On the other hand, there is also evidence that argues against the predation hypothesis, or at least against assuming it is broadly valid outside Scandinavia. In a coniferous plantation in northern England, not only were generalist predators unable to prevent unfenced field vole populations from fluctuating (Lambin et al. 2000), but also specialist predators (common weasels) were neither necessary nor sufficient to drive the fluctuations (Graham 2002). And some populations of northern rodents do not behave as predicted by the predation hypothesis, for example, the population of lemmings in Canada studied by Krebs and his coworkers (1995; Reid et al. 1997), and those on the coast of northern Norway with a mild climate and no permanent snow cover (Strann et al. 2002).
Perhaps, then, the apparent north-to-south variation in rodent dynamics is not due to, or not due only to, the distribution of different kinds of predators, but involves other processes that have not previously been considered. Jgdrzejwski and Jgdrzejewska (1996) and their coworkers suggested that the missing piece in the puzzle is the thickness and productivity of vegetation on the ground— which is, in turn, affected by the length of the winter snow cover. Voles and lemmings live on or under the ground, and their numbers must be determined at least as much by food supplies as by predation.
The vital measures of food supply most relevant to voles are not closely correlated with latitude, because the production of ground vegetation under the thick canopy of a temperate forest is as low as in the tundra, whereas it is high in open grasslands of the temperate zone. In contrast to the Scandinavian school, the Polish team found no significant differences between the impacts of generalist versus specialist predators on rodents. All predators have their greatest impact on rodent populations during the period when numbers are already decreasing, in part because at that time the individual rodents removed are not being replaced.
According to Jgdrzejwski and Jgdrzejewska (1996) and Jgdrzejewska and Jgdrzejwski (1998), even when predation accounts for much of the observed rodent mortality, it does not determine whether the rodent populations will fluctuate substantially across several years (as in "cycles") or merely vary seasonally. Instead, the supply of winter food seems to be the most important precondition permitting rodent populations to increase beyond one season, usually by winter breeding. The simple distinction between the strongly fluctuating northern rodent populations and the more stable southern populations merely represents the two most obvious ends of a continuum governed by an interaction between winter food supplies and predation.
Where the standing crop of ground vegetation is always poor, as in temperate forests, a sudden massive crop of seeds dropped in autumn and lying on the ground all winter certainly stimulates populations of forest rodents and their predators. In Bialowieza Forest, a contingent of generalist predators and high densities of weasels account for almost the entire winter mortality of the resident bank vole and yellow-necked mouse populations, but cannot prevent their numbers from fluctuating wildly between years in response to variations in tree seedfall (Jgdrzejwski et al. 1995). The population irruptions in forests are shortlived, ranging from 8 to 29 rodents per ha, and never reach the extreme densities attained by voles in temperate farmland and steppe (143 to 490 per ha) (Jgdrzejwski & Jgdrzejwska 1996). The Poles concluded that the main prerequisite for the northern vole cycles is not predation by specialist vole hunters such as weasels, but rather a mean standing crop of ground vegetation of over 4,000 kg dry weight per ha in summer.
Bank voles are socially intolerant and, where they are the dominant vole species, their populations can be locally stable, possibly controlled more by social behavior than by weasels. Elsewhere, regular fluctuations are imposed on bank voles by interactions with other rodents, for one of two reasons (Oksanen et al. 2000): (1) weasels switch to bank voles as alternative prey when Microtus voles are scarce, or (2) masting events in deciduous forests supply enough surplus food to disrupt their normal spacing behavior.
The question of whether predators limit prey or vice versa is a simple question to which there is no simple answer. Communities with weasels are good examples of ecological assemblages with several trophic levels (plants, herbivores, primary predators, and secondary predators), whose components respond on different scales to the same environmental changes (Powell 2001). Weasels typically inhabit ecological systems in which food supplies for their main prey vary from one year to the next, as do the foods of voles or mice through a population fluctuation.
These systems consistently generate two kinds of interactions (Ostfeld & Keesing 2000). The bottom-up effect (herbivores reducing plant matter) is accompanied by top-down cascades (predators reducing prey numbers, and thereby releasing plant abundance). Predation usually reduces herbivore populations below the level they could maintain without predators, yet year-to-year changes in productivity of plants will affect year-to-year changes in herbivore populations more than the parallel year-to-year changes in predator efficiency. Obviously, food affects productivity, and predation affects survival, so the density of any given population at a given time is the integrated result of both processes. Consequently, as Jgdrzejwski and Jgdrzejewska's (1996) analysis demonstrates, most communities show complex changes through time that depend at least as much on the vagaries of weather (and its effect on plant production) as on predation.
Whether predation by weasels is or is not central to the population dynamics of rodents, the instability of rodent numbers can lead to dangers for the weasels themselves. Erlinge (1983) found that when the population of rabbits decreased, the stoats had to compete with the generalist predators for rodents. When weasels of any species find themselves in that situation, they are simultaneously both hunters and hunted (Powell 1973, 1982; Korpimaki & Norrdahl 1989a, 1989b).
Working out the consequences of this dilemma at the level of the ecological communities formed by weasels and their prey is not easy, as we noted at the beginning of this chapter. Even manipulative field experiments, which are usually regarded as the most authoritative type of evidence, almost always fail to give definitive answers to questions about the behavior of wild populations (Raffaelli & Moller 2000). We need to concentrate on improving our field techniques, and on using models more effectively to create hypotheses to test in the field, so that we can continue to improve our understanding of predator-prey interactions.
Was this article helpful?