Delayed implantation is more common in some groups of mammals, such as marsupials, mustelids, and cetaceans, than in others. It is controlled by complex physiological feedback interactions with environmental cues (Renfree & Shaw 2000). Explanations for the adaptive advantages of the ability to uncouple the fixed interval between mating and birth by keeping fertilized embryos in limbo for a while are easy to suggest for some species; for example, it allows female fur seals to mate again soon after producing their pups and spend the rest of the year at sea; it allows female wallabies living in arid regions to produce a new joey more rapidly after unexpected rain.
But for the small mustelids there are no such obvious explanations. Rather, there is the puzzling question: Why do stoats and longtails have delayed implantation while common and least weasels do not? Actually, this is really two questions: (1) Is delayed implantation useful to stoats and longtails? and (2) if it is useful, why do common weasels not have it too? Alternatively (looking at the question from the other perspective): (1) Are common and least weasels better off without delayed implantation? and (2) if so, why do stoats and longtails delay?
The ways in which modern animals make their livings are, by definition, ways that work. They are not necessarily the best possible ways, just the best of the ways that have been tried. Delayed implantation may be a useful adaptation original to stoats and longtails, or, alternatively, stoats and longtails may have retained the mechanism controlling delayed implantation, inherited from an ancestral species for which it was useful, because natural selection has not eliminated it from their lineage—even though they are much shorter-lived than all other species with a delay (Mead 1993; Thom et al. 2004).
The second explanation seems to us by far the most likely. Delayed implantation appears to have evolved first in an ancestor common to all the canoid carnivores (Lindenfors et al. 2003), and it has been lost only 11 times since. Stoats and longtails each belong to a lineage (not necessarily the same one) in which delayed implantation has been retained, conserving a mechanism so precise and consistent, there is practically no individual variation. All females of all ages and both species delay implantation every year, everywhere. Clearly, there is no profit in deviation. Surely, the entire suite of adaptations involved in delayed implantation must be maintained because the advantages of the total package outweigh the disadvantages of tinkering with inconvenient bits of it.
It is, in fact, easy to make a list of advantages of delayed implantation to stoats, and we have suggested some (King 1983a, 1984a; Powell 1985b). First, the adult female and her young are all ready for mating at about the same time. This means that any male that can get into a breeding den at the right moment stands to gain several mates all at once. Such extreme polygyny is rare among small carnivores, whose females are usually too widely dispersed for a male to monopolize access to a whole group of them. The precocity and immobility of the nestling females makes stoats the outstanding exception to this general rule, and the consequence for the successful male stoat is a huge payoff in future offspring. Only the old female can put up any resistance to his attentions, so, if she is at home when he calls, she makes the choice of a mate for all the females in the family.
Second, the young female stoats are fertilized before the family breaks up. This means that they are already assured of a litter next season, even if, by then, prospective mates are few. The blastocysts take very little energy to maintain, so pregnancy does not interfere with normal growth; come spring, the young female stoats are free to concentrate on hunting and on finding and establishing a den for giving birth. Young female longtails are also precocious, but not to quite the same extent. They do not mate in their natal dens, but shortly after independence. They still have the advantage of being free to concentrate next spring on hunting and finding a den, as do female stoats, but male longtails do not have the advantage of access to a ready-made harem of infant mates that male stoats have.
The trouble with explanations of evolutionary adaptations is that they are usually proposed in retrospect. As in Rudyard Kipling's famous Just So Stories, adaptation stories may be made to fit the observed outcome. Delayed implantation may well have many advantages for stoats and longtails, but for all we know, both species might have been just as successful without it. As in any theoretical problem, answers are easy; the trick is to find the right question. In this case, we need to know not only why stoats and longtails have retained delayed implantation, but also why (and how) common and least weasels have lost it.
Mikael Sandell (1984) proposed that stoats and common weasels have different patterns of reproduction because the keys to success in contemporary times are different for each. For stoats, he reasoned, the most important consideration is to produce the young as early in the season as possible. Sandell assumed that the young need plenty of time to grow before the next winter, yet the adults need to avoid mating too early in spring, while conditions may still be severe. Delayed implantation allows both. By contrast, King (1983a, 1984a) argued that the only thing that matters to common weasels is a high potential rate of increase, that is, the capacity to turn mouse meat into young weasels faster than anyone else. For this, small size, rapid maturation, and no delayed implantation are the best policies.
Nils Stenseth (1985) was stimulated by the exchange between Sandell (1984) and King (1984a) (in the columns of the journal Oikos) to produce a model incorporating the different reproductive capacities of the two species (stoat, six to
13 young once a year; common weasel, four to eight young once or twice a year). He then addressed Sandell's question on whether these differences could be maintained by the action of contemporary natural selection. He concluded that Sandell's idea is basically correct: The relative difference in litter size and frequency predicted by his model corresponds to the observed difference in nature.
Unfortunately, Stenseth's analysis included some known errors. For example, he assumed that common weasels always have two litters every year, whereas in nature they do this only when food is abundant (McDonald & Harris 2002). More controversially, Stenseth accepted Sandell's assumption that stoats must produce their young early in the year, whereas the advantage of early littering for stoats and longtails seems insufficiently consistent and compelling to account for the remarkable constancy of delayed implantation that both species show.
Stenseth's, Sandell's, and King's exchanges answered some questions but raised new ones, such as how do the tiny Canadian stoats manage to continue to behave like stoats when their body size and food resources are those of common and least weasels? And why must stoats avoid breeding too early in spring when the much smaller least weasels can breed under the snow all winter?
In the fully developed version of his hypothesis, Sandell (1990) proposed that delayed implantation is retained only in those contemporary mammals for whom the period required for active gestation is shorter than the period between the best time for mating and the best time to give birth. If the total gestation period could be lengthened, by inserting a delay in implantation, it might become possible both to breed and to give birth at the respective optimal times. Sandell reviewed the literature and found, as would be expected, that all mammals exhibiting delayed implantation can be interpreted to benefit from the delay.
The beauty of Sandell's hypothesis, however, is that it makes predictions beyond such Just So Stories interpretations. The hypothesis predicts that some mammals that might benefit from a large increase in the length of gestation do not delay because the assumed first step, a short increase, would actually decrease fitness. The hypothesis can also be used to predict delay for mammal species whose biology is poorly known. To date, unfortunately, no one has tested the hypothesis by making real predictions and assessing the outcomes. We hope to see such tests in the future.
With this background, we can search for possible answers to questions about the origins and uneven distribution of delayed implantation in small mustelids by looking more closely at the two best-studied contrasting species, the stoat and the common weasel.
Was this article helpful?