Introduction

Senecio keniodendron (Asteraceae) is an abundant and widespread giant rosette species endemic to the alpine zone of Mount Kenya (lat 0°), Kenya. Along with Senecio brassica, it forms a dominant component of the alpine plant community, and shows a high degree of morphological convergence with giant rosette genera in other tropical alpine areas (Hedberg & Hedberg 1979; Smith & Young 1987; Smith, Chapter 1). Here we summarize studies of S. keniodendron population biology carried out between 1977 and 1985, and briefly compare these results with those both for S. brassica, and for the convergent Andean genus Espeletia (Asteraceae).

Senecio keniodendron occurs most commonly on upper slopes and ridges from c. 3700–4600 m elevation. Adult densities generally range from 1 to 10 plants per 100 m2, with greatest densities on talus slopes between 4000 and 4200 m. The mean height of S.keniodendron plants is significantly positively correlated with slope angle (Figure 15.1; r = +0.60; p < 0.001). Senecio keniodendron is replaced by S. brassica on lower slopes and valley floors; S. brassica has a lower elevational range than S. keniodendron – c. 3400–4400 m. The upper limit of growth for S. keniodendron (4600 m) may be due to a combination of drought and freezing stress (see also Perez 1987, on Andean Espeletia). Figure 15.2 shows the surface response curves for the densities of S. keniodendron (p < 0.04) and S. brassica (p < 0.01) im the Teleki Valley, relative to elevation and slope.

Introduction

Isoetes (Isoetaceae) is a genus of small herbaceous plants often aligned with Lycopodium and Selaginella. There are more than 150 species distributed worldwide, typically in aquatic habitats (Tryon & Tryon 1982). A particularly intriguing aspect of the physiology of these plants is the presence of Crassulacean Acid Metabolism (CAM) (Keeley 1981, 1982), a photosynthetic pathway commonly associated with terrestrial xerophytes. CAM was selected for in these species by the daytime carbon limitation characteristic of their oligotrophic aquatic habitats (Keeley & Busch 1984; Boston & Adams 1985).

Across its range, Isoetes has radiated into a variety of aquatic as well as some terrestrial habitats and these environments have selected for a number of different structural-functional syndromes (Keeley 1987). Aquatic species occur in lacustrine habitats where they are permanently submerged throughout their life cycle and in amphibious environments where they alternate seasonally between aquatic and terrestrial conditions. In general, all aquatic species so far tested possess a well-developed CAM pathway while under water but lose this pathway when grown in an aerial environment. True terrestrial species of Isoetes are few in number although such species are known from most parts of the world. They readily fall into one of two groups: vernally active, summer-deciduous species at relatively low elevations in temperate latitudes; and evergreen species restricted to very high elevations (> 3500 m) in tropical latitudes. The former species show no CAM activity (even if artificially submerged), possess stomata and presumably depend entirely on C3 photosynthesis.

Introduction

Tropical alpine environments are extreme and can generate very powerful selection pressures. Because of their simplicity they can provide valuable models for the study of evolution (Bradshaw 1971). The arborescent rosette growth form is a dominant feature across tropical alpine landscapes and has been the subject of detailed comparative ecological studies (Smith 1979, 1980, 1981; Smith & Young 1982; Young 1985 and Chapter 14). Alpine plant communities of most tropical alpine regions are distinguished from those of temperate latitudes by the presence of these giant rosette plants (Hedberg 1964; Cuatrecasas 1968; Smith 1981), with each major region possessing a unique flora. It has been assumed that the giant rosette form is an adaptive response to tropical alpine environments (Hedberg 1964; Mabberley 1973; Smith 1981), and an example of convergent evolution by different plant families to similar ecological conditions.

A notable feature of tropical alpine systems worldwide is the prevalence of pubescence, particularly among arborescent rosette plants. Most giant rosettes, including Senecio and Lobelia in Afroalpine areas (Hedberg 1964), Lupinus, Espeletia and Puya in the páramos of South America (Heilborn 1925; Smith 1981; Miller 1986), and Argyroxiphium in Hawaii (Carlquist 1974 and Chapter 16), produce dense pubescence on their leaves and/or inflorescences.

With current attention on global problems of biodiversity and climate change, environmental interest in tropical ecosystems has increased tremendously. Very often, nevertheless, tropical biology is focused on lowland humid forests. High mountain systems, however, are also an important feature of tropical landscapes. Compared to lowland tropical forests, there has been surprisingly limited interest in the ecology of organisms in these tropical systems.

Scientific interest in the flora of tropical alpine regions goes back to the middle of the eighteenth century when Joseph de Jussieu and Charles La Condamine collected plants and mapped the high mountain areas of Ecuador as part of a five-year expedition of the French Académie des Sciences. The most vivid early scientific accounts of tropical alpine environments, nevertheless, came from travels of Alexander von Humboldt in South America and Mexico at the beginning of the nineteenth century. Accompanied by a capable young botanist named Aimé Bonpland, Humboldt travelled extensively through the high páramos of Colombia, Ecuador, and northern Peru. The patterns of vegetation zonation which he observed on this trip had a great impact on his thinking, and helped lead to the founding of the modern science of biogeography. The roots of modern ecology can also be traced to this experience which demonstrated to Humboldt the importance of interrelationships between climate, soils and biotic communities. Humboldt saw clearly that the peculiar vegetation of the páramos of the northern Andes was unlike any alpine community in temperate mountain ranges.

Introduction

It is likely that Polylepis (Rosaceae) occurs naturally at higher elevations than any other arborescent angiosperm genus in the world. The 15 species (Simpson 1979) are confined to the South American Andes where they occur primarily in tropical alpine environments. Some Polylepis species tend to form discrete forest stands reaching elevations over 5000 m, well above the upper continuous forest limit (timberline). Throughout their high altitude distribution most members of this genus are exposed to rigorous climatic conditions in which diurnal temperature variations by far exceed seasonal ones and night frosts are frequent.

The genus is exclusively arborescent (trees or shrubs), with individuals ranging in height from 1 m to no more than 30 m. The trees tend to have twisted, crooked stems and branches, particularly in open, exposed habitats. The form and branching pattern of some individuals resembles those of krummholz trees found in temperate alpine regions. The bark is deep red in color and consists of several layers of thin, exfoliating sheets. Although the exfoliating bark is particularly thick at the base of the stem or large branches, the insulating effect is by no means comparable to that of the marcescent leaves that surround the stem of the adjacent giant rosette plants (Smith 1979; Goldstein & Meinzer 1983). The leaves are compound and alternate but often appear whorled owing to the compression of internodes at the branch tips. The leaflets are small, dark green above and are covered with dense, silvery trichomes on the underside in several species.

Introduction

Tropical alpine environments impose unique selection pressures on plants, producing a number of special adaptations. Among these are caulescent and acaulescent rosettes, nyctinasty, semelparity, and resistance to nightly frost. For each of these adaptations there are a number of possible evolutionary explanations. To discriminate between among explanations, and to understand better tropical alpine environments in general, integrated long-term studies of particular species are necessary. Such studies include morphology, ecological physiology, demography and reproductive biology. Long-term population biology studies of tropical alpine plants are rare; most are presented in this volume. These studies are helping us to understand the nature of adaptation in tropical alpine environments.

Giant rosette plants make ideal research subjects for a number of reasons. They are conspicuous and characteristic members of virtually all tropical alpine communities. Their morphologically discrete form makes their growth and individual dynamics easy to quantify. Lastly, they are relatively long-lived, enabling us to examine the effects of long-term changes in local environment.

Since 1977, I have been studying two closely related giant rosette species on Mount Kenya, Lobelia telekii and L. keniensis. I present here a summary of the first 7 years of that study, concentrating on the comparative population biology of these two species.

A number of factors make Mount Kenya lobelias particularly attractive as research subjects. (i) The alpine environment of Mount Kenya is one of the most intensively studied ecosystems in the tropics (Hedberg 1957, 1964; Coe 1967; Coe & Foster 1972; Young & Peacock 1985; Young 1990a).

Introduction

General climatic features of tropical alpine regions have been discussed in Chapter 1 of this volume. To reiterate, the nearly complete lack of temperature seasonality in tropical alpine zones is a key feature in distinguishing them from temperate alpine zones. The Andean páramo zone and similar zones in other tropical high mountains are characterized by high inputs of solar radiation in the presence of low inputs of thermal energy. This characteristic might be expected to present special circumstances from the standpoint of regulation of leaf thermal balance, in contrast to temperate alpine and desert habitats where both solar radiation and thermal energy inputs may be seasonally high. It has been suggested that some of the prominent morphological features found in giant rosette plants (see Chapter 1) represent adaptations for regulation of thermal balance under the special microclimatic conditions encountered in tropical mountains (Hedberg 1964; Larcher 1975).

Numerous studies have dealt with the importance of characteristics such as leaf absorptance to solar radiation, leaf angle and rate of transpirational cooling as determinants of leaf temperature under a given set of environmental conditions (Mooney et al. 1977 Geller & Smith 1982). However, fewer studies have examined the interaction between spatial and temporal changes in environmental variables and plant features thought to be important for regulation of leaf thermal balance (Smith & Nobel 1977; Ehleringer & Mooney 1978).

Historical introduction

Ecological interest in climates of the tropical alpine regions of the world dates back perhaps to the early travels of La Condamine in Ecuador in the mid-19th century, but more dramatically to the remarkable explorations of Alexander von Humboldt in the northern Andes, Central America and Mexico from 1799 to 1804. His contributions to science from these travels, which encompassed geography, biology, geology, climatology, anthropology and other subjects, filled 30 volumes (von Humboldt 1807–39), and had a tremendous influence on the intellectual and economic development of Latin America in the 19th century. No less an authority than Simon Bolivar once remarked, ‘Baron Humboldt did more for the Americas than all of the conquistadors’ (Von Hagen 1948). The scientific studies of von Humboldt served as the foundation of the modern science of biogeography, and his climatological observations in the Andes and on the Mexican volcanoes played a major role in the development of his ideas.

Another notable advance in scientific knowledge of alpine climatology in tropical mountain regions came not from a typical scientist at all but from the noted European alpinist, Edward Whymper. Whymper, who had been the first man to scale the Matterhorn in the Swiss Alps in 1865, came to Ecuador in 1879 to attempt climbs of Chimborazo and other high volcanoes of that region. He was spectacularly successful not only in these ascents, but in the wealth of ecological and climatological data which he collected and published (Whymper 1892).

Introduction

Communities of Espeletia, known locally as ‘frailejones’, constitute the dominant and most striking physiognomic elements of the high montane páramo vegetation of the northern Andes (Figure 13.1). Different species occupy a variety of habitats, ranging from cloud forest, as low as 1500 m in elevation, to close to the upper limit of plant growth, at 4700 m. With over 130 currently recognized species and a wide diversity of life forms and other morphological features, Espeletia has been noted as one of the foremost examples of adaptive radiation in plants (Carlquist 1974).

One of the keys to understanding the ways that groups such as Espeletia have successfully colonized different habitats in high, tropical mountains lies in a study of their reproductive systems. Vegetative reproduction in Espeletia is rare and occurs to a limited extent in just a few species that produce axillary rosettes close to the ground (Cuatrecasas 1979). Seed dispersal is also very restricted, due to the absence of a pappus or other specialized dispersal structure (Cuatrecasas 1976; Smith 1981; Guariguata 1985). Thus, the breeding and pollination systems are the major determinants of gene flow in Espeletia and constitute an important factor affecting the genetic structure of populations.

The little previous information on the reproductive biology of Espeletia was obtained from studies undertaken in the páramos of Edo. Mérida, Venezuela.

The general term ‘tropical alpine’ refers to regions within the tropics occurring between the upper limit of continuous, closed-canopy forest (often around 3500–3900 m) and the upper limit of plant life (often around 4600–4900 m: Hedberg 1951, 1964; Beaman 1962; Troll 1969; Wade & McVean 1969; Wardle 1971; Van der Hammen & Ruiz 1984; Vuilleumier & Monasterio 1986; see Figure 1.1) and is used in preference to regional terms such as ‘páramo’ and ‘jalca’ in the moist Andes from Venezuela to Northern Peru, ‘puna’ in the drier central Andes, and ‘Afroalpine’ and ‘moorland’ in Africa. No clear lower boundary can be defined where natural timberline has been eliminated by man, as in many areas of the Andes and Papua New Guinea (Wade & McVean 1969; Hope 1976; Ellenberg 1979; Ruthsatz 1983), or where the forest is patchy or absent due to low rainfall, as on the north slope of Mount Kenya (Coe 1967; Figure 1.2) and the western slopes of the Peruvian Andes (Weberbauer 1911). In these cases alpine species merge gradually with species of montane pasture, savanna or desert.

Physiognomy of tropical alpine vegetation varies greatly with climatic and edaphic factors; however, certain trends are held in common by many New and Old World tropical alpine areas (see, for example, Hedberg 1964; Coe 1967; Cuatrecasas 1968; Wade & McVean 1969; J. Smith 1977, 1980; Cleef 1978), suggesting convergent evolution (Hedberg & Hedberg 1979; Halloy 1983; Smith & Young 1987).

Introduction

Despite recent interest in alpine ecology, we have little information on anatomy of alpine plants. This is as true of tropical alpine plants as it is of those from temperate mountain areas. The reasons for the lack of studies in anatomy of tropical alpine plants are perhaps surprising.

First, one can cite the tendency for plant anatomists to work little on comparative problems and, when they do, to work in terms of particular taxonomic groups and to express their data in systematic terms rather than in ecological ones. To be sure, interest in ecological plant anatomy has increased in the latter half of the 20th century, and one can cite more studies concerning tropical alpine plants in recent years. The curious climatic regimes of high equatorial mountains make anatomical adaptations of especial interest, as will be seen from the relatively few examples cited in this chapter. Adaptations to frost and to drought are central in tropical plants, but these adaptations are different from those in plants of extremely cold or extremely dry regions.

A second reason for lack of studies on anatomy of tropical alpine species is a by-product of the working habits characteristic of biologists. Ecologists, unless highly theoretical in orientation, tend to use the outdoors as a laboratory. For the plant anatomist, laboratory work must be done indoors, and the habit of working both in the field and in the laboratory apparently does not come easily to most plant anatomists.

Introduction

The savannas of East Africa, perhaps more than anywhere in the world, are known for dramatic plant–animal interactions. Although large herbivores are less common above timberline than below it in East Africa, herbivory is also a powerful force in Afroalpine plant ecology. Many vertebrates and invertebrate herbivore species occur above treeline on Mount Kenya (Moreau 1944; Coe 1967; Jabbal & Harmsen 1968; Coe & Foster 1972; Mulkey et al 1984; Young 1991; Young & Evans 1993). Herbivores influence the distribution of plant species, the size structure of populations, and the success of individual plants on Mount Kenya.

Long-term studies of giant rosette Lobelia and Senecio species have brought to light a number of interesting patterns of herbivory that have major impacts on the biology of these species.

Lobelia

The two high alpine giant rosette lobelias on Mount Kenya are Lobelia telekii and L. keniensis. Mount Kenya lobelias are subject to herbivory from a variety of animal species (Table 18.1). None of the herbivores is host-specific, except perhaps the coleopteran larvae associated with L. telekii roots. Most of these herbivores have only minor effects. However, Lobelia populations near hyrax colonies have suffered severe predation.

The leaves and stems of both Lobelia species are protected by a bitter latex containing anti-herbivore compounds (Mabberley 1975). Although this may limit the activity of their herbivores, it far from renders the plants invulnerable.

Introduction

In tropical alpine regions drought may be the most important seasonal factor in an environment that otherwise lacks significant seasonality (see Smith, Chapter 1). In many of these regions diurnal cycles of physiological drought associated with low soil temperatures may be superimposed on the seasonal changes in soil moisture. From studies of cold temperate zone plants, particularly conifers (Kaufmann 1975, 1977; Running & Reid 1980), it is known that even in soils near field capacity, water uptake by roots may be severely impaired by low soil temperatures (0–5 °C). This high root resistance to water uptake may extend through spring and into early summer in temperate zone coniferous forests with persistent snow cover. In arctic regions, physiological drought may extend throughout the entire summer if the roots are situated above a permafrost layer (Goldstein 1981).

In the tropical alpine zone, where snow cover is not persistent and permafrost does not exist, the potential for physiological drought is nevertheless present. The risk is especially great during the early morning hours when soil temperatures in the root zone are near freezing and potential transpiration is high due to high solar radiation loads. The simultaneous occurrence of low water availability in the absolute sense on a seasonal basis, and in the physiological sense on a diurnal basis, complicates the analysis of drought resistance mechanisms in tropical alpine plants. It also provides an excellent opportunity to study adaptations to drought along both altitudinal and geographical gradients of relative importance of diurnal versus seasonal drought.

Introduction

‘Summer every day and winter every night’ (Hedberg 1964) is a brief but succinct characterization of the tropical alpine climate, pointing to the fact that the amplitude of the daily temperature oscillation by far exceeds that of the monthly mean values. Although cloudiness exerts a mitigating effect on the daily temperature extremes during the rainy seasons, nocturnal frost may occur throughout almost all of the year at altitudes above 4000 m. Therefore tropical alpine plants must maintain mechanisms of permanent frost hardiness which differ considerably from those providing the overwintering plants of temperate climates with seasonal frost resistance. Whereas, for example, in Norway spruce the frost-hardy state is characterized by a high proportion of unsaturated fatty acids in the membrane lipids (Senser 1982), by a shift from photosynthetic starch formation to the production of sucrose and its galactosides (Kandler et al 1979), by a reduced capability of photosynthetic electron transport (Senser & Beck 1979) and by a suspension of growth activity, tropical alpine plants must combine physiological features providing frost resistance with continuously high rates of photosynthesis and growth.

As physiological plant ecologists and population biologists continue to work in tropical alpine environments, our knowledge of the form and functional relationships will surely grow rapidly. These ecosystems present unusual challenges for plant establishment and survival, but remain poorly studied. In this closing chapter, we briefly review what we feel are the important accomplishments to date in tropical alpine ecology and the challenges that remain. We focus on plant growth forms, plant demography, physiological convergence, ecosystem function and global climate change.

Plant growth forms

The most striking aspect of tropical alpine habitats is the diversity of plant growth forms, and the apparent convergence between geographically disjunct tropical mountains with respect to these forms (Hedberg & Hedberg 1979; Rauh 1978). Much progress has been made in understanding the ecological and physiological significance of the giant rosette form, perhaps the most conspicuous and typical form of high tropical mountains. However, the majority of tropical alpine growth forms have not been subject to quantitative and experimental analysis. Sclerophyllous-leaved shrubs, cushion plants, tussock grasses and acaulescent rosette forms have received minimal attention. Species in each of these forms may respond differently to changing environmental conditions, and may contribute in very different ways to edaphic, microclimatic and biotic environments.

Classical ecological paradigms suggest the evolution will lead toward convergence in adaptive traits of morphology, phenology, and physiology which provide the ‘best’ ecological solution to similar environmental stresses in disjunct habitats.

Pachycaul senecios from the upper Afroalpine zone produce stems up to 11 m tall which are coated with marcescent leaves. Each stem or branch is terminated by a huge leaf rosette composed of 30–120 leaves which surround the central cone-shaped leaf-bud. During the course of a year about 50–60 leaves are produced from the bud (Beck et al. 1980) and in principle the same number of mature leaves at the outer periphery of therosette become senescent and die. The leaves are up to 50 cm long and 15–20 cm wide. Senescence commences at the tip which becomes yellow and subsequently necrotic. Yellowing of the lamina spreads from the tip downwards to the base. A senescing leaf, the upper part of which was already partially necrotic, was found to be still capable of photosynthesis in the lower part (E. Beck, unpublished data). This mode of gradual senescence suggests that a substantial portion of the nutrients can be mobilized and exported from the leaf before it dies. Table 11.1 shows the disappearance of nitrogen in the senescing leaf. The plant was located at 4200 m elevation in the Teleki Valley, Mount Kenya. The prolonged viability of the midribs presumably facilitates this mobilization. This type of internal recycling of nitrogen and other nutrients is common in higher plants. However, unlike the majority of the higher plants the pachycaul groundsels do not shed the dead leaves but maintain them as a dense collar around the stem.

Introduction

The nutrient relations of tropical alpine areas are poorly known despite the importance of these processes for understanding spatial and temporal variation in ecosystem productivity (Speck 1983). Nitrogen mineralization – the release of inorganic nitrogen from the substrate – is of particular importance in this respect, and has proved useful in assessing ecosystem dynamics in the European Alps (Rehder 1970, 1976a, b, 1982; Rehder & Schäfer 1978). In this chapter I address the problem of nitrogen mineralization and its impact on primary productivity in the alpine zone of Mount Kenya.

Materials and methods

Samples areas were established in the two main alpine plant communities of the Teleki Valley, on the western flank of Mount Kenya: ‘Senecio keniodendron–Lobelia telekii community’ (Area 1) on steeper slopes and ‘Lobelia keniensis– Senecio brassica community’ (Area 2) on moderate slopes and valley bottoms (see Table 12.1; also Hedberg 1964; Beck et al. 1981; Rehder 1975, 1983; Rehder et al 1981, 1988). The soil at area 1 was a loamy ‘mountain brown gley soil’ while the soil at area 2 was a blackish peaty ‘mountain wet gley soil’ (Beck et al 1981). Mean maximum temperatures were lower at the soil surface in area 2 (Table 12.2).

Nitrogen mineralization was measured with a field incubation test (Ellenberg 1964, 1977; Gerlach 1973). Volumetric soil samples were taken from three depths between 0 and 15 cm, weighed, sieved (4 mm screen), mixed and homogenized by hand.