444 BOOK REVIEWS to the genetics of shell coiling in some gastropods, in such as competition, the mutation involving aD-gene which the direction ofcoiling is determined by a single is therefore also assumed to provide "ecological and Mendelian locus with two alleles, and homeotic mu­ reproductive isolation from the wild-type progenitor" tants in Drosophila. And while one cannot deny the (p. 188) in addition to morphological novelty. Saltation importance of these systems to an understanding of indeed. developmental mechanics, it is not immediately ob­ The main strength of Mechanisms ofMorphological vious why they represent the same mechanism by which Evolution is its primary focus on a problem that typ­ novel phenotypes arise. ically is relegated to a subsidiary role in general treat­ Arthur's down playing of developmental mechanics ments ofevolutionary biology: how do morphological stems from his beliefthat the developmental origin of differences at the level oforder, class, or even phylum "saltational variants" is well established and not, there­ arise and become established? Arthur does a good job fore, a problem; the real problem, he offers, "occurs at of pointing out aspects of this problem that lack con­ the level of the population rather than at that of the vincing explanations derived from the Modem Syn­ individual" (p. 179). This should come as a surprise thesis. He also usefully calls attention to the necessity to many evolutionary biologists. For instance, Futuy­ ofincorporating population-level phenomena in what rna (1979 p. 438) states: "The real problems posed by traditionally has been considered from a narrow mor­ evolution ... lie not so much in the potency ofnatural phological perspective. As an alternative, however, he selection as in the mechanisms by which the variations offers a saltational model that is, at least with respect on which it acts arise .... The problem of how new to certain aspects ofgenetics, development, and ecol­ variations arise falls not within the province of math­ ogy, less realistic than almost any scenario provided ematical genetical theory, but within that of molecular by the synthetic view. The answer to this problem must genetics and developmental biology." The problem, as lie between these two extremes, but determination of Arthur sees it, stems from the fact that his saltational exactly where the answer is still seems far off. variants are not perfect (this would be asking too much ofthe developmental system); they must be sustained LITERATURE CITED through a maladaptive phase lasting several genera­ FUTUYMA, D. J. 1979. Evolutionary Biology. Sinauer, tions until numerous minor mutations bring the novel Sunderland, MA. structure, or body plan, to a new adaptive peak. Thus, MAYR, E. 1963. Animal Species and Evolution. Har­ we are introduced to the concept of "n-selection"-a vard Univ. Press, Cambridge, MA. regime in which an organism's survival seemingly is based solely on whether or not it can breed successfully. Corresponding Editor: J. B. Mitton As this regime entails independence from biotic factors

Evolution. 40(2), 1986, pp. 444-445

THE NEOLITHIC TRANSITION AND THE GENETICS OF POPULATIONS IN EUROPE: A REVIEW'

PETER E. SMOUSE Department ofHuman Genetics and Division ofBiological Sciences University ofMichigan Ann Arbor, MI48109

Received October 14, 1985

One of the more engrossing pastimes that has oc­ second lesson we have learned is that the more infor­ cupied human geneticists over the last 40 years has mation we have from extraneous (non-genetic) sources, been the attempt to make some evolutionary sense out the better job of evolutionary reconstruction we can of the patterns of genetic variation that typify the species do. What this means in practice is that we use the Homo sapiens. The first lesson we have learned from genetic data not so much to infer human history as to this sort of work is that it is seldom possible to infer confirm it. In The Neolithic Transition and the Ge­ the details ofevolutionary history from the genetic data netics ofPopulations in Europe, an archaeologist (A. alone; there are simply too many factors that have J. Ammerman) and a population geneticist (L. L. Cav­ impinged on genetic variation in species to permit un­ alli-Sforza) use the available evidence from genetic ambiguous inference as to causation and timing. The marker frequencies to support their claim that early agriculture spread from southwest Asia across Europe, effecting the neolithic transition in the process. I The Neolithic Transition and the Genetics ofPop­ The book begins with a briefdescription ofthe origins ulations in Europe. A. J. Ammerman and L. L. Cavalli­ ofagriculture in southwest Asia. A distinction is drawn Sforza. Princeton Univ. Press, Princeton, NJ, 1984. xv between the initial development ofdomestic plant and + 176 pp. $25.00. animal species from wild progenitors and the incor- BOOK REVIEWS 445 poration of these species into a well developed agri­ populations as the outnumbered hunter-gatherers be­ cultural system. The domestication phase seems to have came assimilated into the agricultural society. Over occurred between 10,000 and 9,000 BP in southwest 2,500 to 3,000 miles, the net effect would have been a Asia, whereas full blown agriculture was clearly in evi­ gradual diminution ofthe genetic impact ofthe original dence by 8,000 to 7,000 BP. farmers from southwest Asia, thus creating a cline. The neolithic transition in Europe has traditionally The authors substantiate the southeast to northwest been defined in terms of tool-crafting techniques and cline expected from the demic diffusion model by using pottery styles. More recently, emphasis has shifted to principal components analysis on the frequencies of39 the types ofsubsistence economies represented by dif­ alleles, a large number ofthem from the HLA complex. ferent types oftool manufacture. The authors take the Remarkably, this cline accounts for approximately 30% position that the neolithic transition defined in terms of the genetic variation observed among modem pop­ of tools and pottery reflects the development of an ulations. They also extract an additional pair of prin­ agricultural economy throughout Europe; evidence ex­ cipal components that together account for a further ists to suggest that this agricultural system was intro­ 30% of the variation. The first runs roughly east and duced from outside, rather than being developed in west and the second spreads in all directions from a situ and de novo. . focal point north of the Black Sea. They tentatively Archaeological evidence and radio carbon dating postulate large scale gene flow from the east that post­ suggest that agriculture spread across Europe in a gen­ dates the neolithic transition as an explanation of the erally northwesterly direction at the rate of about I east-to-west cline, and point out that the focal zone of km/yr (or 25 km/generation). The total process is the third componentcoincides with the purported source thought to have taken 2,500 to 3,000 years. The central region for the radiation of Indo-European languages question addressed in the book is whether that spread­ (and presumably people). Neither matter is pursued ing process was strictly a matterofcultural (innovation) further, being left for later work. They briefly present diffusion,with agricultural techniques being passed from the results of some rather extensive computer simu­ one group to the next, or whether it was the farmers lations on the processes involved and show that the themselves who moved. The authors argue that, in model hangs together rather nicely. large part, it was the farmers themselves who moved, They are careful to point out that the genetic data and term this process "demic diffusion." are compatible with the demic diffusion hypothesis, Their argument is based on the assumption that cul­ rather than proving it. While it is possible to imagine tural diffusion and demic diffusion would have differ­ other population processes that might have produced ent consequences for the pattern ofgenetic divergence similar genetic results, they have nevertheless made a among current day populations in Europe. Under this good circumstantial case for the demic diffusion model. assumption, extant genetic patterns should provide in­ Future evidence mayor may not support their view. formation on the processes involved in the neolithic Regardless of the ultimate fate oftheir model, however, transition. Given cultural transmission, there should the book will remain a lovely demonstration of what be no long-range genetic continuity, and, barring the can be accomplished when experts from different fields patchy cohesion expected from localized gene flow, pool their intellectual resources. there should be no larger regional trends. Given demic The book represents a synthesis of more than a de­ diffusion, however, there should exist a directional cline cade's collaboration between Ammerman and Cavalli­ in allele frequencies running roughly southeast to Sforza. The presentation is nicely crafted, simple and northwest along an axis paralleling the spread of ag­ coherent, with the esoteric technical details of archae­ riculture. ology, population genetics, and statistics either illus­ Using R. A. Fisher's model for a wave of advance, trated heuristically or referenced. It is an enjoyable they show how genes from southwest Asia would have Sunday afternoon read, yet contains some pithy ma­ swept across Europe. Although the farmers, having a terial. I recommend it heartily. much larger subsistence base, would have greatly out­ reproduced their mesolithic neighbors, a certain amount Corresponding Editor: J. B. Mitton of gene flow would have occurred between the two