When ecologists attempt to understand why a particular species' ("X") population occurs at a given density or exhibits certain changes in numbers through time, they tend to investigate other species that compete with or consume species X. In other words, the focus is on factors that would counteract exponential growth of species X's population as for example the negative feed-back of a predator population might do.

Mutualism as a factor in explaining the observed features of populations is rarely considered to be important. One reason for this may be the simplistic way that mutualism is incorporated into population models (as described in your text). By specifying that the benefit exchanged by mutualists is essentially an increase in carrying capacity for each mutualist, the models predict unlimited expansion of each population until the support system for both is greatly over-taxed and both populations crash. Consequently, theorists have considered mutualism to be a destabilizing force.

Research on ecological systems in tropical habitats suggest that mutualism has been misinterpreted and underplayed as a determining factor in population dynamics.

A. I discussed several major categories of mutualistic interaction:

1. Plant-pollinator relationships

Example: Orchid bees (euglossine) in new world rainforests. Male euglossines of a given species gather odors from specific species of orchid and are the only pollinators of that particular species. Females and males visit a variety of nectar producing plants and are important pollinators for some. Females gather pollen to provision brood chambers from still other plants, some of which rely specifically on these insects to set seed. Still other plants provide resin which euglossine females collect for nest construction and pollinate the provider plant in the process. To understand the population biology of a given species of euglossine bee you would need to understand the specificity and strength of connection to as many as twenty different resource plants in its environment. Conversely to understand whether the population trends of a particular plant in this system are related to changes in euglossine populations you would need to know the specificity and strength of connection this plant has to particular bee species. Since some orchid bees are specific pollinators for some orchids, the extinction of the bee in an area would ultimately spell the loss of the plant by ending juvenile recruitment. However, since orchids can live many decades, it could require long-lived and persistent ecologists to monitor the orchid's decline. Care to make a mathematical model of such a system?

2. Plant-seed disperser mutualism

Adult perennial plants grow where conditions were suitable for seed germination and seedling survival in the past but not when they reproduce. For examples, seed and seedling predators or parasites may build up under a tree dooming babies that try to develope below the parent, or the parent's shade may inhibit growth of its own offspring. Dispersing seeds away from the death zone around the parent and to areas of new disturbance (sunny open ground) is required for many plant species. Fruits are devices which help enlist mobile animals in the dispersal of seeds contained in the fruits. As in pollination mutualisms, the degree of specificity is important in understanding how influential such mutualism might be to driving the dynamics of interacting plant and animal populations. In some extreme case, the loss of a fruit bat species could mean the ultimate loss of a tree species from a Malaysian forest (or vice versa). However most such mutualisms are more diffuse (i.e. trees that rely on several bat species, and bats that rely on fruits of many tree species) and do not boil down to obligate species pairs. Those tend to go extinct!

3. Plant defense mutualisms.

Many plants provide food (food bodies or extra-floral nectar) and/or shelter (hollow stems, petioles, or thorns) to ants, wasps and other predators and parasites in return for protection from plant-feeding animals (herbivores). I gave examples of facultative systems (eg. passionvines and various predaceous ants and parasitoid wasps) and and obligatory ant-plant mutualism, ant acacia and Pseudomyrmex ants, neither of which exist without the other. The acacia provides carbohydrates (extra-floral nectar), proteins and oils (beltian bodies on new leaflets) and domicile (hollow, stipular thorns). The Pseudomyrmex ants protect the host acacia from vertebrate and insect leaf-feeders as well as from other competing plants (they bite, sting, and kill the growth points). This plant drops out coming north in Mexico where cold periods are sufficient to inhibit ant defenses and deer, goats, and cattle are able to feed on the leaves and stems.

4. Cooperative education of predators by aposematic insects

Mullerian mimics are distasteful, warningly colored species that evolutionarily converge on similar warning signals. This is a form of mutualism since the per-capita probability of death to predation within each species' population is lower because they use the same rather that different warning signals. Since no deception is involved, this phenomenon should not technically be termed "mimicry." Heliconius butterflies provide a nice example since species from different clades look more alike within an area than do different populations of the same species from different regions.

B. I discussed the interesting case of Heliconius ethilla in Trinidad whose population dynamics cannot be explained without considering several ways that mutualisms directly or indirectly influence the probabilities of birth, death, and survival of the insect's various life stages .

1. Eggs are placed on new shoots of Passiflora vines. Eggs hatch in 4 days. If all goes well for the larva that subsequently hatches, 5 larval stages or instars spend about 15 days consuming the passionvive leaves. A pupal stage of 9 days ends with the eclosion of a teneral adult. Adults marked in the first day of emergence have been recovered over 6 months later ranking Heliconius as the longest lived butterfly known (excluding species that diapause as adults). To recover a mobile butterfly after 6 months in the same place indicated a remarkable faithfulness to "place" which is associated with birds and mammals, but not insects.

Mark-release-recapture studies showed that for two years, through dry season and wet, two ethilla populations showed remarkable constancy of numbers. (We don't say "stable" because in population biology and in mathematics stability has a specific technical meaning. Experiments would be required to demonstrate "stability.")

2. How does mutualism help explain the observed constancy of Heliconius populations?

a. Plant-pollinator mutualism.

Heliconius adults feed on the pollen of certain plants especially the rainforest cucumber vines (Gurania and Psiguria/Anguria). This is a tight plant-pollinator relationship in which the butterflies are major pollinators for the plants and the plants major food resources for adult maintenance and egg production. 80% of a female's egg production come from amino acids that come from pollen she collects. Only 20% comes from that acquired by the caterpillars feeding on passionvines. Note that in most butterflies and moths, 100% of eggs derive from the efforts of the larval stage and eggs are laid in a quick pulse after adult emergence. In Heliconius, eggs are laid as they are manufactured over the adult's long lifespan.

The butterflies learn the locations of pollen plants and establish home ranges based on pollen foraging routes. It appears that the pollen plants are more significant than larval host in determining a female's assessment of the habitat. Thus, as long as she knows the locations of a network of pollen plants, she will stay in the area, even during periods when new shoots of passionvine hosts are temporarily not available due to weather or defoliation by Heliconius or other competing herbivores. So while most herbivorous insects disperse away when suitable ovipostion sites are scarce, Heliconius females are content to stay put as long as the mutualist plant produces pollen (which is year-around). Moreover the pollen promotes a long reproductive life so that females can wait many weeks for the opportunity to resume egg-laying.

In summary, this mutualism reduces the probability of local extinction due to adult dispersal, increases the probability of individual survival and reproduction across periods when natural perturbations of the environment reduce many other species of insect in the same habitat. Moreover, since pollen supply is 1) limiting and 2) does not increase if the adult butterfly population increases, egg numbers cannot increase as a simple linear function of butterfly population increase. Egg numbers are thus always more constant than number of adult females.

b. Plant defensive mutualisms

Extra-floral nectaries of passionvines encourage a variety of predators (ants, wasps) and parasitoids (flies, wasps) to patrol the shoots and leaves of the vine. Each life history stage, from egg to pupa is exposed to several natural enemies. In the ethilla population over 90% of eggs were killed by egg parasitoids, tiny wasps that feed on extrafloral nectar and hang out on the new shoot. Larvae from eggs that survived wasps face many similar problems in their 15 day developmental period. Ants in particular are important because they patrol whether or not larvae are present (i.e. they are not limited by larvae). Consequently, increases or decreases in caterpillar numbers do not set off predator-prey oscillations as occur when predators have a delayed population response to changes in prey populations.

In the end, each female ethilla butterfly averages one surviving female offspring from the approx. 1000 eggs deposited in her life. The net effect of the entire Passiflora mutualistic defensive system in this case, is to prevent the runaway herbivore outbreak that I described for owl butterflies in the banana plantations of Costa Rica. Otherwise the adult population would not show such constancy as was observed.

c. Mullerian mimicry.

When adult ethilla populations decline to very small numbers, there can be more predaceous birds than adult butterflies. Even though ethilla is bad tasting and aposematic, if each bird tested a butterfly to refresh its memory on what patterns to avoid, the butterfly population would quickly be extinct. It is during these time of rarity that companion aposematic species sharing the same color pattern signals can provide continuing education of local birds and save mullerian partners fron local extinction.

d. Plant-seed disperser mutualism

On a longer time frame the resource plants for the ethilla population are maintained by mutualists of these plants such as pollinators and animals required for effective seed dispersal. These interactions do not account for constancy on the order of a few months in the case of large perennial vines that require years to establish. However the long term trends of Heliconius populations ultimately depend on the mutualistic support of all the key resource plants.

e. Summary diagram showing the interactions discussed above.