Population Viability Analysis

Population viability analysis (PVA) began life as an attempt to answer the question, ``How large must a population be for it to have a reasonable chance of survival for a reasonably long period of time?'' A reasonable chance of survival was often taken as 95%.³ A reasonably long period was often taken to be 100 years.⁴ In its early life population viability analyses were used in an attempt to identify the smallest population that would have a reasonable chance of survival for a reasonable period of time--the minimum viable population size (MVP). Although, as we'll see, assessing the viability of any population is far from easy, it's a useful idea because it crystallizes several other ideas:

• It identifies the population as the critical unit for conservation purposes.⁵ Until about 25 years ago conservationists tended to focus only on protecting land, not on managing the populations of plants and animals that occur there. Moreover, even with the recent shift in emphasis to ``ecosystem-level'' conservation, we mustn't forget that the species that make up an ecosystem are found in populations. So management of some populations may still be necessary.

• The term ``viability'' stresses that we're concerned with persistence of the population over some reasonably long period of time. Furthermore, it emphasizes that we're interested in the prospects that the population can be self-sustaining.

• The idea of ``minimum'' emphasizes the idea that there are certain critical aspects of the interaction of a species with its environment. If the population gets too small, it may no longer be able to cope. There may be a threshold below which a population is doomed to extinction. That it has survived its current decline is not evidence that it can suffer a further decline.

There are, however, several problems associated with the idea of minimum viable populations.

1. The first is the practical problem that if we identify the minimum size of a population that is sufficient for conservation purposes, that's all were likely to get.

2. The second, and more fundamental, is that because of the enormous uncertainties associated with forecasting the fate of populations,⁶ we're unlikely to be able to provide a good estimate of an MVP. Unless MVPs are determined with a large safety factor, specifying an MVP might actually promote extinction, rather than retard it.

3. Morris and Doak [5, p. 43] go so far as to argue that ``no good PVA should attempt to evaluate the risk of utter extinction.'' There are so many things that can go wrong and so many uncertainties that we're better off focusing on quasi-extinction, the number of individuals below which the population is likely to be immediately and critically imperiled.

4. Perhaps most fundamentally of all, populations of a species may cease performing the ecological functions they provide well before they become extinct. Kent Redford [6] points out, for example, that a botanically intact forest will cease to exist if the animal communities responsible for pollination and seed dispersal go extinct or become so rare that pollination and seed dispersal success declines to a point where it is insufficient to allow their persistence.⁷

Today we'll be talking about the principles of population viability analysis. Over the following two lectures we'll examine two applications of the principles: the northern spotted owl and the bay checkerspot butterfly. What we've seen in the last several lectures reminds us that we'll need to take account of stochastic processes in these populations. Before we can start talking about including that stochasticity, however, we have to take a detailed look at the deterministic models to which we will add probabilistic components. Today I'll give an overview of age- and stage-structured demographic models, because they are the most widely used and most widely applicable.⁸ There are several reasons for doing this:

1. You have to have a good understanding of the basic model before you can understand what it's like when randomness is added to it.

2. Recall that many threats to population persistence are systematic. Analysis of deterministic a model helps to identify those systematic threats, especially if they're not immediately obvious.

3. Analysis of a deterministic model also allow us to identify the life-history stages that are most critical in determining individual abundance so that we can focus management efforts where they are likely to be most successful.

4. Even if a complete PVA is not necessary or possible, knowing what one would include helps to structure thinking about management options and leads to the design of management programs that can be expanded into a full PVA if needed.