Effects of Masting on Rodents
Mast production provides an enormous flush of resources for seed consumers in forests dominated by masting trees. In the oak-dominated forests of the eastern United States, as well as those of eastern Europe, mast production causes high overwinter survival rates, and occasionally winter breeding, in both murid and sciurid rodents (Pucek et al. 1993, Ostfeld, Jones, and Wolff 1996, Jones et al. 1998, Wolff 1996). Abundant acorns may improve survival rates by allowing rodents to reduce foraging activities and home range sizes, thereby diminishing their vulnerability to predators. In addition, consuming an abundance of food allows storage of body fat, which may buffer rodents against harsh winter conditions. As a result of high overwinter survival and winter breeding following mast production, rodent populations begin the spring breeding season already at moderate to high density. Several studies have now demonstrated that forest rodent populations reach multiannual peaks in density in the springs or summers following mast production (Pucek et al. 1993, King 1983, Ostfeld, Jones, and Wolff 1996, Ostfeld et al. 1998, Wolff 1996, McShea 2000). Experimental simulation of masting, by providing abundant acorns on three 2.4-ha forest plots, reduced the rate of overwinter decline in white-footed mouse (Pemmyscus leucopus) populations and resulted in spring-summer densities that were approximately
5 times higher than those on unsupplemented control plots (Jones et al. 1998). Studies in Virginia (Wolff 1996) and New York (Ostfeld et al. 1998) have demonstrated that the size of the acorn crop in the fall explains about 80% of the interannual variation in density of white-footed mice in summer.
Many studies have demonstrated the importance of acorns to the autumn and winter diets of white-tailed deer (Odocoileus virginianus) (Healy 1997a, Chapter 14). Although acorn availability does exert demographic effects on deer, as on rodents, population fluctuations among deer are largely independent of variations in acorn production, because of their longer mortality and natality schedules. However, recent studies have revealed a pronounced behavioral response by deer to acorn production. In the autumn of a mast year, deer are attracted to oak-dominated stands and spend a considerable portion of their daily time budgets there. In contrast, during autumns of poor acorn production, deer avoid oak-dominated stands, aggregating instead in forest of other types, such as those dominated by maples (McShea and Schwede 1993; Ostfeld et al. 1998). Because of interactions between mammals and both disease vectors and forest pests, the numerical response to acorns by rodents and the behavioral response by deer have profound implications for human health and forest health (Figure 13.1).
Interaction of Rodents and Deer with Ticks and Lyme Disease
Rodents and deer are crucial hosts for ticks of the genus Ixodes, which are the vectors of the Lyme-disease agent, a spirochete bacterium (Borrelia burgdorferi). Lyme disease is a zoonotic disease, which means that the bacterial pathogen is maintained in wildlife populations and occasionally is transmitted to humans. Unlike some other vector-borne diseases, such as malaria, humans are irrelevant to the maintenance of the Lyme-disease enzootic cycle and only become involved "accidentally" when ticks, which normally feed on wildlife, attack people. Borrelia infections in wildlife hosts, including rodents and deer, appear to be rather benign, resulting in no obvious symptoms and having no detectable effect on survival or reproduction. Because of the mammalian hosts' role in feeding and infecting ticks, the population dynamics and space use of the hosts are critical to the epidemiology of this expanding disease (Lane et al., 1991, Piesman and Gray 1994, Ostfeld 1997).
Lyme disease is by far the most common vector-borne disease in the United States and is increasingly common in Europe. Over the past decade in the United States, between 8,000 and 16,000 cases have been reported to the Centers for Disease Control and Prevention (CDC) each year (Ostfeld 1997). in the United States, Lyme disease is particularly common in the northeastern and northcentral states, where the vector is the black-legged tick, I. scapularis.
Ixodes scapularis typically undergo a two-year life cycle that includes four stages: egg, larva, nymph, and adult (Fish 1993). in autumn of each year, adult ticks feed predominantly on white-tailed deer, mating during a single 3-4-day blood meal. Females drop off after engorging and overwinter in a quiescent state on the forest floor at the site of detachment from the deer. The following late spring or early summer, engorged females produce an egg mass before dying, and the eggs hatch in mid-summer into tiny (~0.5 mm) larvae. Because the adult stage of the tick is specialized to deer, the location of deer in autumn determines the location of newly hatched larvae the following summer. Larvae remain within a few meters of the site of hatching and wait for a host to wander near enough to permit attachment, a behavior called host seeking or questing. Unlike adult ticks, larvae are not specialized in their choice of hosts and may feed from any of a wide array of mammalian, avian, or reptilian hosts.
Because transovarial transmission of B. burgdorferi from female to offspring is highly inefficient, the vast majority of larval ticks hatch from eggs free of the Lyme disease spirochete (Piesman et al. 1986). Therefore, larval ticks are generally harmless. Larvae may become infected with B. burgdorferi if they feed on an infected vertebrate host, but the probability of becoming infected varies strongly with the species of the host. in the northeastern and north central United States, blood meals taken from white-footed mice are by far the most likely to result in infection of the feeding larval tick (e.g., Levine et al. 1985, Magnarelli et al., 1988, Mather 1993). it is for this reason that P. leucopus is considered the principal natural reservoir for Lyme disease in North America. in Europe, several mammalian and avian hosts may be competent reservoirs, resulting in more complex ecological dynamics (Randolph and Craine 1995). Once a larval tick becomes infected, it maintains the in fection through later molts and is capable of transmitting bacteria to subsequent hosts, including humans. Therefore, the population density of white-footed mice in summer, when larval ticks are active, strongly influences the number of ticks that become infected with B. burgdorferi (Mather and Ginsberg 1994). Because the density of infected ticks within areas that people use domestically and recreationally is the primary risk factor for Lyme disease, understanding the dynamics of white-footed mice may allow ecologists to predict and prevent human exposure to the disease (Ostfeld 1997).
After a single 2-3-day blood meal, larval ticks drop off the host and molt into the nymphal stage, which remains quiescent for 10 months or so, only becoming active the following late spring or early summer. Nymphs that acquired B. burgdorferi during their larval meal may transmit the disease agent to their human or nonhuman host during their nymphal meal. Because nymphs are small (~1 mm) and therefore difficult to detect, and because their season of peak activity coincides with that of humans, this life stage is probably responsible for transmitting the majority of Lyme disease cases (Barbour and Fish 1993). At forested sites in southeastern New York State, 25%-35% of nymphs are infected with the Lyme disease spirochete (Van Buskirk and Ostfeld et al., 1998, Ostfeld, unpublished data). Similar to larvae, nymphs do not specialize on any particular host species but instead feed on a wide variety of vertebrates. Feeding to repletion requires 2-3 days, after which nymphs drop off the host and molt into the adult stage, which seeks a deer host a few months later in the fall of the same year.
Acorn production influences Lyme disease risk through two different pathways, one involving deer and the other involving mice. In the autumn of a good mast year, when white-tailed deer are attracted to oak-dominated forest stands, they import their burdens of adult ticks into these habitat types, resulting in peak densities of newly hatched larval ticks the following summer (Ostfeld, Jones, and Wolff 1996, Jones et al. 1998). Because heavy acorn production also causes white-footed mouse populations to reach peaks in density the following summer, mast production results in simultaneous and syntopic peaks of ticks and the most competent natural reservoir for B. burgdorferi. These concurrent events result in a high probability that larval ticks will acquire the Lyme disease agent and molt into an infected nymph. The outcome is a higher than usual risk of Lyme disease during the second summer following heavy masting, given the 1 -year delay before larvae that fed on abundant mice become active as infected nymphs.
The density of white-footed mice is important not only to their parasites and pathogens but also to their prey, which include the gypsy moth (Lymantria dispar). The gypsy moth is a European invader of North American oak forests. In parts of the eastern United States, this species periodically undergoes population outbreaks during which it may defoliate large expanses of oak forest (Chapter 7). Gypsy moth populations tend to remain at low densities for several years before beginning a phase of rapid increase, often spanning five orders of magnitude in egg mass density over 2-3 years (Chapter 7, Campbell 1967, Ostfeld, Jones, and Wolff 1996). After one to several years of peak density, the moth populations then decline steeply, reentering a prolonged low-density phase.
Much attention has been devoted to understanding the causes of fluctuations in gypsy moth populations (Campbell and Sloan 1977, 1978, Doane and McManus 1981, Chapter 7). During the peak phase, moth populations may be regulated by their food supply, particularly when outbreaks result in massive defoliation of oak forests. Evidence indicates that the decline phase is caused by viral pathogens and parasitoids that specialize on gypsy moths and that exhibit a delayed density-dependent response to their moth hosts (Elkinton and Liebhold 1990). Other factors, such as induced chemical defenses by host trees, the use by moths of plant secondary chemicals for defense against pathogens, and delayed effects of high population density on maternal condition and fecundity, are also known to influence gypsy moth populations during both the peak and decline phases (Rossiter et al. 1988, Rossiter 1994, Hunter and Dwyer 1998).
After several larval instars, gypsy moths pupate for about two weeks in midsummer and then eclose into adults. It has long been known that white-footed mice eat gypsy moth pupae, which are a large (~ 2-3 cm), immobile, undefended food source, highly accessible to mice by virtue of their location on the forest floor or low on trunks of trees (Smith 1985, Yahner and Smith 1991). Despite their propensity to attack pupae, however, mice appear to be unimportant in regulating high-density moth populations, largely because neither the functional response nor the numerical response of mice to moths is sufficiently rapid. Nevertheless, recent research has generated strong evidence that mice, via predation on pupae, are responsible for regulating moth populations during the low-density phase (Elkinton et al. 1996, Ostfeld, Jones, and Wolff 1996, Jones et al. 1998).
Studies examining predation on freeze-dried gypsy moth pupae show that mice are the predominant predator in most years and that the proportion of pupae attacked is strongly correlated with mouse density (Elkinton et al. 1996, Ostfeld, Jones, and Wolff 1996). When mouse density exceeds 10-15 individuals ha1, virtually 100% of the experimentally deployed moth pupae were attacked by mice within the 2-week window necessary for eclosion (Ostfeld, Jones, and Wolff 1996). In an experimental field study in which mouse density was reduced by trapping and removal, survival of both experimentally deployed and natural pupae was dramatically higher than in control sites in which mouse density was high. The result was an enormous increase in density of egg masses and caterpillars the following year on plots from which mice had been removed (Jones et al. 1998). Essentially, the reduction of mouse density during the low phase of a gypsy moth cycle released the moth population from regulation by mice and allowed it to begin a phase of rapid growth (Jones et al. 1998). This study, combined with other observational and experimental studies (e.g., Elkinton et al. 1996, Ostfeld, Jones, and Wolff 1996), indicates that moderate- to high-density mouse populations are sufficient to maintain moth populations at low densities in perpetuity, and that a crash in the mouse population when moth populations are sparse is both necessary and sufficient to cause rapid growth toward an outbreak of moths (Jones et al. 1998, Ostfeld and Keesing 2000b). Because crashes in mouse populations are predictable based on mast production (Wolff 1996, Ostfeld and Keesing 2000b), moth outbreaks and defoliation events also may be predictable well in advance.
Gypsy moths do not have a reciprocal effect on population dynamics of white-footed mice. Because moths pupate in midsummer, when food is not limiting to populations of mice (Hansen and Batzli 1978, Wolff 1996), mice do not appear to be affected by the density of gypsy moths.
The potential exists for a positive feedback loop from acorns to mice to gypsy moths to oak trees and masting (Figure 13.1). Gypsy moth defoliation of oaks may delay or prevent the production of mast crops by existing oaks and/or reduce the community dominance of oaks (reviewed by Healy et al. 1997). Temporary or long-term reductions in mast crops are expected to reduce average population densities of white-footed mice (Ostfeld, Jones, and Wolff 1996, Elkinton et al. 1996, Jones et al. 1998), which will relax the suppressive effects of mice on gypsy moths. This in turn will increase the probability of moth outbreaks and defoliation events. The existence of feedback loops adds a level of complexity to forest management, because the impacts of a particular management action may become strongly amplified.
Effects of Mice on Ground-Nesting Songbirds in oak forests of the eastern United States, several species of songbirds, including ovenbirds, worm-eating warblers, veeries, wood thrushes, and dark-eyed juncos, nest at or near ground level. Nests of these species may be vulnerable to attack by various mammalian and avian predators, especially during incubation. indeed, many studies using artificial ground nests suggest the potential for these predators to cause nest failure and even population declines of some passerines (e.g., Leimgruber et al. 1994, Martin 1993). Deployment of artificial ground nests, typically baited with both quail eggs and clay eggs (the latter for acquiring tooth or bill prints useful in identifying nest predators), has suggested that mammals such as raccoons and opposums, and birds such as bluej ays and crows, are the principal predators. However, because quail eggs are larger and have thicker shells than typical songbird eggs, this approach may bias results against the detection of smaller predators such as mice and chipmunks (Maxson and Oring 1978), which typically are unable to handle quail eggs.
Recent studies in southeastern New York State using passerine eggs revealed that the white-footed mouse was responsible for the majority of attacks on artificial ground nests and that eastern chipmunks were the second most frequent predator. Medium-sized mammals and birds were infrequent predators on these nests (K. Schmidt, R. Naumann, J. Goheen, R. Ostfeld, E. Schauber, and A. Berkowitz, unpublished data). in oak-forest plots in which mouse populations were maintained at low densities via removal trapping, attack rates on artificial nests were significantly lower than on control plots supporting high mouse density. in contrast, experimental manipulation of chipmunk densities had no effect on nest-predation rates.
Studies of nest predation using artificial nests have a number of well-recognized potential weaknesses, including lack of parental defense, poor placement by the experimenter, and elevated attractiveness due to scent contamination. Therefore, additional studies, particularly examining attacks on natural nests, will be necessary to determine whether artificial nest experiments accurately represent rates and perpetrators of natural nest predation. However, some evidence suggests that results from artificial-nest studies may accurately reflect processes affecting success of natural nests. in a long-term study of nesting performance of dark-eyed juncos in oak forests of Virginia, Ketterson et al. (1996) showed that the proportion of nests failing to fledge young was strongly correlated with summer density of Peromyscus populations. At these same sites, density of mice in summer was highly correlated with acorn production the prior autumn (Wolff 1996).
The density and structure of understory vegetation may influence the survival of eggs and nestlings of ground-nesting songbirds. Because browsing by deer on forest understories may affect protective cover and the suitability of nesting sites, population size and space use of deer may also strongly influence bird populations indirectly (McShea and Rappole 1997). When acorns are abundant, impacts by deer on under-story vegetation in the autumn and winter may be relaxed, due to reduced browsing, which in turn may enhance protective cover for birds the following summer. On the other hand, dense populations of deer when no acorns are available may have a strongly destructive influence on protective understory vegetation (McShea and Rappole 1997).
Potential Interactions between Mice and Their Predators
Interactions between white-footed mice and their avian and mammalian predators in oak forests have not been well studied. Despite anecdotal reports, ecologists have not yet determined whether raptor or carnivore populations experience unusually high reproductive success during years of high mouse densities. Similarly, little evidence exists to evaluate the possibility that predation by raptors and carnivores is responsible for mouse population declines from high densities. In oak-hornbeam forests of eastern Europe, acorn-caused increases in the population density of rodents, particularly Apodemus sylvaticus and A. flavicollis, appear to induce population growth by their predators, especially mustelids and owls (Jerzejewska and Jerzejewski 1998). These predators, in turn, attack alternative prey, such as nesting songbirds, when rodent populations collapse.
Was this article helpful?