THE STATISTICAL FLUMMERY ABOUT SPECIES LOSS1 Julian L. Simon Species extinction came to prominence in 1979 with Norman Myers's book The Sinking Ark and the 1980 Global 2000 Report to the President. These still are the canonical texts. Global 2000 forecast extraordinary losses of species between 1980 and 2000. "Extinctions of plant and animal species will increase dramatically. Hundreds of thousands of species -- perhaps as many as 20 percent of all species on earth -- will be irretrievably lost as their habitats vanish, especially in tropical forests." The data on the observed rates of species extinction are wildly at variance with these statements, and do not provide support for the various policies suggested to deal with the purported dangers. Furthermore, recent scientific and technical advances -- especially seed banks and genetic engineering, and perhaps electronic mass-testing of new drugs -- have reduced the importance of maintaining a particular species of plant life in its natural habitat. Society properly is concerned about possible dangers to species. Individual species, and perhaps all species taken together, constitute a valuable endowment, and should be guarded just as we guard other physical and esthetic assets. But we should strive for as clear and unbiased an understanding as possible in order to make the best possible judgments about how much time and money to spend in guarding them, in a world in which this valuable activity must compete with guarding and supporting other valuable aspects of civilization. SPECIES LOSS ESTIMATES The original and still-standard forecast for loss of species comes from Lovejoy: What then is a reasonable estimate of global extinctions by 2000? In the low deforestation case, approximately l5 percent of the planet's species can be expected to be lost. In the high deforestation case, perhaps as much as 20 percent will be lost. This means that of the 3-l0 million species now present on the earth, at least 500,000-600,000 will be extinguished during the next two decades. (U.S., l980, II, p. 33l). The basis of any projection should be a body of data collected under a range of conditions that encompass the expected conditions, or that can reasonably be extrapolated to the expected conditions. But none of Lovejoy's references contain any scientifically-impressive body of experience. The only published source given for his key table is Myers's 1979 book. This is Myers's summary of the relevant data: Between the years l600 and l900, man eliminated around seventy-five known species, almost all of them mammals and birds--virtually nothing has been established about how many reptiles, amphibians, fishes, invertebrates and plants disappeared. Since l900 man has eliminated around another seventy-five known species--again, almost all of them mammals and birds, with hardly anything known about how many other creatures have faded from the scene. The rate from the year l600 to l900, roughly one species every 4 years, and the rate during most of the present century, about one species per year, are to be compared with a rate of possibly one per l000 years during the "great dying" of the dinosaurs. From those data, showing a maximum of one known species per year, Myers departed into pure speculation: Since l960, however, when growth in human numbers and human aspirations began to exert greater impact on natural environments, vast territories in several major regions of the world have become so modified as to be cleared of much of their main wildlife. The result is that the extinction rate has certainly soared, though the details mostly remain undocumented. In l974 a gathering of scientists concerned with the problem hazarded a guess that the overall extinction rate among all species, whether known to science or not, could now have reach l00 species per year... Let us suppose that, as a consequence of this man- handling of natural environments, the final one-quarter of this century witnesses the elimination of l million species--a far from unlikely prospect. This would work out, during the course of 25 years, at an average extinction rate of 40,000 species per year, or rather over l00 species per day. The greatest exploitation pressures will not be directed at tropical forests and other species-rich biomes until towards the end of the period. That is to say, the l990s could see many more species accounted for than the previous several dec- ades. But already the disruptive processes are well underway, and it is not unrealistic to suppose that, right now, at least one species is disappearing each day. By the late l980s we could be facing a situation where one species becomes extinct each hour (l979, pp. 4-5). We find these key points in the Myers quotations: (l) The estimated extinction rate of known species is about one every four years between the years from l600 to l900. (2) The estimated rate is about one a year from l900 to the present. (3) Some scientists (in Myers's words) have "hazarded a guess" that the extinction rate "could now have reached" l00 species per year. That is, this number is simply conjecture; it is not even a counter-point estimate but rather an upper bound. The source given for the "some scientists" statement is a staff- written news report. (It should be noted that the subject of this guess is different than the subject of the estimates in (l) and (2), because the former includes mainly or exclusively birds or mammals whereas the latter includes all species. While this difference implies that (l) and (2) may be too low a basis for estimating the present extinction rate of all species, it also implies that there is even less statistical basis for estimating the extinction rate for species other than birds and mammals than it might otherwise seem.) (4) This guessed upper limit in (3) is then increased and used by Myers, and then by Lovejoy, as the basis for the "projec- tions" quoted above. In Global 2000 the language became "are likely to lead" to the extinction of between l4% and 20% of all species before the year 2000. (U.S., l980, II, p. 328) So an upper limit for the present that is pure guesswork has become the basis of a forecast for the future which has been published in newspapers to be read by tens or hundreds of millions of people and understood as a scientific statement. The two historical rates stated by Myers, together with the yearly rates implied by Lovejoy's estimates, are plotted together in Figure 1. It is clear that the Lovejoy extrapolation has no better claim to belief than a rate that is, say, one hundredth as large. Indeed, looking only at the two data points alone, many forecasters would be likely to project a rate close to the past rate, nowhere near Lovejoy's estimate, on the basis of the common wisdom that in the absence of additional information, the best first approximation for a variable tomorrow is its value today, and the best second approximation is that the variable will change at the same rate in the future that it has in the past. Figure 1 Projected change in the amount of tropical forests implicitly underlies the differences between past and projected species-loss rates in Lovejoy's extrapolation. But to connect this element logically, there must be systematic evidence relating an amount of tropical forest removed to a rate of species reduction. The available evidence does not support this theory. For example, Ariel Lugo details the situation in Puerto Rico, where "human activity reduced the area of primary forests by 99%, but, because of coffee shade and secondary forests, forest cover was never below 10 to 15%. This massive forest conversion did not lead to a correspondingly massive species extinction, certainly nowhere near the 50% alluded to by Myers" (1989, p. 28). During the 1980s there was increasing recognition that the rate of species loss really is not known. Indeed, as of 1989 Myers himself wrote, "Regrettably we have no way of knowing the actual current rate of extinction in tropical forests, nor can we even make an accurate guess" (p. 102). One would think that this absence of knowledge would make anyone leery about estimating future extinctions. In response to the sort of questions raised above, the "official" World Conservation Union (IUCN) commissioned a book edited by Whitmore and Sayer (1992) to inquire into the extent of extinctions that appeared after the first draft of this book. The results of that project must be considered amazing. All the authors are ecologists who express concern about the rate of extinction. Nevertheless, they all agree that the rate of known extinctions has been and continues to be very low. This is a sampling of quotations (with emphasis supplied), first on the subject of the estimated rates: ...60 birds and mammals are known to have become extinct between 1900 and 1950 (Reid, 1992,p. 55) [F]orests of the eastern United States were reduced over two centuries to fragments totalling 1-2% of their original extent...during this destruction, only three forest birds went extinct -- the Carolina parakeet ... the ivory-billed woodpecker ... and the passenger pigeon .... Although deforestation certainly contributed to the decline of all three species, it was probably not critical for the pigeon or the parakeet (Greenway, 1967). Why, then, would one predict massive extinction from similar destruction of tropical forest? (Simberloff, 1992, p. 85) IUCN, together with the World Conservation Monitoring Centre, has amassed large volumes of data from specialists around the world relating to species decline ... these and other data indicate that the number of recorded extinctions for both plants and animals is very small ...(Heywood and Stuart, 1992, p. 93) Known extinction rates are very low. Reasonably good data exist only for mammals and birds, and the current rate of extinction is about one species per year (Reid and Miller, 1989). If other taxa were to exhibit the same liability to extinction as mammals and birds (as some authors suggest, although others would dispute this), then, if the total number of species in the world is, say, 30 million, the annual rate of extinc- tion would be some 2300 species per year. This is a very significant and disturbing number, but it is much less than most estimates given over the last decade. (Heywood and Stuart, p. 94) ... if we assume that today's tropical forests occupy only about 80% of the area they did in the 1830s, it must be assumed that during this contraction, very large numbers of species have been lost in some areas. Yet surprisingly there is no clear-cut evidence for this.... Despite extensive enquiries we have been unable to obtain conclusive evidence to support the suggestion that massive extinctions have taken place in recent times as Myers and others have suggested. On the contrary, work on projects such as Flora Meso- Americana has, at least in some cases, revealed an increase in abundance in many species (Blackmore, pers. comm. 1991). (Heywood and Stuart, 1992, p. 96) ... actual extinctions remain low...As Greuter (1991) aptly comments, `Many endangered species appear to have either an almost miraculous capacity for survival, or a guardian angel is watching over their destiny! (Heywood and Stuart, p. 102) ... the group of zoologists could not find a single known animal species which could be properly declared as extinct, in spite of the massive reduction in area and fragmentation of their habitats in the past decades and centuries of intensive human activity. A second list of over 120 lesser-known animal species, some of which may later be included as threatened, show no species considered extinct; and the older Brazilian list of threatened plants, presently under revision, also indicated no species as extinct (Cavalcanti, 1981). (Brown and Brown, 1992, p. 127). Closer examination of the existing data on both well- and little-known groups, however, supports the affirma- tion that little or no species extinction has yet occurred (though some may be in very fragile persist- ence) in the Atlantic forests. Indeed, an appreciable number of species considered extinct 20 years ago, including several birds and six butterflies, have been rediscovered more recently. (Brown and Brown, 1992, p. 128) And here are some comments from that volume on the lack of any solid basis for estimation: It is impossible to estimate even approximately how many unrecorded species may have become extinct (Hey- wood and Stuart, p. 95) While better knowledge of extinction rates can clearly improve the design of public policies, it is equally apparent that estimates of global extinction rates are fraught with imprecision. We do not yet know how many species exist, even to within an order of magnitude. (Reid, 1992,p. 56) Despite this massive evidence, Myers argues: "But we can make substantive assessments by looking at species numbers before deforestation and then applying the analytical techniques of biogeography... According to the theory of island biogeography, we can realistcally reckon that when a habitat has lost 90% of its extent, it has lost half of its species." (p. 43) This is mere speculation, however. And as noted above, Lugo found disconfirming evidence in Puerto Rico. Yet the conservationists go right on pressing for expensive public policies on the unproven assumption that the number of species being extinguished is huge. [OUT IF NECESSARY Many biologists privately agree that the extinction numbers are quite uncertain. But they go on to say the numbers do not matter scientifically. The policy implications would be the same, they say, even if the numbers were different even by several orders of magnitude. But if so, why mention any numbers at all? The answer, quite clearly, is that these numbers do matter in one important way: they have the power to frighten the public in a fashion that smaller numbers would not. I find no scientific justification for such use of numbers.] CONCLUSION There is now no prima facie case for any expensive species- safeguarding policy without further evidence. The existing data on the observed rates of species extinction are inconsistent with the doomsters' claims of rapid disappearance, and they do not support the various extensive and expensive programs they call for. Furthermore, recent scientific and technical advances -- especially seed banks and genetic engineering -- have diminished the importance of maintaining species in their natural habitat. But the question deserves deeper thought, and more careful and wide ranging analysis, than has been given it until now. I do not suggest we ignore extinctions. Rather, we should be as informed as possible. We should separate the available facts from the guesswork and the purposeful misstatements, in order to improve the public decision-making process. And society should take into account -- but in a reasoned fashion -- the economic and non-economic worths of species, in light of our values for human and non-human aspects of nature and other as- pects of life on earth. It is important that we think as clearly as we can about this problem that is indeed difficult to think about sensibly. page 1 /article3 specifor/October 13, 1993 REFERENCES Atkinson, Ian, "Introduced Animals and Extinctions," in Conservation for the Twenty-first Century, Edited by David West- ern and Mary C. Pearl (New York, Oxford: Oxford University Press, 1989). Brown, K. S., and G. G. Brown, "Habitat alteration and species loss in Brazilian forests," in Whitmore, T. C., and J. A. Sayer, 1992, pp. 119-142 Colinvaux, Paul A., "The Past and Future Amazon," in Scien- tific American, May 1989, pp. 102-8. Ehrlich, Paul and Anne, Extinction (New York: Random House, 1981). _____, and Edward O. Wilson, "Biodiversity Studies: Science and Policy", Science, vol.253, August 16, 1991, pp. 758-762. Heywood, V. H., and S. N. Stuart, "Species extinctions in tropical forests," in Whitmore, T. C., and J. A. Sayer (1992), pp. 91-118. Labandeira, Conrad C., and J. John Sepkoski, Jr., "Insect Diversity in the Fossil Record", Science, 16 July, 1993, 310- 314. Lugo, Ariel E., Editor, "Diversity of Tropical Species," in Biology International, Special Issue - 1989. Myers, Norman, The Sinking Ark (New York: Pergamon, l979). Myers, Norman, "A Major Extinction Spasm: Predictable and Inevitable?", in Conservation for the Twenty-first Century, Edited by David Western and Mary C. Pearl (New York, Oxford: Oxford University Press, 1989). Olson, Storrs L., "Extinction on Islands: Man as a Catas- trophe," in Conservation for the Twenty-first Century, Edited by David Western and Mary C. Pearl (New York, Oxford: Oxford Univer- sity Press, 1989). Quinn, James F. and Alan Hastings, "Extinction in Subdivided Habitats," in Conservation Biology, Volume 1, No. 3, October 1987, pp. 198-208. Reid, W. V., "How Many species will there be?," in Whitmore, T. C., and J. A. Sayer 1992), pp. 55-74. Reid, Walter V., and Kenton R. Miller, Keeping Options Alive: The Scientific Basis for Conserving Biodiversity, (World Resources Institute, October 1989). Simberloff, D., "Do species-area curves predict extinction in fragmented forest?," in Whitmore, T. C., and J. A. Sayer, 1992, pp. 75-90. U.S. CEQ and Department of State, The Global 2000 Report to the President, Vol. II, (Washington: GPO, 1980). Whitmore, T. C., and J. A. Sayer (Editors), Tropical Deforestation and Species Extinction, (New York: Chapman and Hall, 1992). Wilson, Edward O., The Diversity of Life (Cambridge: Harvard UP, 1992). Whitmore, T. C., and J. A. Sayer (Editors), Tropical Deforestation and Species Extinction, (New York: Chapman and Hall, 1992). page 2 /article3 specifor/October 13, 1993 NOTES 1. Drawn from the forthcoming A Debate on the Environment by Norman Myers and Julian L. Simon, based on work done with Aaron Wildavsky. page 3 /article3 specifor/October 13, 1993