The theory of regulation in animal populations is fundamental to understanding the dynamics of populations, the causes of mortality, and ecosystem processes. I have tested this theory using long-term records (40 years of data) of two large mammal species in the Serengeti, the African buffalo and the wildebeest (Mduma, Sinclair et al., J. Anim. Ecol., 1999). These studies showed that density dependence was sufficient to regulate the population, and that the cause was ultimately through lack of food acting synergistically with disease. This was the first field study of a mammal population under natural conditions to demonstrate regulation. Other studies on ungulates have subsequently supported these findings (e.g., Clutton-Brock et al., J. Anim. Ecol., 1991; Coulson et al., Science, 2001). My studies are synthesized in three books so far, and we have Serengeti III in process.

Studies of the migrations of large mammals in Serengeti were the first to demonstrate both the cause and function of migration. I showed that migration is an adaptation to track seasonally changing food resources. The significance of migration for understanding ecosystem dynamics is that it results in much larger numbers than would occur with sedentary populations. The consequences of migration are that the population escapes from predator regulation. These results are consistent with subsequent studies on most other migration systems in the world.

These studies also provided the basic understanding for the repeated human famines in northern Africa, migration systems that had become sedentary and unstable (Sinclair & Fryxell 1985). These human migration studies also provided the understanding on the link between natural systems, food supply, and poverty in Africa (Sinclair & Wells 1989).

Animal populations can be regulated by either bottom-up processes such as a shortage of food (or some other resource) or top-down processes through predation. Just when each process occurs has been the subject of some controversy because there is an extensive literature supporting each but no clear rules for predicting them.

In mammals some of these rules are becoming more evident (Sinclair & Krebs 2002, Sinclair 2003). Serengeti studies have shown that limitation through predators is determined by four factors: (i) body size, (ii) the diversity of predators and prey in the system, (iii) whether prey are resident or migratory, and (iv) the presence of alternate prey for predators (Sinclair 1985, 1995)

Analysis of population dynamics in predator-prey systems has confirmed predictions based on the behaviour of predators (e.g., Holling 1959, 1965). First, prey can break away from predator control and become pests, (Sinclair et al. 1990, Pech et al. 1992). Secondly, the characteristics of predator behaviour result in the extinction of populations. This work has provided the basis for a new approach to conservation of endangered mammals and birds in Australia and New Zealand (Sinclair et al. 1998).

In Canada, a large-scale experiment in ecosystem dynamics (with Charles Krebs and others) demonstrated a synergistic interaction of food and predators in determining the abundance of the dominant species, snowshoe hare. As with the Serengeti ecosystem indirect interactions determine trophic dynamics (Krebs et al., Science 1995; Sinclair et al., Oikos, 2000).

Predation is one process that can under special circumstances produce two states in the system. We have tested predictions from the theory of multiple states on several systems around the world, including raptors on mice, and foxes on rabbits in Australia (Sinclair et al. 1990, Pech et al. 1992), and geese grazing on tundra plants (Hik et al. 1992). In Africa, we have demonstrated the existence of two ecosystem states in Serengeti, a woodland state with elephants, and a grassland state also with elephants (Sinclair 1995, Sinclair & Krebs 2002).

I have focused the 10-year biodiversity programme on some specific aspects that are more directly applicable to conservation concerns. With a team lead by my close associate Dr Simon Mduma from Tanzania, we are monitoring changes in vegetation, insects, reptiles, amphibians, birds, and mammals in the context of a series of disturbances to the system that we use as natural experiments.

i) The role of protected areas as baselines for biodiversity.

This aspect addresses the impacts of human exploitation on ecosystems (Arcese & Sinclair 1995, Sinclair 1998). Part of the current Serengeti research uses this approach and I am involved with establishing this in other areas of the world because of the extensive human impact.

If we are to use protected areas as ecological baselines, so as to measure human impacts on ecosystems outside, then we must be able to distinguish between natural succession within the baseline from human impacts outside. In Serengeti a number of major natural changes have taken place, including the savannah tree community, the ungulates, the predators, and the birds, to name some. The first results of this work have now been published (Sinclair et al. 2002) on the birds and insects. They show an extraordinary loss of some 50% of bird species outside of Serengeti that is explained by habitat loss as well as a parallel loss in insect diversity due to human intervention in their systems.

ii) Multiple States and Biodiversity Loss.

Can the system regenerate once a disturbance is removed, or is it locked into a new state, requiring another type of perturbation? To address this question we are using an opportunity presented from changes in a vegetation component of the Serengeti ecosystem. This is the decline that we have observed over 30 years in the Riverine forests. Preliminary results indicate that mammal browsing is the most important factor preventing regeneration, effectively removing all seedlings. This means that forest regeneration in the past was related to the 1890 Great Rinderpest pandemic in Africa. We are now comparing our results with other areas such as Chobe National Park, Botswana. It seems these forests can only establish under certain conditions of low browsing, and that there are possibly two states.

The profound changes I have observed over several decades in one ecosystem, the Serengeti, concerning how systems work, the role of biodiversity in human-dominated systems, and the influences of humans on disturbing ecosystems provide an important example and lesson for the conservation of world ecosystems. The work is relevant to the famines and poverty in Africa, the endangered species of Australia, and the management of pest outbreaks.

1. I have conducted research in Serengeti, Tanzania, since 1965, mainly on the problem of what determines the size of animal populations, particularly vertebrates, and the mechanisms of regulation. This work has expanded to look at the whole ecosystem, documenting how the different components of soils, plants, herbivores, and predators interact. In particular, I am interested in how to provide scientific advice for conservation, and in training for long-term monitoring of biodiversity.
2. Present research is focused on recording the diversity of life in this African savannah system, and comparing human-induced changes with natural change. Research is documenting plant, invertebrate, bird and mammal diversity.
3. Research in Canada has been conducted with a team of others in Kluane National Park and surrounding areas in the Yukon since 1976. This has examined the workings of the northern spruce forests, in particular what has caused the 10-year cycle of snowshoe hare numbers. In addition, the large-scale experiment in Yukon was designed to understand how ecosystems are structured from resource supply or by predation.
4. Experiments in the conservation of endangered marsupial prey and their exotic predators, the red fox and feral cats, have been conducted in Australia since 1992. This was carried out in conjunction with CSIRO, Sustainable Ecosystems division, Commonwealth Government of Australia.
5. Similar experiments are being conducted in New Zealand by their Crown Research Institute, Landcare Research, Inc. I am advising them on their research in the conservation of endangered bird species threatened by exotic weasels, ferrets, and cats; and on ecosystem studies and biodiversity loss.
6. My group is also involved with documenting hotspots of biodiversity in western Canada. Our work in British Columbia includes ways to prioritize conservation sites, and studies of the dynamics of declining populations of endangered species such as the endangered Vancouver Island marmot.

Anthony R.E. SINCLAIR, FRS, FRSC
University of British Columbia
Vancouver, Canada