Measuring Rates of Change in the Fossil Record

Lecture © Mark V. H. Wilson
BIOL 606 Session, University of Alberta, January 19, 2000

Interest in the measurement of rates of evolution in the fossil record was stimulated by Eldredge & Gould's 1972 paper arguing for "punctuated equilibrium" as the dominant mode of speciation in the fossil record. Between speciation events, they argued, species largely exhibited "stasis," meaning they change little or not at all. Studies of change in actual examples of fossil lineages were pioneered by Gingerich (1979 and other papers), who coined the term "stratophenetics" for a method of phylogeny reconstruction involving linkage of fossil samples from stratigraphic layers spaced closely in time. Gingerich's samples of fossil mammal teeth showed evolutionary change at varying rates, but were these rates sufficiently fast to account for major evolutionary events such as origins of species and genera? Rates of change seemed too slow.

In a 1983 paper in Science, Gingerich provided a partial explanation. He compiled data on rates of change in morphological features from laboratory selection experiments, from studies of change since historical dates in introduced species, from studies of post-Pleistocene mammals, and from studies of differences between pairs of fossil species of different ages. He used asa unit of evolutionary rate the darwin = change by a factor of e per million years, proposed earlier by Haldane. Three examples are presented of calculation of rates in darwins, based on studies of horses by MacFadden, trilobites by Sheldon, and stickleback fishes by Bell et al. In his Science paper Gingerich graphed ln(rate in darwins) against ln(interval of measurement in millions of years). He found a strong inverse relationship between measured rate of evolution and the interval of measurement. Generally, rates measured in the fossil record appear much lower than rates measured in post-introduction or laboratory-selection studies.

Possible explanations include the fact that actual evolution proceeds at varying rates and reverses often, so that rates appear lower if measured over longer intervals. In addition, fossil samples typically represent temporally-averaged samples (samples that accumulated over long periods of time), such that within-sample variation is increased, and within-sample evolutionary change is obscured.

Could there be more appropriate units than the darwin? In a 1993 paper Gingerich mentioned the simpson (another unit suggested by Haldane) but proposed a new unit which he termed the haldane = change by a factor of one standard deviation per generation. Use of haldanes may have theoretical advantages, but requires more information about sample variance and generation times than does use of darwins; overall distribution of rates vs intervals is similar in any case, and for the purposes of this talk darwins are used.

A case study of measurement of rates of change is provided by the Eocene fossil-fish deposit at Horsefly, B.C. (e.g. Barton & Wilson, in press), where sequences of varves can be counted to give relative time of death of specimens of the sucker Amyzon aggregatum accurate to within one or a few years. One sequence spans 700 years of varves, and another (studied by Doug Barton for his M.Sc. thesis) spans 10,000 years. In the 700-year sequence, meristics of the fish (e.g. vertebral or fin-ray counts) change significantly over one or several centuries, while in the 10,000 year sequence, meristic features change significantly over several millennia. Environmental changes in the lake can also be estimated by variation in preservation of the fossils (Wilson & Barton, 1996). When rates of change in meristics are calculated in darwins, the rates fall within the general trend on Gingerich's graphs, but within an area of the graph where there is little data (because studies over several hundreds or several thousands of years are rare). The comparison of the changes seen in the 700-year sequence (~ 74 darwins) with those seen in the 10,000-year sequence (~ 7.4 darwins) are a graphic example of the averaging effects of temporal interval on the apparent rates of change in fossil lineages, since in both cases the same meristics and the same dating methods are used.
Barton, D.G., and Wilson, M.V.H. 1999. Microstratigraphic study of meristic variation in an Eocene fish from a 10 000-year varved interval at Horsefly, BC. Canadian Journal of Earth Sciences, 36:2059-2072.
Gingerich, P.D. 1979. The stratophenetic approach to phylogeny reconstruction in vertebrate paleontology. In Phylogenetic Analysis and Paleontology. Edited by J. Cracraft and N. Eldredge. Columbia University Press New York, pp. 41-77.
Gingerich, P.D. 1983. Rates of evolution: effects of time and temporal scaling. Science, 222: 159-161.
Gingerich, P.D. 1993. Quantification and comparison of evolutionary rates. American Journal of Science, 293A: 453-478.
Wilson, M.V.H., and Barton, D.G. 1996. Seven centuries of taphonomic variation in Eocene freshwater fishes preserved in varves: paleoenvironments and temporal averaging. Paleobiology, 22(4): 535-542.


Rapporteur: Curtis Strobeck

The first question asked about the paper "Rates of Evolution: Effects of Time and Temporal Scaling" by Gingerich was why anybody would try to measure the evolutionary rates in mammals and other vertebrates without taking scaling of time into account. Mark answered that the paper was developed partly in response to the statement by S. Stanley that there should be higher rates in animals with advanced social behaviour (vertebrates) than in those animals without advanced social behaviour (many invertebrates). [Note by MW: Gingerich also was responding to questions about the apparently low rates in his fossil mammal studies].

The second criticism was that there are biases in the data that was used. Gingerich recognized that there are biases in the data if the evolutionary rate was very fast over a long period of time, in which case the structures would be so different that they could not be compared. This means that the upper right hand corner of the graph is empty. However, there is probably also a reporting bias for slow evolutionary rates over short periods of time. Since these are unlikely to be reported in the literature, the lower left-hand corner of the graph is also vacant. This leaves you wondering whether, if both of these biases did not exist, you would see any significant slope in the data.

One question the author might be asked is, "What does the paper say about the controversy between punctuated equilibrium and gradualism?" It was decided that it is fairly clear that it says nothing about this controversy, but it was pointed out that it would be differently interpreted if the data were viewed from the perspective of punctuated equilibrium.

Lastly, it is not clear how living fossils, such as the horseshoe crab, fit into the data presented. Living fossils may not be changed because they live in very stable habitats.

The meeting continued at the Power Plant...[off the record]