The subjects of this book are the animals that I will designate as megaherbivores. I define this term to include those plant-feeding mammals that typically attain an adult body mass in excess of one megagram, i.e. ten to the power six grams, 1000 kg, or one metric tonne. This demarcating criterion conveniently encompasses elephant, rhinoceros and hippopotamus among living forms, while giraffe slip marginally into the category. Such animals have been colloquially designated pachyderms; but a thick skin is a minor feature, and it is their very large body size that sets these few species apart from the numerous smaller species of unguligrade herbivore that occupy a wide variety of terrestrial ecosystems today. Paleontologists such as Martin (1967) have used the term ‘megafauna’ to encompass those species attaining a body mass exceeding about 45 kg (100 pounds). However this division is arbitrary and has no functional basis. In this book I show that there are distinctions between animals reaching a mass in excess of 1000 kg, and those of smaller size, in almost all aspects of ecology.

Of course, many whales attain a larger size than any terrestrial mammal, but all whales are carnivorous, feeding on other animals ranging from tiny shrimps to seals. However, among marine mammals there are also the strictly herbivorous sirenians (manatees and dugongs), which feed on submerged plants growing in shallow lagoons and coastal waters. Manatees may weigh up to 1600 kg, while the recently extinct Steller's sea cow weighed up to 6000 kg.

Introduction

In this chapter I consider how the contribution of large herbivores to community and ecosystem processes varies with increasing body size. The ecosystem features to be covered include the biomass levels sustained, energy fluxes and nutrient cycling through this biomass, and the stability of these features over time. The basic question is, how different would these patterns and processes be if megaherbivores were absent from the system?

Biomass levels

Population biomass

The biomass level that a species population sustains represents a relation between the production of food in the environment, and the ability of animals of the species to transform the food into animal biomass. In African savanna regions, vegetation production is proportional to land surface modified by rainfall, while the resting metabolic requirements of an animal per unit of mass are proportional to its body mass raised to the power minus one-quarter. Therefore, if the amount of food available in the vegetation were independent of body size, the population biomass supported per unit of land area should vary in relation to M0.25, i.e. larger species should tend to sustain somewhat higher biomass levels than smaller species.

However, two factors modify the simple relationship developed above. Firstly, the mass-specific metabolic requirements of free-ranging animals, allowing for activity costs, may be scaled in relation to a body mass exponent slightly different from −0.25. For herbivorous mammals, the best available estimate of the scaling exponent is −0.27 (from Nagy 1987, see Chapter 5), i.e. field metabolic requirements scale almost identically to basal metabolic requirements.

Introduction

In this chapter I consider relations between megaherbivore populations and other species, both plant and animal. How different would the community be, in its species composition, habitat physiognomy, productivity and other ecosystem level features, if megaherbivores were absent? The kinds of interactions of importance include (i) the predatory and disturbing impacts of herbivores on vegetation, (ii) competitive and mutualistic interactions among herbivore species, and (iii) the consequences of changes in species representation for ecosystem structure and function.

Impact on vegetation

Larger herbivores exert a direct impact on vegetation by their consumption of plant parts, and by breaking or trampling plants. Plants are damaged by the removal of leaves, bark and other parts, by breakage of branches, which depresses growth, and through being felled or uprooted, causing whole plant mortality. Even low levels of leaf loss may make certain species less successful in competition with other plants, or hold them in the height range within which they are vulnerable to fires. Bark removal may allow attack by wood-boring beetles and pathogens, increasing fire susceptibility. Felling may be fatal to certain plants, but not to others able to regrow by coppice regrowth of the stump. Selective herbivory could suppress the environmental dominance of certain species, providing opportunities for other species to increase. Heavy utilization of all or most species may depress plant biomass and hence primary production. If only the herb layer is heavily impacted, this reduces fire frequency, thereby favoring woody plant invasion.

Introduction

This chapter considers the effects of large body size on social patterns, in particular (i) group size and structure; (ii) male dominance systems; (iii) female mate choice. Except for group size, these features are not readily characterized in numerical terms, and so cannot be related allometrically to body mass. Instead I will employ a cost/benefit analysis, assessing likely gains and losses in the factors determining evolutionary fitness, i.e. survival chances, reproductive contributions, and offspring survival. An inherent shortcoming of such an approach is that it does not adequately allow for possible interactions between these components (Crook & Goss-Custard 1972; Wilson 1975).

Grouping patterns

Jarman (1974) pointed out that among African bovids group size tends to increase with increasing body size. He explained this pattern in terms of the trade-off between the feeding costs of group formation, and the resultant anti-predation benefits. Because of their high specific metabolic rates, small antelopes are selective feeders on high quality plant parts. These are thinly scattered and quickly depleted. Large ungulates in contrast are relatively fiber-tolerant. They experience a much higher density of acceptable food, which is more uniformly distributed and depleted less by other animals foraging in the same area than is the food of small antelope. Intermediate sized species exhibit a gradient between these extremes. Hence the feeding cost of having close companions decreases with increasing body size.

This explains why large ungulates should be more tolerant of nearby conspecifics while foraging than smaller ungulates, but not why they should actively remain in a group. Jarman thus considered predation risks.

Introduction

Eight living species of terrestrial mammal fall into the megaherbivore category in terms of maximum body mass attained. These include two species of elephant, four rhinoceros species, and single species of hippopotamus and giraffe. In this chapter I describe the ecologically important features of their morphology, document their historic and present day geographic distributions, and outline their paleontological origins. This information serves as an essential background to the ecological topics that will form the subject of subsequent chapters.

Morphology

The most basic feature of significance to this book is size. How big do males and females of extant species of megaherbivore grow, in terms of height and weight? Size factors are frequently exaggerated in the literature, especially for large animals that are inconvenient to weigh. A distinction needs furthermore to be made between the asymptotic weights most typically reached by adult animals, and the maximum weights that might be reached by exceptional individuals. Weights may furthermore differ between different subspecies, and within populations in relation to prevailing resource abundance. Animals held in captivity may grow larger or smaller than their wild counterparts, depending on the adequacy of the diets that are provided to them.

Also of fundamental importance are the anatomic features functioning in the procurement and digestion of food. These include in particular the dentition and the structure of the digestive tract.

All large mammalian herbivores are dependent to some degree upon the agency of microbial symbionts for degradation of the cellulose in plant cell walls.

Introduction

Patterns of social organization reflect the cooperative and competitive interactions occurring among animals within local populations relating to survival and reproduction. Generally different age/sex classes differ in their spatial dispersion, i.e. group membership and the spatial relationships both within and between groups. Other social relationships may be evident from the patterns of behavior displayed in encounters, for example those signifying dominant/subordinate relations. Dominance is particularly a feature of adult males, which are inevitably competitors for reproductive opportunities. Females in turn may exert some selection over the sire of their offspring. Anti-predator responses are also appropriately considered in this chapter, since the affiliative relationships established among adult females serve largely to reduce the risks of predation, not only for self but also for progeny.

Group structure

The term group refers to a close spatial association between individuals. However, socially the temporal cohesion of the group is relatively more important than short-term spatial proximity. Groups may vary in size, and in the age/sex classes of animals composing them. Different groups may either move independently of one another, or tend to associate together, or space themselves out with respect to other groups. Grouping patterns may furthermore change seasonally, particularly in relation to variations in reproductive activity.

Elephants

Among African elephants, the nuclear group comprises an adult female together with 1–3 immature progeny of varying ages. However, generally one finds 2–4 mothers plus young associated together to form family units typically numbering about 4 to 12 animals (Fig. 7.1).

Introduction

In this chapter I consider how megaherbivores go about securing their habitat requirements in time and space. What times of day or night are favored for feeding or other maintenance activities, how much time per day is spent foraging, and how does this vary through the seasonal cycle? What size area do animals cover in seeking their food or water needs, and how does this change seasonally?

Temporal patterning of activities

Animals engage in a number of daily activities. These include feeding, travelling between feeding areas and perhaps to and from water, resting, other maintenance behaviors such as drinking, wallowing and grooming, and various forms of social interaction. These need to be scheduled optimally within the diel (day-night cycle), while ensuring that an appropriate amount of time is allocated to each. The animals need to accommodate for variations in temperature, cloud cover, wind and precipitation. Superimposed on these variations is the progression of the seasons, involving changes in the day-night ratio, prevailing temperatures and rainfall, and associated changes in food availability and reproductive physiology.

In the following account I will make a distinction between feeding and foraging. The former is synonymous with eating, i.e. gathering, chewing and swallowing, while the latter also includes movements made while searching for food.

Elephants

African elephants devote roughly equal proportions of the day and night to foraging. They tend to show three peaks in activity, occurring during the early morning, the later part of the afternoon, and around midnight.

Introduction

The decline of megaherbivores did not end with the termination of the Pleistocene. During the nineteenth century, expanding human settlements and continued hunting reduced Asian species to isolated populations, and ivory exploitation led to African elephant becoming rare over most of southern, eastern and western Africa. Following the advent of firearms, white rhino declined in southern Africa from a widespread and abundant species to the brink of extinction over the course of 60 years. In north-east Africa, white rhino recently suffered an even more dramatic decrease, from several thousand animals distributed through three countries in the early 1960s, to a remnant of about 15 restricted to one park in Zaire at the time of writing. Over much of Africa, remaining populations of elephant and black rhino are suffering steady attrition due to continuing human exploitation for ivory and horn. Javan rhino and Sumatran rhino were listed by the IUCN among the world's twelve most threatened animal species; Indian rhino, Asian elephant and black rhino are listed as endangered; and African elephant and white rhino, while currently safe numerically, remain vulnerable to poaching pressures.

Where populations of megaherbivores have been effectively protected, a contrasting conservation problem has arisen. Populations have increased to levels where they have induced vegetation changes such as to threaten the survival of other animal and plant species in these areas. As a result elephant, hippo and white rhino have been culled in the sanctuaries set aside for their protection.

Introduction

In this chapter I cover those ecological attributes that are features of populations rather than of individual animals. Population ecology is generally framed in terms of the logistic model of population growth. The parameters of this model include (i) the maximum or ‘intrinsic’ rate of population growth shown when population density is very low, labelled rmax; (ii) the equilibrium density or ‘carrying capacity’ eventually attained, labelled K.

However, rate of population growth is the difference between recruitment, determined by processes of birth and immigration, and losses, the outcome of deaths and emigration. Since natality and mortality rates vary with age and sex, the realized rate of population increase is influenced by population structure. Strictly, rmax and K are defined only for populations that have attained an equilibrium age and sex composition. However, real populations seldom remain at any equilibrium for long, due to environmental fluctuations. Furthermore, density varies spatially over the population range in relation to habitat suitability.

The ecological features to be considered in this chapter include (i) population composition, in terms of age structure and sex ratio; (ii) rates of population change with time; (iii) population densities attained.

Population structure

In considering age structure, the functional age classes include (i) adults, i.e. animals that have passed the age of socio-sexual maturity and are reproductively active, or at least potentially so; (ii) juveniles, i.e. animals that have not attained the age of independence from their mothers; (iii) subadults, i.e. animals that are intermediate between the above two categories.

Introduction

Much attention has been devoted by ecological theorists to patterns of variability among life history parameters. These not only influence rates of population increase, but are the features most directly associated with genetic fitness. Hence they are likely to be subject to strong natural selection in relation to the ecological conditions prevailing for the species.

A key concept is that of r- and K-selection (MacArthur & Wilson, 1967; Pianka 1970; Horn 1978). These labels are derived from the logistic equation of population growth: r represents the potential rate of population increase, K the equilibrium density or carrying capacity attained by the population. In unstable environments populations are likely to spend much of the time in phases of population increase following catastrophic reductions in density. Under these conditions density dependent factors exert relatively little influence, and the most successful phenotypes should be those which can multiply their descendants at the greatest rate during the increase phase. Such circumstances favor rapid reproduction, i.e. short breeding intervals, large litters and quick growth to maturity. In contrast, in more stable environments populations remain close to carrying capacity for most of the time. In such circumstances intraspecific competition for limiting resources is strong, and the most successful individuals are likely to be those capable of using these limiting resources efficiently towards maintaining a high population density. Selection should favor greater longevity, slower breeding rates and more parental care.

Like the animals it describes, this book has had a long gestation. It started as a Ph.D. thesis on the white rhinoceros, grew into a monograph on rhinoceroses, expanded to include other similarly large herbivores, and then settled on the focus adopted in the pages that follow: the consequences of large size for the ecology of animals such as elephants, rhinoceroses and hippopotami, and by implication extinct species of similar size.

I hope that this work will be of interest to a variety of readers. Firstly, it is written for biologists interested in allometric scaling effects on ecological processes. The correlates of a body mass at the upper limit of the size range among mammalian herbivores are analyzed at various levels, including ecophysiology, behavioral ecology, demography, community interactions and ecosystem processes. Secondly, the book should be an aid to professional conservationists and wildlife managers concerned about the future survival of such large mammals. Scientific facts about these species must be given due cognizance if management actions are to achieve their desired objectives. Thirdly, it is directed towards paleobiologists interested in the ecological roles that similarly large mammals played in the faunas and ecosystems of the past. In particular potential causes of the extinctions of the so-called megafauna during the late Pleistocene are assessed. Finally, I hope that this book will be illuminating to all those who have marvelled at the ways of living of these largest among land animals, whether in the wild or on film.

Introduction

The habitat resources of interest to this chapter are those that individual animals of a species depend upon for their survival. These include food sources, surface water, and refuges from weather extremes.

Food

For large herbivores dietary intake may be characterized either in terms of (i) the plant species eaten, (ii) the plant parts ingested, or (iii) the nutrient contents of the ingested material.

In terms of plant species, the basic classification is in terms of the graminoid: dicotyledon proportions (including non-graminaceous monocots with dicots). The leaves of grasses have higher contents of fibrous cell wall components, and digest more slowly, than the leaves of woody and herbaceous dicots. Silica bodies present in grass leaves further reduce digestibility and also abrade teeth. However, the leaves of woody dicots are ultimately less digestible than those of grasses, due to a higher proportion of indigestible lignin incorporated in the cell wall. Furthermore, the leaves of woody and herbaceous dicots frequently contain toxic or digestibility-reducing compounds, which are much less common in grasses.

In terms of plant parts, the proportions of foliage, stemmy material and fruits in the diet are of interest. Supporting tissues such as stems and bark tend to be high in indigestible fiber, while fruit pulp and seeds contain stores of soluble carbohydrates. Leaves contain the photosynthetic enzymes and are highest in protein and minerals (apart from calcium), although protein content declines as leaves age and fiber contents increase.

Introduction

As animals grow and age they pass through different functional stages in terms of their social relations and contribution to reproduction. These stages may be subdivided as follows (i) infancy and juvenilehood – the period of complete or partial dependence on the mother for sustenance and protection; (ii) adolescence and subadulthood – the early period of independence from the mother, through attainment of physiological sexual maturity; (iii) adulthood – the period following attainment of full social and sexual maturity. Interest lies in the timing of these stages, and in the changing behavioral patterns of animals as they pass through each stage.

Infancy and juvenilehood

This period commences with birth. During the early neonatal period the offspring is completely dependent upon its mother for sustenance in the form of milk. During later infancy the offspring starts supplementing its milk intake with vegetation, but it is some time before nursing ceases and weaning is complete. By use of the term infancy I imply the period during which the young animal could not survive if separated from its mother. Juvenilehood refers to the period of partial dependency on the mother for perhaps some food supplementation, or at least protection from predation. In most species the juvenile period ends when the young animal is driven away by the mother around the time of birth of the next progeny. However, in some species older offspring may remain associated with the mother and her companions through adolescence.

Introduction

It is evident from Chapters 3 and 4 that megaherbivores select high quality green herbage or fruit when these are available, but switch to more fibrous forage during the dry season when choice is restricted. In superficial terms this pattern is not very different from those displayed by smaller ungulates, except for the amount of woody material eaten by elephants at times. Megaherbivores tend to forage for longer over the 24 hour day than do smaller ruminants; but daily foraging times are similar to those of medium-sized non-ruminants like zebras. The home range sizes of megaherbivores are no larger than those of many medium-sized ungulates, again with elephants being a clear exception. To discern body size influences, quantitative data for a range of species of widely varying body size need to be examined.

Diet quality

The nutritional value of food ingested must be adequate to satisfy metabolic demands, otherwise survival chances will be reduced. From the results reported in Chapter 5, total daily metabolic requirements (for maintenance plus activity) are predicted to vary with body size as a function of M0.73. The assimilation rate of nutrients depends both on the capacity of the digestive tract and on the passage rate of its contents. Larger animals can support their lower specific metabolic requirements either by eating less food per day, or by accepting food with lower nutrient concentrations, or some combination of both.

Simulation model of the white rhino population

The model was formulated in PASCAL for implementation on an Apple II microcomputer. The basis of the model is a population made up of 46 year groups, which were grouped into functional age classes differing in their mortality, natality and dispersal rates. The age classes were as follows: old – 36–45 y; adult – 11–35 y; subadult – 6–10 y; immature – 3–5 y; juvenile – 1–2 y; infant – 0 y.

The sex ratio was considered to be 50:50 throughout all age classes. Demographic parameters operated on the year groups in the following order: first emigration, then mortality, then natality. Thus the number of animals entering age group 0 was calculated by multiplying the number of females surviving within the age classes, OLD, ADULT and SUBADULT by the age-class specific natality rates, by a factor of 0.5 to adjust for the sex ratio, and finally by the infant survival rate.

Simple model of expanding and stable populations

In the initial use of the model in Chapters 11 and 13, fixed values were assigned to demographic parameters as in Table II.1 (all rates expressed per annum):

The class of subadults is assumed to operate as a mix of males, still exhibiting immature mortality rates, and females exhibiting adult mortality rates. Subadult females have a varying natality rate depending on the age at first parturition; a natality of 0.07 means that females first give birth in year class 10.

Introduction

Prior to the late Pleistocene, megaherbivores were represented by a wider variety of taxa than occur today, and were present on all continents. Their disappearance from Europe and the Americas took place at the end of the last glacial period of the Pleistocene, around 11,000 years ago, and was synchronous with the extinction of numerous other large mammal forms. The extinctions occurred during a time of rapid climatic change, with associated transformations in habitat conditions. Another important event took place at about the same time: the entry of humans into the Americas, following their expansion through the furthest corners of the Old World. The relative importance of climate and associated habitat changes versus human predation as causal agents in the late Pleistocene extinctions remains an unresolved problem (Martin & Wright 1967; Remmert 1982; Martin & Klein 1984).

Since climatic change and human range expansion are so closely interwoven in time, wider patterns need to be considered in order to understand the causal links in these extinctions. These include the geographic distribution of extinctions, and variations in the incidence of extinctions among genera of differing body size. In the following analysis I focus specifically on large mammalian herbivores, since it is generally accepted that extinctions of carnivores and of large scavenging birds were related to their dependence upon the herbivores as a food source. It is the herbivores that are likely to be most responsive to changing habitats; and it is also such species that were the prime targets as prey for the expanding human population.