3.1. B IODIVERSITY AND UNCERTAINTY
In previous chapters we have encountered many situations where our attempts to assess the value of other species have been hampered by uncertainty. The uncertainties are found on many different levels. Our knowledge and understanding of the species as such, interactions between species, ecosystems functioning and what roles different species play, is still highly incomplete and full of uncertainties.223 The things we do know cannot always be generalised from one ecosystem to another with sustained precision.224 There is also much uncertainty about the long-term effects and side effects of what we do to nature, not least about the threats to the species, and the threats that the loss of species implies for other species and for the ecosystems.225 As we have seen, the uncertainties are not in any way smaller regarding the estimates of the instrumental value that different products and services from nature has for human beings.226 What complicates it even further is that many species have not even been discovered yet. How can we value the services or goods they may supply?227
We will hopefully be able remove some of the uncertainty by more thorough investigations, but to remove all the uncertainties would clearly be very expensive.228 In fact, to get a perfect understanding of what is going on in the biological world, as well as of what will happen in the long term when we make as fundamental changes as removing species from the system, might not even be possible.229 The old view of nature as a machine – a clockwork with mechanic precision where a particular intervention necessarily leads to a particular, foreseeable effect – is replaced by a more modern conception of nature as something dynamic and complex, uncertain and chaotic.230 We have in other words started to realise that the way nature reacts to our treatment is not completely predictable.
Donna Maher talks about a change of science from a situation where
223 Angermeier 2000 p.379, Aniansson 1990 pp.38, 42, Buege 1997 p.7, Cooney 2005 pp.3f, Farber 2000 p.s492, Gamborg & Sandøe 1995 p.16, Ihse 2005 p.71, Norton 1986:2 pp.223, 271, 274, 277, Söderqvist 2005 p.78, Whiteside 2006 pp.xff. The uncertainties are particularly great regarding non-vertebrate and non-terrestrial systems and species. (Cooney 2005 p.3)
224 O’Riordan & Jordan 1995 p.199
225 Cooney 2005 p.3, Gamborg & Sandøe 1995 p.16, O’Riordan & Jordan 1995 p.199, Whiteside 2006 p.33
226 Farber 2000 p.s495, Martinez-Alier 1994 p.xiii, Norton 1986:2 pp.223, 274, Paoletti & Hassall 1999 p.161
227 Randall 1986 p.85
228 Farber 2000 p.s496, McGarvin 2001 p.25
229 Dupré 1993 p.3, McGarvin 2001 p.25, Whiteside 2006 p.34, Östberg 1993 p.232
230 Beltrán 2001 p.4, Herremoës et al 2001 p.193, O’Riordan & Jordan 1995 p.200, Sörlin 1991 p.18, Whiteside 2006 pp.xff
… prediction of system behaviour was a matter of having enough data, to a 'science of surprise', where chaos and unpredictability are endemic, with stability and predictability the exception.231
Sverker Sörlin makes a similar point by referring to chaos theory and catastrophe theory when he tells us that the old fashioned linear models will not help us find out at which point the decreasing ozone layer, or the greenhouse effect, etc. will take an uncontrollable catastrophic turn.232 He does not mention loss of biodiversity, but the same reasoning can probably be used here.
One of these nonlinear phenomena that we have to consider when we are dealing with complex things like living beings or ecosystems is (as we have noted earlier), the existence of threshold values.233 Normally we assume that cause and effect are proportional, and can be described by some linear equation, i.e. a certain change in the cause leads to a corresponding proportional change in the effect. However, in some situations all or most of the effect takes place when the causation power has reached a certain value – the threshold value. In these cases, most changes in the causation power do not have any visible effect at all, but still have the important indirect effect of taking us closer to the threshold value. This climbing closer to the threshold value is in many cases something that takes place invisibly.234 For a long time we will not see any changes at all (or only very minute and seemingly insignificant changes) either of the object we are interested in, or in its surroundings. We will thus not even have any indication of what will happen or when. When the threshold is reached, the next change in the causing power, however small, will mean all the difference in the world. Then the up to now only latent effect will suddenly occur all at once.
In our case, it would mean that the disappearance of a single species, or two or three, from an ecosystem might not result in any discernible effects on, for example, the ecosystem services. This may go on for a while but when a threshold is reached, the results could be dramatic.
Anne and Paul Ehrlich use an analogy about a person who pops rivets from the wings of airplanes. He sells the rivets for 50 cents each and he defends himself by pointing out that:
I’ve already taken 200 rivets out of this wing, and nothing has happened yet. Lots of planes fly with missing rivets. They build a lot of redundancy into jet aircraft, partly because they don’t completely understand the materials and stresses involved, so nobody can prove that taking another rivet out will weaken the wing too much.235
231 Maher 1999-2000
232 Sörlin 1991 p.255
233 Daily 2000 p.335. Clarke 1995 p.41, Herremoës et al 2001 p.193, McGarvin 2001 p.25, Norton 1986:1 p.123, Whiteside 2006 p.33 (Clarke talks about them as “jump effects”.)
234 This is not always the case though. Sometimes it is indicated by something else than the effect we are worried about (and do not see any trace of yet).
235 Ehrlich & Ehrlich 1990 p.95
As we saw in the previous chapter, decreasing redundancy might have unwanted consequences. One consequence that we touched upon was that we might be approaching a threshold. The story about the rivet popper illustrates this problem.
As Bryan Norton points out, the assumption of the ‘rivet popper’ that the absence of any accident so far is an indicator that the risk of an accident in the future is very low, would be true if we were talking about a series of independent events. The problem is that we are not. For every rivet he pops, there are fewer rivets left, which means that the constitution of the plane is constantly getting weaker. The same goes for species: For every species that goes extinct above the speciation rate, there are fewer species left, and the ecosystem – even the global system – is weakened.236
This is typical for threshold effects. Every change in the input takes the system closer to the threshold even though the effect is not noticeable until we reach the threshold.
Margareta Ihse extends the collection of “threshold-analogies” with an analogy about a hammock where the species are the threads of the fabric that will hold us up for a while, but bursts when the fabric gets too thin.237 This is a very good analogy of ecosystems as well as of the circulation of nutrients, etc. in nature. They can be described as a web with many intertwined threads. This gives the system a certain amount of stability and resilience but we do not know when the web gets too thin to uphold its functions. It also illustrates that the resisting power of nature that is due to the redundancy in the systems is never a guarantee against severe changes. It holds back – and hides – the changes for a while and lulls us into a false sense of security. It does not stop the change forever though. When the threshold is reached all the accumulated change occurs at once.
The Ehrlich analogy points at an important difference between the natural disappearance of species and the high extinction rate we are facing at the moment due to anthropogenic interventions: Normally the species that go extinct are replaced by other species just like lost rivets in an airplane are replaced by new rivets.238 At the pace by which species disappear today, the species cannot be replaced fast enough, however, and we face a net loss.
There is one important difference between the analogy with the rivets and the loss of species, however. Unfortunately, this difference also makes the loss of species much more problematic than the loss of rivets. New rivets can be taken from the storeroom and the lost ones can be replaced. Species on the other hand are replaced by evolution. Instead of being taken from a storeroom, they evolve from the genetic basis that already exists in the existing species. This tells us that in order to replace lost species with new species that have a better chance of survival, we need above all a large selection of genes. I.e., we need a large biodiversity, and that is precisely what we are losing.
236 Norton 1986:1 p.122, Norton 1987 p.68
237 Ihse 2005 pp.70f
238 Ehrlich & Ehrlich 1990 p.96
The non-linear aspect can be brought one step further and form another argument to consider: Sometimes a very small change in the input can have a very large effect on the output. If there are effects like this in ecosystems, it must be a very strong argument indeed for extra caution about all interventions in the ecological systems – including interventions that contribute to the extinction of species.
Furthermore, if we take a closer look at the evolutionary process, we will find that one of its inherent features is that it has no predetermined direction. It is not the case that the individuals of a species always get bigger or faster or more intelligent. The direction in which the evolution takes a certain species depends on its environment and on chance. The environment changes all the time, and what “remedy” that evolves in a certain species as an “answer” to a particular change in the environment depends on what its gene pool happens to have in store, and on which re-combinations and mutations that happen to take place.
Which of these “remedies” in turn that eventually are favoured by natural selection, depends not just on one single aspect of the environment in which the species live, but on the total selective pressure that the environment puts on the species. If rabbits (Oryctolagus cuniculus) become faster, foxes (Vulpes vulpes) have to evolve too, but not just in relation to the rabbits. If they evolve a quality that makes them better rabbit hunters but also makes them less resistant to cold or easier prey for the lynx (Felis lynx), they will lose out in the evolutionary game anyway. All species are in fact at any given moment subjected to pressure of many different types from many different directions, and the sources of the pressure are also in their turn subjected to pressure of many different types and from a large number of different directions – including from the species they are exerting pressure on. If we were going to calculate the direction of evolution for the fox, we would have to consider the selection pressure that is placed on the fox by both the lynx and the rabbit, as well as all other species that interact with the fox directly or indirectly as well as all the non-living forces of nature. The rabbit and the lynx and the other species evolve too, however, and that has to be taken into account. The fox is putting both the rabbit and the lynx under selective pressure just as they do with the fox, but that is not all. The Lynx not only eat foxes but also rabbits so we have to look at the pressure they exert on each other.
The Lynx eat other prey too and the rabbit is not just hunted by the fox and the lynx. It therefore does not just evolve in a way that helps the rabbits cope with the threat from these predators, but also as a result of how the golden eagle (Aquila chrysaetos) evolves since they eat rabbits too, etc. The pressure from the lynx and the golden eagle will inevitably also affect what options the genes of the rabbit have when it comes to “dealing” with the threat from the fox and so on.
Then we have to put the result we get for the rabbit back into our equation for the fox together with the results from other prey species for the fox, and so on and so forth – only to find out that while we have been busy doing these calculations the whole scenario has already changed.
In short, we have a problem that is infinitely more difficult to solve than the
“three body-problem” in physics.
What this tells us is that we simply cannot know for sure what will happen in an ecosystem in the long run when we make such a radical alteration as changing the species composition.
Changing the species composition can be done in different ways. It can be done, for example by causing a species to go extinct as we are discussing here, or by putting in a new species that was not there before (but that may well result in other species disappearing).239
The best literary description of the latter is probably Michael Crichton’s book “Jurassic Park”.240 In this book, species of animals and plants that lived more than 65 million years ago are resurrected and introduced in a present day environment. As we know, it did not work out very well in spite of the guarantees from John Hammond and his bio-engineers. This was of course just fiction. We do not know what would really happen in a situation like this, but the point of the story was just that: We do not know, because we cannot know. It is impossible to predict the results from such a project, and therefore we should be more cautious. To recreate pre-historic organisms is quite extreme, but many of the interventions we make are almost as extreme, and as we saw above, our possibilities of foreseeing the results are limited. The best and most frightening illustration in the book is probably the absolute confidence by which Mr Hammond and his staff guarantee the safety of the arrangement. (What is particularly frightening is how easy it is to recognise this unshakable confidence in many people in the real world.)
Michael Crichton’s description is very illustrative and very thought provoking. However, there are also many real-life examples of how we have intervened in nature and ended up very surprised over the results.241 The rabbit population explosion in Australia and the drought catastrophe in Sahel in Africa are both described as examples of catastrophic situations caused by our ignorance about ecology.242 A well-known example of how human beings have deliberately tried to engineer nature to suit our purpose by taking away a species from the system, is the wolves that were hunted virtually to oblivion in North America in order to protect both farm animals and game animals (or to be more precise, to protect human farmers and hunters from the competition). This resulted in an explosive increase in the number of deer, which in turn caused a lot of damage to the ecosystems (and to the deer population). It also had a negative economic effect on the human population since it destroyed the grazing for domesticated animals such as sheep.243
All the examples above confirm the problem of predicting what will happen in an ecosystem as a result of human encroachment. The lesson that seems to emerge from this section, is that we will probably never reach a situation where
239 For a discussion of the handling of uncertainties in relation with the latter, see Cooney & Dickson 2005 p.9.
240 Crichton 1991 passim
241 See e.g. Whiteside 2006 p.11
242 Palmer 1995 p.26f
243 Ricklefs 1997 p. 598
we have enough information to make a fully informed decision as to which course of action is the most rational from an anthropocentric instrumental point of view. The uncertainties are sometimes used as an argument against conservation, and sometimes as an argument in favour of conservation.244 What we need is a strategy that can tell us what the most rational behaviour is from an anthropocentric instrumental viewpoint, given these uncertainties. In the coming sections we will try to find such a strategy, and we will in particular take a closer look at the so called precautionary principle that has been much discussed recently.