Modern animals are descended by a continuous process of change from different animals living in the past. Ancestral characters, or modified remnants of them, are often still discernible in contemporary species, constituting what Stephen Jay Gould called "the footprints of history."
We suggested in Chapter 1 that weasels evolved just before and during the Pleistocene from larger ancestors somewhat similar to martens and polecats. These ancestors probably already had delayed implantation (Lindenfors et al. 2003; Thom et al. 2004), as martens still do. About 53 mammal species are known to have obligate delayed implantation—that is, fewer than 0.05% of all mammals (Thom et al. 2004).The life histories of only 37 of the 60+ species of mustelids are known in any respect; of these, 17 have delayed implantation, a vastly greater proportion (about 34% of all the mustelids) than in mammals in general (Mead 1989; Sandell 1990). How could delayed implantation have gotten started in that far distant ancestral line, and why was it worth keeping?
Some mustelids and their closely related kin, American minks and striped skunks, for example, have short, variable delays of implantation, and sometimes they do not delay at all. These species may hold the clue to the original appearance of delayed implantation. Minks continue to ovulate even when fertilized if more than 6 days elapse between matings—a habit known as superfetation (Shackelford 1952; Mead 1994). Different males can fertilize ova shed in different ovulations, so if a better male comes along after a female mink has bred, she can still mate with the new, better male. Meanwhile, the zygotes resulting from early fertilizations delay implantation so that all can implant together.
This arrangement allows female minks to increase the number of offspring they produce, often fathered by more than one male, and also to improve the quality of their mates (and offspring), yet have all offspring born at the same time at the same developmental stage. If this short delay in implantation of some offspring proved to be beneficial, it could be increased for those species that could gain a further advantage from delaying implantation of all offspring, so increasing the interval between mating and giving birth. The simplest interpretation of mustelid phylogeny suggests that the ancestral mustelid lineage did delay implantation, and that several lines of its descendants have since lost it (Lindenfors et al. 2003). This line of reasoning is plausible only if there were good reasons to abandon such a long-established and successful characteristic.
King (1989a) suggested that the weasels in general became smaller in response to the opportunities presented by the evolution of open grasslands inhabited by voles and lemmings. The characters most favored in predators specializing on these new prey included, of course, the ability to search through miles of rodent tunnels and small burrows (see Figure 2.2). But even more important was the capacity to breed rapidly in response to sudden increases in food supply and to disperse in response to the decreases.
Stoats and common weasels look much alike, and most phylogenetic analyses done so far support the view that they are closely related (Bininda-Edmonds et al. 1999). In the process of adapting their life history characteristics to the uncertainties of life as a specialist predator of voles, both stoats and common weasels have evolved characteristics that substantially increased their reproductive rates compared with their larger bodied ancestors. The most effective ways of doing this, especially for small, short-lived animals, are early maturity, large litter sizes, and multiple litters per year.
On this view, common and least weasels have achieved all these advantages simply by abandoning the delayed implantation inherited from their common ancestor. Analyses of the chromosomes of the two species in Japan confirm the idea that the living least weasel is derived from an ancestral form similar to the stoat (Obara 1985). The problem with this idea is, if this new arrangement was advantageous for nivalis, why didn't erminea do the same?
One possible reason is that stoats developed, instead, a different, additional character, the extraordinary precocity of the juvenile females (King 1984a). The combination of delayed implantation and juvenile precocity allowed simultaneous mating of adult and young females, which in turn meant enormous potential breeding success for dominant males and certain fertilization before dispersal for the young females.
The addition of juvenile precocity turned delayed implantation into an advantage rather than a handicap, which, allied with the larger litters and longer life expectancy of stoats relative to common weasels, made a set of reproductive characters that succeeded. Females inheriting the combination of juvenile precocity and delayed implantation were at an advantage compared with other females. Consider an imaginary population in which most females died at 2.5 years old and delayed implantation was obligatory. A single heritable change allowing a female to mate as a nestling as well as at a year old would double her lifetime fecundity compared with other females who could not mate until they were yearlings—and then could not produce the litter until they were 2 years old. The advantages of the new arrangement to the female, and perhaps even more important to her sons, would ensure that it spread through the population.
One could, of course, consider alternative scenarios. Perhaps stoats and common weasels descended from an ancestor that did not have delayed implantation, and stoats later acquired it. The trouble with this idea is that stoats would have had to acquire juvenile precocity first, since to add delayed implantation without it would have caused a drastic drop in reproductive rate—the opposite effect to that required. Yet juvenile precocity without delayed implantation would have been suicidal for the young females, which could not possibly bear their litters while still nestlings themselves.
One of us (King 1983a) first suggested this hypothesis at a meeting in Helsinki in 1982. Soon afterward, the other (Powell 1985b) published a comparison between real or reasonable life tables for modern stoats and common weasels and for a hypothetical ancestor, addressing the question of how the modern species evolved. He concluded that, as common weasels have such short life spans, they could not have evolved their present small size without the massive extra productivity made possible by direct implantation and opportunistic late summer breeding, first by adults and later, as size decreased, by the rapidly maturing spring-born young. He added that, if stoats are descended from a larger ancestor with good adult survival rates and delayed implantation, the reduction in life expectancy that accompanied their decrease in size "would have required the evolution of juvenile breeding." In other words, these calculations confirmed that the modern stoats could have evolved from an ancestor with delayed implantation, provided they added juvenile precocity later.
The only other logical possibility was that stoats and common weasels are not as closely related as they look; perhaps stoats descended from marten-like ancestors that had delayed implantation, whereas common weasels came from polecat-like ones that did not (King 1983a). At the time it was suggested, this idea had something to recommend it; stoats are more similar to martens than to weasels in several curious characters, for example, in the shape of the bacu-lum and in delayed implantation and the whole physiological mechanism that controls it. This apparently plausible suggestion has since been supported by at least one phylogenetic analysis, which considers nivalis and erminea different enough to be placed in separate subgenera (Abramov 2000), although the general view is usually that stoats and common/least weasels are descended from a fairly recent common ancestor (Chapter 1).
The evolutionary history of common weasels and stoats is therefore equivalent to a natural experiment conducted over several million years. That was plenty of time to test the effect of a single inherited variable—the presence or absence of delayed implantation—on a pair of sympatric species of common origin, adapted to and still living in similar ecological conditions. The results are unusually clear, and their implications have a particular fascination for speculations on the evolution of reproductive strategies: They show how similar, closely related species can evolve different solutions to a common problem.
How do longtails fit into this story? As we saw in Chapter 1, independent lineages of mustelids have lived in North America and Eurasia since the late Miocene, and Mustela frenata has survived in its recognizably modern form much longer than has either erminea or nivalis. Both longtails and stoats have lengthy and obligate delayed implantation and large litters, but while both also have precocial maturation of young females, precocity in longtails is much less extreme than in stoats. Young female longtails breed, not as blind infants, but when they are older, well grown, and independent.
The long evolutionary separation of stoats and longtails suggests that the frenata/erminea story fits a scenario slightly different from the erminea/nivalis comparison. Perhaps longtails inherited delayed implantation from an early common ancestor with stoats and still control it by a similar, invariable genetic code, and later longtails independently added juvenile precocity, but by a slightly different genetic code. If this hypothesis is true, longtails and stoats demonstrate a remarkable degree of convergence—the tendency for similar-looking animals to acquire, regardless of background, adaptations advantageous in competition for particularly abundant resources.
Was this article helpful?