A coupled climate-vegetation-mammal-human model for simulating Late Quaternary megafaunal extinctions

The period 60,000-5,000 years ago saw the extinction of up to a thousand species of large vertebrates (‘megafauna’) across six continents. Understanding the cause of these extinctions is important for several reasons. It is the most recent substantial extinction event in the geological record; there is a background of detailed knowledge about environmental change against which to view the responses of the mammals; and humans are strongly implicated by many researchers as partial or exclusive causal agents.

For all of these reasons, understanding the cause of the extinctions, and the reasons why some species survived while others did not, can provide a unique historical analogue for addressing the current biodiversity crisis. The two main contenders for the megafaunal extinction are vegetation change driven by climate, and hunting by humans, either separately or in combination. Although the extinction was worldwide, we will focus on Europe, northern Asia and North America as these areas have the best data on the distributions and extinction of the mammals. We will first develop computer-based simulations of local climatic conditions across the study area; for the first time climate changes will be modelled on a year-by-year basis over the past 40,000 years. Using this information we will model vegetation types across the entire area. When climate changed, vegetation changed, but our model will be crucially more realistic than previous ones in that we will allow for the lags in vegetational response as plant species expand slowly across large areas (e.g. trees may have taken 1500 years to arrive in northern Europe when climate warmed after a long cold spell). In addition, the model estimates not only the type of vegetation but its productivity, i.e. amount of growth each year, of crucial importance to herbivorous mammals. Many of the mammals that went extinct (such as the woolly mammoth and wooly rhinoceros) were grazing species adapted to the productive grasslands of the last glaciation, and the predators that depended on them. Many of those that survived were browsing (woodland) mammals or those of mixed habitats. We will develop, for both victims and survivors, a biological profile for each species including their body weight, reproductive rate, and preferred foods. These will be determined from living relatives and from direct evidence such as wear on fossil teeth that indicates diet. We will also establish their climatic tolerance from the range of climates they occupied in the past. Adding the mammal fauna to the modelled climate and vegetation, and running the computer model from 40,000 years ago up to the present, the effect of climate changes on the vegetation, and the effect of both on each mammal species, will be evident. Moreover, the model will include feedback from the feeding activities of the mammals to the structure of the vegetation itself. A final element in the model is the addition of variable levels of human hunting, the distribution of people being determined from known archaeological sites. Analysis of all the data will determine if climatic and vegetational change, with or without the addition of hunting, are sufficient to account for the extinction of some megafauna and survival of others. This will be determined by comparing model results with the known pattern of range changes and extinction based on the fossil record. The vegetation model that we develop would also allow prediction of likely responses to future climatic changes. Similarly, the climate simulations will be applicable to other processes (e.g. the changing extent of arctic permafrost). Our results will be directly relevant to various stakeholders, informing landscape management and biodiversity conservation strategies. We will ensure that they are communicated to such stakeholders, as well as to the scientific community and wider public.

Grant reference
Natural Environment Research Council
Total awarded
£602,680 GBP
Start date
30 Sep 2016
3 years 3 months 1 day
End date
31 Dec 2019