Metapopulation Ecology and Extinction

Before we start discussing extinction, it is important to understand exactly how important a minutely detailed study of the ecology of an endangered species is. Such studies will directly impact the eventual survival the species through human efforts. In many cases, the ecologist's evaluation of the direct cause of the endangerment is often an interpretation based upon limited data. An excellent example of this is the case of the Large Blue Butterfly in England.

Maculina arion (the Large Blue Butterfly) is, in its larva stage, an obligate parasite in ant colonies (the ant is Myrmica spp.). This means that the larva must find an ant colony, get into it, and feed on ant larvae.

The butterfly appeared headed towards extinction in southern England in the 1880’s due to adverse weather. From the 1940’s through the 1960’s it became increasingly restricted to southwestern England. By 1974 the number of colonies had fallen from 30 in the 1950’s to only 1 or 2, with the estimated adult population having declined from about 100,000 in 1950 to 250 by 1972.

Two successive droughts during the breeding season lead to the final extinction of the species in England in 1979. The drought apparently caused caterpillars to overcrowd ant nests, which lead to their deaths.

One obvious problem for the butterfly was that its habitat had greatly changed. The initial food for the larva is wild thyme (Thymus drucei) which is a common and widespread colonizer of bare ground in grasslands. However,  a myxomatosis epidemic that killed off rabbits (which grazed on plants), combined with plowing, forestry, and changes in livestock management led to the disappearance of short-turf grasslands.

Aside from the obvious habitat change, other causes for the decline of the butterfly were suggested. The butterfly is large and quite beautiful, so over-collecting was suggested as a major cause of the decline. While this may have initially contributed to the decline of the butterfly, by the 1930’s a reserve had been established which excluded both collectors and domestic livestock in an attempt to protect the butterfly and its habitat. The reserve became overgrown and the butterflies disappeared. This again lead to the suggestion that some change in the habitat had played a role.  However, since the butterfly's main food source (wild thyme) was still present, a number of other causes were proposed, including changes in the weather, pesticides, and inbreeding depression. All of these were simply interpretations of the pattern of extinction, with little hard data to support them.

The species was reintroduced from the European mainland, and an intensive six-year study of the butterfly’s ecology was conducted.

It turned out that one of the most important stages of the butterfly’s life cycle is its parasitism on ant nests during its larval stage. The low survival of caterpillars in ant nests during this stage (4-37%) limits the population. When the caterpillar first emerges from its egg, it feeds on the wild thyme plant, but gains little weight through this feeding. In August, the caterpillar will wander away from the plant and try to entice an ant to adopt it and take it into its nest. The caterpillar does this by producing sugar for the ant and by other tactile and olfactory signals to the ant. It basically tricks the ant into carrying it back to the nest. Once in the nest the caterpillar feeds on ant larvae (it becomes parasitic on the ant nest), while being protected by its thick cuticle, acquired smell, and continued bribes of sugar for the ants.

There are a number of factors required for all this to work. The most important factor is that the right species of ant (Myrmica sabuleti) must be present in the area. At this point the microhabitat comes into play. If the turf increases a scant 2 cm. in height, the microclimate is cooled sufficiently for a second ant species (M. scabrinodes) to enter the area and displace M. sabuleti. This change in ant species results in a drop in caterpillar survival from roughly 15% (in M. sabuleti nests) to 2% (in M. scabinodes nests).

Other factors that affect caterpillar survival are:

the ants must not be short of food, because then they become too discerning to accept the crude mimicry of the caterpillar.
an ant colony must be within 2 m. of the thyme plant, which is the ant’s foraging distance.
there must be at least 400 worker ants in the colony to provide the roughly 230 ant larvae needed by the caterpillar.
most importantly, there should be no queen in the colony – the caterpillar survival rate increases by a factor of three when the queen is absent. If a queen is already present, the nurse ants will attack and kill other queen larvae, which they detect by smell. The caterpillar will get covered with pheromones from the queen larvae while feeding, which will result in it being attacked by the nurse ants. This queen effect is weakest (and butterfly survival highest) when the ant colony is newly established, or has been increasing for only about 3 or 4 years.

Burning and grazing of grasslands tends to create new habitat, which the ant (M. sabuleti) quickly colonizes. In newly burnt grasslands the survival of the caterpillars can reach upwards of 52% (many colonies available, low queen effect within nests), as compared to about 27% in long established turf areas.

Unfortunately, the attempt to create a reserve for the butterfly backfired. By keeping out grazing animals (rabbits, the natural grazers, had been severely reduced by disease; but the livestock which would have replaced them were kept out by fencing the reserve) the habitat changed enough to bring in a new species of ant which the butterfly larvae could not successfully parasitize. So "saving" the butterfly lead to its extirpation (see figure below).

ants.jpg (105290 bytes)


We can now consider the Big Blue Butterfly example in terms of Metapopulation Ecology. A metapopulation is a population which consists of several subpopulations linked together by immigration and emigration. These subpopulations exist in either source areas, where the ecological conditions meet all the needs of the species, so that r, the intrinsic growth rate is positive (births – deaths > 1); or in sink areas, where the individuals can exist, but where some important ecological need is not met with the result that r is negative (births – deaths < 1). The continued presence of the species in a sink area is entirely due to immigration of new individuals from a nearby source area.

     Now consider the figure below. Areas that are shaded red, blue and yellow are grassland areas. The blue areas are those of short turf suitable

meta6.gif (14578 bytes)

for the ant M. sabuleti, red areas are those which have experienced recent burns or cropping and are available for colonization by M. sabuleti, while the yellow areas are those in which the turf has grown high enough to allow the second ant species, M. scabrinodes to displace M. sabuleti. The green arrows indicate dispersal routes of M. sabuleti individuals as they seek to colonize new areas, and white areas are those with trees. Assume that all shaded areas contain wild thyme on which butterflies will deposit eggs. The blue areas will be source areas for the Big Blue Butterfly because they contain all the features needed for the survival of the larvae (initially wild thyme to feed on, followed by ant colonies to parasitize). Yellow areas will be sink areas, because the wrong ant species is dominant, preventing successful colonization. Red areas are recently burned or grazed and are therefore open for colonization by ants (and potential development into source areas for the butterfly); however, until successfully colonized by the ants, these are sink areas for the butterfly. In other words, as long as there are source areas containing nests of M. sabuleti the butterfly should be able to persist in the area.

However, the relationship is dynamic, with areas becoming unsuitable for the necessary ant if the turf is allowed to grow (change of blue to yellow), or becoming suitable for colonization (change of blue or yellow to red), or having been successfully colonized (change of red to blue). Consider the figure below which illustrates hypothetical changes in the environment. While all these changes are for the ant species, they will directly effect the Big Blue Butterfly.


metarole.gif (82036 bytes)

Consider. As long as there are blue areas populated by M. sabuleti, the butterfly will be successful. In those blue areas butterfly births will be greater than butterfly deaths (r is +), producing enough butterflies to both maintain the subpopulation in the blue area, and also to allow individuals to emigrate into the yellow or red areas. In the yellow and red areas the butterfly populations will have a negative r, since there will be little or no successful butterfly reproduction. For the butterflies, the blue areas will be source areas, while the yellow and red areas will be sink areas.

If, as happened in England, the source areas disappear (habitat loss for the ants) due to changes in the microclimate (tall turf), the butterfly will go extinct.

There can be many reasons for an area to be a sink for a given species. In the simplest case we may be dealing with an environment where some aspect of a species' niche is missing. This makes the local environment detrimental to the species in question. Examples of detrimental environmental factors include: lower food availability, lower water availability, poorer nutrient content of food, increased predation, lower available cover, decreased home range defensibility. Any of these would make the environment marginal for the species in question. In the case of the Blue Butterfly, food availability, in terms of the ant colonies which the butterfly larva parasitized, was the limiting factor in the environment.

In other cases, the subpopulation of a species in a given environmental patch may fluctuate in size due to stochastic effects (especially when the population is small – see next set of notes). This may lead to local extinction. However, a local extinction can be prevented by occasional immigrants arriving from neighboring populations. This is known as the Rescue Effect. The Rescue Effect can result in the proportion of habitat patches occupied being relatively constant through time, even though the populations in the individual patches may go extinct relatively frequently.

The Hawaiian Monk Seal – an example of a marginal habitat.

Hawaiian Monk Seals populated both tern and Green Islands. When the U S Coast Guard established bases on those islands the seals moved to nearby islands even though they were not being actively disturbed by the Coast Guard. On the other islands, however, the seal populations went into serious decline. There seemed to be no obvious reason for this until a careful study revealed that the seal population was becoming progressively older. In other words, adult seal survival had not much changed, but there were few young seals. The problem turned out to be the topography of the beaches on the other islands. On both Tern and Green Island the beaches extended high enough topographically to be above the reach of the breaking waves even at the highest tide level; whereas on the other islands the waves would break over the upper portion of the beach too. Seal pups, which are nursed on the beach, could move to higher areas on Tern and Green islands for protection against wave surge. The seal pups are poor swimmers, and on the other islands they were being dragged down the beach face and into the ocean by waves. Once in the water they were easy prey for waiting sharks. The seal population was declining because most of the young were being eaten. When the coast guard abandoned its stations the seals returned to Tern and Green Islands and the population recovered. If they had been forced to remain on the other islands they would have gone extinct due to high juvenile mortality.

Hawseals.jpg (71078 bytes)

Ecological Traps

Another type of sink area is what can be called an ecological trap. This is a habitat, which may be man-made, that attracts animals because it appears to fulfill all the organism’s environmental requirements; but is actually unsuitable for that organism because of some other factor which the animal is unable to recognize. In other words, it is a sink masquerading as a source. Examples of this are corn and soybean fields. These fields were once deeply plowed prior to planting, and that deep plowing made them unsuitable for bird nesting. Now only shallow plowing is used to help limit soil erosion. As a result, birds tend to nest in these fields, and suffer from exposure to herbicides, fertilizers, and ultimately from harvesting machines which destroy the nest before the young birds have grown enough to escape.



It may seem at this point that all we need do to insure the survival of a species is to identify source and sink areas and make sure that source areas are preserved. Unfortunately, that identification may be difficult, because we can also run into cases where habitats are pseudo-sinks. A pseudo-sink is just the opposite of a trap – and area which appears to be a sink, but is actually a source. Mistaken removal of a source habitat would have obviously deleterious effects on the species.

For an example of a pseudo-sink we can examine the case of another butterfly, the checkerspot (Euphydryus editha). This butterfly is found in the Sequoia National Forest in California, and lived on rocky outcrop areas surrounded by coniferous forests. There were two host plants on which the female butterflies laid their eggs (Pedicularis semibarbata and Castilleja disticha). Logging in the 1960’s caused a major habitat loss, and resulted in the butterfly changing their habitat to clear cuts where the females would oviposit on a new host plant (Collinsia torreyi).

By the 1980’s the butterflies occupied a habitat mosaic of both clear cuts and rocky outcrops, with each habitat offering different host plants. Survival, however, was higher on Collinsia torreyi in the clear cuts than it was on Pedicularis semibarbata on the rocky outcrops. It appeared clear that the clear cuts were the source areas and the rocky outcrops the sinks, since the migration rate from clear cut to outcrop was twice that from the outcrops to the clear cuts. Further, the butterfly density was greatest in the clear cuts and then decreased, from a high point on outcrops adjacent to clear cuts, as one moved away from the clear cuts across the outcrops. Again, an obvious source (clear cuts) - sink (outcrops) relationship.

Then between 1989 and 1992 a series of environmental perturbations killed off all of the host plants in the clear cuts. As would be expected, the larval density on the outcrops declined rapidly by about a third, especially on those outcrops near clear cuts which had previously exhibited the greatest larval density. But instead of continuing to local extinction, the population on the rocky outcrops stabilized after a few years. It turned out that the rocky outcrops were not a true sink. Instead it was a pseudo-sink where the population had been maintained above its carrying capacity (K) by immigration from the clear cut areas. Why?

The ecological conditions of the clear cuts and the life history of the host plant, Collinsia torreyi, were the main factors in this case. Collinsia torreyi dies off early, and is not as preferred by butterflies as is Pedicularis semibarbata. However, the early die off of Collinsia torreyi does not prevent the larvae from metamorphosing successfully because the sunny conditions in the clear cut result in an earlier onset of flight season among the butterflies. In other words, the butterflies had left Collinsia torreyi before the plant began to die and dry. These butterflies would then immigrate in large numbers towards the outcrops and their preferred host Pedicularis semibarbata. Butterflies in the outcrop area would begin their flight season later, with little reason to immigrate towards the clear cuts because the main food source there - Collinsia torreyi - would already have dried beyond edibility. So what had appeared to be immigration from source to sink was actually a movement towards a preferred and longer lasting food source.

  Return to Extinction Home Page

  Return to Extinction Notes Directory