Chapter 14 Exploitation：Predation，Herbivory，Parasitism，and Disease
A moose browses intently on the twigs and buds of a willow barely protruding above the deep snow of midwinter (fig. 14.1). With each mouthful it chews and swallows, the moose reduces the mass of the willows and adds to the growing energy store in its own large and complex stomach, energy stores that the moose will need to make it through one more northern winter. Then, a familiar scent catches the moose's attention and startled, it runs off.
FIGURE 14.1 .This moose exploits the twigs and buds of woody plants for the food it needs to survive the cold northern winter. Eventually, wolves may prey upon the moose to meet their needs for food. Suddenly, the clearing where the moose had been feeding is a blur of bounding forms dashing headlong in the direction the moose has gone--a pack of wolves in pursuit of its own meal. A portion of the pack has already run ahead of the moose and is cutting off its retreat. This time, unlike so many times before, the
old moose will not escape. After a fierce straggle, the moose is down and the wolves settle in to feed.
But the wolves are not the only organisms to benefit from this great quantity of food. Within the intestines of the wolves live several species of parasitic worms that will soon claim their share of the wolves' hard-won feast. The worms will turn some of the energy and structural compounds they absorb into the infective stages of their own kind, which after being shed into the environment may attach themselves to other hosts, who will serve as their unwitting providers.
Some of the strongest links between populations are those between herbivore and plant, between predator and prey, and between parasite or pathogen and host. The conceptual thread that links these diverse interactions between species is that the interaction enhances the fitness of one individual-the predator, the pathogen, etc.--while reducing the fitness of the exploited individual--the prey, host, etc. Because of this common thread we can group these interactions under the heading of exploitation.
Let's consider some of the most common means of exploitation. Herbivores consume live plant material but do not usually kill plants. Predators kill and consume other organisms. Typical predators are animals that feed on other animals--wolves that eat moose, snakes that eat 'mice, etc.
Parasites live on the tissues of their host, often reducing the fitness of the host, but not generally killing it. A parasitoid is an insect whose larva consumes its host
and kills it in the process; parasitoids are functionally equivalent to predators. Pathogens induce disease, a debilitating condition, in their hosts.
As clear as all these definitions may seem, they are fraught with semantic problems. Once again, we are faced with capturing the full richness of nature with a few restrictive definitions. For instance, not all predators are animals, a few are plants, some are fungi, and many are protozoans. When an herbivore kills the plant upon which it feeds, should we call it a predator? If an herbivore does not kill its food plants, would it be better to call it a parasite? What do we do with a parasite that kills its host? Is it then a predator or perhaps a pathogen? The point of these questions is not to argue for more terminology but to argue for fewer, less restrictive terms. As is often the case, we are faced with a continuum of interesting and sometimes bewildering interactions involving millions of organisms. Let's recognize the diversity and continuous variation facing the ecologist, put the restrictive definitions aside for the moment, and recognize what is common to all these interactions: exploitation, that is, one organism makes its living at the expense of another.
CONCEPTS
1. Exploitation weaves populations into a web of relationships that defy easy generalization.
2. Predators, parasites, and pathogens influence the distribution, abundance, and structure of prey and host populations.
3. Predator-prey, host-parasite, and host-pathogen relationships are dynamic.
4. To persist in the face of exploitation, hosts and prey need refuges.
CASE HISTORIES: complex interactions
Exploitation weaves populations into a web of relationships that defy easy generalization.
By conservative estimate the number of species in the biosphere is on the order of 10 million. As huge as this number may seem, the number of exploitative interactions between species is far greater. Why is that? Because every one of those 10 million species is food for a number of other species and is host to a variety of parasites and pathogens. In addition, most feed on other species. Exploitative interactions weave species into a tangled web of relationships. For instance, K. E. Havens (1994) estimated that the approximately 500 known species occupying Lake Okeechobee, Florida, are linked by about 25,000 exploitative interactions--50 times the number of species! Exploitation provides much of the detail in the tapestry we call nature. In this Case History section, we try to capture some of the richness of that tapestry by discussing the natural history of a number of interactions.
Parasites and Pathogens That Manipulate Host Behavior
The most obvious form of exploitation is when one organism consumes part
or all of another. Exploitation, however, can assume far more subtle forms. Some species alter the behavior of the species they exploit.
Parasites That Alter the Behavior of Their Hosts
A number of parasites alter the behavior of their hosts in ways that benefit transmission and reproduction of the parasite. Acanthocephalans, or spiny-headed worms, change the behavior of amphipods, small aquatic crustaceans, in ways that make it more likely that infected amphipods will be eaten by a suitable vertebrate host, especially ducks, beavers, and muskrats.
Uninfected amphipods avoid the light, that is, show negative phototaxis. They spend most of their time near the bottoms of ponds and lakes, away from well-lighted surface waters, where the surface-feeding vertebrate hosts of acanthocephalans spend the majority of their time. In contrast, infected amphipods swim toward light, that is, show positive phototaxis, a behavior that' places them near the pond surface in the path of feeding ducks, beavers, and muskrats (Bethel and Holmes, 1977). Interestingly, amphipod behavior remains unaltered until the acanthocephalan has reached a life stage, called a cystacanth, that is capable of infecting the vertebrate host. If eaten earlier, the acanthocephalan would die without completing its life cycle.
Janice Moore (1983, 1984a, and 1984b) studied a similar parasite-host interaction involving an acanthocephalan, Plagiorhynchus cylindraceus, a terrestrial isopod or pill bug, Armadillidium vulgare, and the European starling,
Sturnus vulgaris. In this interaction, the pill bug serves as an intermediate host for Plagiorhynchus, which completes its life cycle in the starling (fig. 14.2).
FIGURE 14.2 The life cycle of Plagiorhynchus cylindraceus, an intestinal parasite of birds. At the outset of her research, Moore predicted that Plagiorhynchus would alter the behavior of Armadillidium. She based this prediction on several observations. One was the relative frequency of infection of Armadillidium and starlings by PIagiorhynchus. Field studies had demonstrated that even where Plagiorhynchus infects only 1% of the Armadillidium population, over 40% of the starlings in the area were infected.
Some factor was enhancing rates of transmission to the starlings, and Moore predicted that it was altered host behavior. Moore thought that the size of
Plagiorhynchus might also be a factor. At maturity, the cystacanth stage of Plagiorhynchus grows to about 3 mm, a substantial fraction of the internal environment of an 8mm pill bug!
Moore brought Armadillidium into the laboratory and established two laboratory populations: an uninfected control group and an infected experimental group. She infected half the laboratory populations of Armadillidium by feeding them pieces of carrot coated with Plagiorhynchus eggs, while keeping the remaining laboratory populations free of Plagiorhynchus. After 3 months, the Plagiorhynchus in the infected populations matured to the cystacanth stage. At this point Moore mixed the infected and uninfected populations.
Because Plagiorhynchus does not alter the outward appearance of Armadillidium, Moore could not determine whether an Armadillidium was infected or not until she dissected it at the completion of an experiment. Consequently, all the behavioral experiments were conducted \"blind\"; that is, without the possibility of observer bias due to prior knowledge of the identity of experimental and control animals.
Moore found that Plagiorhynchus alters the behavior of Armadillidium in several ways. Infected Armadillidium spend less time in sheltered areas and more time in low-humidity environments and on light-colored substrates. These changes in behavior would increase the time an Armadillidium spends in the open, where it could be easily seen by a bird. In summary, infected Armadillidium behave in a way that is likely to increase the probability that they will be discovered by
foraging birds.
In laboratory experiments Moore demonstrated that cap five starlings consistently captured Armadillidium from light rather than dark substrates. She provided caged starlings access to a mixture of 10 infected and 10 uninfected Armadillidium, which wandered freely across the bottom of the cage, half of which was covered by black sand and half by white sand. Under these conditions, starlings ate 72% of the infected Armadillidium but only 44% of the uninfected individuals (fig. 14.3). The starlings took isopods mainly from the surface of white sand, so it seems that the tendency of Armadillidium to seek out light substrates does make them more vulnerable to predation by birds.
FIGURE 14.3.Starling predation on uninfected and infected Armadiltidium
vulgate (data from Moore 1984b). A critical step in this research was to determine whether the changed behavior of infected Armadillidium translates into their being eaten more frequently by wild birds. Moore collected the arthropods that starlings feed to their nestlings and from these collections estimated the rate at which they delivered Armadillidium--about one every 10 hours. Using this delivery rate and the proportion of the Armadillidium population infected by Plagiorhynchus (about 0.4%), she was able to predict the expected rate of infection among starling nestlings if the adults capture Armadillidium at random from the natural population. The proportion of infected nestlings was 32%, about twice the rate of infection predicted if starlings fed randomly on the Armadillidium population. These results support Moore's hypothesis that the altered behavior of infected Armadillidium increases their probability of being eaten by starlings.
Moore emphasized that Plagiorhynchus does not just alter Armadillidium's behavior but alters its behavior in a particular way--in a way that increases the rate at which the final host of the parasite, starlings, is infected.
A Plant Pathogen That Mimics Flowers
Every year the slopes of the southern Rocky Mountains are decorated with the colorful blossoms of wildflowers. Some of these wildflowers, however, are not quite what they seem.
One bright yellow and sweet-smelling \"blossom\" is actually produced by a pathogenic fungus that manipulates the growth of its host plant. This pathogen belongs to a group of fungi called rusts because of their rust-colored spores that appear on the surface of the infected host plant. This particular rust is Puccinia monoica, and its hosts are mustard plants in the genus Arabis. Arabis spp. are herbaceous plants that spend anywhere from a few months to several years as a rosette, a low-growing growth form with a high density of leaves. During the rosette stage Arabis invests heavily in root development and storage of energy in the roots. At the end of the rosette stage, Arabis grows tall quickly, a process called bolting, and flowers (fig. 14.4a). Once pollinated, the flowers form seeds that mature, completing the life cycle of Arabis.
FIGURE 14.4 The effects of the fungus Puccinia monoica on morphology: (a) a normally developed mustard plant, Arabis hoelbollii: (b) a pseudofiower formed by A. Hoelbollii infected by Puccinia
However, Puccinia completely alters the life history of Arabis. It attacks the rosette stage, manipulating its development to produce a growth form that promotes reproduction by the fungus and usually kills the plant. Puccinia infects the rosettes of Arabis in late summer and then invades the meristematic tissue, the actively dividing tissue responsible for plant growth, during the following winter. As it invades the meristematic tissue, Puccinia manipulate future development by the rosette. Infected rosettes elongate rapidly the following spring, maintain a high density of leaves along their entire length, and are topped by a cluster of bright yellow leaves. This cluster of yellow leaves looks very much like the flowers of the buttercup, Ranunculus spp. (fig. 14.4b).
The bright yellow pseudo flowers of infected rosettes are produced by various fungal structures, including spermogonia containing spermatia (fungal reproductive cells), sexually receptive fungal hyphae, and secretions of sticky, sugar-containing spermatial fluid. Most rusts require out-crossing for sexual reproduction, which is accomplished by insects transferring spermatia from one fungus (thallus) to the receptive hyphae of another thallus. Barbara Roy (1993) found that the combination of yellow' color and sugary fluid attracts a wide variety of flower-visiting insects, including butterflies, bees, and flies (fig. 14.5). Flies, the most common visitor to pseudo flowers at her Colorado study site, have been demonstrated to be effective carriers of rust spermatia.
FIGURE 14.5 The pseudoflowers formed by Arabis hoelbollii infected by the fungus Puccinia monoica, such as the one shown here, are attractive to a wide variety of pollinating insects, such as this Polygonia butterfly. Roy's studies demonstrated that Puccinia truncates the life cycle of Arabis and in the process generally kills the host plant. The Arabis that survive attack by Puccinia may go on to flower but none form seeds. Thus, destruction by Puccinia is total.
The Entangling of Exploitation with Competition
We often arrange our thoughts about nature in neat categories like the chapters of this book. In chapter 13 we discussed competition, we now discuss exploitation, and in the next chapter we examine mutualism. Nature itself is not so neatly arranged, nor are natural phenomena so easily isolated. One process is usually connected to several others. The distinction between exploitation and competition is blurred when competitors eat each other.
Predation, Parasitism, and Competition in Populations of Tribolium
Thomas Park and his colleagues (Park 1948, Park et al. 1965) uncovered one of the very first examples of competitors eating each other during their work on competition among flour beetles. As we saw in chapter 13, the outcome of competition between Tribolium castaneum and T. confusum depended upon temperature and moisture. It turns out that the presence or absence of a protozoan parasite of Tribolium, Adelina tribolii, also influences the competitive balance between flour beetle species. The effects of this parasite are also entangled with predation among the flour beetles and cannibalism, which we might think of as a form of intraspecific exploitation.
Park showed that various strains of T. castaneum and T. confusum differ in their rates of cannibalism. Of the two species, T. castaneum is the most cannibalistic but it preys on the eggs of T. confusum at an even higher rate than it cannibalizes its own eggs. In the light of its predatory behavior, it's not surprising that T. castaneum eliminated T. confusum in 84% of 76 competition experiments spanning a period of about 10 years. This predatory strategy works best, however,
in the absence of Adelina.
Adelina invades the cells of its host and lives its life as an intracellular parasite. Beetles become infected when they ingest the oocysts of this parasite, either as they feed on flour or as they consume infected larvae, pupae, or adult beetles. Once in the gut of the beetle, the oocyst eggs rupture, liberating a life stage of Adelina called a sporozoite. The sporozoites penetrate the beetle's gut and enter the body cavity, or haemoc0el. Once in the haemvocoel, the sporozoites invade various cells, where they reproduce asexually and produce a second life stage called the merozoite. The motile merozoites invade yet other host cells eventually, producing male and female sex cells that combine to form zygotes. The zygotes eventually give rise to new sporozoites, which are encased in oocysts. Ingestion of these oocysts by another beetle completes the life
cycle of Adelina.
Several biologists before Park had noted that Adelina caused \"sickness\" and death among Tribolium populations. It was Park, however, who demonstrated that Adelina reduces the density of Tribolium populations and can alter the outcome of competition between T. confusum and T. castaneum. Adelina strongly reduces the population density of T. castaneum populations but has little effect on T. confusum populations. In the absence of the parasite, T. castaneum won 12 of 18 competitive contests against T. confusum. When the parasite was included, however, T. confusum won 11 of 15 contests (fig. 14.6). In other words, parasitism completely reverses the outcome of competitive interactions between the two species. From
insects to African lions, interference escalated to the point of predation appears to be a common occurrence among competitors. However, Park's experiments with Tribolium indicate that parasites may make the outcome of a predaceous competitive strategy difficult to predict.
FIGURE 14.6 The influence of the protozoan parasite Adelina tribolii on competition between the flour beetles Tribolium castaneum and T. confusum (data from Park 1948). CASE HISTORIES:
exploitation and abundance
Predators, parasites, and pathogens influence the distribution, abundance, and structure of prey and host populations.
One of the main reasons ecologists are interested in exploitative interactions between species is that these interactions have the potential to influence prey and host populations. A rapidly growing pool of studies suggest that predators, parasites, and pathogens substantially affect the populations they exploit.
An Herbivorous Stream Insect and Its Algal Food
Gary Lamberti and Vincent Resh (1983) studied the influence of an herbivorous stream insect on the algal and bacterial populations upon which it feeds. The herbivorous insect was the larval stage of the caddisfly (order Trichoptera) Helicopsyche borealis. This insect inhabits streams across most of North America and is most notable for the type of portable shelter it builds as a larva. The larvae cement sand grains together to form a helical portable home that looks just like a small snail shell. In fact, the species was originally described as a freshwater snail. Larval Helicopsyche graze on the algae and bacteria growing on the exposed surfaces of submerged stones. This feeding habit requires that Helicopsyche spend considerable time out in the open, where it would be far more vulnerable to predators were it not for its case.
Lamberti and Resh found that larval Helicopsyche grow and develop
through the summer and fall, attaining densities of over 4,000 individuals per square meter in Big Sulfur Creek, California. At this density, they make up about 25% of the total biomass of benthic animals. A consumer that reaches such high population densities clearly has the potential to reduce the density of its food supply. Lamberti and Resh got an indication of the potential of Helicopsyche to influence its food supply in a preliminary experiment. In this first experiment they placed unglazed ceramic tiles (15.2cm by 7.6cm) on the bottom of the creek and followed colonization of these artificial substrates by algae and Helicopsyche over a period of 7 weeks.
Algae rapidly colonized the tiles, reaching peak density 2 weeks after the tiles were placed in Big Sulphur Creek. The Helicopsyche population reached its highest density 1 week later. Algal biomass decreased from week 2 to week 5 of the study and then rose again during the last 2 weeks, as Helicopsyche numbers declined. These results (fig. 14.7) suggest that the caddisfly larvae depleted their food supply. However, Lamberti and Resh could not be certain. Why is that? First, there are many other benthic invertebrates living in Big Sulphur Creek, some of which might be depleting the algal populations. Second, physical factors could have changed during the 7 weeks of the study, and these changes could have produced the fluctuations in both algal and Helicopsyche populations. This initial experiment provided valuable indications but was not a definitive test.
FIGURE 14.7 Biomass of algae and numbers of the grazing caddisfiy Helicopsyche borealis (data from Lamberti and Resh 1983). In a follow-up study, the researchers used an exclusion experiment to test for the effect of Helicopsyche on its food supply. They placed unglazed ceramic tiles in two 3-by-6 grids of 18 tiles each. One grid was placed directly on the stream bottom, while the other was placed on a metal plate supported by an upside-down J-shaped metal bar. This arrangement, which raised the tiles 15cm above the bottom but still 35 cm below the stream surface, allowed colonization of tiles by algae and most invertebrates while preventing colonization by Helicopsyche. Helicopsyche could not colonize the tiles because their heavy snail-shaped case confines them to the stream bottom. To reach the tiles, Helicopsyche would have to crawl up the J-shaped support bar, out of the water, and then back down, while most other invertebrates could colonize by either drifting downstream with the current or by swimming to the raised tiles. Lamberti
and Resh coated the above-water parts of the bar with an adhesive to prevent adult Helicopsyche from crawling down to the tiles to deposit their eggs. As figure 14.8 shows, the experimental arrangement excluded Helicopsyche while allowing large numbers of other invertebrates to colonize the raised tiles. Such selective manipulations of natural populations are not easy to attain.
FIGURE 14.8 The influence of elevating tiles on colonization by Helicopsyche borealis and other benthic invertebrates (data from Lamberti and Resh 1983). The results of this experiment clearly show that Helicopsyche reduces the abundance of its food supply. Figure 14.9 shows that the tiles without Helicopsyche supported higher abundances of both algae and bacteria. The large effect of Helicopsyche on its food supply is apparent from paired photos of the experimental and control tiles at the beginning and the end of the experiment (fig. 14.10).
FIGURE 14.9 Influence of excluding Helicopsyche borealis on abundance of bacteria and algae (data from Lamberti and Resh 1983). FIGURE 14.10 Effects of excluding Helicopsyche borealis on benthic algal biomass: (a) two sets of tiles at the beginning of experiment,' exclusion tiles in foreground,' (b) same tiles 5 weeks into the experiment.
The influence of exploitation on populations is often best seen when populations are released from exploitation. This is what we saw when Helicopsyche were excluded from experimental habitats. Similar responses occur when exploited populations are introduced into new environments free of significant predators, herbivores, or pathogens.
An Introduced Cactus and an Herbivorous Moth
In the mid-1800s a prickly pear cactus, Opuntia stricta, was introduced to Australia as an ornamental plant and was maintained in gardens for some time. Then, near the turn of the century, Opuntia escaped cultivation and became established in the wild, where it found ideal physical conditions for its growth and reproduction. The plant spread quickly, covering over 20 million ha by the late 1920s. By 1930, it appeared that the spread of Opuntia had begun to slow, but by that time it covered over 24 million ha.
Nowhere in its natural range in North America does the cactus attain the densities reached in Australia. How did this cactus reach such high densities outside its native range?
Biologists trying to control Opuntia suggested that its rapid growth in Australia resulted from the absence of \"natural enemies\"--the herbivores, parasites, and pathogens that attack the cactus in its native environment. The Australian government sent biologists to the native environment to search for biological allies in their fight to control the spread of the cactus.
Biologists eventually identified several insect species that attack the cactus and that might be useful in its control. Of these species, the most effective proved to be a moth eventually described and given the apt name of Cactoblastis cactorum. The appropriateness of the name of this species is apparent if you compare the photos showing Opuntia densities before and after introduction of Cactoblastis (fig. 14.11).
FIGURE 14.11 Collapse of an Opuntia stricta population after introduction of the herbivorous moth Cactoblastis cactorum. Female Cactoblastis deposit eggs on the pads of Opuntia in groups of 70 to 90. When the caterpillars hatch, they burrow into the cactus pad and then spend
their larval lives dining on the inside of the pad. As the caterpillars burrow, they introduce a diversity of fungi and bacteria that also attack the internal tissues. As a consequence of the combined attack by caterpillars and microbes, cactus tissues are quickly reduced to the consistency of mush and whole thickets of succulent Opuntia collapse. The herbivore Cactoblastis was so effective at reducing the population of Opuntia because it was also a dispersal agent for a variety of pathogens.
Following the release of Cactoblastis, the population of cactus collapsed. It took the moth only 2 years to reduce the population of Opuntia by three orders of magnitude, from about 12,000 individuals per hectare to 27 per hectare. During this same time, the area covered by Opuntia fell from 24 million ha to a few thousand.
Eventually, these two populations reached a kind of cyclic equilibrium (Dodd 1940, 1959). The density of the cactus population is low and highly dispersed and therefore not easily found by the moth. Under these conditions local populations of Opuntia can grow. Eventually, however, Cactoblastis finds these local populations and destroys them. Meanwhile there will be an outbreak in another area where Opuntia has temporarily escaped from Cactoblastis. Eventually these are found and also destroyed. Under these conditions, which are more like the conditions under which most species live in their natural environment, it is not obvious what controls the population density of Opuntia. As we shall see, this kind of effective and complex exploitation occurs in other populations.
A Pathogenic Parasite, a Predator, and Its Prey
One of the most impressive aspects of the control of the cactus by the moth Cactoblastis is that the events took place over such a large area. One of the great challenges of ecology is to work at large scales. Ecologists rarely have the opportunity to conduct large-scale experiments, however, nature sometimes provides such opportunities. One such opportunity arose in Sweden when a pathogen severely reduced the population of red foxes, Vulpes vulpes.
Erik Lindstr6m and his colleagues (1994) at the Grims6 Wildlife Research Station in Orebro county reported that mange mites, Sarcoptes scabiei, were first found infesting red foxes in north-central Sweden in 1975. The researchers studied the spread of mange because mange mites are a serious external parasite of foxes that causes hair loss, skin deterioration, and death. Within a decade of its arrival in Sweden, mange had spread over the entire country (fig. 14.12). As it spread, mange reduced the population of red foxes in Sweden by over 70%.
FIGURE 14.12 The spread of mange in red foxes across Sweden from 1975 to 1984 (data from Lindstr6m et al. 1994). As wildlife ecologists, LindstrOm's research team was keenly interested in how the prey of red foxes would respond. Would they find evidence of population control by this predator? From 1972 to 1993, the research team studied several, prey populations as well as red fox populations. They used many sources of information and conducted their studies at local, regional, and national spatial scales.
The results of the study were clear. Red foxes in Sweden reduce the
populations of their prey, including hares, grouse, and roe deer fawns. Figure 14.13 shows the relationship between numbers of red foxes and mountain hares, Lepus timidus, in Sweden. Following the reduction in the red fox population, the number of mountain hares increased two to four times. This is an especially thorough and convincing demonstration of the influence of a terrestrial vertebrate predator on its prey populations. The study also suggests that red foxes may have a significant influence on the cyclic abundance of some of their prey species. The dynamics of prey populations has been the subject of research by ecologists for some time. Studies of predation have been central to this research.
FIGURE 14.13 The numbers of foxes and mountain hares in five counties in Sweden estimated from hunters' harvest records (data from LindstrOim et al. 1994) CASE HISTORIES:
Dynamics
Predator-prey, host-parasite, and host-pathogen relationships are dynamic.
In the last section we saw how some predators, parasites, and pathogens affect the populations they exploit. The picture that emerges from these studies is that the biology of exploitation is complex. As complex as this emerging picture of exploitation may be, it belies an even deeper underlying complexity.
In this section, we add another level of complexity as we take up the topic of temporal dynamics. Populations of a wide variety of predators and prey are not static but cycle in abundance over periods of days to decades.
Cycles of Abundance in Snowshoe Hares and Their Predators
Population cycles are well documented for a wide variety of animals living at high latitudes, including lemmings, voles, muskrats, red fox, arctic fox, ruffed grouse, and porcupines. We have already seen in chapter 10 how periodic outbreaks of voles lead to local increases in the abundance of avian predators due to numerical responses by owls and hawks (Korpimaiki and Norrdahl 1991).
One of the best-studied cases of animal population cycles is that of the snowshoe hare, Lepus americanus, and the lynx, Lynx canadensis, one of the snowshoe ham's chief predators. The population cycles of these two species are especially well documented because the Hudson Bay Company kept trapping
records during most of the eighteenth, nineteenth, and twentieth centuries. Drawing on this unique historical record ecologists were able to estimate the relative abundances of Canadian lynx and snowshoe hare over a period of about 200 years. That record, shown in figure 14.14, demonstrates a remarkable match in the cycles of the two populations.
FIGURE 14.14 Historical fluctuations in lynx and snowshoe hare populations based on the number of pelts purchased by the Hudson Bay Company (data from MacLulich 1937). By the 1950s several hypotheses had been proposed to explain these and other cycles among northern populations. Charles Elton (1924) proposed that cycles of abundance in snowshoe hare and lynx populations are driven by variation in amount of solar radiation as a consequence of sunspot cycles. He proposed that variation in intensity of solar radiation may directly affect snowshoe hares and their food supply and that lynx populations, in turn, respond to the changing abundance of the snowshoe hare, their main prey.
The sunspot hypothesis was rejected by D. MacLulich (1937) and P. Moran (1949), who showed that sunspot cycles do not match snowshoe hare population cycles. The second group of hypotheses, which Lloyd Keith (1963) referred to as \"overpopulation theories,\" suggested that periods of high population growth are followed by (1) decimation by disease and parasitism, (2) physiological stress at high densities leading to increased mortality as a consequence of nervous disorders, and (3) starvation due to reduced quantity and quality of food at high population densities. An alternative to the overpopulation hypothesis was that cycles like that of the snowshoe hare are driven by predators. According to this hypothesis, predators increase in number in response to increasing prey availability and then eventually reduce prey populations.
Keith observed that none of these hypotheses completely accounts for population cycles in snowshoe hare and other northern populations. He went on to say that \"the 10-year cycle is not likely to become better understood by further theorizing. Clearly the present need is for comprehensive 10ng-term investigations by a diversified team of specialists.\" Heeding his own advice, Keith organized such studies. After three decades of research by his team and several other groups 'in North America and Europe, we now have a reasonable picture of the roles played by predators and food supply in producing population cycles in the far north.
The Role of Food Supply
Snowshoe hares live in the boreal forests of North America. As we saw, the
boreal forest is dominated by a variety of conifers such as spruce, Picea spp., jackpines, Pinus banksiana, and tamarack, Larix laricina, and deciduous trees Such as balsam popular, Populus balsamifera, aspen, Populus tremuloides, and paper birch, Betula papyrifera. Within the boreal forest, snowshoe hares associate with dense growths of understory shrubs, which provide both cover and winter food, the most critical portion of the snowshoe hare's food supply.
Snowshoe hares have the potential to reduce the quantity and quality of their food supply. The hares live up to the legendary reproductive capacity of rabbits and hares.
Estimated geometric rate of increase, X (see chapter 10), during the growth phase of a hare population cycle can average as high as 2.0. In other words, snowshoe hare populations can double in size each generation. Keith and his colleagues (1984) have observed snowshoe hare population densities of up to 1,100 to 2,300 per square kilometer. However, local densities are highly dynamic. Keith cites 100-fold fluctuations in snowshoe hare densities in some areas and states that 10- to 30-fold fluctuations are common. Similar densities are sometimes observed in populations of the mountain hare, Lepus timidus, which shows pronounced population cycles across the Eurasian taiga (Keith 1983) and which destroys considerable vegetation at high densities.
Snowshoe hares spend the long northern winter (6 to 8 months) browsing on the buds and small stems of shrubs such as rose, Rosa spp., and willow, Salix spp. Where deep snow provides access, snowshoe hares browse on the saplings of
trees such as spruce and aspen. The most nutritious portions of these shrubs and trees are the small stems (< 4~5 mm diameter). Over the winter, each hare requires about 300 g of these stems each day. In some areas, however, snowshoe hares have been observed to remove over 1,500 g of food biomass per day, possibly wasting a great deal of potential food in the process. Feeding at these rates, one population of snowshoe hares reduced food biomass from 530 kg per hectare in late November to 160 kg per hectare by late March. Many ecologists have demonstrated food shortage during winters of peak snowshoe hare density.
Snowshoe hares also influence the quality of their food supply. Feeding by snowshoe hares induces chemical defenses in their food plants, defenses like those we discussed in chapter 6. Shoots produced after substantial browsing contain elevated concentrations of terpene and phenolic resins, defensive chemicals that repel hungry hares. Elevated concentrations of plant defensive chemicals can persist for up to 2 years after browsing by hares. The effect of these induced chemical defenses reduces usable food supplies during the population decline. Some ecologists suggest that plant defensive responses may be the \"timer\" that produces 10-year population cycles in snowshoe hares.
The Role of Predators
The long historical record of lynx population cycles may have distracted ecologists from the fact that lynxes are only one of several predators that feed on snowshoe hares. Other major predators of snowshoe hares include goshawks, Accipiter gentilis, great homed owls, Bubo virginianus, mink, Mustela vison,
long-tailed weasels, Mustela frenata, red foxes, Vulpes vuIpes, and coyotes, Canis latrans. Populations of these predators are known to cycle synchronously with snowshoe hare populations. Though the lynx is considered to be a specialist on snowshoe hares, the diet of a generalist predator such as the coyote may also be dominated by snowshoe hares. This is particularly tree when snowshoe hare populations are at peak density. A. Todd and L. Keith (1983) report that snowshoe hares made up 67% of the coyote diets in central Alberta, Canada. Ecologists have estimated that predation can account for 60% to 90% of snowshoe hare mortality during peak densities.
In summary, several decades of research provide evidence that both predation and food can make substantial contributions to snowshoe hare population cycles (Haukioja et al. 1983, Keith 1983, Keith et al. 1984). The food availability and predation hypotheses are not mutually exclusive alternatives but rather are complementary. As hare populations increase, they reduce the quantity and quality of their food supply. Reduced food availability, which leads to starvation and weight loss, would itself likely produce population decline. This potential decline is ensured and accelerated by high rates of mortality due to predation. As hare population density is reduced, predator populations decline in mm, plant populations recover, and the stage is set for another increase in the hare population. Thus the cycle repeats itself across the northern forests.
Population Cycles in Mathematical and Laboratory Models
Now let's shift our focus from population cycles in the vast world of the boreal
forest to population cycles in mathematical models and controlled laboratory conditions. Mathematical and laboratory models offer population ecologists the opportunity to manipulate variables that they cannot control in the field. Our question here is whether predator-prey or parasite-host cycles can be produced in mathematical and laboratory models without the complications introduced by factors such as the effects of the prey on its food supply and uncontrolled weather cycles. In other words, can the interactions among exploited populations themselves generate population cycles of the type observed in snowshoe hares? The answer to this question is a qualified yes.
Mathematical Models
The first ecologists to model predator-prey interactions mathematically were Alfred Lotka (1925) and Vito Volterra (1926). Both researchers built their models based on observations of interactions among natural populations. Lotka was impressed by the reciprocal oscillations of populations of moth and butterfly larvae and the parasitoids that attack them. Volterra was inspired by the response of marine fish populations to cessation of fishing during World War I. Volterra observed that the response of fish populations was uneven. Predaceous fish, particularly sharks,-increased in abundance, while the populations upon which they fed decreased. This reciprocal change in numbers suggested that predators have the potential to reduce the abundance of their prey. In this single observation, Volterra somehow saw the potential for predator-prey population cycles and suggested that similar cycles should occur in parasite-host and pathogen-host systems, including those in which humans are involved. With these observations in
mind, Lotka and Volterra then set out to build mathematical models that would produce the cycles that they thought occurred in nature.
The Lotka-Volterra predator-prey equations demonstrated that very simple models will produce cycling of predator and prey populations. The basic Lotka-Volterra model assumes that the host population grows at an exponential rate and that host population size is limited by its parasites, pathogens, or predators:
dNh/dt = rhNh - pNhNp
represents the host population size, and rhNh represents exponential growth by the host population. In the Lotka-Volterra model, exponential growth by the host population is opposed by deaths due to parasitism or predation, which is represented by -pNhNp, where p is the rate of parasitism or predation, Nh is again the number of hosts, and Np is the number of parasites or predators.
On the other side of the parasitoid-host system, the Lotka-Volterra model assumes that the rate of growth by the predator or parasite population is determined by the rate at which it converts the hosts it consumes into offspring (new predators or parasitoids) minus the mortality rate of the parasitoid population:
dNp/ dt =CpNhNp - dpNv
Here again Nh and , Vp are the numbers of hosts and predators or parasites, respectively. The rate at which the predators or parasites convert hosts into offspring is CpNhNp, which is the rate at which the exploiters destroy hosts, pNhNp, times a conversion factor, c, the rate at which hosts are converted to parasite or predator offspring. In the Lotka-Volterra equation, the growth rate of the predator population is opposed by predator deaths, dpNp. Notice that in these equations the only variables are Nh and Np. All the other terms in the Lotka-Volterra model, p, c, dp, and rh, are constants. The Lotka-Volterra predator-prey model is summarized in figure 14.15.
FIGURE 14.15 Anatorny of the Lotka-Volterra equatlans for predator-prey or parasite-host population growth. Now let's reflect on the behavior of this model. Because the host population grows at an exponential rate, its population growth accelerates with increasing population size. However, this tendency to grow faster and faster with increasing Nh is opposed by exploitation. As Nh increases the rate of exploitation, pNhNp, also increases. Consequently, in the Lotka-Volterra model, reproduction by the host is translated immediately into destruction of hosts by the predator. In addition, increased parasitism or predation, pNhNp, is translated directly and immediately into more parasites or predators by CpNhNp. Increased numbers of predators and parasites increase the rate of exploitation since increasing Np increases pNhNp. Growth of the parasite or predator population eventually reduces the host population, which in turn leads to declines in the predator or parasite population. So, like the host, exploiter success carries the seeds of its own destruction.
These reciprocal effects of host and exploiter produce oscillations in the two populations, which we can represent in two ways. In figure 14.16a, population oscillations are presented as we looked at them in snowshoe hare and lynx.
FIGURE 14.16 A graphical view of the Lotkn - Votterra predator-prey model (data from Gause 1934). Populations(see fig.14.13)，while figure 14.16b gives an alter-native representation．The time axis has been eliminated and the two remaining axes represent the numbers Of predator or parasites and hosts．When we Plot population data in this way we see that the Lotka-Volterra model produces oscillations in exploiter and host populations that follow an elliptical path whose size depends upon the initial sizes of host and exploiter populations．Whatever the ellipse size，however, the host and exploiter populations just go round and round on the same path forever.
The prediction of eternal oscillations on a very narrowly defined path is obviously unrealistic. Another unrealistic assumption is that neither the host nor the exploiter populations are subject to carrying capacities. Another is that changes in either population are instantaneously translated into responses in the other population. Despite these unrealistic assumptions, Lotka and Volterra made valuable contributions to our understanding of host-parasite and predator-prey systems. They showed that simple models with a minimum of assumptions produce reciprocal cycles in populations of host and parasite and predator and prey analogous to those that biologists had observed in natural populations. They demonstrated that exploitative interactions themselves can, in theory, produce population cycles without any influences from an outside force such as climatic variation.
Laboratory Models
One of the most successful attempts to produce Lotka-Volterra-type population cycles in the laboratory was that of Syunro Utida (1957) of Kyoto University, Japan. Utida studied interactions between the adzuki bean weevil, Calloso-bruchus chinensis, and a hymenopteran parasitoid wasp, Heterospilus prosopidis, which attacks the bean weevil. Adult weevils lay their eggs on adzuki beans, Paseolus angularis, and upon hatching the larvae feed on the beans until they metamorphose into pupae. When they emerge from the pupal stage, the adult weevils mate and seek out new beans on which to lay their eggs. The entire life cycle, from egg to egg, takes approximately 20 days. While the weevil works at completing its life cycle, the parasitoid wasp searches for weevil larvae and pupae,
where they lay their eggs. The larvae of the wasps feed on the larvae and pupae of the weevils and in the process, kill them. Though the details of their behavior differ, the wasps are predators of the weevils, no less than are lynx predators of snowshoe hares.
Utida's experimental populations lived in petri dishes 1.8 cm tall by 8.5 cm in diameter where temperature was maintained at a constant 30℃ and relative humidity at 75%,Within the petri dishes Utida placed 10 g of adzuki beans with a water content of 15% and added a mixture of adult adzuki bean weevils and parasitoid wasps: either weevils and 8 wasps (population A), 8 weevils and 8 wasps (population C), or 512 weevils and 128 wasps (population E). Every 10 days l0 g of fresh beans were added, and the leavings of the old beans were placed in another dish. Any beetles moved with the spent food were recorded over a period of 20 days.
Utida followed population C for 47 beetle generations, approximately 940 days, after which a mistake in handling killed the population. He followed population E for 82 generations, approximately 1,0 days, after which the weevils died out. Population A was followed the longest, 112 generations, over 6 years, after which the population was accidentally destroyed. It was only by following the beetle and wasp populations for so many generations that Utida was able to see the pattern we look at now.
All three of the experimental populations showed the same cyclic behavior (fig.14.17). For several generations Utida observed reciprocal fluctuations in his
beetle and parasitoid populations that look very similar to those we saw for lynx and hare populations (see fig. 14.13). After an initial phase of high-magnitude oscillations the population cycles were decreased in amplitude, remained in a situation of low-amplitude fluctuations for some time, and then increased in amplitude once again. In population A, high-amplitude cycles continued for the first 20 generations, dampened out until about generation 30, and then resumed high-amplitude cycling until about generation , when the oscillations dampened out once again. FIGURE 14.17 Laboratory populations of a host. the adzuki bean weevil and a parasitoid wasp (data from Utida 1957).
Utida's results are analogous to the patterns of reciprocal fluctuation seen in the Lotka-Volterra model along with some behavior not predicted by the mathematical model. However, despite these differences, like the Lotka-Volterra model, Utida's laboratory model shows that parasitoid-host populations can show reciprocal oscillations without significant temporal variation in the physical environment.
G..Gause (1935) produced similar results when he studied a laboratory population of Paramecium aurelia preying upon yeast. He followed the populations through three cycles. which took only 20 days. Though Gause's experiments were much shorter than Utida's, they also produced oscillations like those predicted by the Lotka-Volterra model.
Utida's and Gause's successes make work with labor story models look far easier than it is. Most attempts to produce Lotka-Volterra-type oscillations in
laboratory populations have failed. Most laboratory experiments have led to extinction of the predator or prey population in a fairly short period of time. To sustain oscillations even for a short period researchers have generally had to provide the prey with refuges of some sort, which indicates another generalization about natural predator-prey systems. It appears that to persist in the face of exploitation by predators, parasites, and pathogens, hosts and prey need refuges.
CASE HISTORIES:
refuges
To persist in the face of exploitation, hosts and prey need refuges.
This section is about refuges, situations in which members of an exploited population have some protection from predators and parasites. When we think of refuges, we generally think of an inaccessible place. There are, however, many other kinds of refuges. Many have nothing to do with places and most do not provide complete security--just enough.
Refuges and Host Persistence in Laboratory and Mathematical Models
Gause's success at producing cycles in populations of Paramecium aurelia and its prey, Saccharomyces exiguus, gives no hint of the difficulties he experienced in his earlier attempts. Gause's first attempts to produce Lotka-Volterra population cycles involved Paramecium caudatum and one of its predators, another aquatic
protozoan called Didinium nasutum. if Ganse grew these organisms in a simple laboratory microcosm, Didinium quickly consumed all the Paramecium (fig. 14.18). The absence of a refuge for the prey led eventually to extinction of the predator and prey populations, Gause responded by putting some sediment on the bottom of his microcosm to provide a refuge for Paramecium. in this case once Didinium had eaten all of the Paramecium not hiding in bottom sediments, it starved and became extinct, Following the disappearance of Didinium and the removal of predation pressure, the population of Paramecium quickly increased. Here, a simple refuge for the prey population led to extinction of the predator.
FIGURE 14.18 Refuges and the persistence of predator-prey oscillation in laboratory populations of prey [Paramecium aurelia) and predators (Didinium nasutum) (data from Gause 1934)
Are these experimental requirements entirely artificial or do they correspond with anything we already know about natural populations? Actually, Gause's experimental results match many of our observations in natural populations. In chapter 9, we saw that on larger scales populations show clumped distributions. Most species are much more common in some parts of their range than in others. Then in chapter 10, we saw how dispersal is an important contributor to population dynamics and that some local populations are maintained entirely by dispersal from other areas. Some biologists have combined observations such as these to hypothesize the existence of population sources and population sinks---local populations maintained by immigration from source populations. In Gause's experiment, the laboratory cultures were population hot spots, or sources, while the microcosms where predator and prey interacted were population sinks. The requirements of Gause's experiment are consistent with the results of later experiments.
C. Huffaker (1958) set out to test whether Gause's results could be reproduced in a situation in which the predator and prey are responsible for their own immigration and emigration among patches of suitable habitat. Huffaker chose the six-spotted mite, Eotetranychus sexmaculatus, a mite that feeds on oranges, as the prey and the predatory mite Typhlodromus occidentalis which attacks E. sexmaculatus, as the predator. Huffaker's experimental setups, or \"universes\" as he called them, consisted of various arrangements of oranges, or combinations of oranges and rubber balls, separated by partial barriers to mite dispersal consisting of discontinuous strips of petroleum jelly.
An important point of natural history is that the predatory mite had to crawl in order to disperse from one orange another, while the herbivorous mill can disperse either by crawling or by \"ballooning,\" a means of aerial dispersal. A mite balloons by spinning a strand of silk that can catch wind currents. Huffaker gave the herbivorous mite the chance to balloon by providing small wooden posts that could serve as launching pads and by having a fan circulate air across his experimental setup.
While Huffaker's simpler experimental universes did not produce predator-prey oscillations, his most elaborate setup of 120 oranges did. These oscillations spanned several months (fig. 14.19). Huffaker observed three oscillations that spanned about 6 months. They were maintained by the dispersal of predator and prey among oranges in a deadly game of hide-and-seek, in which the prey managed to keep ahead of the predator for three full oscillations. These results are similar to those obtained by Gause, but we need to rmember that Huffaker did not directly manipulate dispersal. In Huffaker's experiment both predator and prey moved from patch to patch under their own power.
FIGURE 14.19 Environmental complexity and oscillations in laboratory populations of an herbivorous mite and a predatory mite (data from Huffaker 1958). The importance of refuges was recognized by Lotka (1932) and incorporated into his mathematical theory of predator-prey relations. The starting point for his discussion were the Lotka-Volterra predator-prey equations that we discussed previously:
dNh/dt =rhNh－pNhNp and dNh/dt =cpNhNp－dpNp
The part of this equation that provided the starting point for Lotka's discussion was p, the capture or consumption rate of the predator. Lotka pointed out that while it may be reasonable to assume that p is a constant for a particular
environment, its value should change from one environment to another if the environments differ structurally, particularly if there is a difference in the availability of refuges in the two environments. Specifically, p should be lower where the prey or hosts have access to more refuges. This refinement of the Lotka-Volterra predator-prey model anticipated recent theoretical analysis of the role that refuges and spatial diversity in general play in the persistence of predator-prey and parasite-host systems. While Lotka's analysis concentrated on physical refuges that could shelter terrestrial prey, he recognized the wide variety of forms that refuges could take. He pointed out, for instance, that flight is a refuge for birds from terrestrial predators.
Exploited Organisms and Their Wide Variety of \"Refuges\"
Space
Most of our discussion has focused on what we might call \"spatial\" refuges, places where members of the exploited population have some protection from predators and parasitoids. Many forms of spatial refuge are familiar: burrows, trees, air, water (if faced with terrestrial predators), and land (if faced with aquatic predators). However, some spatial refuges differ in subtle ways from other areas.
As we saw previously, the cactus Opuntia stricta, which had completely covered vast areas of Australia, was controlled by a combination of an herbivorous insect, Cactoblastis cactorum, and pathogenic microbes. The introduction of the insects did not drive the cactus to ex6ncfion, however. One reason for the
persistence of the cactus is that it has a number of spatial refuges. As we saw, small, isolated cactus populations are difficult for Cactoblastis to find. This is a spatial refuge much like that designed into Huffaker's experimental arrangement of oranges. In addition, the insects do not vigorously attack the cactus where it grows on nutrient-poor soils and/or above 600 to 900 m elevation, due to low quality of the cactus tissues or low temperatures.
St. John's Wort, Hypericum perforatum, persists in similar refuges in the face of attacks by the beetle Chrysolina quadrigemina, one of the chief enemies of Hypericum in the Pacific Northwest region of the United States. Hypericum was introduced into areas along the Klamath River around 1900,and its population quickly grew to cover about 800,000 ha by 1944. Following the release of the beetles, the area covered by St. John's Wort was reduced to less than 1% of its maximum coverage. This remnant population of the plant was concentrated in shady habitats, where, though it grows more poorly than in sunny areas, it is protected from the beetles, which avoid shade.
Protection in Numbers
Living in a large group provides a type of refuge. Aside from the potential of social groups to intimidate would-be predators, numbers alone can reduce the probability of an individual prey or host being eaten. We can make this prediction based solely on the work of C. S. Holling (1959) on the responses of predators to prey density. In chapter 6 we looked at the functional responses of several predators and herbivores. Briefly, predator functional response results in
increasing rate of food intake as prey density increases. Eventually, however, the predator's feeding rate levels off at some maximum rate. In chapter 10, we looked at numerical response, a second component of predator response to prey density that results in increased predator density as prey density increases. As with functional response, the numerical response eventually levels off at the point where further increases in prey density no longer produce increased predator density.
Now let's put functional response and numerical response together to predict the predator's combined response to increased prey density. We can combine the two responses by multiplying the number of prey eaten per predator times the number of predators per unit area:
By dividing the prey consumed per unit area by the population density of the prey (prey consumed/area), we can determine the percentage of the prey population consumed by the predator. If we plot percentage of the prey consumed against prey density over a broad range of prey densities, the prediction is that the percentage of the prey population consumed will be lower at high prey densities (fig. 14.20).
FIGURE 14.20 Prey density and he percentage of prey consumed due to combined functional and numerical responses (data from Holling 1959). Why should the percentage of the prey consumed by the predator decline at high prey densities? The answer to this question, which may not be obvious at first, lies in the predator functional and numerical responses. We see this effect because both numerical and functional responses level off at intermediate prey densities; that is, beyond a certain threshold, further increases in prey density do not lead to either higher predator densities or increased feeding rates. Meanwhile, the density of the prey population continues to increase and the proportion of the prey eaten by predators declines. This work by Holling suggests that prey can reduce their individual probability of being eaten by occurring at very high densities. It appears that this defensive tactic, which is called predator satiation, is employed by a wide variety of organisms from insects and plants to marine invertebrates and African antelope.
Predator Satiation by an Australian Tree
A great number of plants flower and produce seeds synchronously over large areas, including bamboo, pines, beech, and oak. This phenomenon of synchronous widespread seed and fruit production is called masting. Daniel Janzen (1978) proposed that a major selective force favoring the production of mast crops is seed predation. He suggested that mast crops may lead to satiation of seed predators, allowing some seeds to escape predation, germinate, and successfully establish.
Many Australian trees in the genus Eucalyptus disperse their seeds in large numbers following forest fires. Seeds are produced each year but mostly remain stored in closed seed capsules that are retained on the tree. Following a tire, a massive, synchronized release floods the forest floor with seeds.
Synchronous seed release by Eucalyptus may be a defense against seed predators, but which ones? The chief seed predators in Australian forests appear to be ants. We usually think of Australia as the continent of kangaroos and koalas, but the region could be just as well known for its ants. Australia harbors a tremendous diversity of ants. For instance, while North American deserts support about 160 species of ants, the deserts of Australia contain an estimated 2.000 species and the continent as a whole may harbor nearly 4,000 species. The ants in Australian forests have been reported to prevent forest regeneration by removing up to 80% of the seeds broadcast by foresters.
D.O'Dowd and A.Gill(1984)used field experiments to determine whether synchronous .seed dispersal by Eucalyptus might be a means to reduce seed losses to ants. They set up study plots in two forests of Australian alpine ash, Eucalyptus delgatensis, a gigantic tree found in the Australian Alps and Tasmania, where it commonly grows to 50 to 60 m in height. One of the study sites was to serve as a reference site and the second ms an experimental site. E. delgatensis constituted nearly 90% of the tree biomass at both study sites. The researchers monitored a number of physical and biological variables at the control and experimental sites and then set a controlled high-intensity fire at the experimental site. The area burned was approximately 98 ha; 93% of the trees were killed.
The results of O'Dowd and Gill's field experiments support their hypothesis that synchronous seed dispersal by Eucalyptus reduces losses of seeds to ants. As expected, the fire stimulated the release of massive numbers of seeds. During the 3 weeks following the fire, seed fall was approximately 405 fertile seeds per square meter, compared with a peak seed fall of 10 seeds per square meter per week at the control site, which was not burned. The fire also seemed to stimulate ant activity. Prior to the fire, researchers trapped an average of 176 ants belonging to 14 species each week. After the fire, the number of ants trapped each week rose to an average of 680 individuals belonging to 23 species. Despite this strong numerical response, the rate at which ants removed seeds dropped following the fire. The rate of seed removal from seed trays at the experimental site dropped from an average of about 65% per week during the 5 weeks prior to the fire to an average of about 14% per week during the 5 weeks following the fire. This result is consistent with the predator satiation hypothesis proposed by O'Dowd and Gill
and is consistent with Holling's prediction of reduced predator combined response at high prey densities.
So, what does O'Dowd and Gill's experiment tell us? We know that E. delgatensis stores seeds in closed seed capsules, that these seeds are released synchronously following intense fires, and that a substantial proportion of these seeds escape predation even in the face of strong numerical response by seed-eating ants. Is this information sufficient to conclude that the apparent predator satiation strategy of E. delgatensis provides an effective \"refuge\" from predation? It seems that we should also know whether greater seed survival is translated into greater seedling establishment, an evolutionary bottom line. O'Dowd and Gill also showed that seedling establishment was greater on the burned experimental plot than on the control plot (fig. 14.21). About 1.5 years after the experimental fire, seedling survival at the experimental site was approximately 2 individuals per square meter, or 20,000 individuals per hectare. This seems a respectable level of reproductive success for trees that could eventually reach 50 to 60 m in height. Predator satiation also provides protection to many kinds of insects, but nowhere is this more apparent than among the cicadas.
FIGURE 14.21 Seedling establishment by Eucalyptus delgatensis at burned and unburned sites (data from O'Dowd and Gill 1984). Predator Satiation by Periodical Cicadas
Periodical cicadas, Magicicada spp., emerge as adults once every 13 years in the southern part of their range in North America and once every 17 years in the northern part of their range. Though these insects emerge only once every 13 or 17 years in any particular area, virtually every year sees a brood emerging somewhere in eastern North America. An emergence of periodical cicadas produces a sudden flush of singing insects whose density can approach 4 x 106 individuals per hectare, which translates into a biomass of 1,900 to 3,700 kg of
cicadas per hectare, the highest biomass of a natural population of terrestrial animals ever recorded.
Periodical cicadas are insects of the order Homoptera, which includes the leafhoppers and aphids. Like their relatives, cicadas make their living by sucking the fluids of plants and spend either 13 or 17 years of their life as nymphs underground, where they feed on the xylem fluids in roots. When mature, nymphs dig their way to the soil surface, where they shed their nymphal skin and emerge as winged adults. Among periodical cicadas this emergence is so synchronized that millions of adults emerge over a period of only a few days. Following emergence males fly to the treetops, where they sing the mating songs to which females are attracted. After they mate, females lay their eggs in living twigs of shrubs and trees. When the nymphs hatch in about 6 weeks, they immediately drop to the ground and burrow down lo a root, where they begin to feed, moving around very little for the next 13 or 17 years. A mass emergence of periodical cicadas, one of the most memorable biological phenomena nature has to offer, appears aimed at predator satiation.
K. Williams and his colleagues (1993) tested the effectiveness of predator satiation in a population of 13-year periodical cicadas in northwest Arkansas. They monitored emergence of cicadas using conical emergence traps constructed of plastic mesh and inverted their traps to measure predation rates (fig. 14.22). Nymphs emerging from the ground below the traps could be counted to estimate the numbers of emerging nymphs. Then, as adult cicadas died from a variety of factors, including physical factors, senescence, and pathogens, they fell from the
trees to the ground, where stone were caught in the inverted traps. Because the major predators were birds, predation rates could be estimated because birds discard the wings of cicadas s they feed upon them. The wings falling into the inverted traps gave an estimate of predation rates.
FIGURE 14.22 Estimating cicada population site and predation rates by birds Patterns of mortality and predation rates relative to population size support the predator satiation hypothesis. Williams and his colleagues estimated that 1,063,000 cicadas emerged from their 16 ha study site and that 50% of these emerged during four consecutive nights. Cicada abundance peaked in late May and then declined rapidly during the first 2 weeks of June. Part of this decline was due to mortality from severe thunderstorms during the first week of June. Figure 14.23 shows that losses due to birds were low throughout the period of peak cicada abundance and then climbed to I00% as cicada populations declined during June. These results indicate that the predator satiation tactic was
sufficiently effective to reduce cicada losses to birds to only 15% of the total population.
FIGURE 14.23 Cicada population density and their percent mortality due to predation (data from Williams. Smith. and Stephan 1993). Size as a Refuge
We first encountered size-selective preda6on in chapter 6 among bluegills, Lepomis macrochirus, and pumas, Fells concolor. However, many other organisms select their prey by size, In fact, average prey size shows a significant correlation with predator size across taxa ranging from lizards to small mammals. The reason for size selective predation among such a diverse array of organisms is that prey capture and consumption are mechanical pm61~ms, and as we .saw in chapter 6, size can influence the time required to handle prey and therefore the rate of energy intake. The bottom line is that for a given predator some prey are simply
too large to be profitable and so are not attacked.
Now let's look at size from the perspective of the prey. If large individuals are ignored by predators, then large size may offer a form of refuge. An obvious example is on the African savanna. While a variety of predators may attack the calves of elephants or rhinoceros, the same predators avoid the adults, which have been observed to kill adult lions (fig. 14.24). On a smaller scale, Robert Paine (1976) found that the seastar Pisaster ochraceus does not consume the largest individuals in populations of one of its chief prey species, the mussel Mytilus californianus. Figure 14.25 shows that the maximum size of mussels eaten by seastars is a function of seastar size. Notice that most of the successful predation observed by Paine involved small to medium-sized seastars attacking mussels less than 11cm long. Most seastars cannot eat the largest mussels, and the largest seastars that can were limited to a few areas of coastline in the study area. What this means is that if a mussel can manage to escape predation long enough to reach 10 to 12cm in length, it will be immune from attack by most seastars. When Paine removed the seastars from an area of the intertidal zone, resident mussels survived at higher rates and therefore grew to a larger average size (fig. 14.26). When Paine allowed seastars to recolonize the area, many of the mussels were large enough that they effectively escaped predation by seastars. This result has implications that reach far beyond higher survival within a single prey population.
FIGURE 14.24 Large size can provide a refuge from predators. While young African elephants may be vulnerable to predation by African lions mature elephants are not. FIGURE 14.25 Large mussels are eaten infrequently by the seastar Pisaster
ochraceus (data from Paine 1976). FIGURE 14.26 Growth by mussels in an intertidal area from which the seastar. Pisaster ochraceus, was excluded (data from Paine 1976). If predators pass up prey above a particular size threshold, might natural selection favor organisms that project a \"large\" body size to some would-be predators? It appears that some aquatic insects have been selected to do just that. Barbara Peckarsky (1980, 1982) observed that mayflies in the family Ephemerellidae would \"stand their ground\" in the face of a foraging predatory stonefly, la fact, they would not only stand their ground, they would curve their abdomens over their backs and point the tips of their abdominal cerci into the face and antennae of a stonefly, a behavior Peckarsky called a \"scorpion\" posture (fig. 14.27). Usually a stonefly greeted in this way does not attack. While many other
stream ecologists had seen this behavior in epbemerellid mayflies, Peckarsky was the first to suggest that the scorpion posture was a defensive tactic in which the mayfly projected a larger image to a tactile, size-selective predator.
FIGURE 14.27 Posturing b, an ephemerellid mayfly confronted by a predaceous stonefly. Why should a large stonefly avoid large ephemerellid mayflies? Large ephemerellids have been observed attacking stoneflies trying to prey on them, and so like lions that avoid rhinoceros, stoneflies that avoid ephemerellids may be protecting themselves from injury. Most ephemerellids, however, present no danger to large predaceous stoneflies, so self-protection only partially answers our question. For the bulk of encounters between stoneflies and ephemerellid mayflies, large apparent size would probably indicate low profitability, low E/T in terms of optimal foraging theory (see chapter 6), and send the predator looking for a prey that would yield a higher energy return. It may be that the display by ephemeral lids is not a bluff, however, since they require an exceptionally long handling time
for a prey of their size. The scorpion posture of ephemerellids may be a case of \"truth in advertising.\"
While we may think of predators as threats to ourselves or to livestock or crops, many predators and parasites have been used to control populations of insects that attack crops or to control invasive weeds. As we shall see in the Applications and Tools section, predators are increasingly used to control parasites that infect humans.
APPLICATIONS AND TOOLS:
using predators to control a parasite
Parasitic diseases afflict approximately 600 million people across the planet, particularly in tropical and subtropical countries. Despite intensive efforts at control, many parasitic diseases are spreading and the number of cases appears to be increasing. A key factor in this increase appears to be human population growth, which puts additional pressure on sanitation and health care systems and increases the number of hosts for human parasites. The leading parasitic disease in humans is malaria, which is transmitted to humans by mosquitoes and infects an estimated 250 million people. The second most prevalent parasitic disease is schistosomiasis, which infects approximately 200 million people. Schistosomiasis is a debilitating infection caused by blood flukes of the genus Schistosoma. Infections by this parasite are particularly debilitating to children. The scope and intensity of infections by Schistosoma and other parasites challenge the world
health community to develop systems for their control. However, the problem of control is essentially an ecological one and therefore complex.
Much of this complexity is due to the life cycle of the parasite (fig. 14.28), Schistosoma spends its larval phase as a parasite in aquatic snails and its adult phase in humans. The cercariae, the stage of Schistosoma that infects humans, are released by snails into the water, Cercariae penetrate the skin of humans in streams, lakes, or ponds containing infected snails. Some Schistosoma infect the human digestive tract while others infect the urinary tract. Humans that either urinate or defecate in water complete the parasite's life cycle by facilitating the infection of snails.
FIGURE 14.28 The llfe cycle of Schistosoma. Schistosomiasis can now be treated with a variety of drugs. However, treated patients are often reinfected. Consequently, schistosomiasis control programs that rely solely on treatment of infected individuals cannot control the spread of infection. Effective control must also include the populations of snails that serve as
intermediate hosts. Methods of snail control have included applications of chemicals that kill snails, introduction of other snails that competitively displace the snail species that serve as intermediate hosts for Schistosoma, or, increasingly, using predators to control snail populations. This work draws upon one of the main concepts of this chapter: predators, parasites, and pathogens influence the distribution, abundance, and structure of prey and host populations.
Scientists in several East African countries are researching the potential of a variety of predators to control the host snails infected by Schistosoma. One of those countries is Kenya. Kenya's population includes I to 2 million people infected with Schistosoma. Health officials are concerned that the number of cases may increase due to increased pollution of freshwater environments by untreated sewage and increased construction of clams and irrigation systems, which favor the growth of snail populations. In the face of these threats, Kenyan health officials are developing a comprehensive, multifaceted plan for control of schistosomiasis. One of the elements in their plan is to use predators to control the parasite's host snails.
One of the predators being tested for its effectiveness at snail control is the crayfish Procambarus clarkii, which is native to North America, Procambarus was introduced into Kenya during the 1970s. Health officials became interested in the potential of the crayfish to control snail populations when snail surveys showed that habitats with Procambarus lacked host snails. However, because Procambarus is not native to Kenya, ecologists and health officials are proceeding cautiously. They point out that Procambarus, which has been introduced to 24 countries
worldwide, is a highly invasive species with the potential to disrupt native populations and ecosystems. It has already spread rapidly across Kenya. In addition, the crayfish may threaten rice cultivation in Kenya. Its burrows have damaged rice fields in other regions and it eats rice seedlings. The environments where Procambarus may be most useful and where it will likely cause little damage are the thousands of artificial ponds that dot the rural Kenyan landscape. These ponds are used to water livestock and as domestic water supplies.
Ecologists have used a combination of laboratory and field studies to test the potential of Procarnbarus to control host snail populations. In one of the early studies Bruce Hofldn and Sam Loker of the University of New Mexico joined Gerald Mkoji and Davy Kocch of the Kenya Medical Research Institute in a survey of the distributions of Procarabarus and host snails in Kenya (Hofkin et al. 1991a). The research team restricted their survey to areas where both the crayfish and host snails were known to occur and to habitats within those areas that could support both organisms. The survey, which included 53 sites, revealed a highly significant negative association between Procambarus and host snails. Nine sites had neither snails nor crayfish. In the 44 other sites, 19 had snails only, 21 had crayfish but no snails, and only 4 sites had both crayfish and snails (fig. 14.29). The snails at these sites were present in small numbers and consisted of species incapable of transmitting human schistosomiasis. This survey indicates clearly that Procambarus has the potential to eliminate host snails from a variety of aquatic habitats in Kenya.
FIGURE 14.29 Distributions of shall hosts of Schistosoma and crayfish in Kenyan ponds (data from Hofkin et al. 1991a). Based on the results of their survey, Hot'kin and his colleagues (1991 b) conducted a series of laboratory and field trials to test directly for the capacity of crayfish to eliminate host snail populations. The first step in these studies was to determine if Procambarus eats the snail species infected by Schistosoma. Laboratory feeding studies in l0L aquaria showed that the crayfish readily eat the main species of host snails. Because the aquaria were bare and no alternative food was provided for crayfish, the conclusions that we can draw from these experiments are limited. Hofkin and his colleagues concluded only that Procambarus has the potential to control host snails.
To more closely simulate field conditions, Hofkin and his colleagues next
increased the size and complexity of their laboratory environments and provided crayfish with alternative food. In one of these experiments, they used large plastic tanks to which they added l0cm of soil and then filled with water to 30 cm depth. Each tank contained 50 host snails and four water lilies, which provided both food and shelter for the snails. They added two or four crayfish to each tank and counted the number of remaining snails after 5, 15, or 30 days. Procambarus significantly reduced the number of snails after just 5 days and eliminated them entirely after 30 days in i
tanks containing four crayfish. Crayfish also ate significant quantities of water lily leaves, reducing the habitat and food I available to snails. Therefore, Procambarus is both a predator and competitor of host snails.
With these results, Kenyan health officials were encouraged to begin testing the effectiveness of crayfish to control host snail populations in the field. However, the potential of Procambarus to damage native ecosystems suggested caution. Consequently, Gerald Mkoji of the Kenya Medical Research Institute and a research team from several other institutions suggested that the crayfish be used only in small artificial impoundments, where they would pose no threat to Kenyan wildlife. These ponds are a major source of infection by Schistosoma. It appears that Procambarus has eliminated host snail populations from the ponds where it has been introduced, Used judiciously, the crayfish appear to have the potential to reduce the impact of a serious human pathogen without causing significant environmental damage. This result comes directly from the growing understanding of the ecology of exploitation, the central theme of this chapter.
SUMMARY CONCEPTS
The diversity of interactions between herbivores and plants, between predators and prey, and between parasites, parasitoids, pathogens, and hosts can be grouped under the heading of exploitation--interactions between species that enhance the fitness of one individual at the expense of another.
Exploitation weaves populations into a web of relationships that defy easy generalization. The number of exploitative interactions between species far exceeds the number of species in the biosphere, and the nature of exploitation goes far beyond the typical consumption of one organism by another. For instance, many parasites and pathogens manipulate host behavior to enhance their own fitness at the expense of the host. Spiny-headed worms alter the behavior of a variety of crustacean hosts in a way that increases the probability that the one host species will be eaten by another. A pathogenic fungus manipulates the growth program of its host plant in a way to produce \"pseudoflowers,\" structures aimed at promoting the reproduction of the pathogen. In the process the pathogen usually kills the host plant and always renders it sterile. Predation by one flour beetle species on another can be used as a potent means of interference competition except in the presence of a protozoan parasite, which seems to give a competitive advantage to less predaceous species.
Predators, parasites, and pathogens influence the distribution, abundance, and structure of prey and host populations. Herbivorous stream insects have been shown to control the density of their agal and bacterial food.
The herbivorous moth larva Cactoblastis cactorum combined with pathogenic microbes reduced the coverage of prickly pear cactus in Australia from millions of hectares to a few thousand. A parasitic infestation reduced the red fox population in Sweden by 70%, which in turn led to increases in the abundance of several prey species eaten by foxes. This parasitic disease revealed the influence of a predator on its prey populations.
Predator-prey, parasite-host, and pathogen-host relationships are dynamic. Populations of a wide variety of predators and prey show highly dynamic fluctuations in abundance ranging from days to decades. A particularly well-studied example of predator-prey cycles is that of snowshoe hares and their predators, which have been shown to result from the combined effects of the snowshoe hares on the quantity and quality of their food and of the predators on the snowshoe hare population. Mathematical models of predator-prey interactions by Lotka and Volterra suggest that exploitative interactions themselves can produce population cycles without any influences from outside forces such as weather. Predator-prey cycles have also been observed in a few laboratory populations under restricted circumstances.
To persist in the face of exploitation, hosts and prey need refuges. The refuges that promote the persistence of hosts and prey include secure places to which the exploiter has limited access. However, living in large groups can be considered as a kind of refuge since it reduces the probability that an individual host or prey will be attacked. It appears that predator satiation is a defensive tactic used by a wide variety of organisms from rain forest trees to temperate insects.
Growing to large size can also represent a kind of refuge when the prey species is faced by size-selective predators, Size is used as a refuge by prey species ranging from stream insects and intertidal invertebrates to rhinoceros.
Predators and parasites have been used to control populations of insects that attack crops or to control invasive weeds. Recent research in Kenya has shown that a crayfish, Procambarus clarkii, controls the snails that act as intermediate hosts for Schistosoma. a highly pathogenic human parasite. Preliminary results indicate that crayfish successfully control host snails in the artificial impoundments used for livestock watering and domestic water, important sources of infection by Schistosoma.
REVIEW QUESTIONS
1. Predation is one of the processes by which one organism exploits another. Others are herbivory, parasitism, and disease. What distinguishes each of these processes, including predation, from the others? We can justify discussing these varied processes under the heading of exploitation because each involves one organism making its living at the expense of another. By what \"currency\" would you measure that expense (e.g., energy, fitness)?
2. How are manipulation of host behavior by spiny-headed worms and manipulation of plant growth by the rust Puccinia monoica the same? How are they different? The details of these parasitic interactions are very different in many ways from the predatory behavior of lions on the savannas of Africa. How are
the activities of spiny-beaded worms, rusts, and lions similar?
3. Predation by one flour beetle species on another can be used as a potent means of interference competition. However, the predatory strategy seems to fail consistently in the presence of the protozoan parasite Adelina tribolii. Explain how the predatory strategy works in one environmental circumstance and fails in another.
4. In this chapter we have seen how an herbivorous stream insect controls the density of its food organisms, how an herbivorous moth larva and pathogenic microbes combine to control an introduced cactus population, and how decimation of a red fox population led to increases in the populations of the foxes' prey. We do not know the specific environmental factors controlling most populations. Explain why such factors must exist, (Hint: Think back to our discussions of geometric and exponential growth in chapterⅡ).
5. Early work on exploitation focused a great deal of attention on predator-prey relations, However, parasites and pathogens represent a substantial part of the case histories in this chapter. Is this representation by parasites and pathogens just the result of biased choices by the author or do you think that parasites and pathogens have the potential to exert significant controls on natural populations? Justify your answer.
6. Researchers have suggested that predators could actually increase the population density of a prey species heavily infected by a pathogenic parasite
(Hudson, Dobson, and Newborn 1992). Explain how predation could lead to population increases in the prey population.
7. Explain the roles of food and predators in producing cycles of abundance in populations of snowshoe hare. Populations of many of the predators that feed on snowshoe hares also cycle substantially. Explain population cycles among these predator populations.
8. What contributions have laboratory and mathematical model made to our understanding of predator-prey population cycles? What are the shortcomings of these modeling approaches? What are their advantages?
9. We included spatial refuges, predator satiation, and size in c discussions of the role played by refuges in the persistence exploited species. How could time act as a refuge? Explain how natural selection could lead to the evolution of temporal \"refuges.\"
Joseph Culp and Gary Scrimgeour (1993) studied the timing of feeding by mayfly larvae in streams with and without mayflies feed by grazing on the exposed surfaces of stones, where they are vulnerable to predation by fish, which in the streams studied are size-selective feeders and feed predominantly during the day. In the study streams without fish, both a small and large mayflies have a slight tendency to feed during the day but feed at all bouts of the day and night, in the streamed with abundant fish populations, small mayflies fed around the ~1 clock, while large mayflies fed mainly at night. Explain these patterns in terms of
time as a refuge and size-selective predation
I0. When applying ecological concepts to practical problems, there are often trade-offs between the benefits and costs of a management decision. Review the costs and benefits of using the crayfish Procambarus clarkii for control of schistosomiasis in Kenya.
SUGGESTED READINGS
Moore, J. 1984. Parasites that change the behavior of their host. Scientific American 250:108-15. An excellent review of the influences of parasites on prey behavior. Exceptionally well written and illustrated.
Roy, B. A. 1993. Floral mimicry by a plant pathogen. Nature 362:56-58.
Fascinating example of manipulation of a plant and its pollinators by a pathogenic fungus.
Lamberti, G. A. and V.H. Resh. 1983. Stream periphyton and insect herbivores: an experimental study of grazing by a caddisfly population. Ecology : 112~--35.
A classic field experiment that reveals the controlling influence of a benthic invertebrate grazer on a stream community. A model for the design of field experiments.
Lindstrǒm, E. R., H. Andren, P. Angelstam, G6ran Cederlund. B. Hǒmfeldt, L.
Jaderberg, P.-A. Lenmell. B. Martinsson, K. Skǒld, and J. E. Swenson. 1994. Disease reveals the predator: sarcoptic mange, red fox predation, and prey populations. Ecology 75:1042－49.
Large-scale, long-term study of interactions between a pathogenic parasite, a predator, and its prey. An example of how to use natural experiments to extract in formation from nature.
Utida, S. 1957. Cyclic fluctuations of population density intrinsic to the host-parasite system. Ecology 38:44249. A concise, readable report of one of the longest-running experiments on parasitoid-host population cycles－a landmark study in the field.
Huffaker, C.B. 1958, Experimental studies on predation: dispersion factors and predator-prey oscillations. Hilgardia 27:343-83.
Huffaker, C.B., K.P. Shea, and S.G. Herman. 1963. Experimental studies on predation: complex dispersion and levels of food in an acarine predator-prey interaction. Hilgardia 34:305-30.
Two classic papers on predator-prey oscillations, involving complex laboratory studies. Difficult in some places but worth the effort. These two papers have had a major impact on the study of exploitative interactions.
Haukioja, E., K. Kapiainen, P. Niemela, and J. Tuomi. 1983. Plant availability
hypothesis and other explanations of herbivore cycles: complementary or exclusive ahematives? Oikos 40:419--32.
Keith, L. B. 1983. Role of food in hare population cycles. Oikos 40:385-95.
These two papers provide comprehensive reviews of the hypotheses proposed to explain hare population cycles.
Williams, KS., KG. Smith, and F.M Stephen. 1993. Emergence of 13-yr periodical cicadas (Cicadidae: Magicicada): phenology, mortality, and predator satiation. Ecology 74:1143-52.
Beautifully designed field study of a complex problem. Demonstrates predator satiation by periodical cicadas.
ON THE NET
Visit our website at http://www.mhhe.com/ecology for links
to the following topics:
Animal Population Ecology
Parasitism, Predation, Herbivory
Field Methods for Studies of Populations Parasitic Protists
Human Diseases Caused by Nematode Coevolution
因篇幅问题不能全部显示，请点此查看更多更全内容
Copyright © 2019- igbc.cn 版权所有 湘ICP备2023023988号-5
违法及侵权请联系：TEL:199 1889 7713 E-MAIL:2724546146@qq.com
本站由北京市万商天勤律师事务所王兴未律师提供法律服务