Theory and methods in carnivore conservation: who limits whom: predators or prey?

Abstract

Animal populations can be limited by the availability of food (limited from the bottom of the food chain up), by predators (limited from the top of the food chain down), or by the interaction of these two processes. Whether carnivores, in particular, limit the populations of their prey, or are limited by those prey, has long been controversial and is critical to conservation of prey and predatory species. I return to the question because it is a good question that we wish to have answered and, in part, because it has no simple answer. Our knowledge of ecological communities has matured to the point that we can tease the question apart, look at its pieces, and find conditional answers. Predator-prey models suggest that predators may limit prey populations on one scale while food limits prey, and prey limit predators on another scale. Predator populations are always limited by the availability of their prey. Data from the literature supports action on two time scales.

Introduction

"Who limits whom: predators or prey?" is an old question that resurfaces regularly because, in part, it is a good question that we wish to have answered and, in part, because it has no simple answer. I address the question once again because our knowledge of ecological communities has matured to the point that we can tease the question apart, look at its pieces, and find conditional answers. These answers provide a basis for managing populations of predators and prey, for conserving endangered species, for understanding what kinds of population fluctuations we can affect and what kinds we can not, and for knowing where in an ecological system our management energies should be targeted.

The debate regarding who limits whom has been viewed from two predominant perspectives. The Bottom-Up perspective emphasizes that plant defenses against herbivores and herbivore defenses against predators are so strong that communities are limited at each level predominantly by food (Murdoch 1966; Polis 1999; Polis & Strong 1996; Strong 1992; White 1978). From this perspective, predators in species-rich, terrestrial communities rarely have large effects on the populations of theft prey but, instead, are limited by them. Herbivore populations limit the populations of their predators because predator populations are small and herbivores are well adapted to avoid their predators.

In contrast, a Top-Down perspective was outlined by Elton (1927), then made more rigorous by Hairston et al. (1960) in their "The World is Green" paper, and has been explored extensively by Fretwell, Oksanen and coworkers (Fretwell 1977 ,1987; Oksanen et al. 1981; Oksanen and Oksanen 2000) as the Exploitation Ecosystem Hypothesis. From this perspective, plants are vulnerable to folivores but are seldom severely defoliated because of predation pressure on herbivores. The Exploitation Ecosystem Hypothesis is not limited to productive, species-rich systems and is most often tested using communities with vertebrate predators and prey.

Here, I ignore the abilities of plants to defend themselves and explore the effects on herbivores and predators of varying the productivity of plants that is available to herbivores. For convenience, I call this available productivity of plants "primary productivity."

Model predators and prey

Using a difference equation model similar to that used by Boyce and Anderson (1999), incorporating Type 3 functional responses (Holling 1959) and density-dependent population growth for both predators and prey, I developed a system with three trophic levels. The responses of this system to changes in primary productivity resemble the classic Top-Down trophic pattern (Figure 1a) except that the herbivore graph has a "hump" (Figure 1b). The model predicts that at high levels of primary productivity, predators limit their prey to lower population sizes than they do at intermediate levels of primary productivity. Similarly, when herbivores are not limited by predators, they limit the standing crop of vegetation. The model also predicts a most basic characteristic of herbivore and predator populations: primary productivity establishes the conditions that allow herbivores, or herbivores and predators to exist. If primary productivity is not great enough, herbivore populations are not large enough to support predators and predator populations can not become established. Hence, predators are ultimately limited by their prey (Figure 2).

[FIGURES 1-2 OMITTED]

Admittedly, most predator-prey systems involve more than one prey and one predator species, yet some well-studied systems are quite simple. On Isle Royale in Lake Superior, for example, moose (Alces alces) have been basically the only prey for wolves (Canis lupus) for nearly five decades (McLaren and Peterson 1994). Similarly, muskoxen (Ovibus moschatus) generally are prey only for wolves, or wolves and Man (Gray 1987; Mech 1992). In the Upper Midwest states, wolves and white-tailed deer (Odocoileus virginianus) form the dominant predator-prey system (Mech 2000).

To introduce a bit more realism into my model, I varied the annual primary productivity and the annual predation rate on herbivores stochastically (Figure 3a, 3b). Such stochastic variation in primary productivity mimics annual variation due to weather. I modelled the stochastic variation in availability of prey as a variation in wolf carrying capacity, which could result from changes in snow conditions. For parameter values that lead to realistic relative densities of predator and prey, stochastic variation of primary productivity or predation rate, or both, affects all three trophic levels (Figure 3b). If stochasticity in primary productivity is superimposed on a long-term cycle, which might be caused by cyclical variation in climate, for example, the cycle may be seen in the predator and prey populations (Figure 3c). A researcher studying such a predator-prey system might conclude that interactions between predator and prey populations lead to population cycles when the cycles are actually driven from below by variation in primary productivity.

[FIGURE 3 OMITTED]

Consistent with the results of Boyce and Anderson (1999), I found that variation in primary productivity had a profoundly greater effect on year-to-year changes in herbivore populations than did variation in predation efficiency. The vast majority of the year-to-year variation in the model herbivore population could be explained by the productivity of vegetation during the preceding two years, while only about a quarter could be explained by the variation in predation rates. On a year-to-year basis, the model predicts that variation in food affects population dynamics of herbivores more than does variation in predation (Figure 4).

[FIGURE 4 OMITTED]

Stochastic variation of predation rate leads to an approximate 10% decrease in the long-term, average predator populations. This result is expected for predators with a Type 3 functional response. Because predation rate increases in a concave-downward fashion with high and increasing prey populations, variations in prey population size below the average population size lead to larger changes in predation than do variation in prey populations above the average. For the same reason, stochastic variation in primary productivity leads to a similar, though smaller, decrease in long-term, average prey populations. Stochastic variation in both primary productivity and predation, however, leads to a modest (ca. 2%) increase in prey populations, because of the decreased predator populations.

Who controls whom, predators or prey? The model predicts that each controls the other but the control acts at different levels, or on different scales. Ultimately, predator populations are controlled by the populations of their prey, which are, in turn, dependent on the primary productivity of their foods. When primary productivity can support an herbivore population large enough to support carnivores, however, predation then decreases the long-term, average size of the population of its prey (Figure 2). In contrast, on a year-to-year basis, variation in productivity of food causes more variation in herbivore population sizes than does variation in predation rates (Figure 4).

Real predators and prey

Do real populations of predators and prey act as the model predicts, with control acting on two scales? Results from field studies support this hypothesis.

The data for communities with vertebrates in arctic and north temperate ecosystems reviewed by Oksanen and Oksanen (2000) are consistent with control on two scales. Similarly, data for communities with herbivorous or predatory insects and small, vertebrate predators reviewed by Schmitz et al. (2000) are consistent with control on two scales.

McLaren and Peterson (1994) argued from data on wolves, moose and balsam fir (Abies balsamia) on Isle Royale, that predation by wolves controlled the moose population. In addition, moose browsed their major winter food, balsam fir, so heavily that production of fir was limited when the moose populations was high. Reanalysis of their data yields results consistent with model results: production of fir explains a significant amount of the year-to-year variation in the moose population, while year-to-year variation in the wolf population does not.

I studied fishers (Martes pennanti) and their prey, particularly porcupines (Erethizon dorsatum), in Michigan's Upper Peninsula in the 1970s (reviewed by Powell 1993). Fishers had been extirpated in the Upper Peninsula by the 1930s, and by the 1960s, porcupine populations had grown to previously unknown sizes. Fishers were reintroduced in the 1960s, and thus I studied a growing population. Porcupines had two limiting resources: winter food and safe winter dens. During winter, porcupines denned near their food trees. During the years of high population density, porcupines browsed hemlocks and white pines heavily during winters, often stunting tree growth. Where small stands of hemlocks dwindled and died, porcupines abandoned the associated dens. In my study area, most dens were hollow trees or hollow logs. Hollow logs with holes at both ends were acceptable to porcupines before fishers were reintroduced (Brander 1973; Brander and Books 1973; Powell and Brander 1977). As the fisher population grew, however, fishers killed porcupines in log dens with two entrance holes. A safe den near good winter food became in short supply. Fishers caused a significant decrease in the porcupine population (Figure 5).

[FIGURE 5 OMITTED]

Wehausen (1996) also documented a resurgent predator population causing the distinct decease of a formerly healthy prey population, in this case mountain lions (Puma concolor) causing the decrease of a mountain sheep (Ovis canadensis) population.

Thus, field data support the hypothesis that populations of real predators and prey exhibit controls on two scales. Most annual variation in populations of prey appears controlled by annual variation in food supplies. Yet, when primary productivity is high enough to support communities of prey and their predators, predators are capable of limiting prey populations to levels below the levels they would reach without predation. In addition, when prey populations do not support predators, they limit the abundance of their food.

Management implications

Reduction of a predator's population size to increase the size of a prey population is often considered by management agencies (Gasaway et al. 1983). The models presented here indicate that reducing the population of a predator will have little effect on the year-to-year fluctuations in the populations of its prey but may affect the long-term mean population size of prey. Figures 1, 2 and 3a illustrate how, when primary productivity is high enough, predators might limit prey from reaching the population size the food supply can support. Consequently, reducing the population of a predator can be one option to consider, with caution, when faced with a threatened or endangered population of a prey species.

Reduction of a predator population must be considered with caution for several reasons. First, Figure 3c illustrates that cyclic and stochastic fluctuations in primary productivity can lead to large fluctuations in an herbivore population that might be confused with a predator-prey cycle. The models presented here suggest that the year-to-year variation in abundance of herbivores are caused more by fluctuations in food supply than by changes in predation. The same may be true of long-term cycles of prey populations. Second, most predators prey on several prey species, which, in turn, have more than one predator, yet the models presented here deal with single predator and prey populations. The switching of prey by predators may prevent any one predator from limiting any one prey population (Murdoch and Oaten 1975), and the removal of one predator species may simply offer opportunity to another. Finally, long-term control of a predator population when primary productivity varies stochastically may lead to extirpation of the predator, especially when the prey is also subject to control, as through harvest. If long-term harvest of a prey species exceeds 10% of its equilibrium population size, and primary productivity fluctuates as in my models, any control of the predator population could lead to its extirpation within decades.

Literature cited

Boyce, M.S. and E.M. Anderson. 1999. Evaluating the role of carnivores in the greater Yellowstone ecosystem. Pp 265-284 in T.W. Clark, E.P. Curlee, S.C. Minta, and P.M. Kareiva, eds. Carnivores in ecosystems: The Yellowstone experience. Yale University Press, New Haven, Connecticut.

Brander, R.B. 1973. Life history notes on the porcupine in a hardwood-hemlock forest in Upper Michigan. Michigan Academician 5(4):425-433.

Brander, R.B. and D.J. Books. 1973. Return of the fisher. Natural History 82(1):52-57.

Elton, C. 1927. Animal Ecology. Sedgwick & Jackson, London.

Fretwell, S.D. 1977. The regulation of plant communities by food chains exploiting them. Perspectives in Biology and Medicine 20:169-185.

Fretwell, S.D. 1987. Food chain dynamics: The central theory of ecology? Oikos 50:291-301.

Gasaway, W.C., R.O. Stephenson, J.L. Davis, P.E.K. Shepard, and O.E. Burris. 1983. Interrelationships of wolves, prey, and man in interior Alaska. Wildlife Monographs 84:1-50.

Gray, D. 1987. The muskoxen of Polar Bear Pass. Fitshenry & Whiteside, Markham, Ontario.

Hairston, N.G., F.E. Smith, and L.B. Slobodkin. 1960. Community structure, population, and competition. American Naturalist 94:421-425.

Holling, C.S. 1959. The components of predation as revealed by a study of small mammal predation of the European pine sawfly. Canadian Entomology 91:293-320.

McLaren, B.E. and R.O. Peterson. 1994. Wolves, moose and tree rings on Isle Royale. Science 266:1555-1558.

Mech, L.D. 1992. Wolves of the High Arctic. Voyageur Press, Stillwater, Minnesota.

Mech, L.D. 2000. The Wolves of Minnesota: Howl in the Heartland. Voyageur Press. Stillwater, Minnesota.

Murdoch, W.W. 1966. Community structure, population control, and competition B a critique. American Naturalist 100:219-226.

Murdoch, W.W. and A. Oaten. 1975. Predation and population stability. Advances in Ecological Research 9:1-125.

Oksanen, L,S. D. Fretwell, J. Arruda, and P. Niemela. 1981. Exploitation ecosystem in gradients of primary productivity. American Naturalist 118:240-261.

Oksanen, L. and T. Oksanen. 2000. The logic and realism of the hypothesis of exploitation ecosystems. American Naturalist 155:703-723.

Polis, G.A. 1999. Why are parts of the world green? Multiple factors control productivity and the distribution of green biomass. Oikos 86:3-15.

Polis, G.A. and D.R. Strong. 1996. Food web complexity and community dynamics. American Naturalist 147:813-846.

Powell, R.A. 1993. The fisher: Life history, ecology and behavior. 2nd edition. University of Minnesota Press, Minneapolis.

Powell, R.A. and R.B. Brander. 1977. Adaptations of fishers and porcupines to their predator-prey system. Pp 45-53 in R.L. Phillips and C. Jonkel, eds. Proceedings of the 1975 Predator Symposium. Montana Forest & Conservation Experiment Station, University of Montana, Missoula.

Schmitz, O.J., P.A. Hamback and A.P. Beckerman. 2000. Trophic cascades in terrestrial systems: A review of the effects of carnivore removals on plants. American Naturalist 155:141-153.

Strong, D.R. 1992. Are trophic cascades all wet? Differentiation and donor-control in speciose ecosystems. Ecology 73:747-754.

Wehausen, J.D. 1996. Effects of mountain lion predation on bighorn sheep in the Sierra Nevada and Granite Mountains of California. Wildlife Society Bulletin 24:471-479.

White, T.C.R. 1978. The importance of a relative shortage of food in animal ecology. Oecologia 3:71-86.

Roger A. Powell
Department of Zoology, North Carolina State University, Raleigh, North
Carolina 27695-7617