Come to Me, My Melancholic Baby!

Evolutionary Anthropology 3

CROTCHETS & QUIDDITIES

KENNETH WEISS

Some famous textbook cases of evolution aren’t unambiguously true. Does it CATCHING EVOLUTION matter? IN THE ACT Of Birds and Bugs here’s not much point in writ- the comparison was wrong. He said ing a column about things we the horse was bigger. Of course, we Come to me, my melancholy baby, Talready know, but there is value have to presume that he himself had Cuddle up and like your hue ... in looking at things we thought we actually seen a fox terrier. All your fears are foolish fancy, already knew. One way to do this is to As so often happens, something maybe see how textbooks illustrate basic similar came up coincidentally in an You know dear, that I’ll be tree to principles. I was led to write this col- unrelated conversation, and is closer you. umn when my daughter took an intro- to anthropology. In 1940 the promi- ductory bioanthropology course this nent geneticist A. H. Sturtevant sug- Every smog may have a silver lining, year and asked me something about gested that tongue-rolling ability illus- But until the sun shines through, natural selection. Her class had been trated the genetic control of human Smile, my honey dear, discussing the example of industrial traits, and we’ve used it as a classroom While I kiss away each tear, melanism and the peppered moth. I example ever since (are you TT or tt?). ‘Cause I’ll be staying melancholic found that curious, because I thought It costs nothing to test in students, too!’ I had seen several years ago that that and if you assume a dominant gene in example had been undermined by Hardy-Weinberg equilibrium you can contradictory evidence from color always estimate the allele frequency. A quick check of the web confirmed changes in the same moths in North However, in 1965 Sturtevant retracted my recollection that numerous prob- America. How could it still be in text- his suggestion that the trait was ge- lems had been raised about industrial books? netic, saying “I am still embarrassed melanism as an example of natural This brought to mind Stephen Jay to see it listed in some current works selection observed in action (and Gould’s making fun in Natural History as an established Mendelian case.” touted by creationists as disproving of the classic description of the early Tongue rolling has still not been com- the evolutionary edifice) (see Majerus horse Eohippus (or Hyracotherium)as pletely displaced as a classroom ex- and B. S. Grant references). What are being the size of a fox terrier (Figure ample of Mendelian inheritance, but these issues? 1). This comparison was first used in if it is genetic at all, it’s neither a sim- In 1896 J. W. Tutt suggested protec- 1904 by the paleontologist H. F. Os- ple single-locus trait nor even a single tive coloration against bird predation born in reconstructing horse evolu- biological trait, and may in fact be as an explanation of the rapid recent tion. The simile was then copied from something most people can acquire rise in the frequency of melanic forms textbook to textbook by authors—and during a critical developmental pe- of the peppered moth Biston betularia for readers—who, like me, probably riod. in Britain. The selective force was pre- had little idea even what a fox terrier These whimsical examples of text- dation facilitated by industrial air pol- looked like. And according to Gould, book echoing illustrate Lewis Carroll’s lution. Most moths had been of the promise in The Hunting of the Snark light-colored typica form with black that “what I tell you three times is speckles, and enjoyed protective col- true.” But this may not always be triv- oration on lichen-covered trees where Ken Weiss is Evan Pugh Professor of An- ial, so it is worth looking at cases in they were hard for birds to see, be- thropology and Genetics at Penn Univer- sity. which questions have been raised cause they dramatically resemble the about the two classic examples used mottled pattern of lichens (Figure 2). to buttress core concepts of evolution- But soot from the industrial revolu- Evolutionary Anthropology 12:3–6 (2003) ary biology, which could have serious tion blackened the trees in the 19th DOI 10.1002/evan.10040 Published online in Wiley InterScience practical consequences if the exam- century, killed the lichens, and de- (www.interscience.wiley.com). ples are wrong. prived the moths of their protective 4 Evolutionary Anthropology CROTCHETS & QUIDDITIES

is also evidence of differential mortal- ity in the expected direction, but of pre-adult moths, which cannot be at- tributed to protective coloration. Fi- nally, I was able to confirm my recollection that there was a parallel change in color variants in North American moths, and parallel pat- terns of industrialization, but lichens were not involved here and the change may be continuing in regions that do not share comparable pollution differ- ences. It is disturbing that so many aspects of what seemed such a classically simple exemplar of evolution could be cred- Figure 1. Horse and Hounds? (Not to scale.) Courtesy www.nature.ca/nature_e.cfm and ibly questioned. However, it appears www.akc.org. that the problems, though serious, do not entirely invalidate the story. If soot and lichens were not involved in the cover. This favored a formerly rare (and hence most exposed) places on color pattern changes in North Amer- melanic variant whose peppering the tree trunks; normally these moths ica, there was a correlation here, as in black spots were so large that the hide in the branches, and are active Britain, with air pollutants like SO2 that affect reflectance on trees. That’s not wings were essentially all black, con- only at night when birds aren’t. Ket- the same as mimicking lichens, but it is ferring reversed protective coloration tlewell stressed the role of lichens be- a plausible source of protective colora- against the newly blackened trees. cause they appeared to provide an ex- tion. Although several studies showing The carbonaria form rose rapidly in ceptionally cryptic background for the differential predation by birds of con- frequency. As pollution waned after peppered form, but there has been de- spicuous moths had various design Clean Air Acts were passed in the 20th bate about whether moths preferen- century, the variant frequencies re- problems, their findings have been con- tially seek lichens that match their sistently in the expected direction. It is verted towards the lighter peppered own color, which would obviate pre- form. probably true that protective coloration dation based on color. Even the cor- This selective hypothesis was exam- is at most one of several factors that will relation of morph frequency with the ined theoretically by the leading pop- explain this story. For example, the fact presence of lichens is imperfect in ulation geneticist J. B. S. Haldane and that color changes in two continents Britain, so geographic anomalies be- others in the 1920s. In the 1950s Ber- have paralleled the history of air pollu- tween pollution and melanic fre- nard Kettlewell undertook a series of tion is compelling, but we have to be- famous experiments to demonstrate it quency have had to be explained after ware that we not fall into the common directly. Among other tests, he placed the fact by invoking gene flow. There trap of equating correlation with causa- moths on trees and observed cryptic to non-cryptic predation rates by birds. This became, and remains, the classic direct demonstration of evolution by natural selection. However, many aspects of this work have been questioned. I mention some of them to illustrate the problems. To the human eye the moths seemed to enjoy cryptic protection, but evidence suggests that birds can see into the UV light spectrum, which means that protective coloration may be less effective to them than it is to the scientists suggesting the hypothesis. Kettlewell performed his experiments under various artificial conditions. He used unnatural moth densities in his experiments, and he released the moths in daylight, when Figure 2. Liking lichens? Kettlewall’s Biston betularia. Lichen covered tree with dark and light they quickly landed on the nearest (lower left) variants. (From Kettlewell, 1956.) CROTCHETS & QUIDDITIES Evolutionary Anthropology 5

time their study began. Strong selection in times of drought also did not exhaust variation, as was shown by the birds’ ability to respond when rains returned. Overall, the phenotype changes or states over these 30 years were not predictable from the changes that could have been observed in a single season or short time span. For example, data from the first 5 years could not have been used in population genetics models to predict the 30-year results. Selection certainly occurs, and if your beak is too small you can’t eat big tough seeds and won’t do well Figure 3. Darwin’s finches. (From Darwin’s The Voyage of the Beagle, 1845.) if there aren’t any small seeds around. The Grants’ major point was a plea for more long-term observational studies, tion. Air pollution has been associated Rosemary Grant have observed the because typical small, short-term ob- with many things, and we may not Galapagos finches in detail. During servational studies may be insufficient know them all. this time there have been changes in to understand how evolution is actu- Above all, however, the changes in climate, including droughts that af- ally working. This is clearly appropri- color frequency have been too rapid, fect plant survival and hence the dis- ate, but there may be two deeper widespread, and consistent to be due tribution of available seed sizes. The points. First, the kinds of traits Dar- to any known factor other than some Grants systematically collected data win observed are still bouncing form(s) of natural selection, whether on survival and census counts of two around evolutionarily. How confi- or not we know the exact factors in- species (Geospiza fortis, medium dently can we infer the past selective volved. At the very least, the melanism ground finch; and Geospiza scandens, forces that got the traits to where they story does illustrate micro-evolution, cactus finch). They recently summa- are (temporarily) today? And second, even if this old chestnut of evolution rized the changes of body size, beak no observational study is long relative turns out to be more subtle than the size, and beak shape they observed to the evolutionary time scale, which simple textbook tale. over this period (Grant and Grant, is often what we want to understand. 2002). On the micro-evolutionary Would even a 100 year study be enough? scale, they provide elegant studies of To Speak of Beaks selection, as defined as statistically significant change in trait frequencies The oddity of a brief episode of in- COMPLEXITY BURIED tense selection induced by human in- from one generation to the next, di- WITHIN SIMPLICITY dustrial pollution may have parallels rectly observed in action. in more “natural” natural selection. A major finding was that selection The true evolutionary story is more Climate change has also been involved occurred in repeated short bursts. complex than the textbooks depiction in another classic story, related to the When appropriate circumstances ex- of a long, steady convergence to an evolution of Darwin’s famous Galapa- isted, the birds showed the expected adaptive peak. To many biologists, it gos finches. These birds are so well favorable relative survival of birds may suffice just to know that selection established as defining instances of with beak characteristics suited for can be demonstrated in the wild. But adaptive evolution that the Galapagos the seed size distribution that existed it is at best less satisfying to have to Islands are on the must see list of bi- in a given year, and this really was invoke constantly and rapidly chang- ologists the way a hajj to Mecca is in “selection” in the evolutionary sense ing mixes of selective forces. We may Islam. Everyone knows that after ar- that is, it changed the trait distribu- be getting at least the main selective riving from South America the finches tions in the next generation. But the factors right, but so many things in evolved by adaptation to local condi- direction, magnitude, and consistency evolutionary biology dissolve on close tions on the various islands. It is al- of these bursts varied among traits inspection, that I’ll be crotchety and most mandatory to include this exam- and between the species. Stochastic ask how we know that the other fac- ple even in anthropology textbooks. (statistically nonsignificant) changes tors were just nuisances, and that re- Darwin himself illustrated this adap- also occurred, and traits also changed sults would have been the same if they tive radiation with a drawing of their significantly over this period as a re- weren’t there. There are many prece- beak variation (Figure 3). But here, sult of hybridization and gene flow. dents to give us pause about such an too, the story is not so simple, in in- There were reversals in the direction assumption. teresting ways. of selection, but they did not restore a Diethard Tautz (2000) has ex- Over the past 30 years, Peter and population to its initial state at the pressed the challenge in a memorable 6 Evolutionary Anthropology CROTCHETS & QUIDDITIES way. The outcome of selection de- the conservative assumption that ary biology, it would be nicer if all we pends on its strength and duration, harm is harm. That’s roughly compa- had to worry about were tiny horses and on population size. If these fac- rable to extrapolating the results of or rolled-up tongues. tors have roughly quantitatively com- strong short-term selection in finches, parable impact, which may be the rule so can we trust the results even of NOTES more than the exception, evolutionary well-done but necessarily short-term outcomes are fundamentally probabi- natural-conditions field experiments? Send any comments on this column listic, which Tautz refers to as a “pop- The issue is more than just academic, to: [email protected]. I maintain Crotchety ulation genetic uncertainty principle,” because transgenic insecticide genes Comments on my web page: www. reminiscent of the Heisenberg princi- may already have escaped from do- anthro.psu.edu/rsrch/weiss lab.htm. I ple in physics. This could happen even mestic GM corn to wild maize in Oax- thank Anne Buchanan, Sam Sholtis, if selection were strong but irregular aca, Mexico, as was reported by an- and John Fleagle for critically reading or episodic, or involved many compet- other study (Chapela and Quist, this manuscript; Bruce Grant for moth ing factors. To infer a selection regime 2001). Despite some methodological help; and Chris Carlson for pointing out under such conditions would require problems (Nature, April 11, 2002), the the tongue-rolling example. studies that were themselves on an authors maintain that their basic find- My Melancholy Baby by E. Burnet, evolutionary or species wide scale. ing is true and it has been confirmed. music; G. Norton, lyrics; 1912 (e.g., ϳ If so, this will constitute evolution in http://home.swipnet.se/ w-134536/ mel27.htm). Butterflies That Won’t Migrate action, but how can we know whether this is something to be concerned and Genes That Do? about? THINGS TO READ These issues, of course, apply in Unfortunately, as I was finishing Many things discussed here can be principle to our need to explain hu- this column the peppered moth story profitably explored by web searching. man and primate evolution from a was made the subject of highly publi- mix of fragmentary fossil data and cized scandal-mongering, with the Gateshouse AMR, Ferry N, Raemaekers RJM. 2002. The case of the monarch butterfly: a verdict is short-term observation, little if any of implication that no evidence re- returned. Trends in Genetics 18: 249–251. it as clear as the situation in moths mained that industrial melanism was Gould SJ. 1988. The case of the creeping fox and finches. But I think that more an example of evolution in action (see terrier clone. Natural History 97: 16–24. than just the richness of the real evo- Grant, 2002). Sadly, this is likely to Grant BS, Owen DF, Clarke CA, Owen DF. 1996. Parallel rise and fall of melanic peppered moths lutionary story is lost in the uncritical change textbooks, which we can pre- in America and Britain. J Heredity 87: 351–357. repetition of simplistic textbook ex- dict will make much of the scandal, Grant BS. 1999. Fine tuning the peppered moth amples. The subtleties that actually and that that, too will be repeated paradigm. Evolution 53: 980–984. occur have implications for social pol- from book to book. We can only hope Grant BS, Wiseman LL. 2002. Recent history of melanism in American peppered moths. J Hered- icy as can be seen by an example of that despite the uncertainties the evi- ity 93: 86–90. interest to North American anthropol- dence that this is a documented evo- Grant BS. 2002. Sour grapes of wrath. Science ogists because it involves one of our lutionary story will also proliferate. 297: 941–942. favorite species: maize. It is important to raise all appropri- Grant PR, Grant BR. 2002. Unpredictable evolution in a 30-year study of Darwin’s finches. Sci- A few years ago it was reported that ate questions about dogma in science. ence 296: 707–711. pollen from genetically modified That in both the cases of beaks and Hansen-Jesse LC, Obrycki JJ. 2000. Field deposi- (GM) corn was dangerous to monarch spots it has been shown directly that tion of Bt transgenic corn pollen: lethal effects on the monarch butterfly. Oecologia 125: 241–248. butterflies (Losey et al., 1999; Hansen- selection works, and that we’ve prob- Kettlewell HBD. 1956. Further selection experi- Jesse and Obrycki, 2000). This chal- ably identified at least some of the ments on industrial melanism in the Lepidop- lenged heavily vested interests, and re- forces, is a triumph of evolutionary tera. Heredity 10: 287–301. sulted in several studies to see if the biology. But we should also smile and Losey J, Rayer LS, Carter ME. 1999. Transgenic pollen harms monarch larvae. Nature 399: 214. reports were true. Generally, the new not be melancholy that weaknesses Majerus MEN. 1998. Melanism-evolution in ac- studies found little if any ill effects have been discovered. Weaknesses are tion. New York: Oxford University Press. (Gatehouse et al., 2002). The original not failures. Digesting the message Quist D, Chapela IH. 2001. Transgenic DNA in- findings in part reflected the intensi- that evolution even in these classic trogressed into traditional maize landraces in Oaxaca, Mexico. Nature 414: 541–543. fied experimental exposures that were cases is not so clear after all will lead Sturtevant AH. 1940. A new inherited character applied in order to get a useful result to deeper understanding of evolution in man. Proc Natl Acad Sci USA 26: 100–102. in a reasonable time period. This has and better applications to situations Sturtevant HA. 1965. A History of Genetics. New been standard practice in testing that may really matter, as the maize York: Harper and Row. drugs or food additives in mice fed example may illustrate. Tautz D. 2000. A genetic uncertainty principle. Trends in Genetics 16: 475–477. doses many times larger than any hu- But rather than always having to Tutt JW. 1986. British Moths. London: Routledge. man will normally consume, under fight rearguard actions in evolution- © 2003 Wiley-Liss, Inc.