Extinction vortices, population viability analyses, IUCN risk categories

In reality the species that inhabit the earth are simultaneously subjected to the whole multiplicity of the extinction causes considered thus far: habitat destruction and fragmentation, overexploitation, genetic deterioration, global climate change, demographic and environmental stochasticity. Very often an initial reduction of a population size due for example to a catastrophic event (fire, spillage of pollutants, etc.), increases the risk of extinction due to another cause (e.g. competition with alien species) leading to a further reduction of the population abundance and thus making it more prone to being impacted by yet another cause of extinction (e.g. demographic stochasticity). This chain process has been called extinction vortex (Gilpin and Soulé, 1986) in a somewhat fanciful but effective way and is graphically represented in Fig. 17.

Figure 17: Graphical representation of the so-called vortex of extinction. When a species enters the vortex, a sequence of phenomena gradually reduces a population size down to possible local extinction.

To conduct a sufficiently realistic risk analysis for a population it is necessary to include all the factors that, to our knowledge, may affect its future dynamics. The techniques used for this purpose go generally under the name of Population Viability Analysis (PVA, Boyce (1992)) and are often coded into specially developed computer programs, such as VORTEX , RAMAS, ALEX, NEMESIS. Currently the term PVA is used to describe the set of statistical tools and simulation techniques that allow the estimation of the probability that a population or group of populations may persist for a certain time in a given territory. The example in Fig. 18 reports the results of a PVA conducted for a critically endangered species: the African elephant. The goal of a PVA is often linked to management; in this case the authors of the analysis have tried to provide an idea about ​​the influence of the size of possible reserves so that they can better protect the elephant populations.

Figure: Cumulated probability of extinction as a function of time for populations of African elephant (Loxodonta africana) in reserves of various sizes. The scale is semi-logarithmic. The initial elephant density is supposed to be anyway equal to 12 elephants per 10 km$^{2}$. For instance, an elephant population in a ​​125 km$^{2}$ protected area has a 40% probability of becoming extinct in 700 years. Modified after Armbruster and Lande (1993).

The accuracy and reliability of a PVA critically depends on the available data regarding demographic and genetic parameters and the spatial distribution of individuals, the availability of present and future habitats, the influence of climatic factors, and so on. In most cases, the available data are extremely scarce, also because the species at greatest risk of extinction are often rare and poorly studied. For this reason it has been necessary to propose, in addition to formal analyses such as PVA, more empirical and speditive methods which allow the classification of the risk status of a species. Therefore, Mace and Lande (1991) proposed classification methods that are based on the type of information that is realistically available. The criteria make use of different indices that can be measured or quantified in just a few years. Table 6 shows the proposal by Mace and Lande (1991) which essentially is based on indices of observed or predicted decline, on the observed or predicted distribution range, on the size of single or fragmented population, on the extinction probability within a given time interval as obtained from a formal PVA. These criteria form the basis for the classification of IUCN (International Union for Conservation of Nature, http://www.iucn.org/) which is published under the name of Red List. After extensive discussion the classification methods of Mace and Lande were slightly changed, making them substantially more flexible. The current classification scheme (see Fig. ) and the pertaining methods can be found in the latest edition of the IUCN manual (IUCN, 2012).

Table 4: Mace and Lande's criteria for the classification of species endangered by extinction. A species is placed into one of three categories (Critically Endangered, Endangered, Vulnerable) if it meets any one of the five criteria. $N$ indicates the total population size, while $N_s$ specifies the maximum size of subpopulations that make up the population, if it is fragmented.

Demographic characteristic	Critical	Endangered	Vulnerable
Observed decline	80% in 10 years or 3 generations	50% in 10 years or 3 generations	20% in 10 years or 3 generations
Geographical range	$<$ 100 km$^2$, single location	$<$ 5000 km$^2$, $<$ 5 locations	$<$ 20000 km$^2$, $<$ 10 locations
Total population	$N < 250$ $N_s < 50$	$N < 2,500$ $N_s < 250$	$N < 10,000$ $N_s < 1000$
Projected decline	$>$ 25% in 3 years or 1 generation	$>$ 20% in 5 years or 2 generations	$>$ 20% in 10 years or 3 generations
Extinction probability	$>$ 50% in 10 years or 3 generations	$>$ 20% in 20 years or 5 generations	$>$ 10% in 100 years

Figure: The different categories of the current species classification according to the IUCN criteria that are the basis of the so-called Red List.