains of large rivers on the Atlantic and Indian Ocean coasts. At the interface of fresh and saline waters, a limited number of woody and herbaceous genera have adapted to the challenging aquatic environment, including the mangrove genera Avicennia, Brugueria, Ceriops, Laguncularia, Rhizophora and Sonneratia and the sea-grasses Halodule and Cymodocea.

Afromontane Biome

(1,271,000 km², 4.3%). Most of the Afromontane Biome occurs at altitudes of 1200–2200 m, with important forests in the highlands of Cameroon, Angola, South Africa, Ethiopia and East Africa where the highest peak (Kilimanjaro) reaches 5895 m. The Afromontane grasslands, shrublands and woodlands cover 80% of the biome, with forests and alpine communities in moister and higher areas. Rainfall ranges from 1000–2500 mm per annum. Afromontane forests have a distinctive floristic composition, including diagnostic genera such as Podocarpus, Olea, Juniperus, Apodytes and Philippia, shared across the archipelago-like chain of forests that follow the escarpments and highlands of eastern and southern Africa and the isolated mountains of western Africa.

Guineo-Congolian Forest/Savanna Mosaics

(2,536,000 km², 8.7%). Lying between the Guineo-Congolian rain forests and the Sudano-Zambezian mesic savannas is a broad belt of transitional forest/savanna mosaics. Much of the forest/savanna mosaic in West Africa has been highly transformed by agriculture to secondary grasslands and croplands. Fire-adapted tallgrass savannas dominate these mosaics, with rain forests constituting less than 10% of the area.

East African Coastal Forest/Savanna Mosaics

(344 km², 1.1%). Along the lowlands and interior escarpment of the East African coast, and extending from Somalia southwards to South Africa, is a band of grasslands, savannas and fragments of moist forests, today largely transformed by human activities, but with a rich flora and fauna. The forests have both Guineo-Congolian and Afromontane elements, while the savannas comprise a mix of both Arid and Mesic Zambezian floristic elements.

4 Classification of Angola’s Biomes and Ecoregions

Representatives of six of the seven African biomes and one of the mosaics described above are found in Angola. The founders of Angola’s vegetation classification and mapping systems, Gossweiler and Mendonça (1939) and Barbosa (1970) respectively described and mapped 19 and 32 main vegetation types, each with many subtypes. Here we will not try to initiate the learning process by describing every vegetation type. Rather, we will focus on the biomes and ecoregions, following the pattern of Africa’s biomes described above. Understanding the distribution of vegetation relies on maps, of which the latest covering the vegetation of the whole of Angola is that of Barbosa (1970). More recently, the World Wildlife Fund (WWF) has produced a map of the Ecoregions of Africa (Burgess et al., 2004); updated by Dinerstein et al. (2017) and available at https://ecoregions2017.appspot.com.

For the purpose of delineating the ecoregions of Angola, the WWF map of African biomes and ecoregions has been adapted, but in addition, recognises the distinction of the Arid and Mesic savannas biomes in Fig. 2.2. Furthermore, Angolan ecoregion 12—Coastal Arid Savannas—is recognised as a separate ecoregion. The sharp mapping lines separating the ecoregions are artificial. In reality, Angola’s vegetation is an ever-changing continuum of species abundances, densities and physiognomies, with few sharp boundaries. Biomes and ecoregions do not end at a given point, but rather merge in a complex mosaic pattern across their interfaces.

Fig. 2.2

Biomes and mosaics of Angola. I Guineo-Congolian rain forest (dark green); II Afromontane forests and grasslands (blue); III Mesic savanna (yellow-green); IV Arid savanna (brown); V Desert (red); VI Mangrove (purple); F/S Guineo-Congolian rain forest/mesic savanna transitional mosaic (Turquoise)

Angola’s biomes and their 16 constituent ecoregions will be described in detail in Part IV. These groupings (with ecoregion numbers), plus conservation areas or sites where typical examples can be studied in the field, are listed below. The classification and distribution of Angola’s biomes (Fig. 2.2) follows that of African biomes and closely correlates with the regional centres of endemism of White (1971, 1983) reflecting their close evolutionary and ecological relationships. A simplified profile of biome structure is presented in Fig. 2.3.

Fig. 2.3

A profile of vegetation physiognomy with increasing rainfall from desert to semi-desert shrublands, to arid and mesic savannas and woodlands, and to closed rain forest. Redrawn after Shorrocks (2007) The Biology of African Savannahs. Oxford University Press, Oxford

Angola’s biomes, transitional mosaics and escarpment zone, mapped in Fig. 2.2, and with examples of where they may be studied, comprise:

• I Guineo-Congolian Rain Forest Biome and Forest/Savanna Transitional Mosaic: (Ecoregions 1–3) Maiombe, Pingano, Cumbira, Lagoa Carumbo

• II Afromontane Forest and Grassland Biome: (Ecoregions 4 and 5) Moco, Namba

• III Mesic Savanna Biome: (Ecoregions 6–11) Cangandala, Luando, Bicuar, Cameia

• IV Arid Savanna Biome: (Ecoregions 12–14) Iona, Namibe, Chimalavera, Quiçama

• V Namib Desert Biome: (Ecoregion 15) Iona

• VI Mangrove Biome: (Ecoregion 16) Quiçama, Ilha dos Passaros

An idealised profile of vegetation physiognomy along a rainfall gradient is presented in Fig. 2.3, while Fig. 2.4 demarcates the distribution of Angolan biomes and ecoregions, using Barbosa’s (1970) vegetation map as a template. The mapped ecoregions closely follow the concepts of ecoregions as defined in Burgess et al. (2004) and developed further during field surveys in Angola. Table 2.3 provides summarised characteristics of each ecoregion, plus estimates of the area occupied by each as a percentage of Angola’s land area. The key plant genera found within the Barbosa vegetation units included in each ecoregion provide an idea of their floristic composition. Chapter 3 provides outlines of the main vegetation types and habitats typically found within each ecoregion, providing context to the detailed account of environmental drivers of pattern, given in Part II.

Fig. 2.4

Ecoregions of Angola, modified from Burgess et al. (2004). Boundaries of Ecoregions are based on the delineation of vegetation types by Barbosa (1970). Ecoregion numbers follow Table 2.3. Provincial boundaries are indicated in white

Table 2.3 Biomes and Ecoregions of Angola. (MAP—Mean Annual Precipitation)

Box 2.1: Terminology of Structural, Compositional and Functional Attributes

The terminology used in classifying and describing the biomes and ecoregions of Angola can be challenging, as terms used for biogeographic and structural concepts and even terms used in local languages are frequently intermixed in the literature. Some terms, such as savanna, are used both in a strictly structural sense, and also in a broad biome sense, as discussed later (Box 2.2). In order to better understand the identity (structure, composition and physiognomy) of the key vegetation components found in Angola, some definitions of structural terms are essential.

• A forest is a plant community with a closed tree canopy of touching or interlocking crowns (50% or more, typically above 75% of projected vertical cover) in two or more layers (strata) and usually with a shrub and sapling layer and a discontinuous herb layer. Climbers and epiphytes may also be present. In Portuguese literature, the term floresta densa is used to distinguish woodlands from closed forest.

• A thicket is a very dense, often impenetrable plant community of large multi-stemmed shrubs and trees. Climbers can be abundant, but grasses are sparse. Thickets can be evergreen or deciduous, thorny or non-thorny. The terms balcedo or brenha are used in Portuguese papers for thicket.

• A woodland is a stratified plant community with an open tree layer (less than 50% projected canopy cover) with crowns less than one diameter apart or touching, but usually not overlapping. Woody species might be trees or shrubs, usually single-stemmed but occasionally multi-stemmed. Some Portuguese researchers use the terms floresta clara or floresta aberta for woodland, but the term mato is recommended for the savanna woodlands of Angola.

• A savanna is a plant community with a discontinuous layer of woody species (trees or shrubs) whose individuals are spaced more than one crown diameter apart and with a usually dense ground layer of non-woody species of grasses and forbs. Savannas might be dominated by trees (tree savanna—savana arborizada) shrubs (shrub savanna—savana com arbustos) or clumps of woody species usually associated with termitaria (clump savanna).

• A grassland is a plant community dominated by grasses (tufted, stoloniferous or wiry) and forbs (broad-leaved herbaceous, non-graminoid flowering plants), with few if any woody species. Grasslands are termed prados or savana herbácea in Portuguese.

These definitions relate to physiognomic structure. They do not describe any floristic or ecological characteristics that relate to different evolutionary relationships or environmental conditions. With a flora of over 6800 plant species in Angola, it is essential that a small group of indicator species, genera or families be used to characterise vegetation types and ecoregions. It is usual to choose indicators that are either dominant in numbers or area occupied within a vegetation type (such as species of Acacia or Brachystegia) or which through their size are most prominent (such as Adansonia or Baikiaea) or which are indicators of a particular biome (such as Podocarpus in Afromontane forest). Some biomes can be characterised at the level of sub-families of common species such as the Detarioideae (including Brachystegia, which characterise the Mesic Savanna Biome) or the Mimosoideae (including Acacia which characterise the Arid Savanna Biome).

Ecological characteristics are as important as physiognomic and floristic in defining biomes and are influenced by biotic as well as abiotic factors as described below.

Seasonality of temperature and rainfall will determine whether trees lose their leaves annually (deciduous) or every few years (evergreen). Light penetration through the canopy influences grass cover and in turn determines whether fires will enter wooded communities. The combination of soil, water and nutrient conditions often determines whether the woody species have fine pinnate leaves (microphyllous) or broad leaves (mesophyllus). Soil nutrient availability for plant growth is poor (dystrophic) or rich (eutrophic). The pH of soils ranges from acid, neutral to alkaline. The degree of herbivory is associated with the dystrophic/eutrophic dichotomy. Rich soils produce nutritious grasses, herbs and woody plants. The plants of eutrophic soils frequently have thorny, spiny or other defence mechanisms against browsing. These mechanisms have evolved over millions of years in communities populated by numerous species of large mammal herbivores. Grasses can translocate nutrients from above ground to root systems at the end of summer (‘sour grasses’) or retain nutrients in the foliage into winter (‘sweet grasses’). Most southern African grasses are physiologically adapted to sunny, warm dry conditions (C₄ grasses). Grasses adapted to growing in shaded and cooler, moister conditions, such as forest floors, follow the C₃ photosynthetic pathway typical of grasses of temperate climates (Sect. 3.3.2). Many African savanna trees and shrubs have evolved mechanisms to survive regular natural fires, such as thick corky bark or the ability to coppice (re-sprout) after fires. The extreme of such fire adaption is found in species of geoxyles—woody members of dozens of plant families that escape fire damage with roots, trunks and branches below the soil surface. After the passage of fires, geoxyles produce short branchlets which carry leaves, flowers and fruit before grasses emerge and shade them out following the first rains (Box 15.1). Further plant structural and functional adaptations will be described in the detailed accounts of Angolan biomes (Part IV).

Box 2.2: When is a Forest a Forest, and when is a Savanna a Savanna?

Considerable confusion has developed in recent decades regarding the use of the term ‘forest’ and ‘savanna’ by ecologists working in different African countries. In particular, the Food and Agricultural Organization (FAO) of the United Nations has confounded the classification of open and closed vegetation by defining all land with more than 10% tree canopy cover as ‘forest’. The ecological definition used in this book recognises the transition from wooded savannas to closed forests at projected canopy cover of 50%. It is of importance to clarify these concepts before proceeding with the description of Angolan biomes and ecoregions.

Huntley and Walker (1982) defined the tropical savanna biomes of the world on the basis of their co-dominance of trees and a near-continuous layer of heliophilous (shade-intolerant) grasses that follow the C₄ photosynthetic pathway. The crowns of savanna trees and the canopies of savanna woodlands are more permeable to light than those of forests. The grasslands associated with the savanna woodlands support a rich diversity of forbs (non-grass herbs) that make up an important component of their botanical diversity. Fires penetrate the woodlands of the savanna biomes, but due to the sparse cover of grasses, are seldom damaging to the fire-tolerant trees and shrubs. Savanna tree saplings are fast growing, rising above the ‘kill-zone’ of fires within a few years. In summary, savanna plants are fire-tolerant but shade-intolerant. Huntley (1982) further recognised two distinctive savanna biomes—arid/eutrophic and mesic/dystrophic—occupying different landscapes across Africa with distinctive floras and faunas. The two savanna biomes have a similar codominance of trees and grasses responding to periods of soil water stress during the dry season (Table 2.4). The dichotomy in African savanna biomes, and the distinction between closed forests and savannas, is especially important to recognise in Angola, and is outlined below.

Table 2.4 General characteristics of arid/eutrophic, mesic/dystrophic savannas and closed-canopy rain forest biomes of Angola

In Africa, the transition from the mesic savanna biome to the rain forest biome is characterised by a mosaic of forest and savanna communities, each very different in their response to fire, and existing as bi-stable states, as described in Sect. 3.2.7. While the transition from Mesic/Dystrophic to Arid/Eutrophic savannas can be gradual and appear as a continuum, the interface between the closed forest and mesic savanna components of the mosaics is very sharp (Fig. 12.5). Because the forest patches are often of limited size and diverse shape, they are seldom mapped as distinct communities, and ecologists tend to designate them as ‘mosaics’ (Barbosa, 1970) or ‘transitions’ (White, 1983). In the map of African biomes (Fig. 2.1) to the north of the Guineo-Congolian Rain Forests, a belt of Guinean forest/savanna mosaics lie between the forest blocks and the Sudanian Savannas. To the south of the Congo Basin, a belt of Guineo-Congolian forest/savanna mosaics separates the forest block from the Mesic Savannas. The sharp lines marking the biome boundaries are artificial. In reality, the boundaries between most biomes are gradual, as rainfall and landscape changes interdigitate.

The characteristics of the two savanna biomes and the closed canopy rain forest biome are summarised in Table 2.4 and described below.

Closed Canopy Forests. The forests of Angola have a continuous and dense tree canopy, with sparse shade-tolerant grasses, herbs and shrubs on the forest floor. Shade-tolerant species are adapted to low-light environments, in contrast to shade-intolerant species, which are adapted to high-light environments. Shade-adapted plants have low photosynthetic, respiratory, metabolic, and growth rates. Sun plants (shade-intolerant) generally have higher photosynthetic, respiratory, and growth rates but lower survival rates under shaded conditions. The closed canopy forests have a distinct flora to that of both arid and mesic savannas. Forest trees and shrubs are fire-intolerant. C₄ grasses are typically very sparse or absent from the forest interior. The few grasses that do occur are shade-tolerant, C₃ species. The low biomass of grasses and herbs on the forest floor and its shaded moist microclimate limits the entry of fire into the forest interior. Where fires do penetrate, during extended dry spells, or where strong winds drive hot fires into the forest, damage and death of tree and understorey species can occur.

In Angola, four distinct closed forest types on well-drained (terra-firme) sites can be recognised.

• The most extensive and species-rich closed forests are those of Guineo-Congo floristic affinity (Ecoregions 1, 2 and 3).

• Second, extremely limited areas of Afromontane forests occur on the highest mountains (Ecoregion 4).

• A third forest type, comprising elements of both Guineo-Congolian and Zambezian floras occurs along the foothills of the Angolan Escarpment, as dry deciduous forests within Ecoregions 6 and 14.

• Finally, the evergreen Cryptosepalum forests of Moxico (Ecoregion 10) have a closed canopy and many forest traits, and physiognomically meet several forest criteria, but floristically are part of the Mesic Savanna Biome and are placed there in this account.

In addition, two specialised edaphically determined habitats (inland swamps and coastal esturies) support limited areas of closed-canopy swamp forests and mangroves.

Forest/Savanna Interface. Shade-intolerant savanna species fail to establish under the canopy of forest. In the tall grasslands surrounding forest patches in the Congolian Forest/Savanna mosaic (Ecoregions 2 and 3), fires usually die at the forest margin, unless this has been weakened by logging or other disruptive human activities. Where the forest margin is disturbed, fast-growing invasive alien species, such as Chromolaena odorata and Inga vera out-compete indigenous species and become established. Hot fires in the tall grasslands kill any seedlings of forest species that might have expanded into adjoining grassland. The contrasting influences of fire and shade result in very sharp boundaries between the distinct floras of the two systems of the Forest/Savanna mosaic, as discussed in detail in Chaps. 12 and 14.

Mesic Savannas. Across Angola’s extensive interior plateau, miombo woodlands, savannas and associated grasslands dominate, occupying 68% of Angola (Ecoregions 6–11). To the north of the miombo, where it transitions into evergreen and semi-deciduous forests of the Congo Basin, a tallgrass savanna, which totals 16% of Angola, forms a mosaic with the forest patches, (Ecoregions 2 and 3). These tall grasslands are highly flammable. The trees and shrubs that are found in the mesic savannas have evolved fire-adaptations such as a thick corky bark that protects the inner cambium or growing tissues of the stems and branches, robust underground storage organs, and the ability to re-sprout after fire from epicormic buds that lie under the protective cork bark. The Angolan miombo, and the Congolian Forest/Savanna mosaics occupy leached soils that are poor in nutrients, in areas receiving an annual precipitation exceeding 650 mm. They are Mesic/Dystrophic savannas.

Arid Savannas. In the arid southwest of Angola, and along the arid coastal lowlands, the savannas are distinctive from those of the miombo and the tallgrass savannas of the Congolian Forest/Savanna mosaic. These are the Arid/Eutrophic savannas, which occupy 13% of Angola (Ecoregions 12–14). They lie on richer soils, are subject to soil water stress for up to eight months of the year, are less productive of biomass than the mesic savannas, and as a consequence are less prone to the regular fires that characterise the mesic savannas. Profiles illustrating the contrasts between Mesic/Dystrophic and Arid/Eutrophic savannas are presented in Figs. 2.5 and 2.6. The ecological relationships between the two savanna biomes and of the transitional mixed savannas and forest/savanna mosaics are presented in Fig. 2.7.

Fig. 2.5

Physiognomic profile: Woodland of the Mesic Savanna Biome. Note the near-continuous grass cover under the light canopy of trees. Redrawn after Shorrocks (2007) The Biology of African Savannahs. Oxford University Press, Oxford

Fig. 2.6

Physiognomic profile: Arid Savanna Biome. Note continuous grass cover below broken tree and shrub cover. Redrawn after Shorrocks (2007) The Biology of African Savannahs. Oxford University Press, Oxfords

Fig. 2.7

A generalised model of the savanna types of southern Africa. The environmental space is defined by Plant Available Nutrients (PAN) and Plant Available Moisture (PAM) and by the importance of disturbance by fire and herbivory. Indicator sub-families and genera for dominant trees and grasses are included. Plant Available Moisture integrates rainfall, water infiltration, evapo-transpiration, soil texture and hydrologic regime in a single function. The two extremes of Arid/Eutrophic and Mesic/Dystrophic savannas are often separated by mixed savannas, while the transition from Mesic savannas to Guineo-Congolian forests is occupied by mosaics of forest patches and galleries within a matrix of tallgrass savannas with fire-tolerant trees. After Frost et al. (1986) and Scholes (1997)