Heated debates on the conservation management of elephants often are dominated by personal opinion, hearsay, anecdotal statements and personal interpretations of research findings. Selective understanding assumes that personal interpretations cannot be questioned, especially when based on science. Doctrines enforced by generalisations such as 'there are too many elephants' or 'elephants destroy biodiversity' motivate calls for the management of elephant numbers without consideration of what the consequences of such management will be. Media reports often boost such calls and define the 'elephant problem' in terms of numbers and the rates at which numbers change over time. The selective use of words such as 'destroy, destruction, extreme, overabundance', also by some scientists, adds to some of the misconceptions and it therefore is not surprising that fables replace facts when proposing solutions for the 'elephant problem'.
In this document we attempt to set the record straight by addressing some generalisations, facts and fables that dominate discussions on elephant management. Much of the interpretation is based on recent scientific papers and on 10 years of intensive academic research on elephant populations across a gradient of environmental and management conditions across southern Africa. This research covered populations in South Africa, Namibia, Botswana, Zambia, Malawi and Mozambique. The research findings have either been published in peer-reviewed scientific journals, or have been accepted for publication, or are presently under peer review.
Conservation management of elephants should be dictated by science
This is a myth - science and scientists can only contribute to management as a process that is dictated by societal norms as expressed through value systems. Science does, however, provide a foundation and set of tools that should be used in framing, developing and evaluating conservation management decisions/approaches. We fully appreciate that the management of elephants is connected to differences of opinion, especially where science and ethics are concerned. We are in support of the scientific method and approach as a tool to guide and develop ecologically meaningful, site specific conservation management protocols for elephants.
The African elephant is endangered
The precarious status of Africa's elephants is not really a matter of debate and the cautious standing to its conservation status makes sense. The apparent killing of about half of Africa's elephants in the 1980's motivated the Conference of Parties (CoP) to the Convention on the International Trade in Endangered Species of Wild Flora and Fauna (CITES) to list African elephants on Appendix 1 of the convention in 1989. This banned international trade in ivory. Present opinions on this status differs across the continent and some conservation officials from several southern African countries level arguments to favour trade in ivory as a reward for their management efforts that led to the increase in local elephant numbers.
Most East African and a growing number of West and Central African countries feel differently and officials believe that the resumption of legitimate trade will fuel illegal ivory poaching and threaten dwindling populations elsewhere on the continent. Present estimates suggest that some 470,000-600,000 elephants live in Africa, some of which are better protected than others. Their fragmented distribution in Central and West Africa is due to habitat loss and increased landscape degradation. Such losses follow from developmental initiatives that are fuelled by the extraction of natural resources. The East African populations are also losing habitat due to increasing demands from rapidly increasing human populations. These threats do not seem to be apparent across most of southern Africa where elephants in several countries are extending their ranges, albeit it in response to either good or bad management practises. Based on numbers it may well be argued that the African elephant is not locally endangered, but that several local populations are threatened. However, accepting scientific reasoning that motivates the distinction between savanna and forest elephants as distinct species may alter this viewpoint, especially when considering the lack of reliable information on the status of forest elephants.
Past management of elephants is based on scientific research
This is a myth - an evaluation of past management of elephants does not support this notion. Science, in some cases, did inform management but was not the basis for management decisions. Management decisions seem to have been based on personal experiences, opinion and advocacy.
This is no different for elephants than for the management of a large number of other species - managers often seem to give more weight to experience than to formal scientific research findings. Some of the reasons for this apparent anomaly are the inaccessibility of research findings, be it for research either not addressing specific management issues, or scientists presenting their findings through channels and/or terminologies that are inaccessible to managers or, a lack of cooperation between scientists and managers. Management decisions therefore tend to be based on experience, some of which has been gained by others than the managers themselves. To date, word-of-mouth and causation (based on observation) have more than often dictated management decisions.
In southern Africa the increase in elephant numbers is a measure of conservation success
This is only partly true. Conservation measures such as the control of poaching certainly provided for safe havens where elephant numbers could increase, so did the fencing of conservation areas and the provision of water through waterholes. Fencing and water provision, however, created a new set of problems. Fences hinder the natural limitation of elephant numbers by interfering with dispersal movements and water provision reduces natural die-offs during dry spells that are typical of the savannas where elephants live. A lack of dispersal and high survival drive high rates of increases in numbers. Conservation management actions such as these thus may be the cause of the so-called 'elephant-problem'.
Elephants destroy trees
Elephants certainly do kill trees by uprooting, debarking and/or breaking branches (thereby making them more susceptible to fire and/or infestation by wood-boring insects). Savannah trees, however, are also killed by fire and droughts and it is difficult to distinguish and or separate the causes of trees dying off.
Scientists working in South Africa's Kruger National Park and in Botswana's Chobe National Park noted (in independent publications) the consequences that the rinderpest (infectious viral disease effecting both cattle and other large herbivores, which reached epidemic proportions in southern Africa in the 1890's) epidemic may have had for tree survival - their notion is that the dramatic fall in herbivore numbers through this exotic disease provided trees the opportunity to survive to adulthood in abnormally high numbers. This skewed tree numbers in the savannas, which, in more recent years, through the local recoveries in elephant numbers, are being reduced, consequently recovering the former state of savannas.
High elephant numbers explain their impact on vegetation
This is not entirely true - elephant numbers per unit area varies greatly in conservation areas across southern Africa - some places that experience high impact do have elephant densities (number of individuals per unit area) lower than some places that experience little impact. It is more likely that impact results from fences and the provision of water obstructing and altering seasonal and between-year movements of elephants. Such interferences may force elephants to utilise the same areas throughout the year and from year to year, thereby not giving plants the seasonal and periodic relief that enables them to recover naturally. Impact therefore may be the outcome of spatial limitations rather than numbers. This is also supported by the influence of elephants on vegetation being dictated by rainfall patterns rather than elephant numbers per unit area. The management of the so-called "elephant problem" therefore should focus on the impact of elephants and not on their numbers. Management strategies therefore should focus on restoring the way elephants utilise natural landscapes.
There are too many elephants
This is not true. In Africa there are presently only about half the number of elephants we had some 40 years ago - between 1970 and 1989 illegal poaching reduced elephant numbers to around 500 000. Most populations in southern Africa escaped this abnormal incidence and numbers have since increased, albeit not in all countries, to former numbers. For instance Zambia's Luangwa valley presently has only about a quarter of the elephants that lived there in the early 1980's. At about 150, 000 elephants, Botswana's population is also about a third lower than in the early 1900s. That said, it should also be noted that elephants are notoriously difficult to count. Consequently scientists have developed surveying methods that yield estimates of population size as indices of real numbers.
These indices are mere mathematical abstractions based on surveys of which the precision varies with numbers and with survey intensities. Due to costs, surveys seldom cover more than ten per cent of the area of land where elephants occur in protected areas. Inaccuracies due to thick vegetation cover that limits elephant sightings, the characteristics of the survey areas, observer bias and experience or lack thereof, as well as the number of elephants that live in an area all contribute to levels of precision. In general, precision is extremely poor with an average value of 65 per cent for the 596 estimates that have been assessed. This means that a population estimated to comprise 1,000 elephants may have as few 350 or as many as 1,650 elephants.
Most estimates are therefore by no means a true reflection of the actual numbers of elephants that live in a specific area. For southern African conservation areas that protect some 270,000 elephants there may be anything from 81,000 to 459,000 elephants in reality. The statement that there are too many elephants therefore has a poor foundation and little scientific support.
Despite the difficulty of counting elephants a comparison of year-to-year estimates show that elephant numbers are increasing rapidly in most protected areas in southern Africa
This over-simplification is not true. Nearly half of the estimates of elephant numbers from across Africa are of low quality due to extreme extrapolations, field conditions, low survey intensities and high aircraft speed during surveys. Only nine of the 596 estimates from surveys completed over the past 40 years (1966 to 2006) and that reported sufficient detail were precise enough to be used to detect a five per cent yearly rate of increase in numbers.
Furthermore, statistical methods that are routinely used to estimates rates of changes in population numbers may need as many as 17 yearly estimates to confidently detect a five per cent yearly rate of change in elephant population size. Only one out of the 85 time series that we constructed from sample survey estimates had estimates that were good enough to confidently detect a population trend. Thus, for elephants, the low precision of estimates and the limited number of estimates per time series make most estimates from sample surveys insufficient for the detection of growth rates in a scientifically acceptable manner.
Nearly 70 per cent of 156 elephant populations in Africa for which relatively good census data exist, are stable. However, only 30 per cent of these time series had sufficient statistical power to deduce that populations were stable (mostly from southern and East Africa). To conclude, elephant populations in Africa may be increasing, decreasing, or be stable in size over time. The majority of populations showed no significant change over time. These inconclusive trends should thus not motivate management decisions.
Without management elephant numbers will double in ten years
This statement is false and based on unrealistic assumptions. Under the most ideal conditions elephant populations that increase through births and deaths may double in about 13 years as an outcome of a maximum intrinsic rate of increase of 5.5 per cent per year, only under the unlikely assumption that environmental conditions will be constant over the period of increase. Higher growth rates can only be achieved through skewed immigration, often in response to water provision, or may occur in newly founded populations that through chance events experience synchronous births soon after having been established, or in breeding herds with an unstable age structure, as has been noted in several newly founded populations in South Africa. The calculation of doubling time is based on the unlikely assumption that resources cannot limit population increase.
Culling effectively controls elephant numbers
This is not true. If implemented continuously and at a rate higher than the population's growth rate, the culling of elephants in confined populations will reduce their numbers However, culling may also stimulate the local immigration of elephants into such areas, where competition for resources is minimal as has been the case in South Africa's Kruger National Park and Zimbabwe's Hwange National Park. Moreover, by lowering elephant numbers that can relieve vital rates from limitations enforced by factors associated with high numbers (density dependent forces), culling may effectively stimulate population growth rate. All culled populations that we are aware of increased in numbers after the cessation of culling, sometimes at higher rates than what can be explained by births alone, thus supporting the notion that dispersal movements enhances population growth under certain conditions. Furthermore, and more importantly, we have no evidence that culling has reduced the impact that elephants have had on vegetation. For instance, in the Kruger National Park, where some 17,000 were culled over a period of 27 years, tree numbers declined by some 38 per cent for most of the time when culling kept population numbers relatively stable. This provides further support for our idea that impact is the consequence of spatial utilisation patterns rather than number per se.
Contraception can address the elephant problem
Contraceptives can be applied to elephants and can limit individual reproductive output. Contraception can even reduce population growth rates when applied to at least 75 per cent of all reproducing females in a population. However, contraception does not reduce elephant numbers instantaneously, nor does it stop elephants from feeding on trees - contraception thus does not address the impact that elephants may have on vegetation and other species. Contraception may be used to control the sizes of breeding herds in small and intensely managed areas. The long-term consequences of contraception for the social well-being of individuals are not known.
High birth and low death rates allow elephants to increase rapidly
This is not true. Under ideal conditions the average female elephant gives birth when 12 years old and produces 12 calves over an ideal life-time of 60 years. When considering that half of these calves will be females, each of whom, if exposed to conditions the same as those that their mothers experience or have experienced, will produce the same number of calves. Under these theoretical conditions the population could increase at about seven per cent per year through births and deaths. These conditions never prevail in nature and not all calves survive to reproduce. Consequently some populations may increase at 5.5 per cent per year but most populations increase at lesser rates, some stabilise in numbers and some decrease, albeit in response to poaching, habitat loss or dispersal movements out of their range.
Elephant populations can limit their own growth rates
Theoretically all animal populations tend towards a state where numbers remain relatively stable over time - consequently their long-term growth rates centre on zero. The statement that elephants can limit their growth rates holds true only under special conditions where management, such as the provision of water and the placement of fences, are not interfering with processes that limit growth rates. New research supports earlier notions that both the calving interval and the age at which a heifer will calve for the first time increases with an increase in elephant densities. The survival of these calves and of other elephants seems to decrease with increased variability in rainfall. Reduced survival and reproductive outputs reduces population growth rates and some of the largest elephant populations in southern Africa (notably that of northern Botswana) have stabilised, probably due to reduced birth rates, although increased dispersal and the extension of their range into neighbouring countries may also explain stabilisation. Increased dispersal movements also accounts for the recent stabilisation of elephant numbers in Zimbabwe's Hwange National Park.
The birth and death rates of elephants change with numbers
This is partly true, as increasing numbers give rise to increased densities, which can reduce birth rates but not always increase death rates. In some savannahs conception rates vary with rainfall and primary plant productivity. In others death rates apparently increase with increasing variability in rainfall due to extended droughts that are typical of the savannas where elephants live. Rapid changes in numbers, however, are more likely driven by dispersal movements where such opportunities exist.
Elephant management needs to be based on numbers
This is not true. Management that focuses on numbers only ignores that the impact that elephants may have on other species primarily results from the way that elephants utilise land and the space available to them. The way they utilise space in fenced-off protected areas is dictated by the distribution of water and by fences that hinders roaming behaviour, emigration and immigration. Management should focus on impact and therefore on spatial utilisation and not on elephant numbers per se. Furthermore, where managers are concerned about small and decreasing populations, it may be more feasible to monitor size and age structures to determine trends in elephant mortality and their causes.
Dispersal can limit elephant numbers
This notion is supported by recent research on populations in Botswana and Zimbabwe. Dispersal (immigration and emigration) adjusts population numbers to short-term changes in food and water supplies, thereby decreasing or increasing the growth rates of populations locally. Although there is relatively little data on dispersal in elephants and the effects thereof, there is evidence that the construction of fences around conservation areas may lead to an increase in numbers, because elephants are no longer able to disperse out of these areas. Furthermore, in northern Botswana, while elephant numbers increased over 20 years, their densities remained relatively stable over the same period, probably because of elephants moving to neighboring countries. The stabilisation of elephant numbers in Zimbabwe's Hwange National Park also recently has been ascribed to dispersal movements. Here, water provided by management attracts elephants while the lack thereof due to dysfunctional water holes, repels them. Dispersal movements induced by water management therefore may be important for the management of elephant numbers and the impact that elephants have on vegetation.
Time related variability in environmental conditions may limit elephant populations
This is likely the case, as all animals respond to prevailing environmental conditions. Environmental conditions change continuously and dictate the availability of food and water, both which dictate habitat utilisation over time and space. Given the opportunity, elephants will select certain habitats over others and usually choose those habitats that provide best for their needs. They will then move away from deteriorating habitats or habitats that do not meet their needs. Such movement can reduce numbers locally.
Elephants are catholic in their environmental needs
This is true for elephants living successfully in landscapes that range from deserts to forests, as long as water and food are available. Elephants feed on a variety of plants and plant material, usually preferring grasses and sedges during rainy seasons and the leaves, twigs and bark of trees during dry seasons. Food preferences vary between landscapes and tend to be dictated by availability more than selection. Elephants are relatively inefficient feeders for they digest only 60 to 70 per cent of the material they consume.
Elephants are adapted to cope with environmental change and variability
This is indeed the case for elephants who can cope with extreme conditions that are typical of the landscapes they inhabit. Behavioural and physiological adaptations keep their body temperatures relatively constant and for this they are dependent on water to drink and cool off during the heat of the day. They also make use of shade under trees and during the cold of the night also actively seek relatively warm areas under tree cover. During hot summers they apparently offload heat by seeking relatively cool areas during the night. Elephants also feed on a great variety of trees and grasses and often their preferences are dictated by availability rather than choice. In a few cases their preferences for certain rare plants may induce local extinctions. In many cases such extinctions do not occur because some plants may escape from depredation by growing on places inaccessible to elephants, such as on steep mountain slopes or in hilly terrains.
Water is a key resource for elephants
Elephants depend on drinking water. Water is a key resource and the distribution thereof dictates their use of habitat. Consequently, elephants usually roam within relatively close proximity to water and breeding herds seldom wander more than 10 kilometres from water.
Elephants are long-lived and typically experience several droughts during a life time, especially when living in savannahs where three to four-year long droughts may recur at about 12 year intervals. They typically do lose body condition during the dry season but gain body reserves during the wet season when the nutritional value of their preferred food is relatively high. Their large bodies enable them to withstand nutritional stresses, as do the wide choice of plants they feed on. Death due to starvation seldom occurs for elephants are capable of moving relatively long distances in search of food, albeit that roaming distances of breeding herds are limited by the abilities of calves.
Elephants have a negative effect on other species
Elephants do destroy individual plants by uprooting trees and/or breaking their branches. They also remove bark from trees, thereby making trees susceptible to insect infestation and fire. More trees close to water are destroyed by elephants than trees further away from water. Broken tree stems are known to enhance the presence of certain lizard species while the lack of trees is speculated to influence hole nesting birds. We have no proof for tree damage affecting bird distribution as has been speculated by some proponents of culling. A study in Zimbabwe showed that reduced tree cover induced a reduction in the number of ant and bird species in miombo woodlands. The destruction of trees also goes hand in hand with elephants being confined for extended periods to relatively small areas.
The African savanna needs elephants
Yes, for we know that the structure of African savannahs is driven by the interactions between climatic and biological interactions. The so-called arid and transitional savannas that receive less than 600mm of rainfall per year persist as a mixture of grasses and trees through the disturbance imposed by rainfall, while the wet savannahs, which on average receive more than 800mm of rainfall per year, persist in response to the disturbance invoked by herbivores feeding on grasses and trees. In terms of biomass, elephants often dominate the herbivore guild in savannahs and consequently play a major role in modifying the structure and function of these landscapes. Their feeding on trees modifies canopy shape as well as the survival of saplings and adult trees, thereby ensuring micro-environmental conditions that allows for the co-existence of grasses and woody species that dictates the savannah composition. Without elephants this composition may be altered and favour bush encroachment. This may disfavour other herbivores that graze on sun-loving grasses and sedges. Elephants thus may be considered important components of the savannah ecosystem, their presence favouring a diverse cohort of herbivores.
Elephants are ecological engineers that maintain savannas
Elephants do indeed have the ability to shape the structure and function of the systems of which they are part of. They have been allocated ecological statuses such as 'keystone species', 'umbrella species' and 'ecological engineers', some of these as a matter of speech rather than as motivated by the roles they fulfil in their natural environs. They have also had the misfortune of being referred to by some as a 'problem species' and even a 'pest' while others have coined phrases such as 'flagship species' and 'conservation ambassadors'. None of these have a scientific or ecological foundation.
Ivory poaching continues to threaten the survival of elephants in some southern African populations
This is true. Zambian populations have few large and old elephants, herds are small and individuals are often tuskless. The imbalances in age structures suggests that most populations in Zambia have not recovered from past intense poaching and may still be exposed to some poaching, while censuses imply a continuing, albeit slow, decline in some populations. This supports independent genetic assessment that suggests Zambia as a focal point in southern Africa of supplying illegal ivory to international markets.
The elephant problem is one of limited space, not numbers
Several Peer-reviewed research papers in leading international journals support the notion. Re-current local colonisation and extinction events drive the dynamics of many animal species that exist as so-called 'meta-populations'. The variability across space induces variability in population responses and this contributes significantly to regional stability in numbers and provides opportunity for vegetation and elephants to co-exist. Both short- and long-term dispersal movements benefit such co-existence.
Elephant population's exist in a metapopulation structure
Metapopulations are collections of populations that exchange individuals and that repeatedly colonise land that has become vacant through local extinction. To exist as a metapopulation the demography of its populations must differ, be in asynchrony, individuals must disperse and at a given point in time, the continuity in landscapes must provide for both occupied and vacant habitats. Given these requirements elephants certainly may be defined as having a metapopulation structure when living without the confines imposed by artificial borders and boundaries.
Recent assessments show that the demography of elephants does change across landscapes and that longer-term changes in numbers of different populations do differ, with some populations being stable, some decreasing, and some increasing. Elephants are also known to colonise vacant land and the varying mosaic of landscapes across their southern African distributional range allows for repeated colonisation. The likelihood of elephant populations functioning as a metapopulation is thus high and research on the topic continues.
Metapopulation management can address the 'elephant problem'
Metapopulation management is a component of spatial and habitat management. It focuses management on landscapes rather than populations and on impact rather than numbers. It provides for the manipulation, albeit natural or artificial, of short- and long-term spatial occupation and hence for varying impact on vegetation quantity and quality. Because the 'elephant problem' is that of impact on other species, such management addresses the problem directly, rather than indirectly.