In the following discussion we address the question of what "trigger mechanism" might bring about the first-order immigration episodes that we have tallied above. An unexpected result of this analysis is the close association of the six first-order immigration episodes into three pairs. This is not conceivably a random pattern. These three pairs of immigration episodes pack a majority of North American immigrants into a time frame equal to only about one-tenth of the Cenozoic. Perhaps the second episode in each pair represents delayed discovery, due to poor recovery of rare fossils during early phases of colonization. Thus, the second immigration wave could be an "echo" of the first. After careful review of the stratigraphic and geographic details of each instance, we find this hypothesis unlikely, although it is worthy of more rigorous analysis. We are left with the straightforward view that the physical and ecological conditions that triggered the first of each pair (or cluster) of major immigration episodes were reiterated within one or two million years, thus bringing about paired and clustered sets of immigration episodes. If correct, this pattern provides an important clue as to the sustained (or recurrent) force of whatever conditions "triggered" the immigration episodes into North America.
In exploring possible causal mechanisms for immigration and rapid turnover episodes on the North American continent, we turned to two other large bodies of correlative evidence. First, we scanned the record of land mammal faunal turnover on other continents in search of corroborative evidence. Secondly, we looked at the marine record of global climatic change (as represented by oxygen isotope signals) for correlative patterns during the Cenozoic.
As the land mammal successions on other continents become more continuously documented and their chronologies become more fully correlated with the absolute time scale, it is important to determine whether they respond synchronously with the immigration and rapid turnover episodes observed in the North American record. We summarize a few highlights from other continental mammal faunas in Europe, China, the Indian subcontinent, and Africa. South America and Australia were almost completely isolated during most of the Tertiary; therefore, their land mammal faunas need not be involved in this immigration analysis.
The European record of Cenozoic land mammals closely approaches that of North America both in chronological continuity and in taxonomic diversity. Unfortunately, no complete revision of European mammal invasions is available. Mein (1989) lists many generic first appearances for Neogene mammal ages, but the lists do not distinguish between allochthonous and autochthonous first appearances. Such distinctions are more challenging in western Europe than in North America because it is simply a portion of Eurasia and the proportion of immigrant genera in successive faunas is probably much greater in Europe than in North America. Thus we are unable to derive an authoritative analysis of immigrants in the European record to compare with that for North America.
Opdyke (1990) recognized a strong correlation between Neogene faunal turnover episodes on the two continents, using a somewhat indirect approach. Since the land mammal ages were established independently in Europe and North America, on the basis of empirical biostratigraphy, he compared the biochronological boundaries. Synchronous boundaries indicate synchronous pulses of faunal turnover, which must be governed by global climatic events. If, on the other hand, the mammalian turnover (and immigration) episodes were independent and governed only by local effects, there should be little or no correlation between the two systems. As Opdyke (1990) suggests, there is a remarkable degree of correspondence between the two independently derived continental land mammal chronologies during the Neogene.
The status of land mammal ages in Europe was recently reviewed by Lindsay et al. (1990). Although the methods of mammal age definition differ considerably from those used in North America (Woodburne, 1987), the succession of mammal ages on the two continents appears nearly isochronous from Early Miocene through Late Pliocene. We highlight some of the more favorable comparisons below.
Legendre (1986) indicates three major immigration episodes among Paleogene land mammals of Europe. First is the strong continuity with North America in the earliest Eocene; this has long been recognized as the time of arrival of modern orders such as Primates, Perissodactyls, and Artiodactyls. The second is a small immigration wave in the Lutetian about 45 Ma. The third and most important is the rapid turnover episode long known as the "Grande Coupure," at approximately 32 to 34 Ma. What began mainly as an extinction event, led soon to a modernized fauna, characterized for example by new families of perissodactyls. The landscape in western Europe shifted from predominantly rain forest to savanna during this Early Oligocene transition (Legendre, 1986).
In the European Neogene system, the beginning of the Orleanian (MN 4) is marked by the immigration of the browsing horse, Anchitherium; two genera of proboscidea from Africa; and at least five genera of ruminants, possibly from Africa (Mein, 1989; Ginsburg, 1989; Tassy, 1989).