Abstract

For more than a century collections and observations have been made in the Kenyan and Tanzanian coastal forests.

Coastal forest is defined in terms of its geographical position and geomorphological association. Environmental and human influences are discussed.

Major patterns in the flora are discussed in terms of five ecogeographical elements, relating to distribution of species within and adjacent to coastal forests. The distribution of species outside the coastal region is summarised as well.

Forests are described informally, emphasising local variation and complex patterns. The range of association within both Moist and Dry forests locally is compounded by Northern and Southern elements, defined by species of restricted distribution.

Origins of the observed patterns are discussed. Apart from environmental and human influences, ‘Gleasonian’ factors are likely to have been important. Priorities for conservation measures are suggested.

Historical background

Botanical collections from East Africa's coastal regions started with missionary activity in the mid-19th century. The first mission station on mainland East Africa was established in 1855 close to Mombasa (Krapf, 1860, 1882), and the Rev. Thomas Wakefield soon sent plant specimens to England from this area (see Brewin, 1879; Wakefield, 1904). A few years later, Rev. Charles New provided a vivid, first account (New, 1873) of the natural history of the coastal vegetation (nyika) and of the relationship between the forests and the Mijikenda (the ‘Wanyika’) who lived in and around them.

Introduction

Although much of East Africa is arid savanna, about 2% of Tanzania contains rich tropical rain forest. The majority of this is montane forest, found on a chain of ancient (80 million years old) block-fault mountains; these mountains, which stem from the Pare Mountains in the north to the Southern Highlands in the south, have been termed the Eastern Arc Mountain Chain (Lovett, 1985). Most mountains in the chain have been isolated from one another since the Pleistocene (Hamilton, 1982; Kingdon, 1989) by a sea of arid woodland savanna. The isolation of the montane forests, in conjunction with a relatively stable climate (Hamilton, 1982), has resulted in significantly fewer overall species of nearly all taxa examined when compared with the more continuously distributed Guineo–Congolian forest (Stuart, 1981; Rodgers, Owen & Homewood, 1982; Lovett, 1985; Lovett, Bridson & Thomas, 1988; Kingdon, 1989 and several chapters in this volume). [The Guineo–Congolian forest is the main forest block extending across Central Africa, from Lake Victoria to Liberia: White, 1981.] The isolation of the Eastern Arc mountains also has produced high rates of endemism in almost every major taxonomic group (Stuart, 1981; Rodgers et al., 1982; Hamilton, 1988; Lovett et al., 1988; Kingdon, 1989 and several chapters in this volume). This endemism is particularly striking among the plants: over 25% of the 2000 plant species found in Tanzanian forests are endemic (Lovett, 1985).

Introduction

The evolution and zoogeography of the montane forest avifaunas of Africa have long been a source of fascination to ornithologists (e.g. Chapin, 1923, 1932; Moreau, 1933, 1952, 1954, 1963, 1966; Dowsett, 1971, 1980a,b; Hamilton, 1976; Diamond & Hamilton, 1980; Stuart, 1981a, 1983, 1986; Jensen & Stuart, 1985). The montane forests are characterised by many altitudinally restricted species which usually have disjunct distributions reflecting the patchiness of the available habitat. Many lowland species occur in Africa's montane forests, but do not necessarily do so as isolated populations; these species are not considered here. For the purposes of this chapter we have defined montane forest species as those which do not normally occur below a certain altitude (in eastern Tanzania, few such species occur below 700–900 m a.s.l. during the breeding season), and which seem to be dependent upon forest for their survival (usually for nesting sites and food). Thus species restricted to forest at intermediate elevations (usually between 700 and 1500 m), such as the Banded Green Sunbird Anthreptes rubritorques, are considered here to be montane.

It is not always easy to decide what is a forest species and what is a montane species, since the distinctions are not necessarily hard and fast ones (see Stuart, 1983 for a fuller discussion of this problem). One complication is that many forest dependent species are able to survive in cleared areas within a few kilometres of forest (Stuart, 1983).

Introduction

Aims

It is the purpose of this chapter to list and identify the mammalian forest fauna of eastern Africa, updating the list with new information. We provide some simple measures of affinity and distinctness in the forest mammal communities of far eastern Africa. We consider to what extent the affinities might reflect eastward dispersal from a westerly focus or a residual forest fauna impoverished by climatic degradation. Two principal levels of endemism are identified: those species that are broadly distributed in the forests of eastern Africa and those that have a very restricted distribution within this region. Further categories of endemics are also identified, restricting species to (i) montane, (ii) lowland and (iii) coastal forest types, including those found on Zanzibar, Pemba and Mafia Islands. The possible importance of colonisation, competition and isolation in the evolution of distinct species or subspecies is discussed.

A list of forest-dependent mammals from eastern and southwestern Tanzania has been assembled and is compared with a list of comparable mammals from western Uganda (Budongo, Kibale, Kalinzu, Kayonza and Bwindi). Those species held in common are listed in Table 11.1, with local subspecies shown separately under their regional heading. From these lists some simple measures of endemism have been estimated.

The interest of mammals

There are indisputably far fewer species in the East African forests than in Central and West Africa. Furthermore, the shared occurrence of various rodent, insectivore and primate species, blue duiker Cephalophus monticola, palm civet Nandinia binotata, and the anomalure Anomalorus derbianus has encouraged the view that these forests are merely impoverished outliers of the Guineo–Congolian block.

Introduction

In the eastern half of Africa the forests are mainly restricted to mountains and surrounded by savanna or even semi-desert. They have very appropriately been compared with an archipelago by White (1981). Their isolation and the high degree of endemism of flora and fauna raise questions about the evolutionary and geographic history of the forests as a whole and of the species living there. As with real islands the central questions are:

1. 1. From the geographic point of view, have these islands always been isolated or is there a history of interconnections?

2. 2. From the biological point of view, how are the species distributed and related, and where do their sister species live?

Starting from an allopatric speciation model there is a causative correlation between the two questions in such a way that the geographic history of the islands must have influenced the evolution of the species living in the islands. Biogeography is concerned with this correlation. To put it in a simple way, the biogeographic question is:

1. 3. Have the species of a particular island originated on the spot (and if so, how about the ancestors) or are they colonists from elsewhere (either by jump dispersal or following a range expansion of the habitat), or a mixture of both (and if so, what are the proportions)?

With real islands the ecological difference between the island and its surroundings (the sea) is so extreme that an origin of the great majority of the terrestrial island organisms from the sea is most unlikely if not out of question.

Plant pathogenic bacteria are typically motile, single-celled organisms, for which a range of light and electron microscope techniques are available to investigate aspects of structure and morphology (Sigee, 1989). Determination of the chemical and structural organisation of plant pathogenic bacteria is important to an understanding of cell function and host—pathogen interactions, and is also an important factor in bacterial taxonomy and pathogen identification. Phytopathogenic bacteria can be examined either during growth in sterile medium (in vitro culture, Fig. 2.1) or in association with higher plants, where they may be present on the plant surface or within infected tissue (growth in planta, see Fig. 2.3c).

Characteristic morphology and fine structure

The small size (0.5–2.0 μm diameter) of plant pathogenic bacteria places them close to the limits of resolution of the light microscope, and the examination of bacterial preparations with this instrument normally involves the use of an oil immersion objective to obtain maximum detail. These organisms are also quite difficult to see in terms of their optical contrast, and light microscope examination normally involves the use of stained preparations or phase-contrast microscopy (see Fig. 6.17a). Under optimal conditions of light microscopy, general features of morphology such as size, shape, and the presence of flagella and a capsule may be resolved, but little further detail can be determined.

The ability of plant pathogenic bacteria to survive and multiply outside and inside plants, and to cause disease, is determined to a large extent by their genetic constitution. The genetic analysis of plant pathogenic bacteria currently involves the application of molecular techniques for the identification and investigation of bacterial genes that are important in all of these aspects, and will be considered first. Following sections discuss the role of specific genes and gene systems in the activity of plant pathogenic bacteria in relation to the determination of compatibility and incompatibility, disease virulence, and non-pathogenic characteristics. The final part of this chapter deals with the occurrence and role of plasmids in these bacterial cells.

Molecular genetics: identification and investigation of bacterial genes

Bacterial genes, occurring on either chromosomal or plasmid DNA, are involved in the determination of a wide range of phenotypic characteristics. In recent years new techniques of molecular biology have been particularly successful in the genetic analysis of plant pathogenic bacteria (Daniels et al., 1988), and have been described in detail in a number of recent texts (e.g. Brown, 1986; Sambrook et al., 1989). The major objectives of molecular genetics are:

1. Identification and isolation (cloning) of specific genes with defined functions.

2. […]

Infection of plants by pathogenic bacteria can generally be considered in terms of three interrelated phases:

1. Population build-up, competition and migration of bacteria at the plant surface.

2. Bacterial entry into plant tissue.

3. Migration of bacteria within the plant to and from regions of multiplication.

Build-up and activity of epiphytic populations

Population level

The presence of epiphytic pathogens on host plants does not imply that disease will necessarily develop, and many cases have been reported where quite high levels of pathogenic bacteria were present on symptomless foliage. This has been noted, for example, for Pseudomonas syringae pathovars on red maple (Malvick & Moore, 1988) and snap beans (Legard & Schwartz, 1987) and for Erwinia amylovora on apple and pear blossom (see later).

In other situations, the presence of epiphytic bacteria does lead to disease development. This was initially noted by Crosse (1957), who reported the presence of Pseudomonas syringae pv. mors-prunorum as an epiphyte on cherry foliage, leading to canker formation. The relationship between epiphytic occurrence and disease development has subsequently been investigated for a wide range of bacterial pathogens by monitoring naturally occurring populations and carrying out experimental inoculations of plant surfaces. These studies have shown that plant infection and disease development depend on a number of factors, including the particular host—pathogen combination, critical environmental conditions, physiological stress of the host plant and the attainment of minimal threshold levels of the pathogen.

Although compatible phytopathogenic bacteria share a common ability to spread and multiply within the host plant, the manner in which they do this and the effect they have on the host plant (disease) vary considerably. This chapter considers general aspects of disease induction, different types of disease that are caused by plant pathogenic bacteria and the range of bacterial characteristics that are important in disease development.

The induction of bacterial disease

The ability of plant pathogenic bacteria to cause disease in a particular host plant depends on many features, including environmental aspects, plant physiology and development, and the expression of pathogenicity and virulence factors by the bacterial cells.

Environmental and physiological factors affecting disease development

Environmental factors are important in the development of plant disease for their direct effects on infection (Chapter 5) and for their indirect effects in determining the physiological status of the plant.

The various aspects of the plant which affect disease development are discussed by Lozano & Zeigler (1990) and include nutritional status, photoperiodic conditioning and stage of maturity and development.

Levels of macronutrients have been shown to be important in plant susceptibility to Erwinia stewartii, where elevated levels of N and P increase susceptibility and high levels of Ca and K increase resistance.

The taxonomy of plant pathogenic bacteria, with its three interrelated aspects of classification, nomenclature and identification, is a central aspect of bacterial plant pathology. It is clearly important to be able to establish the identity of an isolated plant pathogenic bacterium, so that the agent causing a particular disease — or with the potential to cause disease — can be clearly defined. Phytopathogenic bacteria may be isolated from a wide range of sources, including infected plant tissue, seed surfaces, soil and water environments, and the identification of bacteria from such sites has implications not only for disease diagnosis and pathogenicity but also for studies on disease epidemiology and aetiology. In addition to defining bacteria as agents of disease, taxonomic studies can also provide useful insights into phylogenetic relationships between the phytopathogens.

The taxonomy of plant pathogenic bacteria will be considered in relation to two major aspects:

1. The establishment of a clearly defined and internationally accepted system of classification and nomenclature.

2. Bacterial identification, including: isolation and identification from different sites, pathogenicity testing, in vitro diagnosis, and computer identification by numerical analysis. There is now a wide range of features on which bacterial classification can be based, and which can be used for identification.

General principles of bacterial taxonomy, with details of the various characteristics that can be used for classification, are discussed in Bergey's Manual of Systematic Bacteriology (Kreig & Holt, 1984).

In natural environments, where a particular host species occurs within mixed vegetation, the development and spread of disease is probably limited to some extent by the separation of individual plants within the area. This constraint does not apply in the crop situation, where localised infection and progression of disease within the homogeneous plant population can occur rapidly. In this artificial situation, where the natural balance between pathogen and host does not apply, special control measures often have to be adopted if the large scale occurrence of disease and consequent major crop loss are to be avoided. These measures fall into four main categories: chemical control, biological control, breeding of resistant cultivars and sanitary procedures.

Chemical control

Chemical control agents are of two main types: bactericides (synthetic organic and inorganic compounds) and antibiotics (naturally occurring microbial products). The use of these two types of control agent is considered in the first part of this section, with a final discussion on general aspects of chemical control.

Bactericides

The range of compounds used as bactericides has recently been reviewed by Sekizawa and Wakabayashi (1990) who divide these compounds into four main categories: synthetic bactericides formerly used for crop protection, currently used synthetic bactericides, traditional inorganic compounds and soil nitrification inhibitors.

Plant pathogenic bacteria are not restricted in their occurrence to infected plant tissue, but are widely dispersed throughout the external environment. This chapter will consider the general occurrence of plant pathogenic bacteria in the aerial and soil/water environments, environmental interactions at the micro-level and the association of these bacteria with invertebrates (vectors).

The aerial environment

The aerial occurrence of plant pathogenic bacteria is clearly of particular relevance to those pathogens that infect aerial parts of plants, including leaves, flowers and fruit. The aerial environment includes both physical aspects (e.g. occurrence of bacteria in rain and aerosols) and biotic aspects (occurrence of bacteria on plant surfaces and aerial dispersal by vectors).

Occurrence of bacteria in rain and aerosols

The aerial environment presents a potentially important medium for both survival and transmission of plant pathogenic bacteria, particularly where cells are contained in rain water or fine water droplet dispersions (aerosols).

Rain-water from infected foliage may contain high levels of phytopathogenic bacteria, and may be important in the spread of bacteria both within and between plants. A good example of rain dispersal of pathogen within single plants is provided by the studies of Miller on Erwinia amylovora (reported in Van der Zwet & Keil, 1979), who showed that if a source of inoculum was present in the upper part of a tree, the region of secondary infection below was cone-shaped due to downward dispersal of bacteria by rain-splash.

The entry of bacteria into the plant during the infection process leads to various types of interaction, observable at the level of the whole plant, constituent tissues or individual cells. These interactions have been investigated experimentally by artificial infiltration of intact plants (‘Inoculation of intact plants’, this page) or by the use of in vitro systems (including micropropagates, excised organs and cell suspensions; see ‘Use of in vitro systems’, p. 132).

Inoculation of intact plants

The effect of different bacteria in determining the nature of the plant response was initially demonstrated by Klement et al. (1964), who artificially infiltrated leaves of tobacco with a range of bacterial species (Fig. 6.1) and observed three main types of result:

1. Hypersensitive reaction (HR): where there is typically a rapid death of the plant cells, with no spread of bacteria to surrounding tissues. This reaction was induced by a range of bacteria comprising various pathovars of Pseudomonas syringae.

2. Disease reaction: involving a delayed host cell response, with spread of bacteria to other parts of the plant. This reaction was induced by Pseudomonas syringae pv. tabaci and resulted in wildfire disease.

3. No observable reaction, after infiltration of the saprophytic bacterium Pseudomonas fluorescens.

The results obtained by Klement et al.

Bacteria—plant associations

The origin and evolutionary development of higher plants has occurred in environments that were already colonised by bacteria, resulting in the co-evolution of a range of bacteria—plant associations. The associated microbes may be broadly considered in two main categories: epiphytic bacteria (present on the outside of the plant) and internal bacteria (infecting the plant tissue).

Epiphytic bacteria

These are associated with the plant surface, which is generally divided into root (rhizosphere) and aerial (phyllosphere) regions. A wide range of bacteria are adapted to various microenvironments at the soil and air interface, and are important in such aspects as nutrient uptake, frost damage, and biological control of plant pathogens. Many of these epiphytic bacteria are saprophytes, obtaining complex nutrients from the plant. Some epiphytic bacteria are also parasites, spending part of their life cycle on the plant surface, and part within the plant tissue.

Infective bacteria: parasites and symbionts

Parasitic bacteria are able to invade plant tissue, where they grow and multiply, and cause localised or general deterioration in the health of the plant. The great majority of these parasites are extracellular, multiplying within intercellular spaces but not penetrating plant cell walls or entering protoplasts. The relatively few parasitic bacteria that are able to penetrate the higher plant cell include members of the genus Agrobacterium (with the ability to transfer part of the genome into the plant cell) and Rhizobium (where the whole organism enters the plant cell).

INTRODUCTION

The cohesion-tension theory (Dixon & Joly, 1894) is widely accepted as the mechanism of the ascent of sap in tall trees. However, some aspects of this mechanism remain incompletely explained. One of these aspects is the effect of freezing and thawing on the conductivity of conifer xylem.

During freezing gases dissolved in water come out of solution and, at ice/water interface velocities above 2.5 mm s−1, gas bubbles nucleate (Bari & Hallet, 1974). Lybeck (1959), Sucoff (1969) and Robson, McHardy and Petty (1988) observed gas bubbles in frozen xylem tracheids. If bubbles are present in the tracheid lumen when the xylem sap is in tension (Le. at pressures below absolute vacuum) they may expand and cavitate individual tracheids.